0% found this document useful (0 votes)

43 views

Atlas Users1

Uploaded by

Алексей Завгородний

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

43 views

Atlas Users1

Uploaded by

Алексей Завгородний

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 1533

ATLAS User’s Manual

DEVICE SIMULATION SOFTWARE

SILVACO, Inc.
4701 Patrick Henry Drive, Bldg. 2 September 13, 2012
Santa Clara, CA 95054
Phone: (408) 567-1000
Web: www.silvaco.com
Notice

The information contained in this document is subject to change without notice.

SILVACO, Inc. MAKES NO WARRANTY OF ANY KIND WITH REGARD TO THIS
MATERIAL, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTY
OF FITNESS FOR A PARTICULAR PURPOSE.
SILVACO, Inc. shall not be held liable for errors contained herein or for incidental or
consequential damages in connection with the furnishing, performance, or use of this
material.
This document contains proprietary information, which is protected by copyright laws of the
United States. All rights are reserved. No part of this document may be photocopied,
reproduced, or translated into another language without the prior written consent of
SILVACO, Inc.
ACCUCELL, ACCUCORE, ACCUMODEL, ACCUTEST, ATHENA, ATHENA 1D, ATLAS,
BLAZE, C-INTERPRETER, CATALYSTAD, CATALYSTDA, CELEBRITY, CELEBRITY C++,
CIRCUIT OPTIMIZER, CLARITYRLC, CLEVER, DECKBUILD, DEVEDIT, DEVEDIT3D,
DEVICE3D, DISCOVERY, EDA OMNI, EDIF WRITER, ELITE, EXACT, EXPERT,
EXPERT200, EXPERTVIEWS, FERRO, GATEWAY, GATEWAY200, GIGA, GIGA3D, GUARDIAN,
GUARDIAN DRC, GUARDIAN LVS, GUARDIAN NET, HARMONY, HIPEX, HIPEX-C,
HIPEX-NET, HIPEX-RC, HYPERFAULT, LASER, LED, LED3D, LISA, LUMINOUS,
LUMINOUS3D, MAGNETIC, MAGNETIC3D, MASKVIEWS, MC DEPO/ETCH, MC DEVICE, MC
IMPLANT, MERCURY, MIXEDMODE, MIXEDMODE3D, MOCASIM, MODELLIB, NOISE,
NOMAD, OLED, OPTOLITH, ORGANIC DISPLAY, ORGANIC SOLAR, OTFT, PROMOST,
QUANTUM, QUANTUM3D, QUEST, REALTIMEDRC, RESILIENCE, S-PISCES, S-SUPREM3,
S-SUPREM4, SCOUT, SDDL, SFLM, SIC, SILOS, SIMULATION STANDARD, SMARTLIB,
SMARTSPICE, SMARTSPICERF, SMARTSPICE200, SMARTVIEW, SOLVERLIB, SPAYN, SPDB,
SPIDER, SPRINT, STELLAR, TCAD DRIVEN CAD, TCAD OMNI, TFT, TFT3D,
THERMAL3D, TONYPLOT, TONYPLOT3D, TWISTER, TWISTERFP, UTMOST, UTMOST III,
UTMOST IV, UTMOST IV- FIT, UTMOST IV- MEASURE, UTMOST IV- OPTIMIZATION,
VCSEL, VERILOG-A, VICTORY, VICTORY CELL, VICTORY DEVICE, VICTORY PROCESS,
VICTORY STRESS, VIRTUAL WAFER FAB, VWF, VWF AUTOMATION TOOLS, VWF
INTERACTIVE TOOLS, VWF MANUFACTURING TOOLS, and VYPER are trademarks of
SILVACO, Inc.
All other trademarks mentioned in this manual are the property of their respective owners.
Copyright © 1984 - 2012, SILVACO, Inc.

2 ATLAS User’s Manual

How to Read this Manual

Style Conventions

Font Style/Convention Description Example

• This represents a list of items or • Bullet A

terms. • Bullet B
• Bullet C
1. This represents a set of directions To open a door:
2. to perform an action. 1. Unlock the door by inserting
3. the key into keyhole.
2. Turn key counter-clockwise.
3. Pull out the key from the
keyhole.
4. Grab the doorknob and turn
clockwise and pull.
 This represents a sequence of FileOpen
menu options and GUI buttons to
perform an action.
Courier This represents the commands, HAPPY BIRTHDAY
parameters, and variables syntax.
Times Roman Bold This represents the menu options File
and buttons in the GUI.
New Century Schoolbook This represents the variables of x+y=1
Italics equations.
This represents the additional
Note: important information. Note: Make sure you save often when
working on a manual.

TIMES NEW ROMAN IN SMALL This represents the names of the ATHENA and ATLAS.
CAPS SILVACO products.

3 ATLAS User’s Manual

Table of Contents

15 ATLAS User’s Manual

1.2.2: Fully Integrated Capabilities

ATLAS works well with other software from SILVACO. For example, ATLAS
• Runs in the DECKBUILD interactive run-time environment.
• Is interfaced to TONYPLOT, the interactive graphics and analysis package.
• Accepts input from the ATHENA and SSUPREM3 process simulators.
• Is interfaced to UTMOST parameter extraction and device modeling software.
• can be used in experiments with the VWF AUTOMATION TOOLS.

SILVACO, Inc. 21
Features And Capabilities of ATLAS Introduction

1.2.3: Sophisticated Numerical Implementation

ATLAS uses powerful numerical techniques, including:
• Accurate and robust discretization techniques.
• Gummel, Newton, and block-Newton nonlinear iteration strategies.
• Efficient solvers, both direct and iterative, for linear subproblems.
• Powerful initial guess strategies.
• Small-signal calculation techniques that converge at all frequencies.
• Stable and accurate time integration.

22 ATLAS User’s Manual

Using ATLAS With Other Silvaco Software Introduction

1.3 Using ATLAS With Other Silvaco Software

ATLAS is best used with the VWF INTERACTIVE TOOLS. These include DECKBUILD,
TONYPLOT, DEVEDIT, MASKVIEWS, and OPTIMIZER. DECKBUILD provides an interactive run
time environment. TONYPLOT supplies scientific visualization capabilities. DEVEDIT is an
interactive tool for structure and mesh specification and refinement. MASKVIEWS is an IC
Layout Editor. The OPTIMIZER supports black box optimization across multiple simulators.
ATLAS, however, is often used with the ATHENA process simulator. ATHENA predicts the
physical structures that result from processing steps. The resulting physical structures are
used as input by ATLAS, which then predicts the electrical characteristics associated with
specified bias conditions. The combination of ATHENA and ATLAS makes it possible to
determine the impact of process parameters on device characteristics.
The electrical characteristics predicted by ATLAS can be used as input by the UTMOST
device characterization and SPICE modeling software. Compact models based on simulated
device characteristics can then be supplied to circuit designers for preliminary circuit design.
Combining ATHENA, ATLAS, UTMOST, and SMARTSPICE makes it possible to predict the
impact of process parameters on circuit characteristics.
ATLAS can also be used as one of the simulators within the VWF AUTOMATION TOOLS.
VWF makes it convenient to perform highly automated simulation-based experimentation.
VWF is used in a way that reflects experimental research and development procedures using
split lots. It therefore links simulation very closely to technology development, resulting in
significantly increased benefits from simulation use.

23 ATLAS User’s Manual

The Nature Of Physically-Based Simulation Introduction

1.4 The Nature Of Physically-Based Simulation

ATLAS is a physically-based device simulator. Physically-based device simulation is not a
familiar concept for all engineers. This section will briefly describe this type of simulation.
Physically-based device simulators predict the electrical characteristics that are associated
with specified physical structures and bias conditions. This is achieved by approximating the
operation of a device onto a two or three dimensional grid, consisting of a number of grid
points called nodes. By applying a set of differential equations, derived from Maxwell’s laws,
onto this grid you can simulate the transport of carriers through a structure. This means that
the electrical performance of a device can now be modeled in DC, AC or transient modes of
operation.
There are three physically-based simulation. These are:
• It is predictive.
• It provides insight.
• It conveniently captures and visualizes theoretical knowledge.
Physically-based simulation is different from empirical modeling. The goal of empirical
modeling is to obtain analytic formulae that approximate existing data with good accuracy
and minimum complexity. Empirical models provide efficient approximation and
interpolation. They do not provide insight, or predictive capabilities, or encapsulation of
theoretical knowledge.
Physically-based simulation has become very important for two reasons. One, it is almost
always much quicker and cheaper than performing experiments. Two, it provides information
that is difficult or impossible to measure.
The drawbacks of physically-based simulation are that all the relevant physics must be
incorporated into a simulator. Also, numerical procedures must be implemented to solve the
associated equations. These tasks have been taken care of for ATLAS users.
Those who use physically-based device simulation tools must specify the problem to be
simulated. In ATLAS, specify device simulation problems by defining:
• The physical structure to be simulated.
• The physical models to be used.
• The bias conditions for which electrical characteristics are to be simulated.
The subsequent chapters of this manual describe how to perform these steps.

24 ATLAS User’s Manual

Chapter 2
Getting Started with ATLAS
Overview Getting Started with ATLAS

2.1 Overview
ATLAS is a physically-based two and three dimensional device simulator. It predicts the
electrical behavior of specified semiconductor structures and provides insight into the internal
physical mechanisms associated with device operation.
ATLAS can be used standalone or as a core tool in SILVACO’s VIRTUAL WAFER FAB
simulation environment. In the sequence of predicting the impact of process variables on
circuit performance, device simulation fits between process simulation and SPICE model
extraction.
This chapter will show you how to use ATLAS effectively. It is a source of useful hints and
advice. The organization of topics parallels the steps that you go through to run the program.
If you have used earlier versions of ATLAS, you will still find this chapter useful because of
the new version.
This chapter concentrates on the core functionality of ATLAS. If you’re primarily interested
in the specialized capabilities of a particular ATLAS tool, read this chapter first. Then, read
the chapters that describe the ATLAS tools you wish to use.

26 ATLAS User’s Manual

ATLAS Inputs and Outputs Getting Started with ATLAS

2.2 ATLAS Inputs and Outputs

Figure 2-1 shows the types of information that flow in and out of ATLAS. Most ATLAS
simulations use two input files. The first input file is a text file that contains commands for
ATLAS to execute. The second input file is a structure file that defines the structure that will
be simulated.
ATLAS produces three types of output files. The first type of output file is the run-time
output, which gives you the progress and the error and warning messages as the simulation
proceeds. The second type of output file is the log file, which stores all terminal voltages and
currents from the device analysis. The third type of output file is the solution file, which
stores 2D and 3D data relating to the values of solution variables within the device at a given
bias point.

DevEdit

(Structure and
Mesh Editor) Runtime Output

Structure Files

ATHENA ATLAS Log Files

Device Simulator
(Process Simulator)
TonyPlot

Command File (Visualization

Tool)
Solution Files

DeckBuild

(Run Time Environment)

Figure 2-1 ATLAS Inputs and Outputs

27 ATLAS User’s Manual

Modes of Operation Getting Started with ATLAS

2.3 Modes of Operation

ATLAS is normally used in conjunction with the DECKBUILD run-time environment, which
supports both interactive and batch mode operation. We strongly recommend that you always
run ATLAS within DECKBUILD. In this section, we present the basic information you need to
run ATLAS in DECKBUILD. The DECKBUILD USER ’S MANUAL provides a more detailed
description of the features and capabilities of DECKBUILD.

2.3.1 Interactive Mode With DeckBuild

To start ATLAS in DECKBUILD, type:
deckbuild -as
at the UNIX system command prompt. The command line option, -as, instructs DECKBUILD
to start ATLAS as the default simulator.
If you want to start from an existing input file, start DECKBUILD by typing:
deckbuild -as <input filename>
The run-time output shows the execution of each ATLAS command and includes error
messages, warnings, extracted parameters, and other important output for evaluating each
ATLAS run. When ATLAS runs in this mode, the run-time output is sent to the output section
of the DeckBuild Window and can be saved as needed. Therefore, you don’t need to save the
run-time output explicitly. The following command line, however, specifies the name of a file
that will be used for storing the run-time output.
deckbuild -as <input filename> -outfile <output filename>
In this case, the run-time output is sent to the output file and to the output section of the
DeckBuild Window.

2.3.2 Batch Mode With DeckBuild

To use DECKBUILD in a non-interactive or batch mode, add the -run parameter to the
command that invokes DECKBUILD. A prepared command file is required for running in batch
mode. We advise you to save the run-time output to a file, since error messages in the run-
time output would otherwise be lost when the batch job completes. For example:
deckbuild -run -as <input filename> -outfile <output filename>
Using this command requires a local X-Windows system to be running. The job runs inside a
DECKBUILD icon on the terminal and quits automatically when the ATLAS simulation is
complete. You can also run DECKBUILD using a remote display. For example:
deckbuild -run -as <input file> -outfile <output file> 
-display<hostname>:0.0

28 ATLAS User’s Manual

Modes of Operation Getting Started with ATLAS

2.3.3 No Windows Batch Mode With DeckBuild

For completely non-X Windows operation of DECKBUILD, use the -ascii parameter. For
example:
deckbuild -run -ascii -as <input filename> -outfile <output
filename>
This command directs DECKBUILD to run the ATLAS simulation without the DeckBuild
Window or icon. This is useful for remote execution without an X Windows emulator or for
replacing UNIX-based ATLAS runs within framework programs.
When using batch mode, use the UNIX command suffix, &, to detach the job from the current
command shell. To run a remote ATLAS simulation under DECKBUILD without display and
then logout from the system, use the UNIX command, nohup , before the DECKBUILD
command line. For example:
nohup deckbuild -run -ascii -as <input filename> 
-outfile <output filename> &

2.3.4 Running ATLAS inside Deckbuild

Each ATLAS run inside DECKBUILD should start with the line:
go atlas
A single input file may contain several ATLAS runs each separated with a go atlas line.
Input files within DECKBUILD may also contain runs from other programs such as ATHENA
or DEVEDIT along with the ATLAS runs.
Running a given version number of ATLAS
The go statement can be modified to provide parameters for the ATLAS run. To run version
5.14.0.R, the syntax is:
go atlas simflags=”-V 5.14.0.R”
Starting Parallel ATLAS
The -P option is used to set the number of processors to use in a parallel ATLAS run. If the
number set by -P is greater than the number of processors available or than the number of
parallel thread licenses, the number is automatically reduced to this cap number. To run on 4
processors, use:
go atlas simflags=”-V 5.14.0.R -P 4”

29 ATLAS User’s Manual

Modes of Operation Getting Started with ATLAS

2.3.5 Batch Mode Without DeckBuild

You can run ATLAS outside the DECKBUILD environment. But this isn’t recommended by
SILVACO. If you don’t want the overhead of the DeckBuild Window, use the No Windows
Mode. Many important features such as variable substitution, automatic interfacing to process
simulation, and parameter extraction are unavailable outside the DECKBUILD environment. To
run ATLAS directly under UNIX, use the command:
atlas <input filename>
To save the run-time output to a file, don’t use the UNIX redirect command (>). Simply
specify the name of the output file. For example:
atlas <input filename> -logfile <output filename>

Note: The standard examples supplied with ATLAS will not run correctly outside of DECKBUILD.

2.3.6 TMA Compatibility Mode

You can add the -TMA command line flag to the atlas command to direct the ATLAS
simulator to operate in TMA Compatibility Mode. In this mode, the default material models
and parameters are altered to closely agree with those of TMA’s Medici simulator.

2.3.7 ISE Compatibility Mode

You can add the -ISE command line flag to the atlas command to direct the ATLAS
simulator to operate in ISE Compatibility Mode. In this mode, the default material models
and parameters are altered to closely agree with those of ISE’s Dessis simulator.

30 ATLAS User’s Manual

Accessing The Examples Getting Started with ATLAS

2.4 Accessing The Examples

ATLAS has a library of standard examples that demonstrate how the program is used to
simulate different technologies. These examples are a good starting point for creating your
own simulations. The examples are accessed from the menu system in DECKBUILD. To select
and load an example:
1. Start DECKBUILD with ATLAS as the simulator, which is described in the previous
section.
2. Use left mouse button to pull down the Main Control menu.
3. Select Examples. An index will then appear in a Deckbuild Examples Window (see
Figure 2-2).

Figure 2-2 Examples Index in DeckBuild

The examples are divided by technology or technology group. For instance, the most
common technologies are individually listed (e.g., MOS are under BJT), while others
are grouped with similar devices (e.g., IGBT and LDMOS are under POWER, and
solar cell and photodiode are under OPTOELECTRONICS).

31 ATLAS User’s Manual

Accessing The Examples Getting Started with ATLAS

4. Choose the technology by double clicking the left mouse button over that item. A list of
examples for that technology will appear. These examples typically illustrate different
devices, applications, or types of simulation.
You can also search for an example by selecting the Index button. Wildcards can be
used in the search.
5. Choose a particular example by double clicking the left mouse button over that item in the
list. A text description of the example will appear in the window. This text describes the
important physical mechanisms in the simulation and the details of the ATLAS syntax
used. You should read this information before proceeding.
6. Press the Load example button. The input command file for the example will be copied
into your current working directory together with any associated files. A copy of the
command file will be loaded into DECKBUILD. Note that the Load example button
remains faded out until this step is performed correctly.
7. Press the run button in the middle frame of the DECKBUILD application window to run
the example. Alternatively, most examples are supplied with results that are copied into
the current working directory along with the input file. To view the results, select
(highlight) the name of the results file and select Tools-Plot. See the TONYPLOT USER ’S
MANUAL for details on using TONYPLOT.

32 ATLAS User’s Manual

The ATLAS Syntax Getting Started with ATLAS

2.5 The ATLAS Syntax

An ATLAS command file is a list of commands for ATLAS to execute. This list is stored as
an ASCII text file that can be prepared in DECKBUILD or in any text editor. Preparing an input
file in DECKBUILD is preferred. You can create an input file by using the DeckBuild
Commands menu in the DeckBuild Window.

2.5.1 Statements and Parameters

The input file contains a sequence of statements. Each statement consists of a keyword that
identifies the statement and a set of parameters. The general format is:
<STATEMENT> <PARAMETER>=<VALUE>
With a few exceptions, the input syntax is not case sensitive. One important exception is that
commands described in this manual as being executed by DECKBUILD rather than ATLAS are
case sensitive. These include EXTRACT, SET, GO, and SYSTEM. Also, filenames for input and
output under UNIX are case sensitive.
For any <STATEMENT>, ATLAS may have four different types for the <VALUE> parameter.
These are: Real, Integer, Character, and Logical.
An example of a statement line is:
DOPING UNIFORM N.TYPE CONCENTRATION=1.0e16 REGION=1 OUTFILE=my.dop
The statement is DOPING. All other items are parameters of the DOPING statement. UNIFORM
and N.TYPE are Logical parameters. Their presence on the line sets their values to true.
Otherwise, they take their default values (usually false). CONCENTRATION is a Real
parameter and takes floating point numbers as input values. REGION is an Integer parameter
taking only integer numbers as input. OUTFILE is a Character parameter type taking strings as
input.
The statement keyword must come first but the order of parameters within a statement is
unimportant.
You only need to use enough letters of any parameter to distinguish it from any other
parameter on the same statement. Thus, CONCENTRATION can be shortened to CONC. REGION.
It can’t be shortened to R, however, since there’s a parameter called RATIO associated with the
DOPING statement.
You can set logical parameters to false by preceding them with the ^ symbol. Any line
beginning with a # is ignored. These lines are used as comments.
ATLAS can read up to 256 characters on one line. But it is better to spread long input
statements over several lines to make the input file more readable. The \ character at the end
of a line indicates continuation.
For more information about statements and parameters in ATLAS, see Chapter 21
“Statements”.

33 ATLAS User’s Manual

The ATLAS Syntax Getting Started with ATLAS

2.5.2 The Order of ATLAS Commands

The order in which statements occur in an ATLAS input file is important. There are five
groups of statements that must occur in the correct order (see Figure 2-3). Otherwise, an error
message will appear, which may cause incorrect operation or termination of the program. For
example, if the material parameters or models are set in the wrong order, then they may not be
used in the calculations.
The order of statements within the mesh definition, structural definition, and solution groups
is also important. Otherwise, it may also cause incorrect operation or termination of the
program.
.

Group Statements
MESH
1. Structure Specification REGION
ELECTRODE
DOPING

MATERIAL
2. Material Models Specification MODELS
CONTACT
INTERFACE

3. Numerical Method Selection METHOD

LOG
4. Solution Specification SOLVE
LOAD
SAVE

5. Results Analysis EXTRACT

TONYPLOT

Figure 2-3 ATLAS Command Groups with the Primary Statements in each Group

2.5.3 The DeckBuild Command Menu

The DeckBuild Command Menu (Command Menu) can help you to create input files. This
menu is found under the Commands button on DECKBUILD’s main screen. The Commands
Menu is configured for ATLAS whenever ATLAS is the currently active simulator in
DECKBUILD. When ATLAS is active, which is indicated in the lower bar of the DeckBuild
Window, an ATLAS command prompt will appear in the DECKBUILD output section.
The Command Menu gives you access to pop-up windows where you type information.
When you select the Write button, syntactically correct statements are written to the
DECKBUILD text edit region. The DeckBuild Command Menu does not support all possible
ATLAS syntax, but aims to cover the most commonly used commands.

34 ATLAS User’s Manual

The ATLAS Syntax Getting Started with ATLAS

2.5.4 PISCES-II Quick Start

This section is a quickstart for those who may be familiar with the syntax and use of the
Stanford University PISCES-II program or other device simulators derived from this
program.
The major differences between ATLAS and PISCES-II are:
• all graphics are handled by a separate interactive graphics program, TONYPLOT. By using
TONYPLOT, you no longer need to run the device simulator simply to plot or alter
graphics.
• no need to separate individual ATLAS simulations into separate input files. Multiple runs
of ATLAS are possible in the same input file separated by the line go atlas. There’s also
no need to separate process and device simulation runs of SILVACO products into
separate input files. A single file containing ATHENA and ATLAS syntax is permitted in
DECKBUILD.
• the interface from process to device simulation is handled though a single file format
compatible with other programs. The file read by ATLAS is the default output file format
of ATHENA. No special file format for the interface is required.
• when defining a grid structure within ATLAS, the NODE and LOCATION syntax to define
exact grid line numbers in X and Y is not recommended. A more reliable and easier to use
syntax using LOCATION and SPACING is available.
• using the REGRID command is not recommended due to the creation of obtuse triangles. A
standalone program, such as DEVEDIT, can be used as a grid pre-processor for ATLAS.
• all numerical method selection commands and parameters are on the METHOD statement.
The SYMBOLIC statement is not used. Historically, SYMBOLIC and METHOD were used as a
coupled pair of statements, but it is more convenient to use a single statement (METHOD)
instead. Most of the old parameters of the SYMBOLIC statement have the same meaning
and names, despite this move to a single statement. One notable change in ATLAS is that
you can combine numerical methods together.
See the “Pisces-II Compatibility” on page 79 for more information concerning the
translation of PISCES-II numerics statements.
• various general purpose commands are actually part of the DECKBUILD user environment.
These include SET, EXTRACT, GO, SYSTEM, and SOURCE. These commands can be
interspersed inside ATLAS syntax.
• Variable substitution is supported for both numerical and string variables using the SET
statement and the $ symbol. To avoid confusion, the # symbol is preferred over the $
symbol for comment statements.
In addition to these changes, the physical models are generally different in ATLAS. Most of
the original PISCES-II models have been preserved but often are not the default or the
recommended models to use. See the on-line examples for technology specific information
about models.

35 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

2.6 Defining A Structure

There are three ways to define a device structure in ATLAS.
The first way is to read an existing structure from a file. The structure is created either by an
earlier ATLAS run or another program such as ATHENA or DEVEDIT. A MESH statement
loads in the mesh, geometry, electrode positions, and doping of the structure. For example:
MESH INFILE=<filename>
The second way is to use the Automatic Interface feature from DECKBUILD to transfer the
input structure from ATHENA or DEVEDIT.
The third way is create a structure by using the ATLAS command language. See Chapter 21
“Statements” for more information about the ATLAS syntax.

2.6.1 Interface From ATHENA

When ATHENA and ATLAS are run under DECKBUILD, you can take advantage of an
automatic interface between the two programs. Perform the following steps to load the
complete mesh, geometry, and doping from ATHENA to ATLAS.
1. Deposit and pattern electrode material in ATHENA.
2. Use the ELECTRODE statement in ATHENA to define contact positions. Specify the X and
Y coordinates as cross-hairs to pin-point a region. The whole region is then turned into
electrode. In many cases, only the X coordinate is needed. For example:
ELECTRODE NAME=gate X=1.3 [Y=-0.1])
There is a special case to specify a contact on the bottom of the structure. For exam-
ple:
ELECTRODE NAME=substrate BACKSIDE
3. Save a structure file while ATHENA is still the active simulator. For example:
STRUCTURE OUTF=nmos.str
4. Start ATLAS with the go atlas command written in the same input deck. This will
automatically load the most recent structure from ATHENA into ATLAS.
If you need to load the structure saved in step 4 into ATLAS without using the auto-interface
capability, use the MESH command. For example:
MESH INF=nmos.str
ATLAS inherits the grid used most recently by ATHENA. With a careful choice of initial
mesh or by using the grid manipulation techniques in ATHENA, you can produce a final
mesh from ATHENA that will give good results in ATLAS. But, a grid that is appropriate for
process simulation isn’t always appropriate for device simulation. If the final ATHENA mesh
is inappropriate for ATLAS, use either DEVEDIT to re-mesh the structure or the REGRID
command.

Note: There’s no need to specify a MESH command in ATLAS when using the Automatic Interface of
DECKBUILD.

36 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

2.6.2 Interface From DevEdit

A 2D or 3D structure created by DEVEDIT can be read into ATLAS using the following
statement.
MESH INF=<structure filename>
This statement loads in the mesh, geometry, electrode positions, and doping of the structure.
ATLAS will automatically determine whether the mesh is 2D for S-PISCES or BLAZE, or
3D for DEVICE3D or BLAZE3D.
If the structure coming from DEVEDIT were originally created by ATHENA, then define the
electrodes in ATHENA as described in the previous section. If the structure is created in
DEVEDIT, the electrode regions should be defined in the Region/Add region menu of
DEVEDIT.

2.6.3 Using The Command Language To Define A Structure

To define a device through the ATLAS command language, you must first define a mesh. This
mesh or grid covers the physical simulation domain. The mesh is defined by a series of
horizontal and vertical lines and the spacing between them. Then, regions within this mesh
are allocated to different materials as required to construct the device. For example, the
specification of a MOS device requires the specification of silicon and silicon dioxide
regions. After the regions are defined, the location of electrodes is specified. The final step is
to specify the doping in each region.
When using the command language to define a structure, the information described in the
following four sub-sections must be specified in the order listed.
Specifying The Initial Mesh
The first statement must be:
MESH SPACE.MULT=<VALUE>
This is followed by a series of X.MESH and Y.MESH statements.
X.MESH LOCATION=<VALUE> SPACING=<VALUE>
.
Y.MESH LOCATION=<VALUE> SPACING=<VALUE>
.
The SPACE.MULT parameter value is used as a scaling factor for the mesh created by the
X.MESH and Y.MESH statements. The default value is 1. Values greater than 1 will create a
globally coarser mesh for fast simulation. Values less than 1 will create a globally finer mesh
for increased accuracy. The X.MESH and Y.MESH statements are used to specify the locations
in microns of vertical and horizontal lines, respectively, together with the vertical or
horizontal spacing associated with that line. You must specify at least two mesh lines for each
direction. ATLAS automatically inserts any new lines required to allow for gradual transitions
in the spacing values between any adjacent lines. The X.MESH and Y.MESH statements must
be listed in the order of increasing x and y. Both negative and positive values of x and y are
allowed.

37 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

Figure 2-4 illustrates how these statements work. On the left hand plot, note how the spacing
of the vertical lines varies from 1 m at x=0 and x=10 m to 0.5 m at x=5 m. On the right
hand plot, note how specifying the SPACE.MULT parameter to have a value of 0.5 has doubled
the density of the mesh in both the X and Y directions.
You can specify the PERIODIC parameter on the MESH statement to mean that the structure
and mesh are periodic in the x direction.

Figure 2-4 Non-uniform Mesh Creation using ATLAS Syntax

After an initial mesh has been defined, you can remove grid lines in specified regions. This is
typically done in regions of the device where a coarse grid is expected to be sufficient such as
the substrate. The removal of grid lines is accomplished using the ELIMINATE statement. The
ELIMINATE statement removes every second mesh line in the specified direction from within
a specified rectangle. For example, the statement:
ELIMINATE COLUMNS X.MIN=0 X.MAX=4 Y.MIN=0.0 Y.MAX=3
removes every second vertical grid line within the rectangle bounded by x=0, x=4, y=0 and
y=3 microns.
Specifying Regions And Materials
Once the mesh is specified, every part of it must be assigned a material type. This is done
with REGION statements. For example:
REGION number=<integer> <material_type> <position parameters>
Region numbers must start at 1 and are increased for each subsequent region statement. You
can have up to 200 different regions in ATLAS. A large number of materials is available. If a
composition-dependent material type is defined, the x and y composition fractions can also be
specified in the REGION statement.
The position parameters are specified in microns using the X.MIN, X.MAX, Y.MIN, and
Y.MAX parameters. If the position parameters of a new statement overlap those of a previous
REGION statement, the overlapped area is assigned as the material type of the new region.
Make sure that materials are assigned to all mesh points in the structure. If this isn’t done,
error messages will appear and ATLAS won’t run successfully.

38 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

You can use the MATERIAL statement to specify the material properties of the defined regions.
But you must complete the entire mesh and doping definition before any MATERIAL
statements can be used. The specification of material properties is described in Section 2.7.2
“Specifying Material Properties”.
Cylindrical Coordinates
Cylindrical coordinates are often used when simulating discrete power devices. In this mode,
ATLAS operates with x=0 as the axis of symmetry around which the cylindrical geometry is
placed. Many of the default units change when cylindrical coordinates are used. The
calculated current is in Amps rather than the usual Amps per micron. External elements are
specified in absolute units (e.g., Farads, not Farads/micron for capacitors).
The MESH statement must be used to specify cylindrical symmetry. The following statement
creates a mesh, which contains cylindrical symmetry.
MESH NX=20 NY=20 CYLINDRICAL
There are 20 mesh nodes along the X axis and 20 mesh nodes along the Y axis.
The following statement imports a mesh, which contains cylindrical symmetry.
MESH INF=mesh0.str CYLINDRICAL

Note: The CYLINDRICAL parameter setting isn’t stored in mesh files. Therefore, this parameter must be
specified each time a mesh file, which contains cylindrical symmetry, is loaded.

Specifying Electrodes
Once you have specified the regions and materials, define at least one electrode that contacts
a semiconductor material. This is done with the ELECTRODE statement. For example:
ELECTRODE NAME=<electrode name> <position_parameters>
You can specify up to 50 electrodes. The position parameters are specified in microns using
the X.MIN, X.MAX, Y.MIN, and Y.MAX parameters. Multiple electrode statements may have
the same electrode name. Nodes that are associated with the same electrode name are treated
as being electrically connected.
Some shortcuts can be used when defining the location of an electrode. If no Y coordinate
parameters are specified, the electrode is assumed to be located on the top of the structure.
You also can use the RIGHT, LEFT, TOP, and BOTTOM parameters to define the location. For
example:
ELECTRODE NAME=SOURCE LEFT LENGTH=0.5
specifies the source electrode starts at the top left corner of the structure and extends to the
right for the distance LENGTH.
Specifying Doping
You can specify analytical doping distributions or have ATLAS read in profiles that come
from either process simulation or experiment. You specify the doping using the DOPING
statement. For example:
DOPING <distribution_type> <dopant_type> <position_parameters>

39 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

Analytical Doping Profiles

Analytical doping profiles can have uniform, gaussian, or complementary error function
forms. The parameters defining the analytical distribution are specified in the DOPING
statement. Two examples are shown below with their combined effect shown in Figure 2-5.
DOPING UNIFORM CONCENTRATION=1E16 N.TYPE REGION=1
DOPING GAUSSIAN CONCENTRATION=1E18 CHARACTERISTIC=0.05 P.TYPE \
X.LEFT=0.0 X.RIGHT=1.0 PEAK=0.1
The first DOPING statement specifies a uniform n-type doping density of 1016 cm-3 in the
region that was previously labelled as region #1. The position parameters X.MIN, X.MAX,
Y.MIN, and Y.MAX can be used instead of a region number.
The second DOPING statement specifies a p-type Gaussian profile (see Equation 21-3) with a
peak concentration of 1018 cm-3. This statement specifies that the peak doping is located along
a line from x = 0 to x = 1 microns. Perpendicular to the peak line, the doping drops off
according to a Gaussian distribution with a standard deviation of ( 0.05  2 ) µm. At x < 0 or
x > 1, the doping drops off laterally with a default standard deviation that is ( 70  2 )% of
CHARACTERISTIC. This lateral roll-off can be altered with the RATIO.LATERAL parameter. If
a Gaussian profile is being added to an area that was already defined with the opposite dopant
type, you can use the JUNCTION parameter to specify the position of the junction depth
instead of specifying the standard deviation using the CHARACTERISTIC parameter.

Figure 2-5 Analytical specification of a 2D Profile

40 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

The other analytical doping profile available is the complementary error function. This is
defined as

z exp  –y
2 2
erfc  z  = -------  dy 2-1


where the z variable is the distance scaled by the characteristic distance defined by the CHAR
parameter.
The following example show DOPING statements that use this analytical form.
DOPING ERFC N.TYPE PEAK=0.5 JUNCTION=1.0 CONC=1.0E19 X.MIN=0.25 \
X.MAX=0.75 RATIO.LAT=0.3 ERFC.LAT
DOPING P.TYPE CONC=1E18 UNIFORM

This sets up a donor profile with a peak concentration of 1.0E19 cm-3 at X = 0.5 microns.
The CHAR parameter, which determines the rate of change of the doping level with distance, is
not directly set on the DOPING profile. Instead, it is calculated so that the net doping level at
the position given by the JUNCTION parameter is zero. In this example, the acceptor
concentration is 1.01018 cm-3 everywhere and so we require a donor density of 1.01018 cm-
3
at a position of 1 micron to create the p-n junction there. The value of CHAR is calculated
from the formula
erfc   JUNCTION – PEAK   CHAR  = 0.1 2-2
which results in a value of CHAR of approximately 0.43 microns.
Additionally, the donor concentration falls off in the lateral direction outside the range of 0.25
to 0.75 microns. The lateral falloff parameter is defined to be 0.3 times the principal falloff
parameter and has the shape of the complementary error function.
Importing 1D SSUPREM3 Doping Profiles
One-dimensional doping profiles can be read into ATLAS from a SSUPREM3 output file.
The doping data must have been saved from SSUPREM3 using the statement:
STRUCTURE OUTFILE=<output filename>
at the end of the SSUPREM3 run.
In ATLAS, the MASTER parameter of the DOPING statement specifies that a SSUPREM3 file
will be read by ATLAS. Since this file will usually contain all the dopants from the
SSUPREM3 simulation, the desired dopant type must also be specified. For example, the
statement:
DOPING MASTER INFILE=mydata.dat BORON REGION=1
specifies that the boron profile from the file mydata.dat should be imported and used in
region #1. SSUPREM3 profiles are imported into ATLAS one at a time (i.e., one DOPING
statement is used for each profile or dopant). The statements:
DOPING MASTER INFILE=mydata.dat BORON OUTFILE=doping.dat
DOPING MASTER INFILE=mydata.dat ARSENIC X.RIGHT=0.8 RATIO=0.75
DOPING MASTER INFILE=mydata.dat ARSENIC X.LEFT=2.2 RATIO=0.75

41 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

offset the arsenic doping from boron to create a 2D doping profile from a single SSUPREM3
result.
It is advisable to include the OUTFILE parameter on the first DOPING statement to create a
2D3D doping file. This file will then be used in the next section to interpolate doping on a
refined mesh after a REGRID. This file, however, can’t be plotted in TONYPLOT. The position
parameters and the RATIO.LATERAL parameter are used in the same manner as for analytical
doping profiles to set the extent of the 1D profile.

2.6.4 Automatic Meshing (Auto-meshing) Using The Command Language

Automatic meshing provides a simpler method for defining device structures and meshs than
the standard method described in Section 2.6.3 “Using The Command Language To Define A
Structure”. Auto-meshing is particularly suited for epitaxial structures, especially device
structures with many layers (for example, a VCSEL device). Auto-meshing unburdens you
from the delicate bookkeeping involved in ensuring that the locations of mesh lines in the Y
direction are consistently aligned with the edges of regions. This is done by specifying the
locations of Y mesh lines in the REGION statements. The following sections will show how
auto-meshing is done, using a few simple examples.
Specifying The Mesh And Regions
In the first example, we use a simple device to show you the fundamental concepts of auto-
meshing. The first statements in this example are as follows:
MESH AUTO
X.MESH LOCATION=-1.0 SPACING=0.1
X.MESH LOCATION=1.0 SPACING=0.1
These statements are similar to the ones used in the standard method that described a mesh
using the command language as shown in Section 2.6.3 “Using The Command Language To
Define A Structure”. There are, however, two key differences. The first difference is the
inclusion of the AUTO parameter in the MESH statement. You need this parameter to indicate
that you want to use auto-meshing. The second and more important difference is that in this
example we will not specify any Y.MESH statements. This is because the locations of Y mesh
lines will be automatically determined by the parameters of the REGION statements.
You can still specify one or more Y.MESH statements. Such defined mesh lines will be
included in the mesh. But including Y.MESH statements is optional in auto-meshing.
In the next few statements of the example, we will show you several new concepts that will
explain how auto-meshing works. The following four lines describe the regions in the
example device:
REGION TOP THICKNESS=0.02 MATERIAL=GaN NY=5 DONOR=1E16
REGION BOTTOM THICKNESS=0.1 MATERIAL=AlGaN NY=5 DONOR=1E17
X.COMP=0.2
REGION TOP THICKNESS=0.08 MATERIAL=AlGaN NY=4 ACCEPTOR=1E17
X.COMP=0.2
REGION BOTTOM THICKNESS=0.5 MATERIAL=AlGaN NY=10 DONOR=1E18
X.COMP=0.2

42 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

New Concepts
First, it appears that composition and doping are being specified in the REGION statement.
This is the case for the DONOR, ACCEPTOR, X.COMPOSITION, and Y.COMPOSITION parameters
in the REGION statement that specify uniform doping or composition or both over the
specified region. These parameters are also available to the standard methods described in
Section 2.6.3 “Using The Command Language To Define A Structure” but are more
amenable to specification of epitaxial structures such as we are describing in this example.
Next, you should notice several other new parameters. These are the TOP, BOTTOM,
THICKNESS, and NY parameters. All of these are used to describe the relative locations and
thicknesses of the layers as well as the locations of the Y mesh lines. The most intuitive of
these parameters is the THICKNESS parameter, which describes the thickness in microns, in
the Y direction of each layer. As for the extent in the X direction, in the absence of any
specified X.MIN or X.MAX parameters, it is assumed that the region extends over the full range
of the X mesh described above in the X.MESH statements.
The NY parameter describes how many Y mesh lines are contained in the region so that the Y
mesh lines are evenly spaced over the region. You can use the SY parameter instead of NY to
specify the spacing in microns between Y mesh lines in the region. Make sure the value of SY
does not exceed the value of THICKNESS. Generally, the relationship between SY, NY and
THICKNESS can be expressed by the following:
SY = THICKNESS/NY
Figure 2-6 shows the meanings of the TOP and BOTTOM parameters.

43 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

a) Structure after 1st REGION Statement b) Structure after 2nd REGION Statement

c) Structure after 3rd REGION Statement d) Structure after 4th REGION Statement
Figure 2-6 Simple Example of Auto-meshing Showing Sequence of Structure Development
This figure shows a progression of representations of how the structure’s mesh and region
outlines appear after each REGION statement. This figure should give you an intuitive feel of
how the regions are alternately placed on the top or bottom of the structure according to the
specification of the TOP or BOTTOM parameters. It is important to keep in mind that the
ATLAS coordinate convention for the Y axis is that positive Y is directed down into the
device. This is similar to using the TOP and BOTTOM parameters of the ELECTRODE statement.
One thing you might notice in this figure is that the number of Y mesh lines in each region
does not always match the number specified. This is because at each interface between
regions, the Y mesh line spacing is ambiguously defined and the auto-meshing algorithm will
always pick the smaller spacing between the two at each interface. Then, the spacing between
Y mesh lines varies continuously between subsequent interfaces in a similar way as it does
for mesh spacings specified by the LOCATION and SPACING parameters in the X.MESH and
Y.MESH statements.

44 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

The auto-meshing algorithm maintains the "notion" of the current Y locations of the "top" and
"bottom". Let’s call these locations "Ytop" and "Ybottom".
Before any REGION statements are processed, "Ytop" and "Ybottom" are both defined to be
equal to zero. As the REGION statements are processed, the following cases are addressed:
• If you place the region on the top, as specified by the TOP parameter, the region will
extend from "Ytop" to "Ytop"-THICKNESS (remember positive Y points down) and
"Ytop" will move to the new location "Ytop"-THICKNESS.
• If you place the region on the bottom, as specified by the BOTTOM parameter, the region
will extend from "Ybottom" to "Ybottom"+THICKNESS and "Ybottom" will move to
the new location "Ybottom"+THICKNESS.
The auto-meshing algorithm will ensure that all regions are perfectly aligned to their
neighboring regions and there are no inconsistencies between the locations of Y mesh lines
and the region edges that they resolve.

45 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

Non-uniformity In The X Direction and Auto-meshing

In some cases, you may want to define a device with material discontinuities in the X
direction. Such discontinuities may represent etched mesa structures or oxide apertures.
There are a couple of ways to do this in auto-meshing. For example, you want to etch out a
region from the structure of the previous example, from X=0 to the right and from Y=0 to the
top. You can add the statement
REGION MATERIAL=Air X.MIN=0.0 Y.MAX=0.0
In this statement, we are not using the auto-meshing features but are using syntax of the
standard meshing methods described in Section 2.6.3 “Using The Command Language To
Define A Structure”. ATLAS supports mixing the standard syntax with auto-meshing but be
careful when doing this as we will explain shortly. Figure 2-7 shows the resulting structure
and mesh after adding this statement. In this figure, we see that we have obtained the desired
result.

Figure 2-7 Structure Etch Example in Auto-meshing

The potential pitfall in using absolute Y coordinates in auto-meshing is that the location you
choose, for example, by summing up thicknesses may not match the location your computer
has calculated with its inherent numerical imprecision. The result is not only that the resulting
structure may not exactly match what you desire. More importantly, you may end up
accidently creating a mesh with Y mesh lines closely spaced (usually in the order of the
machine precision). This can cause numerical instability (poor convergence) and can
overflow or underflow in certain models. What’s worse is that this situation is difficult to
detect.

46 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

There is, however, one situation where we can absolutely predict the location of an edge or Y
mesh line. That is at Y=0. This is exactly what we have done in our example. So if you want
to use auto-meshing with specifications of an absolute value of Y, then arrange your device
structure so that the specification of Y will be at zero.
This also applies to the location of a recessed electrode. Make sure it is located at Y=0 if you
are using auto-meshing.
There is another method of providing for discontinuities in material in the X direction that is
absolutely safe from the previously discussed problems. In this approach, we use another
parameter called STAY in the REGION statement in conjunction with the TOP or BOTTOM
parameters.
The following describes the effect of the STAY parameter in the REGION statement.
• If you place the region on the top, as specified by the TOP parameter, and the STAY
parameter is specified, the region will extend from "Ytop" to "Ytop"-THICKNESS and
"Ytop" will remain in its current position.
• If you place the region on the bottom, as specified by the BOTTOM parameter, and the STAY
parameter is specified, the region will extend from "Ybottom" to
"Ybottom"+THICKNESS and "Ybottom" will remain in its current position.
The use of the STAY parameter can best be illustrated by the following example. In this
example, we will reproduce the same structure discussed in the last example but this time
using only STAY parameters and by not using any specification of absolute Y coordinates. The
new REGION specifications are as follows:
REGION BOTTOM THICKNESS=0.1 MATERIAL=AlGaN NY=5 DONOR=1E17
X.COMP=0.2
REGION BOTTOM THICKNESS=0.5 MATERIAL=AlGaN NY=10 DONOR=1E18
X.COMP=0.2
REGION TOP STAY THICKNESS=0.02 MATERIAL=GaN NY=5 DONOR=1E16
REGION TOP THICKNESS=0.02 MATERIAL=Air NY=5 X.MIN=0.0
REGION TOP STAY THICKNESS=0.08 MATERIAL=AlGaN NY=4 ACCEPTOR=1E17
X.COMP=0.2
REGION TOP THICKNESS=0.08 MATERIAL=Air NY=4 X.MIN=0.0

47 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

In this example, we slightly rearranged the REGION statements for clarity and split one region
into two. Figure 2-8 shows the results.

Figure 2-8 Structure Etching Example Using The STAY Parameter

The following describes the operation of the STAY parameter in this example:
• The STAY parameter in the third REGION statement means that the fourth REGION will start
at the same Y coordinate as the third.
• Since the THICKNESS and NY parameters are the same in the third and fourth regions, they
will have the same range of Y values and the same Y mesh.
• The specification of X.MIN in the fourth REGION statement means that the region will not
completely overlap the third region but will only overlap to a minimum X value of 0.
• The lack of a STAY parameter in the fourth REGION statement means that the fifth region
will lie directly atop the third and fourth regions.
• The STAY parameter in the fifth REGION statement means that the sixth REGION will start
at the same Y coordinate as the fifth.
• Since the THICKNESS and NY parameters are the same in the fifth and sixth regions, they
will have the same range of Y values and the same Y mesh.
• The specification of X.MIN in the sixth REGION statement means that the region will not
completely overlap the fifth region but will only overlap to a minimum X value of 0.
As you can see, using the STAY parameter carefully avoids the problems of specifying values
of Y coordinates and the potential problems involved. By doing so, you can specify arbitrary
stepped structures.

48 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

Grading of Composition and Doping

We have provided another method for specifying composition or doping or both in the
REGION statement that is available for both auto-meshing and meshing using the standard
method described in Section 2.6.3 “Using The Command Language To Define A Structure”.
This method allows linear grading of rectangular regions. Eight new parameters of the
REGION statement support this function. They are ND.TOP, ND.BOTTOM, NA.TOP, NA.BOTTOM,
COMPX.TOP, COMPX.BOTTOM, COMPY.TOP, and COMPY.BOTTOM. In the syntax of these
parameter names, "TOP" refers to the "top" or extreme Y coordinate in the negative Y
direction, and "BOTTOM" refers to the "bottom" or extreme Y coordinate in the positive Y
direction. With this in mind, the following rules apply:
• If you specify ND.TOP and ND.BOTTOM in a REGION statement, the donor doping in the
region will vary linearly from the value specified by ND.TOP at the "top" of the device to
the value specified by ND.BOTTOM at the "bottom" of the device.
• If you specify NA.TOP and NA.BOTTOM in a REGION statement, the acceptor doping in the
region will vary linearly from the value specified by NA.TOP at the "top" of the device to
the value specified by NA.BOTTOM at the "bottom" of the device.
• If you specify COMPX.TOP and COMPX.BOTTOM in a REGION statement, the X composition
fraction in the region will vary linearly from the value specified by COMPX.TOP at the
"top" of the device to the value specified by COMPX.BOTTOM at the "bottom" of the device.
• If you specify COMPY.TOP and COMPY.BOTTOM in a REGION statement, the Y composition
fraction in the region will vary linearly from the value specified by COMPY.TOP at the
"top" of the device to the value specified by COMPY.BOTTOM at the "bottom" of the device.

Note: Any subsequent DOPING statements will add to the doping specified by the doping related parameters on
the REGION statement.

Superlattices and Distributed Bragg Reflectors DBRs

For auto-meshing, we have provided a convenient way of specifying a certain class of
superlattices or as they are commonly used distributed Bragg reflectors (DBRs). This class of
superlattice includes any superlattice that can be described as integer numbers of repetitions
of two different layers. By different, we mean that the two layers may have different
thicknesses, material compositions, or dopings, or all. These "conglomerate" structures are
specified by the DBR or SUPERLATTICE statement. Actually, SUPERLATTICE is a synonym
for DBR. Therefore in the following discussion, we will use the DBR syntax and recognize that
SUPERLATTICE and DBR can be used interchangeably.
Most of the syntax of the DBR statement is similar to the syntax of the REGION statement,
except the syntax is set up to describe two regions (or two sets of regions). As you will see,
we differentiate these regions (or sets of regions) by the indices 1 and 2 in the parameter’s
name.
The following example should make this concept clear.
MESH AUTO
X.MESH LOCATION=-1.0 SPACING=0.1
X.MESH LOCATION=1.0 SPACING=0.1

49 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

DBR TOP LAYERS=4 THICK1=0.1 THICK2=0.2 N1=2 N2=3 MAT1=GaAs

MAT2=AlGaAs \
X2.COMP=0.4
DBR BOTTOM LAYERS=3 THICK1=0.05 THICK2=0.075 N1=3 N2=3 MAT1=AlGaAs
\
MAT2=GaAs X1.COMP=0.4
In this example, we can see we are using auto-meshing because of the appearance of the AUTO
parameter in the MESH statement. The DBR statements should be familiar since the
functionality and syntax are similar to the REGION statement. We will not discuss those
similarities here. We will only discuss the important differences. For more information about
the similarities, see Chapter 21 “Statements” .
The main new parameter is the LAYERS parameter. This parameter specifies the total number
of regions added. The region indexed "1" is always added first, then the region indexed "2"
is added, and depending on the value of LAYERS, the sequence of regions continues 1, 2, 1, 2,
1, ...
Figure 2-9 shows the functionality of the DBR statement, which gives the resulting structure
and mesh from the example.

Figure 2-9 Supperlattice Example Structure

50 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

2.6.5 Modifying Imported Regions

If you import a device structure from a file using the INFILE parameter of the MESH
statement, you may want to modify some of the characteristics of one or more of the regions
in the structure. To do this, specify a REGION statement with the MODIFY parameter and the
NUMBER or NAME parameter assigned to the region number of interest. You can specify/
respecify any of the following REGION statement parameters: STRAIN, WELL.CNBS,
WELL.VNBS, WELL.GAIN, POLAR.SCALE, QWELL, LED, WELL.FIELD, and POLARIZATION.

2.6.6 Remeshing Using The Command Language

It can be difficult to define a suitable grid when specifying a structure with the command
language. The main problem is that the meshes required to resolve 2D doping profiles and
curved junctions are quite complicated. Simple rectangular meshes require an excessive
number of nodes to resolve such profiles. If a device structure only includes regions of
uniform doping, then there’s usually no need to regrid. But when realistic 2D doping profiles
are present, a regrid may be necessary.

Note: The recommended solution for defining complex mesh structures for ATLAS is to use the standalone
program, DEVEDIT.

Regrid On Doping
ATLAS includes a regridding capability that generates a fine mesh only in a localized region.
You specify a quantity on which the regrid is to be performed. The mesh is then refined in
regions where the specified quantity varies rapidly. Whenever a specified quantity (usually
doping) changes quickly, the regridding will automatically grade the mesh accordingly. You
can get a regrid on doping before any solutions are obtained. You can do this by using the
statement:
REGRID LOGARITHM DOPING RATIO=2 SMOOTH.KEY=4 DOPFILE=<filename1> \
OUTFILE=<filename2>
This statement must be used after the MESH, REGION, MATERIAL, ELECTRODE, and DOPING
statements described previously. The effects of this REGRID statement on a simple diode
structure are shown in Figure 2-10.

51 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

Figure 2-10 Regrid on doping provides improved resolution of junction

In this statement, the regridding will resolve doping profiles to two orders of magnitude in
change.
The doping file, filename1, must be specified in the first DOPING statement with the
OUTFILE parameter. The results of the regrid are saved in the file, filename2. The
SMOOTH.KEY parameter value selects a smoothing algorithm. A value of 4 is typically best as
this algorithm tends to produce the fewest obtuse triangles. For a complete description of the
various smoothing algorithms, see Chapter 20 “Numerical Techniques”.
Regrid Using Solution Variables
The REGRID statement can use a wide range of solution variables as the basis for mesh
refinement. A regrid on solution variables can only be used after a solution has been obtained.
After a regrid on a solution variable, the solution must be re-solved at the same bias in
ATLAS.
For example:
REGRID POTENTIAL RATIO=0.2 MAX.LEVEL=1 SMOOTH.K=4
DOPFILE=<filename1>
SOLVE PREV
Regrid on potential is often used for high voltage power devices.

52 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

Note: You can use the REGRID statement any number of times on a structure. But we advise you to quit and
restart ATLAS between regrids on electrical quantities. You can use the go atlas statement to do this.
This should be followed by a MESH statement, loading the output file of the REGRID command, and a re-
setting of all material and model parameters.

2.6.7 Specifying 2D Circular Structures

In some situations, it may be desirable to construct a device with a circular cross section.
This is possible by using DEVEDIT and by using the MESH statement in the ATLAS command
language. Make sure to put the MESH statement first in the input deck and specify the
CIRCULAR parameter.
The radial mesh spacing is then given by a series of R.MESH statements and the angular mesh
given by a series of A.MESH statements. The angles are specified in degrees between 0 and
360.
For example
MESH CIRCULAR

R.MESH LOC=0.0 SPAC=0.05

R.MESH LOC=0.2 SPAC=0.05
R.MESH LOC=0.35 SPAC=0.025
R.MESH LOC=0.39 SPAC=0.01
R.MESH LOC=0.4 SPAC=0.01

A.MESH LOC=0 SPAC=30

A.MESH LOC=360 SPAC=30
would create a circular mesh with major angular spacing of 30° and a radial spacing, which is
relatively coarse near to the origin and becomes finer towards the edge of the device.
The angular spacing defines a set of major spokes radiating out from the origin with the
number of elements between each spoke increasing with distance from the origin. The mesh
spacing may be irregular, although a regular mesh spacing suffices for most problems.
The spacing in the radial direction will in general be irregular. This can result in the creation
of obtuse elements, particularly if the spacing decreases rapidly with increasing radius. The
MINOBTUSE parameter on the MESH statement deploys an algorithm to reduce the occurence of
obtuse elements, particularly obtuse boundary elements. It is set to true by default, but you
can clear it by using the syntax ^MINOBTUSE.
For some device structures, you can reduce the size of the solution domain by using
symmetry arguments. ATLAS allows you to create a wedge shaped device by one of two
methods. In the first method, you specify more than one A.MESH statement, and the maximum
angle specified in these statements is the angle of the wedge. In the second method, you only
specify one A.MESH statement with a location of zero degrees and the required angular
spacing. The MAX.ANGLE parameter specified on the MESH statement gives the angle of the
wedge.

53 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

For example
MESH CIRCULAR
...
A.MESH LOC=0 SPAC=30
A.MESH LOC=150 SPAC=30
or alternatively
MESH CIRCULAR MAX.ANGLE=150
...
A.MESH LOC=0 SPAC=30
would both create an angular wedge with an angle at the origin of 150° and a major angular
spacing of 30°. The R.MESH statements have been omitted for clarity.

Note: MAXANGLE should not exceed 180°.

Once a circular mesh has been created, it may be subdivided into regions using the R.MIN,
R.MAX, A.MIN and A.MAX parameters on the REGION statement.
R.MIN is the inner radius of the region. R.MAX is the outer radius of the region. Both are in the
units of microns. A.MIN and A.MAX define the minimum and maximum angular limits of the
region respectively, both in degrees.
For example
REGION NUM=1 MATERIAL=SILICON A.MIN=0 A.MAX=360.0 R.MAX=0.4
REGION NUM=2 MATERIAL=OXIDE A.MIN=30 A.MAX=150.0 R.MIN=0.35
R.MAX=0.4
will enforce the area between angular limits of 30 and 150 degrees and radial limits of 0.35
and 0.4 microns to be an oxide region, and the rest of the mesh defined above to be a silicon
region.
Similarly, the ELECTRODE, DOPING, and INTERFACE statements have R.MIN, R.MAX, A.MIN,
and A.MAX parameters. Support for doping of circular meshes is extended from that described
in Section 2.6.3 “Using The Command Language To Define A Structure” to allow analytical
doping profiles in the radial direction (UNIFORM, GAUSS, ERFC) with optional lateral fall off in
the angular (i.e., at constant radius) direction. The lateral falloff is GAUSSIAN unless
ERFC.LAT is specified to change it to the complementary error function. The usual parameters
for specifying the analytical profile apply with the principal direction being the radial
direction.
If you set the maximum angle to 180° and add the following lines, you will obtain the
structure as shown in Figure 2-11.
ELECTRODE NAME=DRAIN R.MIN=0.35 A.MIN=0.0 A.MAX=30.0
ELECTRODE NAME=SOURCE R.MIN=0.35 A.MIN=150.0 A.MAX=180.0
ELECTRODE NAME=GATE A.MIN=60.0 A.MAX=120.0 R.MIN=0.39 R.MAX=0.4
If you further add the DOPING statement below, then you will obtain the doping profile as
shown in Figure 2-12.

54 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

DOPING GAUSS N.TYPE CONC=1.0e19 CHAR=0.05 R.MIN=0.15 R.MAX=0.2

A.MIN=0.0 A.MAX=180.0

Figure 2-11 Semi-circular structure created using ATLAS command syntax

55 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

Figure 2-12 Gaussian doping profile applied to the semi-circular structure of Figure 2-11

Note: To obtain the best results, align region and electrode boundaries with existing meshlines (radial or major
spokes).

56 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

2.6.8 Specifying 3D Structures

The syntax for forming 3D device structures is an extension of the 2D syntax described in the
previous section. The MESH statement should appear as:
MESH THREE.D
The THREE.D parameter tells ATLAS that a three dimensional grid will be specified. The
other statements used to specify 3D structures and grids are the same as for 2D with the
addition of Z direction specifications. The statements:
MESH THREE.D
X.MESH LOCATION=0 SPACING=0.15
X.MESH LOCATION=3 SPACING=0.15
Y.MESH LOCATION=0 SPACING=0.01
Y.MESH LOCATION=0.4 SPACING=0.01
Y.MESH LOCATION=3 SPACING=0.5
Z.MESH LOCATION=0 SPACING=0.1
Z.MESH LOCATION=3 SPACING=0.1
define a 3D mesh that is uniform in the X and Z directions and varies in the Y direction.
Position parameters for the Z direction (Z.MIN and Z.MAX) are also used on REGION,
ELECTRODE, or DOPING statements.

2.6.9 Specifying 3D Cylindrical Structures

As mentioned in Section 2.6.3 “Using The Command Language To Define A Structure”, you
can model a quasi-3D cylindrical structure in ATLAS2D by specifying the CYLINDRICAL
parameter on the MESH statement. This has the drawback that the resulting device structure
and solution has no dependence on the angle around the axis of rotation. In ATLAS3D, the
CYLINDRICAL parameter enables you to create a general cylindrical structure. The MESH
statement must appear as:
MESH THREE.D CYLINDRICAL
where the THREE.D parameter informs the simulator to create a fully 3D mesh.
When specifying the CYLINDRICAL parameter, you must now specify the structure in terms of
radius, angle, and cartesian Z coordinates.
There are three mesh statements analogous to those used for general structures that are used
to specify mesh in radius, angle and Z directions. The R.MESH statement is used to specify
radial mesh. The A.MESH statement is used to specify angular mesh. The Z.MESH statement is
used to specify mesh in the Z direction.
For example:
MESH THREE.D CYLINDRICAL

R.MESH LOCATION=0.0 SPACING=0.0004

R.MESH LOCATION=0.0028 SPACINg=0.00005

57 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

R.MESH LOCATION=0.004 SPACING=0.0002

A.MESH LOCATION=0 SPACING=45

A.MESH LOCATION=360 SPACING=45

Z.MESH LOCATION=-0.011 SPACING=0.001

Z.MESH LOCATION=0.011 SPACING=0.001
Here, the R.MESH lines are similar to the familiar X.MESH, Y.MESH, and Z.MESH except the
R.MESH locations and spacings are radial relative to the Z axis in microns.
The locations and spacings on the A.MESH lines specify locations and spacings in degrees of
rotation about the Z axis. The Z.MESH lines are exactly the same as have been already
discussed.
For 3D cylindrical structure, specifying the REGION and ELECTRODE statements also are
modified to allow specification of ranges in terms of radius, angle and z location. The range
parameters are R.MIN, R.MAX, A.MIN, A.MAX, Z.MIN and Z.MAX. R.MIN, R.MAX, Z.MIN and
Z.MAX are all in units of microns. A.MIN and A.MAX are in degrees.
To continue the example, we will specify a surround gate nano-tube as follows:
REGION NUM=1 MATERIAL=silicon \
Z.MIN=-0.1 Z.MAX=0.1 A.MIN=0 A.MAX=360.0 R.MAX=0.0028

REGION NUM=2 MATERIAL=oxide \

R.MIN=0.0028 R.MAX=0.004 Z.MIN=-0.011 Z.MAX=0.011 A.MIN=0
A.MAX=360.0

ELECTRODE NAME=DRAIN Z.MIN=-0.011 Z.MAX=-0.01 R.MAX=0.0028

ELECTRODE NAME=SOURCE Z.MIN=0.01 Z.MAX=0.011 R.MAX=0.0028

ELECTRODE NAME=GATE Z.MIN=-0.004 Z.MAX=0.004 R.MIN=0.0038 

MATERIAL=aluminum
Here, we defined a central core of silicon surrounded by an oxide cladding. The device has a
surrounding gate with a source and drain at either end.

58 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

Figure 2-13 shows a cutplane normal to the Z axis at z=0 for this device.

Figure 2-13 Cutplane of 3D Cylindrical Simulation

In some instances, you can reduce the size of the cylindrical device by using symmetry
arguments. So if the device has mirror symmetry about some plane through its axis of
rotation, it suffices to model only half of the cylinder. ATLAS allows you to achieve this by
one of two methods. In the first method, you specify more than one A.MESH statement, and
the maximum angle specified in these statements is the maximum angle of the device
structure. In the second method, you only specify one A.MESH statement with a location of
zero degrees and the required angular spacing. The MAX.ANGLE parameter specified on the
MESH statement gives the angle of the wedge.
For example, to make a semi-cylinder, you use either of the following syntaxes
MESH CYLINDRICAL THREE.D
...
A.MESH LOC=0 SPAC=30
A.MESH LOC=180 SPAC=30
...
or alternatively
MESH CYLINDRICAL THREE.D MAX.ANGLE=180
...
A.MESH LOC=0 SPAC=30
...

59 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

where statements for radial and vertical meshes are omitted for clarity. The major angular
spacing is 30°.
The value of the maximum angle should not be greater than 180° and must satisfy the
condition that the angular range of the device must start and end on major spokes of the radial
mesh. In other words, the maximum angle must be equal to a integral multiple of angular
mesh spacing. For example, if the angular mesh spacing was 45°, then the maximum angle
can be 45, 90, 135, or 180°. ATLAS will automatically adjust the value of the maximum
angle if it is necessary.
If the mesh spacing in the radial direction is regular, then the resulting mesh will usually have
no obtuse elements. If the mesh spacing in the radial direction decreases with increasing
radius, then it is possible that some obtuse triangular elements will be formed. Use the
parameter MINOBTUSE to try to reduce the number of obtuse triangular elements.
The DOPING statement has been modified to accept the parameters R.MIN, R.MAX, A.MIN and
A.MAX. These apply to the analytic doping PROFILES, namely UNIFORM, GAUSSIAN, and
ERFC. R.MIN and R.MAX specify the minimum and maximum radial extents in microns.
A.MIN and A.MAX specify the minimum and maximum angular extents in degrees. The
principal direction for the GAUSSIAN and ERFC dependence is the Z direction. Thus, the usual
DOPING parameters, such as CHAR, PEAK, DOSE, START and JUNCTION all apply to the Z
direction with R.MIN, R.MAX, A.MIN, A.MAX defining the extents at which lateral fall off
occurs.
The lateral fall-off in the radial and angular directions can be controlled by the LAT.CHAR or
RATIO.LAT parameters.
For example:
DOPING GAUSSIAN N.TYPE START=0.25 CONC=1.0e19 CHAR=0.2 \
LAT.CHAR=0.2 \
R.MIN=0.4 R.MAX=0.6 A.MIN=0.0 A.MAX=360
produces a gaussian doping profile in the Z direction, with a peak at Z=0.25 microns, and
with a lateral gaussian fall-off in the radial direction when it is outside a radius of between 0.4
and 0.6 microns. Figure 2-14 shows the radial and angular distribution. Figure 2-15 shows
the radial and z-variation. The Z direction corresponds to the vertical direction on the
cutplane.

60 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

Figure 2-14 Cylindrical gaussian doping profile on a cutplane perpendicular to the Z axis

61 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

Figure 2-15 Cylindrical gaussian doping profile on a cutplane containing the Z axis

2.6.10 Extracting 2D Circular Structures From 3D Cylindrical Structures

To simulate circular structures in 2D that are analogous to the 3D meshing described above,
use the cutplane feature in structure saving. To do this, specify the CUTPLANE and
Z.CUTPLANE parameters of the SAVE statement. The logical parameter CUTPLANE specifies
that you are interested in outputting a 2D cutplane from a 3D structure. The Z.CUTPLANE
specifies the Z coordinate value in microns of the location where you want the cutplane.
For example, Figure 2-13 was generated by the following syntax:
SAVE CUTPLANE Z.CUTPLANE=0.0

62 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

2.6.11 General Comments Regarding Grids

Specifying a good grid is a crucial issue in device simulation but there is a trade-off between
the requirements of accuracy and numerical efficiency. Accuracy requires a fine grid that
resolves the structure in solutions. Numerical efficiency is greater when fewer grid points are
used. The critical areas to resolve are difficult to generalize because they depend on the
technology and the transport phenomena. The only generalization possible is that most
critical areas tend to coincide with reverse-biased metallurgical junctions. Typical critical
areas are:
• High electric fields at the drain/channel junction in MOSFETs
• The transverse electric field beneath the MOSFET gate
• Recombination effects around the emitter/base junction in BJTs
• Areas of high impact ionization
• Around heterojunctions in HBT’s and HEMTs.

The CPU time required to obtain a solution is typically proportional to N, where N is the
number of nodes and varies from 2 to 3 depending on the complexity of the problem. Thus,
the most efficient way is to allocate a fine grid only in critical areas and a coarser grid
elsewhere.
The three most important factors to look for in any grid are:
• Ensure adequate mesh density in high field areas
• Avoid obtuse triangles in the current path or high field areas
• Avoid abrupt discontinuities in mesh density
For more information about grids, see Chapter 20 “Numerical Techniques”, Section 20.3
“Meshes”.

2.6.12 Maximum Numbers Of Nodes, Regions, and Electrodes

ATLAS sets some limits on the maximum number of grid nodes that can be used. But this
shouldn’t be viewed as a bottleneck to achieving simulation results. In the default version, 2D
ATLAS simulations have a maximum node limit of 100,000. 3D ATLAS simulations have an
upper limit of 40,000,000 nodes with no more than 100,000 nodes in any one Z plane and a
maximum number of Z planes of 2,000.
This limit is high enough that for almost all simulations of conventional devices, running out
of nodes is never an issue. For most 2D simulations, you can obtain accurate results with
somewhere between 2,000 and 4,000 node points properly located in the structure.
If you exceed the node limits, error messages will appear and ATLAS will not run
successfully. There are two options to deal with this problem. The first option is to decrease
the mesh density because simulations with the maximum nodes take an extremely long time
to complete. The second option is to contact your local SILVACO office
([email protected]).
A node point limitation below these values might be seen due to virtual memory constraints
on your hardware. For each simulation, ATLAS dynamically allocates the virtual memory.
See the SILVACO INSTALLATION GUIDE for information about virtual memory requirements.
The virtual memory used by the program depends on the number of nodes and on items, such
as the models used and the number of equations solved.

63 ATLAS User’s Manual

Defining A Structure Getting Started with ATLAS

Also, there is a node limit for the number of nodes in the X or Y directions in 2D and 3D
ATLAS. In the standard version, this limit is 20,000 nodes. This is applicable to meshes
defined in the ATLAS syntax using X.MESH and Y.MESH statements.
The maximum number of regions defined in both 2D and 3D ATLAS is 1,000. The maximum
number of definable electrodes is 100. Again, if you need to include more than the maximum
number of regions or electrodes, contact your local SILVACO office ([email protected]).

2.6.13 Quadrilateral REGION Definition

In addition to the rectangular shaped regions, which we have considered so far, you can create
regions with sloping sides. Although it is possible to make complicated overall region shapes
using rectangular regions as building blocks, this task is made easier by being able to specify
the co-ordinates of each of the four corners of the REGION. To do this, use the P1.X, P1.Y,
P2.X, P2.Y, P3.X, P3.Y, P4.X, and P4.Y parameters on the REGION statement. Then, (P1.X,
P1.Y) are the XY coordinates of the first corner of the region and so on.
The region must lie entirely within the device limits as defined by the mesh. To obtain best
results, these corner coordinates must coincide with valid mesh points as created by the
X.MESH and Y.MESH statements.
Because the overall mesh shape will still be rectangular, you may need to use VACUUM
regions as padding to get a non-rectangular shaped device. Alternatively, use DEVEDIT to get
a non-rectangular overall device shape.
Example:
REGION NUMBER=3 MATERIAL=SILICON P1.X=0.0 P1.Y=1.0 P2.X=3.0
P2.Y=1.0 P3.X=3.5
P3.Y=2.5 P4.X=0.0 P4.Y=2.0

Table 2-1 Quadrilateral Shaped REGION Definition

Statement Parameter Type Default Units

REGION P1.X Real Microns

REGION P1.Y Real Microns
REGION P2.X Real Microns
REGION P2.Y Real Microns
REGION P3.X Real Microns
REGION P3.Y Real Microns
REGION P4.X Real Microns
REGION P4.Y Real Microns

64 ATLAS User’s Manual

Defining Material Parameters And Models Getting Started with ATLAS

2.7 Defining Material Parameters And Models

Once you define the mesh, geometry, and doping profiles, you can modify the characteristics
of electrodes, change the default material parameters, and choose which physical models
ATLAS will use during the device simulation. To accomplish these actions, use the CONTACT,
MATERIAL, and MODELS statements respectively. Impact ionization models can be enabled
using the IMPACT statement. Interface properties are set by using the INTERFACE statement.
Many parameters are accessible through the SILVACO C-INTERPRETER (SCI), which is
described in Appendix A “C-Interpreter Functions”. This allows you to define customized
equations for some models. For more information about the INTERFACE and MODELS
statements, see Sections 21.24 “INTERFACE” and 21.37 “MODELS”.

2.7.1 Specifying Contact Characteristics

Workfunction for Gates or Schottky Contacts
An electrode in contact with semiconductor material is assumed by default to be ohmic. If a
work function is defined, the electrode is treated as a Schottky contact. The CONTACT
statement is used to specify the metal workfunction of one or more electrodes. The NAME
parameter is used to identify which electrode will have its properties modified.
The WORKFUNCTION parameter sets the workfunction of the electrode. For example, the
statement:
CONTACT NAME=gate WORKFUNCTION=4.8
sets the workfunction of the electrode named gate to 4.8eV. The workfunctions of several
commonly used contact materials can be specified using the name of the material. You can
specify workfunctions for ALUMINUM, N.POLYSILICON, P.POLYSILICON, TUNGSTEN, and
TU.DISILICIDE in this way. The following statement sets the workfunction for a n-type
polysilicon gate contact.
CONTACT NAME=gate N.POLYSILICON
Aluminum contacts on heavily doped silicon is usually ohmic. For this situation, don’t
specify a workfunction. For example, MOS devices don’t specify:
CONTACT NAME=drain ALUMINUM /* wrong */
The CONTACT statement can also be used to specify barrier and dipole lowering of the
Schottky barrier height. To enable barrier lowering, specify the BARRIER parameter, while
specifying dipole lowering using the ALPHA parameter. For example, the statement:
CONTACT NAME=anode WORKFUNCTION=4.9 BARRIER ALPHA=1.0e-7
sets the work function of the Schottky contact named anode to 4.9eV enables barrier
lowering and sets the dipole lowering coefficient to 1 nm.

Note: When a Schottky barrier is defined at a contact, we recommend that a fine y mesh is present just beneath
the contact inside the semiconductor. This allows the Schottky depletion region to be accurately simulated.

65 ATLAS User’s Manual

Defining Material Parameters And Models Getting Started with ATLAS

Setting Current Boundary Conditions

The CONTACT statement is also used to change an electrode from voltage control to current
control. Current controlled electrodes are useful when simulating devices, where the current
is highly sensitive to voltage or is a multi-valued function of voltage (e.g., post-breakdown
and when there is snap-back).
The statement:
CONTACT NAME=drain CURRENT
changes the drain electrode to current control. The BLOCK or NEWTON solution methods are
required for all simulations using a current boundary condition. For more information about
these methods, see Section 20.5 “Non-Linear Iteration”.
Defining External Resistors, Capacitors, or Inductors
Lumped resistance, capacitance, and inductance connected to an electrode can be specified
using the RESISTANCE, CAPACITANCE, and INDUCTANCE parameters in the CONTACT
statement. For example, the statement:
CONTACT NAME=drain RESISTANCE=50.0 CAPACITANCE=20e-12
INDUCTANCE=1e-6
specifies a parallel resistor and capacitor of 50 ohms and 20 pF respectively in series with a 1
H inductor. Note that in 2D simulations, these passive element values are scaled by the
width in the third dimension. Since in 2D ATLAS assumes a 1m width, the resistance
becomes 50 -m.
Distributed contact resistance for an electrode can be specified using the CON.RESIST
parameter. For example, the statement:
CONTACT NAME=source CON.RESISTANCE=0.01

specifies that the source contact has a distributed resistance of 0.01 cm2.

Note: Simulations with external resistors, capacitors, or inductors must be solved using the BLOCK or NEWTON
solution method.

Floating Contacts
The CONTACT statement is also used to define a floating electrode. There are two distinctly
different situations where floating electrodes are important. The first situation is for floating
gate electrodes used in EEPROM and other programmable devices. The second situation is
for contacts directly onto semiconductor materials such as floating field plates in power
devices.
To enable floating gates, specify the FLOATING parameter on the CONTACT statement. For
example, the statement:
CONTACT NAME=fgate FLOATING
specifies that the electrode named fgate will be floating and that charge boundary conditions
will apply.

66 ATLAS User’s Manual

Defining Material Parameters And Models Getting Started with ATLAS

For contacts directly onto semiconductor, the FLOATING parameter cannot be used. This type
of floating electrode is best simulated by specifying current boundary conditions on the
CONTACT statement. For example, the statement:
CONTACT NAME=drain CURRENT
specifies current boundary conditions for the electrode named drain. On subsequent SOLVE
statements, the drain current boundary condition will default to zero current. Therefore,
floating the contact.
You can also make a floating contact to a semiconductor using a very large resistor attached
to the contact instead. For example:
CONTACT NAME=drain RESIST=1e20
Note that extremely large resistance values must be used to keep the current through the
insignificant contact. Using a lumped resistor will allow the tolerance on potential to move
slightly above zero. For example, if the tolerance is 10-5V and the defined resistance was only
10M.m, then a current of 10-12 A/m may flow through the contact, which may be
significant in breakdown simulations.
Shorting Two Contacts Together
It is possible in ATLAS to tie two or more contact together so that voltages on both contacts
are equal. This is useful for many technologies for example dual base bipolar transistors.
There are several methods for achieving this depending on how the structure was initially
defined.
If the structure is defined using ATLAS syntax, you can have multiple ELECTRODE statements
with the same NAME parameter defining separate locations within the device structure. In this
case, the areas defined to be electrodes will be considered as having the same applied voltage.
A single current will appear combining the current through both ELECTRODE areas.
Also, if two separate metal regions in ATHENA are defined using the ATHENA ELECTRODE
statement to have the same name, then in ATLAS these two electrodes will be considered as
shorted together.
If the electrodes are defined with different names, the following syntax can be used to link the
voltages applied to the two electrodes.
CONTACT NAME=base1 COMMON=base
.
SOLVE VBASE=0.1

Here, the electrode, base1, will be linked to the electrode, base. The applied 0.1V on base
will then appear on base1. ATLAS, however, will calculate and store separate currents for
both base and base1. This can be a useful feature. But in some cases, such as where
functions of the currents are required in EXTRACT or TONYPLOT, it is undesirable. You can
add the SHORT parameter to the CONTACT statement above to specify so that only a single
base current will appear combining the currents from base and base1. Note that the
electrode specified by the COMMON parameter should preferably have a name that exactly
matches one of those specified in the ELECTRODE statement (see the NAME description in
Section 21.15 “ELECTRODE”). Additionally, the electrode specified by the NAME parameter

67 ATLAS User’s Manual

Defining Material Parameters And Models Getting Started with ATLAS

should not have a bias applied directly to it. You should always apply the bias to the electrode
specified by the COMMON parameter.
You can also specify the voltages on the linked electrodes to be different, but have a constant
OFFSET relative to the bias on the electrode specified by the COMMON parameter.
The statement
CONTACT name=base1 COMMON=base FACTOR=0.5
ensures the bias on base1 will be equal to the bias on base + 0.5 V. If you also specify the
MULT parameter on this statement, then FACTOR will be interpreted as multiplicative rather
than additive. In this case, the bias applied to base1 will be one-half of the bias applied to
base. If FACTOR (and optionally MULT) is applied to a contact for which current boundary
conditions are being used, then it has no effect. Any existing bias differences between the
linked electodes are, however, maintained under current boundary conditions. To apply
additive or multiplicative relationships between currents on linked electrodes, remove the
CONTACT statements linking them. Then, directly specify the required values of the currents
on the SOLVE statement.
When loading a structure from ATHENA or DEVEDIT, where two defined electrode regions
are touching, ATLAS will automatically short these and use the electrode name that was
defined first.
Making an Open Circuit Contact
It is often required to perform a simulation with an open circuit on one of the defined
electrodes. There are three different methods to make an open circuit contact. The first
method is to entirely deleting an electrode from the structure file. The second method is to
add an extremely large lumped resistance. For example, 1020 onto the contact to be made
open circuit. The third method is to switch the boundary conditions on the contact to be made
open circuit from voltage controlled to current controlled. Then, specifying a very small or
zero current through that electrode.
Each of these methods are feasible. If a floating region, however, is created within the
structure, then numerical convergence may be affected. As a result, we normally recommend
that you use the second method because it ensures better convergence.

2.7.2 Specifying Material Properties

Semiconductor, Insulator, or Conductor
All materials are split into three classes: semiconductors, insulators and conductors. Each
class requires a different set of parameters to be specified. For semiconductors, these
properties include electron affinity, band gap, density of states and saturation velocities.
There are default parameters for material properties used in device simulation for many
materials.
Appendix B “Material Systems” lists default material parameters and includes a discussion
on the differences between specifying parameters for semiconductors, insulators, and
conductors.

68 ATLAS User’s Manual

Defining Material Parameters And Models Getting Started with ATLAS

Setting Parameters
The MATERIAL statement allows you to specify your own values for these basic parameters.
Your values can apply to a specified material or a specified region. For example, the
statement:
MATERIAL MATERIAL=Silicon EG300=1.12 MUN=1100
sets the band gap and low field electron mobility in all silicon regions in the device. If the
material properties are defined by region, the region is specified using the REGION or NAME
parameters in the MATERIAL statement. For example, the statement:
MATERIAL REGION=2 TAUN0=2e-7 TAUP0=1e-5
sets the electron and hole Shockley-Read-Hall recombination lifetimes for region number two
(see the “Shockley-Read-Hall (SRH) Recombination” section on page 3-204 for more
information about this type of recombination). If the name, base, has been defined using the
NAME parameter in the REGION statement, then the statement:
MATERIAL NAME=base NC300=3e19
sets the conduction band density of states at 300 K for the region named base.
The description of the MATERIAL statement in Section 21.32 “MATERIAL” provides a
complete list of all the material parameters that are available.
Heterojunction Materials
The material properties of heterojunctions can also be modified with the MATERIAL statement.
In addition to the regular material parameters, you can define composition dependent material
parameters. For example, composition dependent band parameters, dielectric constants, and
saturation velocities.
For heterojunction material systems, the bandgap difference between the materials is divided
between conduction and valence bands. The ALIGN parameter specifies the fraction of this
difference applied to the conduction band edge. This determines the electron and hole barrier
height and overrides any electron affinity specification. For example, the statement:
MATERIAL MATERIAL=InGaAs ALIGN=0.36
MATERIAL MATERIAL=InP ALIGN=0.36
specifies that 36% of the bandgap difference between InGaAs and InP is applied to the
conduction band and 64% is applied to the valence band. For example, if the band gap
difference (  Eg) for this material system is 0.6 eV, then the conduction band barrier height is
0.216 eV and the valence band barrier height is 0.384 eV.
For heterojunction devices, the transport models may be different for each material. You can
specify these models and their coefficients for each material using the MODELS statement. See
Section 2.7.4 “Specifying Physical Models” for a description of this option.

69 ATLAS User’s Manual

Defining Material Parameters And Models Getting Started with ATLAS

2.7.3 Specifying Interface Properties

The INTERFACE statement is used to define the interface charge density and surface
recombination velocity at interfaces between semiconductors and insulators. For example, the
statement:
INTERFACE QF=3e10
specifies that all interfaces between semiconductors and insulators have a fixed charge of
3.1010cm-2. In many cases, the interface of interest is restricted to a specific region. This can
be accomplished with the X.MIN, X.MAX, Y.MIN, and Y.MAX parameters on the INTERFACE
statement. These parameters define a rectangle, where the interface properties apply. For
example, the statement:
INTERFACE QF=3e10 X.MIN=1.0 X.MAX=2 Y.MIN=0.0 Y.MAX=0.5
restricts the interface charge to the semiconductor-insulator boundary within the specified
rectangle. In addition to fixed charge, surface recombination velocity and thermionic
emission are enabled and defined with the INTERFACE statement. For more information about
this statement, see Section 21.24 “INTERFACE”.

2.7.4 Specifying Physical Models

Physical models are specified using the MODELS and IMPACT statements. Parameters for these
models appear on many statements including: MODELS, IMPACT, MOBILITY, and MATERIAL.
The physical models can be grouped into five classes: mobility, recombination, carrier
statistics, impact ionization, and tunneling. Section 3.6 “Physical Models” contains details for
each model.
Tables 2-2 to 2-6 give summary descriptions and recommendations on the use of each model.
Table 2-7 is a guide for compatibility between models.
All models with the exception of impact ionization are specified on the MODELS statement.
Impact ionization is specified on the IMPACT statement. For example, the statement:
MODELS CONMOB FLDMOB SRH FERMIDIRAC
IMPACT SELB
specifies that the standard concentration dependent mobility, parallel field mobility,
Shockley-Read-Hall recombination with fixed carrier lifetimes, Fermi Dirac statistics and
Selberherr impact ionization models should be used.
ATLAS also provides an easy method for selecting the correct models for various
technologies. The MOS, BIP, PROGRAM, and ERASE parameters for the MODELS statement
configure a basic set of mobility, recombination, carrier statistics, and tunneling models. The
MOS and BIP parameters enable the models for MOSFET and bipolar devices, while PROGRAM
and ERASE enable the models for programming and erasing programmable devices. For
example, the statement:
MODELS MOS PRINT
enables the CVT, SRH, and FERMIDIRAC models, while the statement:
MODELS BIPOLAR PRINT

70 ATLAS User’s Manual

Defining Material Parameters And Models Getting Started with ATLAS

enables the CONMOB, FLDMOB, CONSRH, AUGER, and BGN models. If LAT.TEMP is also
specified on the MODELS statement, or the TEMPERATURE parameter differs from 300 Kelvin
by more than 10 Kelvin, then the ANALYTIC model is used instead of CONMOB.

Note: The PRINT parameter lists to the run time output the models and parameters, which will be used during the
simulation. This allows you to verify models and material parameters. We highly recommend that you
include the PRINT parameter in the MODELS statement.

Physical models can be enabled on a material by material basis. This is useful for
heterojunction device simulation and other simulations where multiple semiconductor regions
are defined and may have different characteristics. For example, the statement:
MODEL MATERIAL=GaAs FLDMOB EVSATMOD=1 ECRITN=6.0e3 CONMOB
MODEL MATERIAL=InGaAs SRH FLDMOB EVSATMOD=1 \
ECRITN=3.0e3
change both the mobility models and critical electric field used in each material. For devices
based on advanced materials, these model parameters should be investigated carefully.
Energy Balance Models
The conventional drift-diffusion model of charge transport neglects non-local effects, such as
velocity overshoot and reduced energy dependent impact ionization. ATLAS can model these
effects through the use of an energy balance model, which uses a higher order approximation
of the Boltzmann Transport Equation (see Section 3.1.3 “The Transport Equations”). In this
equation, transport parameters, such as mobility and impact ionization, are functions of the
local carrier temperature rather than the local electric field.
To enable the energy balance transport model, use the HCTE, HCTE.EL, or HCTE.HO
parameters in the MODELS statement. These parameters enable the energy transport model for
both carriers, electrons only, or holes only respectively. For example, the statement:
MODELS MOS HCTE
enables the energy balance transport model for both electrons and holes in addition to the
default MOSFET models.

2.7.5 Summary Of Physical Models

Concentration Dependent CONMOB Lookup table valid at 300K for Si and GaAs
only. Uses simple power law temperature
dependence.
Concentration and Temperature ANALYTIC Caughey-Thomas formula. Tuned for 77-
Dependent 450K.
Arora’s Model ARORA Alternative to ANALYTIC for Si.
Carrier-Carrier Scattering CCSMOB Dorkel-Leturq Model. Includes n, N and T
dependence. Important when carrier
concentration is high (e.g., forward bias
power devices).
Parallel Electric Field FLDMOB Si and GaAs models. Required to model any
Dependence type of velocity saturation effect.
Tasch Model TASCH Includes transverse field dependence. Only
for planar devices. Needs very fine grid.
Watt Model WATT Transverse field model applied to surface
nodes only.
Klaassen Model KLA Includes N, T, and n dependence. Applies
separate mobility to majority and minority
carriers. Recommended for bipolar devices
Shirahata Model SHI Includes N, E . An alternative surface
mobility model that can be combined with
KLA.
Modified Watt MOD.WATT Extension of WATT model to non-surface
nodes. Applies constant E effects. Best
model for planar MOS devices

72 ATLAS User’s Manual

Defining Material Parameters And Models Getting Started with ATLAS

Table 2-3 Mobility Models

Model Syntax Notes

Lombardi (CVT) Model CVT Complete model including N, T, E//, and

E effects. Good for non-planar devices.

Yamaguchi Model YAMAGUCHI Includes N, E// and E  effects. Only

calibrated for 300K.

Table 2-4 Recombination Models

Model Syntax Notes

Shockley-Read-Hall SRH Uses fixed minority carrier lifetimes. Should be

used in most simulations.
Concentration Dependent CONSRH Uses concentration dependent lifetimes.
Recommended for Si.
Auger AUGER Direct transition of three carriers. Important at
high current densities.
Optical OPTR Band-band recombination. For direct
materials only.
Surface S.N Recombination at semiconductor to insulator
S.P interfaces. This is set in the INTERFACE
statement.

Table 2-5 Impact Ionization

Model Syntax Notes

Selberherr’s Model IMPACT SELB Recommended for most cases. Includes

temperature dependent parameters.
Grant’s Model IMPACT Similar to Selberherr’s model but with different
GRANT coefficients.
Crowell-Sze IMPACT Uses dependence on carrier
CROWELL scattering length.
Toyabe Model IMPACT Non-local model used with Energy Balance. Any
TOYABE IMPACT syntax is accepted.

Concannon N.CONCAN Non-local model developed in Flash EEPROM

P.CONCAN technologies.

73 ATLAS User’s Manual

Defining Material Parameters And Models Getting Started with ATLAS

Table 2-6 Tunneling Models and Carrier Injection Models

Model Syntax Notes

Band-to-Band (standard) BBT.STD For direct transitions. Required with very high
fields.
Concannon Gate Current N.CONCAN Non-local gate model consistent with
Model P.CONCAN Concannon substrate current model.
Direct Quantum tunneling QTUNN.EL Quantum tunneling through conduction band
(Electrons) barrier due to an insulator.
Direct Quantum tunneling QTUNN.HO Quantum tunneling through valence band barrier
(Hole) due to an insulator.
Fowler-Nordheim FNORD Self-consistent calculation of tunneling through
(electrons) insulators. Used in EEPROMs.
Fowler-Nordheim (holes) FNHOLES Same as FNORD for holes.
Klaassen Band-to-Band BBT.KL Includes direct and indirect transitions.
Hot Electron Injection HEI Models energetic carriers tunneling through
insulators. Used for gate current and Flash
EEPROM programming.
Hot Hole Injection HHI HHI means hot hole injection.

Note: In the notes in Tables 2-2 through 2-6, n is electron concentration, p is hole concentration, T is lattice
temperature, N is dopant concentration, Ell is parallel electric field, and Eis perpendicular electric field.

Table 2-7 Model Compatibility Chart

CONMOB FLDMOB TFLDMB2 YAMAGUCHI CVT ARORA ANALYTIC CCSMOB SURMOB LATTICE H E.BALANCE

CONMOB [CM] — OK OK YA CV AR AN CC OK OK OK
FLDMOB [FM] OK — TF1 YA CV OK OK OK OK OK OK
TFLDMB2 [TF] OK TF1 — YA CV OK OK TF TF OK OK
YAMAGUCHI YA YA YA — CV YA YA YA YA NO NO
[YA]

CVT [CV] CV CV CV CV — CV CV CV CV OK OK
ARORA [AR] AR OK OK YA CV — AR CC OK OK OK
ANALYTIC [AN] AN OK OK YA CV AR — CC OK OK OK
CCSMOB [CC] CC OK TF YA CV CC CC — OK OK OK

74 ATLAS User’s Manual

Defining Material Parameters And Models Getting Started with ATLAS

Table 2-7 Model Compatibility Chart

CONMOB FLDMOB TFLDMB2 YAMAGUCHI CVT ARORA ANALYTIC CCSMOB SURMOB LATTICE H E.BALANCE

SURFMOB [SF] OK OK TF YA CV OK OK OK — OK OK
LATTICE H [LH] OK OK OK NO OK OK OK OK OK — OK
E.BALANCE OK OK OK NO OK OK OK OK OK OK 2

[EB]

Key To Table Entries

MODEL ABBREVIATION = The model that supersedes when a combination is specified. In some cases, a warning message is issued when a model is
ignored.
OK = This combination is allowed.
NO = This combination is not allowed.
NOTES:

1.Uses internal model similar to FLDMOB

2.When models including a parallel electric field dependence are used with energy balance the electric field term is
replaced by a function of carrier temperature.

Using the C-Interpreter to Specify Models

One of the ATLAS products is a C language interpreter that allows you to specify many of the
models used by ATLAS. To use these functions, implement the model in C as equations in a
special file called an ATLAS lib file. You can access the default ATLAS template file by
typing:
atlas -T <filename>
at the UNIX command prompt. This creates a default template file with the name specified by
the <filename> parameter. See Appendix A “C-Interpreter Functions” for a listing of the
default C-INTERPRETER functions. To use the interpreter functions, give the corresponding
parameters in the statements containing the name of the C language file with the model given
as the parameter value. For example, the statement:
MATERIAL NAME=Silicon F.MUNSAT=munsat.lib
specifies that the file, munsat.lib, contains the C-INTERPRETER function for the specification
of the parallel field dependent electron mobility model.

75 ATLAS User’s Manual

Choosing Numerical Methods Getting Started with ATLAS

2.8 Choosing Numerical Methods

2.8.1 Numerical Solution Techniques
Several different numerical methods can be used for calculating the solutions to
semiconductor device problems. Numerical methods are given in the METHOD statements of
the input file. Some guidelines for these methods will be given here. For full details, however,
see Chapter 20 “Numerical Techniques”.
Different combinations of models will require ATLAS to solve up to six equations. For each
of the model types, there are basically three types of solution techniques: (a) decoupled
(GUMMEL), (b) fully coupled (NEWTON) and (c) BLOCK. The GUMMEL method will solve for each
unknown in turn keeping the other variables constant, repeating the process until a stable
solution is achieved. The NEWTON method solve the total system of unknowns together. The
BLOCK methods will solve some equations fully coupled while others are de-coupled.
Generally, the GUMMEL method is useful where the system of equations is weakly coupled but
has only linear convergence. The NEWTON method is useful when the system of equations is
strongly coupled and has quadratic convergence. The NEWTON method may, however, spend
extra time solving for quantities, which are essentially constant or weakly coupled. NEWTON
also requires a more accurate initial guess to the problem to obtain convergence. Thus, a
BLOCK method can provide for faster simulations times in these cases over NEWTON. GUMMEL
can often provide better initial guesses to problems. It can be useful to start a solution with a
few GUMMEL iterations to generate a better guess. Then, switch to NEWTON to complete the
solution. Specification of the solution method is carried out as follows:
METHOD GUMMEL BLOCK NEWTON
The exact meaning of the statement depends on the particular models it applied to. This will
be discussed in the following sections.
Basic Drift Diffusion Calculations
The isothermal drift diffusion model requires the solution of three equations for potential,
electron concentration, and hole concentration. Specifying GUMMEL or NEWTON alone will
produce simple Gummel or Newton solutions as detailed above. For almost all cases, the
NEWTON method is preferred and it is the default.
Specifying:
METHOD GUMMEL NEWTON
will cause the solver to start with GUMMEL iterations. Then, switch to NEWTON if convergence is
not achieved. This is a robust but a more time consuming way of obtaining solutions for any
device. This method, however, is highly recommended for all simulations with floating
regions such as SOI transistors. A floating region is defined as an area of doping, which is
separated from all electrodes by a pn junction.
BLOCK is equivalent to NEWTON for all isothermal drift-diffusion simulations.

76 ATLAS User’s Manual

Choosing Numerical Methods Getting Started with ATLAS

Drift Diffusion Calculations with Lattice Heating

When the lattice heating model is added to drift diffusion, an extra equation is added. The
BLOCK algorithm solves the three drift diffusion equations as a NEWTON solution. Follows this
with a GUMMEL solution of the heat flow equation. The NEWTON algorithm solves all four
equations in a coupled manner. NEWTON is preferred once the temperature is high. BLOCK,
however, is quicker for low temperature gradients. Typically, the combination used is:
METHOD BLOCK NEWTON
Energy Balance Calculations
The energy balance model requires the solution of up to 5 coupled equations. GUMMEL and
NEWTON have the same meanings as with the drift diffusion model. In other words, GUMMEL
specifies a de-coupled solution and NEWTON specifies a fully coupled solution.
But BLOCK performs a coupled solution of potential, carrier continuity equations followed by
a coupled solution of carrier energy balance, and carrier continuity equations.
You can switch from BLOCK to NEWTON by specifying multiple solution methods on the same
line. For example:
METHOD BLOCK NEWTON
will begin with BLOCK iterations. Then, switch to NEWTON if convergence is still not achieved.
This is the most robust approach for many energy balance applications.
The points where the algorithms switch is pre-determined, but can be changed in the METHOD
statement, the default values set by SILVACO work well for most circumstances.
Energy Balance Calculations with Lattice Heating
When non-isothermal solutions are performed in conjunction with energy balance models, a
system of up to six equations must be solved. GUMMEL or NEWTON solve the equations
iteratively or fully coupled respectively. BLOCK initially performs the same function as with
energy balance calculations, then solves the lattice heating equation in a de-coupled manner.
Setting the Number of Carriers
ATLAS can solve both electron and hole continuity equations, or only for one or none. You
can make this choice by using the CARRIERS parameter. For example:
METHOD CARRIERS=2
specifies a solution for both carriers is required. This is the default. With one carrier, the
ELEC or HOLE parameter is needed. For example, for hole solutions only:
METHOD CARRIERS=1 HOLE
To select a solution for potential, only specify:
METHOD CARRIERS=0

Note: Setting the number of carriers using the syntax, MODEL NUMCARR=<n>, is obsolete and should not be
used.

77 ATLAS User’s Manual

Choosing Numerical Methods Getting Started with ATLAS

Important Parameters of the METHOD Statement

You can alter all of the parameters relevant to the numerical solution process. This, however,
isn’t recommended unless you have expert knowledge of the numerical algorithms. All of
these parameters have been assigned optimal values for most solution conditions. For more
information about numerical algorithms, see Chapter 20 “Numerical Techniques”.
The following parameters, however, are worth noting at this stage:
• CLIMIT or CLIM.DD specify minimal values of concentrations to be resolved by the
solver. Sometimes, you need to reduce this value to aid solutions of breakdown
characteristics. A value of CLIMIT=1e-4 is recommended for all simulations of
breakdown, where the pre-breakdown current is small. CLIM.DD is equivalent to CLIMIT
but uses the more convenient units of cm-3 for the critical concentration. CLIM.DD and
CLIMIT are related by the following expression.

CLIM DD = CLIMIT N c N v 2-3

• DVMAX controls the maximum update of potential per iteration of Newton’s method. The
default corresponds to 1V. For power devices requiring large voltages, you may need an
increased value of DVMAX. DVMAX=1e8 can improve the speed of high voltage bias ramps.
• CLIM.EB controls the cut-off carrier concentration below, which the program will not
consider the error in the carrier temperature. This is applied in energy balance simulations
to avoid excessive calculations of the carrier temperature at locations in the structure
where the carrier concentration is low. Setting this parameter too high, where ATLAS
ignores the carrier temperature errors for significant carrier concentrations, will lead to
unpredictable and mostly incorrect results .
Restrictions on the Choice of METHOD
The following cases require METHOD NEWTON CARRIERS=2 to be set for isothermal drift-
diffusion simulations:
• current boundary conditions
• distributed or lumped external elements
• AC analysis
• impact ionization
Both BLOCK or NEWTON or both are permitted for lattice heat and energy balance.

Note: Simulations using the GUMMEL method in these cases may lead to non-convergence or incorrect results.

78 ATLAS User’s Manual

Choosing Numerical Methods Getting Started with ATLAS

Pisces-II Compatibility
Previous releases of ATLAS (2.0.0.R) and other PISCES-II based programs use the
SYMBOLIC command to define the solution method and the number of carriers included in the
solution. In this version of ATLAS, the solution method is specified completely on the
METHOD statement.
The COMB parameter, which was available in earlier ATLAS versions, is no longer required. It
has been replaced with either the BLOCK method or the combination of GUMMEL and NEWTON
parameters. Table 2-8 identifies direct translations of old syntax to new.

Note: These are direct translations and not necessarily the best choices of numerical methods.

Table 2-8 Parameter Syntax Replacements

Old Syntax (V2.0.0.R) New Syntax

symbolic newton carriers=2 method newton

symbolic newton carriers=1 elec method newton carriers=1 electron

symbolic gummel carriers=0 method gummel carriers=0

symbolic newton carriers=2 method gummel newton

method comb

models lat.temp models lat.temp

symbolic newton carriers=2 method block
method comb

models hcte models hcte

symbolic gummel carriers=2 method block
method comb

79 ATLAS User’s Manual

Obtaining Solutions Getting Started with ATLAS

2.9 Obtaining Solutions

ATLAS can calculate DC, AC small signal, and transient solutions. Obtaining solutions is
similar to setting up parametric test equipment for device tests. You usually define the
voltages on each of the electrodes in the device. ATLAS then calculates the current through
each electrode. ATLAS also calculates internal quantities, such as carrier concentrations and
electric fields throughout the device. This is information that is difficult or impossible to
measure.
In all simulations, the device starts with zero bias on all electrodes. Solutions are obtained by
stepping the biases on electrodes from this initial equilibrium condition. As will be discussed,
due to the initial guess strategy, voltage step sizes are limited. This section concentrates on
defining solution procedures. To save results, use the LOG or SAVE statements. Section 2.10
“Interpreting The Results” on how to analyze and display these results.

2.9.1 DC Solutions
In DC solutions, the voltage on each electrode is specified using the SOLVE statement. For
example, the statements:
SOLVE VGATE=1.0
SOLVE VGATE=2.0
solves a single bias point with 1.0V and then 2.0V on the gate electrode. One important rule
in ATLAS is that when the voltage on any electrode is not specified in a given SOLVE
statement, the value from the last SOLVE statement is assumed.
In the following case, the second solution is for a drain voltage of 1.0V and a gate voltage of
2.0V.
SOLVE VGATE=2.0
SOLVE VDRAIN=1.0
When the voltage on a particular electrode is never defined on any SOLVE statement and
voltage is zero, you don’t need to explicitly state the voltage on all electrodes on all SOLVE
statements. For example, in a MOSFET, if VSUBSTRATE is not specified, then Vbs defaults to
zero.
Sweeping The Bias
For most applications, a sweep of one or more electrodes is usually required. The basic DC
stepping is inconvenient and a ramped bias should be used. To ramp the base voltage from
0.0V to 1.0V with 0.05V steps with a fixed collector voltage of 2.0V, use the following
syntax:
SOLVE VCOLLECTOR=2.0
SOLVE VBASE=0.0 VSTEP=0.05 VFINAL=1.0 NAME=base
The NAME parameter is required and the electrode name is case-sensitive. Make sure the
initial voltage, VSTEP and VFINAL, are consistent. A badly specified ramp from zero to 1.5V
in 0.2V steps will finish at 1.4V or 1.6V.

80 ATLAS User’s Manual

Obtaining Solutions Getting Started with ATLAS

Generating Families of Curves

Many applications such as MOSFET Id/Vds and bipolar Ic/Vce simulations require that a
family of curves is produced. This is done by obtaining solutions at each of the stepped bias
points first, and then solving over the swept bias variable at each stepped point. For example,
in MOSFET Id/Vds curves, solutions for each Vgs value are obtained with Vds=0.0V. The
output from these solutions are saved in ATLAS solution files. For each gate bias, the solution
file is loaded and the ramp of drain voltage performed.
The family of curves for three 1V gate steps and a 3.3V drain sweep would be implemented in
ATLAS as follows:
SOLVE VGATE=1.0 OUTF=solve_vgate1
SOLVE VGATE=2.0 OUTF=solve_vgate2
SOLVE VGATE=3.0 OUTF=solve_vgate3

LOAD INFILE=solve_vgate1
LOG OUTFILE=mos_drain_sweep1
SOLVE NAME=drain VDRAIN=0 VFINAL=3.3 VSTEP=0.3

LOAD INFILE=solve_vgate2
LOG OUTFILE=mos_drain_sweep2
SOLVE NAME=drain VDRAIN=0 VFINAL=3.3 VSTEP=0.3

LOAD INFILE=solve_vgate3
LOG OUTFILE=mos_drain_sweep3
SOLVE NAME=drain VDRAIN=0 VFINAL=3.3 VSTEP=0.3
The LOG statements are used to save the Id/Vds curve from each gate voltage to separate files.
We recommend that you save the data in this manner rather than to a single LOG file (see
Section 2.10 “Interpreting The Results” ).

81 ATLAS User’s Manual

Obtaining Solutions Getting Started with ATLAS

2.9.2 The Importance Of The Initial Guess

To obtain convergence for the equations used, supply a good initial guess for the variables to
be evaluated at each bias point. The ATLAS solver uses this initial guess and iterates to a
converged solution. For isothermal drift diffusion simulations, the variables are the potential
and the two carrier concentrations. If a reasonable grid is used, almost all convergence
problems in ATLAS are caused by a poor initial guess to the solution.
During a bias ramp, the initial guess for any bias point is provided by a projection of the two
previous results. Problems tend to appear near the beginning of the ramp when two previous
results are not available. If one previous bias is available, it is used alone. This explains why
the following two examples eventually produce the same result. The first will likely have far
more convergence problems than the second.
1. SOLVE VGATE=1.0 VDRAIN=1.0 VSUBSTRATE=-1.0
2. SOLVE VGATE=1.0
SOLVE VSUBSTRATE=-1.0
SOLVE VDRAIN=1.0
In the first case, one solution is obtained with all specified electrodes at 1.0V. In the second
case, the solution with only the gate voltage at 1.0V is performed first. All other electrodes
are at zero bias. Next, with the gate at 1.0V, the substrate potential is raised to -1.0V and
another solution is obtained. Finally, with the substrate and the gate biased, the drain potential
is added and the system solved again. The advantage of this method over the first case is that
the small incremental changes in voltage allow for better initial guesses at each step.
Generally, the projection method for the initial guess gives good results when the I-V curve is
linear. But it may encounter problems if the IV curve is highly non-linear or if the device
operating mode is changing. Typically, this may occur around the threshold or breakdown
voltages. At these biases, smaller voltage steps are required to obtain convergence. As will be
described, ATLAS contains features such as the TRAP parameter and the curve tracer to
automatically cut the voltage steps in these highly non-linear area.
Numerical methods are described in Section 2.8 “Choosing Numerical Methods” and Chapter
20 “Numerical Techniques”.
In many cases, these methods are designed to overcome the problems associated with the
initial guess. This is particularly important in simulations involving more than the three drift
diffusion variables. Generally, coupled solutions require a good initial guess, whereas de-
coupled solutions can converge with a poor initial guess.
The Initial Solution
When no previous solutions exist, the initial guess for potential and carrier concentrations
must be made from the doping profile. This is why the initial solution performed must be the
zero bias (or thermal equilibrium) case. This is specified by the statement:
SOLVE INIT
But if this syntax isn’t specified, ATLAS automatically evaluates an initial solution before the
first SOLVE statement. To aid convergence of this initial guess, the solution is performed in the
zero carrier mode solving only for potential.

82 ATLAS User’s Manual

Obtaining Solutions Getting Started with ATLAS

The First and Second Non-Zero Bias Solutions

From experience with ATLAS, it is found that the first and second non-zero bias solutions are
the most difficult to obtain good convergence. Once these two solutions are obtained, the
projection algorithm for the initial guess is available and solutions should all have a good
initial guess.
These first two solutions, however, must use the result of the initial solution as the basis of
their initial guess. Since the initial solution is at zero bias, it provides a poor initial guess.
The practical result of this is that the first and second non-zero bias solutions should have
very small voltage steps. In the following example, the first case will likely converge whereas
the second case may not.
1. SOLVE INIT
SOLVE VDRAIN=0.1
SOLVE VDRAIN=0.2
SOLVE VDRAIN=2.0
2. SOLVE INIT
SOLVE VDRAIN=2.0
The Trap Parameter
Although ATLAS provides several methods to overcome a poor initial guess and other
convergence problems, it is important to understand the role of the initial guess in obtaining
each solution. The simplest and most effective method to overcome poor convergence is by
using:
METHOD TRAP
This is enabled by default. Its effect is to reduce the bias step if convergence problems are
detected. Consider the example from the previous section:
SOLVE INIT
SOLVE VDRAIN=2.0
If the second SOLVE statement does not converge, TRAP automatically cuts the bias step in
half and tries to obtain a solution for Vd = 1.0V. If this solution does not converge, the bias
step will be halved again to solve for Vd = 0.5V. This procedure is repeated up to a maximum
number of tries set by the METHOD parameter MAXTRAPS. Once convergence is obtained, the
bias steps are increased again to solve up to 2.0V. The default for MAXTRAPS is 4 and it is not
recommended to increase it, since changing the syntax to use smaller bias steps is generally
much faster.
This trap facility is very useful during bias ramps in overcoming convergence difficulties
around transition points such as the threshold voltage. Consider the following syntax used to
extract a MOSFET Id/Vgs curve.
SOLVE VGATE=0.0 VSTEP=0.2 VFINAL=5.0 NAME=gate

83 ATLAS User’s Manual

Obtaining Solutions Getting Started with ATLAS

Assume the threshold voltage for the device being simulated is 0.7V and that ATLAS has
solved for the gate voltages up to 0.6V. The next solution, at 0.8V, may not converge at first.
This is because the initial guess was formed from the two sub-threshold results at Vgs=0.4V
and 0.6V and the solution has now become non-linear. The trap facility will detect the
problems in the 0.8V solution and cut the bias step in half to 0.7V and try again. This will
probably converge. The solution for 0.8V will then be performed and the bias ramp will
continue with 0.2V steps.

2.9.3 Small-Signal AC Solutions

Specifying AC simulations is a simple extension of the DC solution syntax. AC small signal
analysis is performed as a post-processing operation to a DC solution. Two common types of
AC simulation in ATLAS are outlined here. The results of AC simulations are the
conductance and capacitance between each pair of electrodes. Tips on interpreting these
results is described in Section 2.10 “Interpreting The Results”.
Single Frequency AC Solution During A DC Ramp
The minimum syntax to set an AC signal on an existing DC ramp is just the AC flag and the
setting of the small signal frequency. For example:
SOLVE VBASE=0.0 VSTEP=0.05 VFINAL=1.0 NAME=base AC FREQ=1.0e6
Other AC syntax for setting the signal magnitude and other parameters are generally not
needed as the defaults suffice. One exception is in 1D MOS capacitor simulations. To obtain
convergence in the inversion/deep depletion region, add the DIRECT parameter to access a
more robust solution method.
Ramped Frequency At A Single Bias
For some applications, such as determining bipolar gain versus frequency, you need to ramp
the frequency of the simulation. This is done using the following syntax:
1. SOLVE VBASE=0.7 AC FREQ=1e9 FSTEP=1e9 NFSTEPS=10
2. SOLVE VBASE=0.7 AC FREQ=1e6 FSTEP=2 MULT.F NFSTEPS=10
The first case ramps the frequency from 1GHz to 11GHz in 1GHz steps. A linear ramp of
frequency is used and FSTEP is in Hertz. In the second example, a larger frequency range is
desired and so a geometrical step of frequency is used. The MULT.F parameter is used to
specify that FSTEP is a unitless multiplier for the frequency. This doubles the frequency in
successive steps from 1MHz to 1.024GHz.
You can combine the following syntax for ramping the frequency of the AC signal with the
syntax for ramping the bias. The frequency ramps are done at each bias point during the DC
ramp.
SOLVE VBASE=0.0 VSTEP=0.05 VFINAL=1.0 NAME=base AC FREQ=1.0e6 \
FSTEP=2 MULT.F NFSTEPS=10

84 ATLAS User’s Manual

Obtaining Solutions Getting Started with ATLAS

2.9.4 Transient Solutions

Transient solutions can be obtained for piecewise-linear, exponential, and sinusoidal bias
functions. Transient solutions are used when a time dependent test or response is required. To
obtain transient solutions for a linear ramp, specify the TSTOP, TSTEP, and RAMPTIME
parameters.
Figure 2-16 shows the syntax of the transient voltage ramp. The RAMPTIME specifies the time
that the linear ramp should obtain its final value. TSTOP specifies the time that solutions will
stop. TSTEP specifies the initial step size.
Subsequent time steps are calculated automatically by ATLAS. For example, the statement:
SOLVE VGATE=1.0 RAMPTIME=1E-9 TSTOP=10e-9 TSTEP=1e-11,
specifies that the voltage on the gate electrode will be ramped in the time domain from its
present value to 1.0V over a period of 1 nanoseconds.
Time domain solutions are obtained for an additional 9 nanoseconds. An initial time step of
10 picoseconds is specified. Note that if you specify subsequent transient solutions, don’t
reset the time to zero.

RAMP TIME TSTOP

1.0

VGATE

TSTEP

1 2 3 4 5 6 7 8 9 10
time (ns)
Figure 2-16 Diagram showing syntax of Transient Voltage Ramp in ATLAS

85 ATLAS User’s Manual

Obtaining Solutions Getting Started with ATLAS

2.9.5 Advanced Solution Techniques

Obtaining Solutions Around The Breakdown Voltage
Obtaining solutions around the breakdown voltage can be difficult using the standard ATLAS
approach. It requires special care when choosing voltage steps and interpreting the results.
The curve tracer described in “The Curvetrace Capability” on page 87 is the most effective
method in many cases.
A MOSFET breakdown simulation might be performed using the following standard syntax
for ramping the drain bias. Note that the setting of CLIMIT is recommended for breakdown
simulations when the pre-breakdown leakage is low.
IMPACT SELB
METHOD CLIMIT=1e-4
SOLVE VDRAIN=1.0 VSTEP=1.0 VFINAL=20.0 NAME=drain
If the breakdown were 11.5V, then convergence problems will be expected for biases higher
than 11.0V. Although it depends on technology used, it is common for the breakdown curve to
be flat up to a voltage very close to breakdown and then almost vertical. The current changes
by orders of magnitude for very small bias increments.
This produces some problems for ATLAS using the syntax described above. If the breakdown
occurs at 11.5V, there are no solutions for voltages greater than this value. ATLAS is trying to
ramp to 20.0V. Therefore, it is inevitable that ATLAS will fail to converge at some point. This
is usually not a problem since by that point the breakdown voltage and curve have been
obtained.
Above 11V, bias step reduction will take place due to the TRAP parameter. ATLAS will
continually try to increase the drain voltage above 11.5V and those points will fail to
converge. But it will solve points asymptotically approaching Vds=11.5V until it reaches the
limit set by the MAXTRAPS parameter. If you use the default of 4 traps, the minimum allowed
voltage step will be 1.0(0.5)4 or 0.004V. This is normally enough accuracy for determining
the breakdown point. But the simulation might not allow the current to reach a sufficiently
high level before MAXTRAPS is needed.
Typically in device simulation, the breakdown point is determined once the current is seen to
increase above the flat pre-breakdown leakage value by two orders of magnitude in a small
voltage increment. If you want to trace the full breakdown curve up to high current values,
apply more advanced techniques than the simple voltage ramp. Two of these techniques are
described in the following subsections. These methods may use extra CPU time.
Using Current Boundary Conditions
In all of the examples considered in the basic description of the SOLVE statement, it was
assumed that voltages were being forced and currents were being measured. ATLAS also
supports the reverse case through current boundary conditions. The current through the
electrode is specified in the SOLVE statement and the voltage at the contact is calculated.
Current boundary conditions are set using the CONTACT statement as described in Section 2.7
“Defining Material Parameters And Models”.
The syntax of the SOLVE statement is altered once current boundary conditions are specified.
SOLVE IBASE=1e-6

86 ATLAS User’s Manual

Obtaining Solutions Getting Started with ATLAS

The syntax above specifies a single solution at a given current.

SOLVE IBASE=1e-6 ISTEP=1e-6 IFINAL=5e-6 NAME=base
This sets a current ramp similar in syntax to the voltage ramp described earlier.
SOLVE IBASE=1e-10 ISTEP=10 IMULT IFINAL=1e-6 NAME=base
This is similar to the previous case. The IMULT parameter, however, is used to specify that
ISTEP should be used as a multiplier for the current rather than a linear addition. This is
typical for ramps of current since linear ramps are inconvenient when several orders of
magnitude in current may need to be covered.
Important points to remember about current boundary conditions are that the problems of
initial guess are more acute when very small (noise level) currents are used. Often, it is best to
ramp the voltage until the current is above 1pA/m and then switch to current forcing.
When interpreting the results, it is important to remember the calculated voltage on the
electrode with current boundary conditions is stored as the internal bias. In other words, base
int.bias in TONYPLOT or vint.base in DECKBUILD’s extract syntax.
The Compliance Parameter
Compliance is a parameter used to limit the current or voltage through or on an electrode
during a simulation. You can set an electrode compliance. After reaching this limit, the bias
sweep will stop. This is similar to parametric device testing when we stop a device from being
over-stressed or destroyed. The compliance refers to the maximum resultant current or
voltage present after obtaining a solution. If you set an electrode voltage, the compliance will
then refer to the electrode current. If there are current boundary conditions, you can set a
voltage compliance.
The statements:
SOLVE VGATE=1.0
SOLVE NAME=drain VDRAIN=0 VFINAL=2 VSTEP=0.2 COMPL=1E-6 CNAME=drain
solve for IV on the gate and then ramps the drain voltage towards 2V in 0.2V steps. The
simulation will stop if it reaches 1µA/µm of drain current before Vd=2V. Thus, as in
parametric testing, you can define a particular level and set the simulation to solve up to that
point. Once ATLAS reaches the compliance limit, it will simulate the next statement line in
the command file.
The Curvetrace Capability
The automatic curve tracing algorithm can be invoked to enable ATLAS to trace out complex
IV curves. The algorithm can automatically switch from voltage to current boundary
conditions and vice versa. You can use a single SOLVE statement to trace out complex IV
curves, such as breakdown curves and CMOS latch-up including the snapback region and
second breakdown. The algorithm is based upon a dynamic load line approach.
For example, typical CURVETRACE and SOLVE statements to trace out an IV curve for the
breakdown of a diode would look like:
CURVETRACE CONTR.NAME=cathode STEP.INIT=0.5 NEXTST.RATIO=1.2 \
MINCUR=1e-12 END.VAL=1e-3 CURR.CONT
SOLVE CURVETRACE

87 ATLAS User’s Manual

Obtaining Solutions Getting Started with ATLAS

CONTR.NAME specifies the name of the electrode, which is to be ramped. STEP.INIT specifies
the initial voltage step. NEXTST.RATIO specifies the factor used to increase the voltage step in
areas on the IV curve away from turning points. MINCUR sets a small current value above that
activates the dynamic load line algorithm. Below the MINCUR level, using the STEP.INIT and
NEXTST.RATIO determines the next solution bias. END.VAL stops the tracing if the voltage or
current of the ramped electrode equals or exceeds END.VAL. Using VOLT.CONT or CURR.CONT
specify whether END.VAL is a voltage or current value.
When you plot the log file created by the CURVETRACE statement in TONYPLOT, select the
internal bias labeled int.bias for the ramped electrode instead of plotting the applied bias,
which is labeled Voltage.

2.9.6 Using DeckBuild To Specify SOLVE Statements

The DeckBuild Solve menu can be used generate SOLVE statements. The menu has a
spreadsheet style entry. To access this menu, select the Command/Solutions/Solve... button
in DECKBUILD. To define a test, click the right mouse button in the worksheet and select the
Add new row option. This will add a new row to the worksheet. This procedure should be
repeated once per electrode in your device structure. The entry for each cell can then be edited
to construct a SOLVE statement.
Some cells require the selection using a pop menu or the entry of numerical values. The
electrode name is specified in the first cell. To edit the electrode name, use a popup menu by
pressing the right mouse button on the cell. The second cell specifies whether the electrode
will be a voltage (V), current (I) or charge (Q) controlled. The third cell specifies whether the
SOLVE statement is to be a single DC solve (CONST), a swept DC variable (VAR1), a stepped
DC variable (VAR2), or a transient solution (PULSE). The remaining cells specify the
parameter values that are required for the type of solution desired.
The pop-up window to specify the solution file names are accessed through the Props...
button. You can construct several SOLVE statements to create solve sequences, which define a
test. This test may be saved in a file and read in using the Save... and Load... buttons.

88 ATLAS User’s Manual

Interpreting The Results Getting Started with ATLAS

2.10 Interpreting The Results

As already described in Section 2.2 “ATLAS Inputs and Outputs”, ATLAS produces three
different types of output files. These files are also described in the following sections.

2.10.1 Run-Time Output

Run-time output is provided at the bottom of the DeckBuild Window. If it’s run as a batch job,
the run-time output can be stored to a file.
Errors occurring in the run-time output will be displayed in this window. Note that not all
errors will be fatal (as DECKBUILD tries to interpret the file and continue). This may cause a
statement to be ignored, leading to unexpected results. We recommend that you check the
run-time output of any newly created input file the first time it runs to intercept any errors.
If you specify the PRINT option within the MODELS statement, the details of material
parameters and constants and mobility models will be specified at the start of the run-time
output. This is a useful way of checking what parameters values and models are being applied
in the simulation. We recommend that you always specify MODELS PRINT in input files.
During SOLVE statements, the error numbers of each equation at each iteration are displayed
(this is a change from the previous ATLAS version). It isn’t vital for you to understand the
iteration information. It can, however, provide important insights in the case of convergence
problems.
The output can be interpreted as follows:

proj psi n p psi n p

direct x x x rhs rhs rhs

i j m -5.00* -5.00* -5.00* -26.0* -17.3* -17.3*

- - - - - - - - - - - - - - - - - - - - - - - - - - -

1 N - -2.934 -1.932 -25.2 -10.19 -9.876

1.932

2 N - -5.64* -4.267 -28.8* -16.67 -15.47

4.741

3 A - -11.7* -9.63* -28.8* -16.67 -18.0*

11.3*

Electrode Va(V) Jn(A/um) Jp(A/um) Jc(A/um) Jt(A/um)

=======================================================================

gate 0.000e+00 -0.000e+00 -0.000e+00 0.000e+00 0.000e+00

source 0.000e+00 -3.138e-13 -1.089e-35 -3.138e-13 -3.138e-13

drain 1.000e-01 3.139e-13 1.076e-23 3.139e-13 3.139e-13

substrate 0.000e+00 -6.469e-19 -8.853e-17 -8.918e-17 -8.918e-17

89 ATLAS User’s Manual

Interpreting The Results Getting Started with ATLAS

The top left value, proj, indicates the initial guess methodology used. The default projection
method is used here. Alternatives are previous, local, or init. The second value, direct,
indicates the solver type. This will either be direct or iterative.
The first three column headings: i, j, and m indicate the iteration numbers of the solution and
the solution method. i indicates the outer loop iteration number for decoupled solutions. j
indicates the inner loop number. m indicates the solution method by a single letter, which are:
• G = gummel
• B = block
• N = newton
• A = newton with autonr
• S = coupled Poisson-Schrodinger solution
The remaining column headings indicate which column lists the XNORM and RHSNORM errors
for the equations being solved. See Chapter 20: “Numerical Techniques”, Section 20.5.7
“Error Measures” for a full description of these errors. The values printed in each error
column under the hashed line are the logarithm to base 10 of the error. Earlier PISCES
versions would print the floating point value. The values printed above the dashed line in each
column are the tolerances used.
When the star * symbol appears as the least significant digit in the number, it means this error
measure has met its tolerance.
After achieving convergence, ATLAS lists the results by electrode. The column, Va, lists the
voltage at the contact surface. This will differ from the applied voltage if external resistors or
the curvetracer are used. All relevant current components are listed. Here, only electron, hole,
conduction, and total currents are given. In other modes of simulation, these columns may
differ.
The output of AC analysis, MIXEDMODE, and 3D simulations differ from this standard. See
Chapter 12 “MixedMode: Mixed Circuit and Device Simulator” for more information about
MIXEDMODE.
ATLAS may produce a very large amount of run-time output for complex simulations. You
can save run-time output to a file as shown in Section 2.3 “Modes of Operation”.

2.10.2 Log Files

Log files store the terminal characteristics calculated by ATLAS. These are current and
voltages for each electrode in DC simulations. In transient simulations, the time is stored. In
AC simulations, the small signal frequency and the conductances and capacitances are saved.
For example, the statement:
LOG OUTF=<FILENAME>
is used to open a log file. Terminal characteristics from all SOLVE statements after the LOG
statement are then saved to this file along with any results from the PROBE statement.
To not save the terminal characteristics to this file, use another LOG statement with either a
different log filename or the OFF parameter.
Typically, a separate log file should be used for each bias sweep. For example, separate log
files are used for each gate bias in a MOS Id/Vds simulation or each base current in a bipolar
Ic/Vce simulation. These files are then overlaid in TONYPLOT.

90 ATLAS User’s Manual

Interpreting The Results Getting Started with ATLAS

Log files contain only the terminal characteristics. They are typically viewed in TONYPLOT.
Parameter extraction of data in log files can be done in DECKBUILD. Log files cannot be
loaded into ATLAS to re-initialize the simulation.
Units of Currents In Log files
Generally, the units of current are written into the log file. In TONYPLOT, it is Amperes per
micron. This is because ATLAS is a two-dimensional simulator. It sets the third dimension (or
Z direction) to be one micron. Thus, if you compare ATLAS 2D simulation results for a
MOSFET versus the measured data from a MOSFET of width 20 micron, you need to
multiply the current in the log file by 20.
There are four exceptions:
• In the 3D modules of ATLAS, the width is defined in the 3D structure. Therefore, the
units of the current are Amperes.
• In MIXEDMODE, set the width of devices. The current is also in Amperes.
• When cylindrical coordinates are used, the current written to the log file is integrated
through the cylinder and is also in Amperes.
• When the WIDTH parameter in the MESH statement is used, the current is scaled by this
factor and is in Amperes.
Similar rules apply for the capacitance and conductance produced by AC simulations. These
are usually in 1/(ohms.microns) and Farads/micron respectively.
UTMOST Interface
ATLAS log files can be read directly into the batch mode, UTMOST. The following
commands in UTMOST are used to read in a set of IV curves stored in separate log files.
INIT INF=<filename> MASTER
INIT INF=<filename> MASTER APPEND

Note: The UTMOST statement in ATLAS is no longer recommended for interfacing to UTMOST.

2.10.3 Parameter Extraction In DeckBuild

The EXTRACT command is provided within the DECKBUILD environment. It allows you to
extract device parameters. The command has a flexible syntax that allows you to construct
specific EXTRACT routines. EXTRACT operates on the previous solved curve or structure file.
By default, EXTRACT uses the currently open log file. To override this default, supply the
name of a file to be used by EXTRACT before the extraction routine. For example:
EXTRACT INIT INF="<filename>"
A typical example of using EXTRACT is the extraction of the threshold voltage of an MOS
transistor. In the following example, the threshold voltage is extracted by calculating the
maximum slope of the Id / Vg curve, finding the intercept with the X axis and then subtracting
half of the applied drain bias.
EXTRACT NAME="nvt" XINTERCEPT(MAXSLOPE(CURVE (V."GATE",
(I."DRAIN"))) \
-(AVE(V."DRAIN"))/2.0)

91 ATLAS User’s Manual

Interpreting The Results Getting Started with ATLAS

The results of the extraction will be displayed in the run-time output and will be by default
stored in the file results.final. To store the results in a different file at the end of EXTRACT
command, use the following option:
EXTRACT....DATAFILE="<filename>"
Cut-off frequency and forward current gain are of particular use as output parameters. These
functions can be defined as follows:
# MAXIMUM CUTOFF FREQUENCY
EXTRACT NAME="FT_MAX" MAX(G."COLLECTOR""BASE"/
(6.28*C."BASE""BASE"))

#FORWARD CURRENT GAIN

EXTRACT NAME="PEAK GAIN" MAX(I."COLLECTOR"/ I."BASE")

Note: Over 300 examples are supplied with ATLAS to provide many practical examples of the use of the
EXTRACT statement.

AC Parameter Extraction
You can perform basic analysis of the capacitance and conductance matrix produced by
ATLAS using DECKBUILD or TONYPLOT. The capacitance between gate and drain will be
labeled as Cgate>drain in TONYPLOT or c."gate""drain" in DECKBUILD’S EXTRACT.
The total capacitance on any electrode is defined as Celectrode>electrode. Thus, the
magnitude of Cgate>gate is the total gate capacitance.
The LOG statement also includes options for small-signal, two-port RF analysis including s-
parameter extraction. The solutions are saved into the log file and in the run-time output. The
list of options for RF analysis are:
s.param, y.param, h.param, z.param, abcd.param gains
Terminal impedance and parasitics are accounted for by adding any of the following:
impedance=<val>, rin=<val>, rout=<val>, rcommon=<val> or
rground=<val>, lin=<val>, lout=<val>,
lcommon=<val> or lground=<val>, width=<val>
The width defaults to 1m and impedance defaults to 50. All parasitics default to zero.
The Stern stability factor, k, is calculated along with current gain (h21), GUmax, and GTmax
when the GAINS option is added to the LOG statement.
The run-time output for AC analysis has been modified to only list the analysis frequency and
electrode conductance/capacitance values. If one of the two-port options is added to the LOG
statement (e.g., S.PARAM), the two-port parameters are also included in the run-time output.

Note: AC parameter conversion utilities in the UTMOST statement have been discontinued.

See Appendix C “RF and Small Signal AC Parameter Extraction” for more information.

92 ATLAS User’s Manual

Interpreting The Results Getting Started with ATLAS

Additional Functions
EXTRACT has two additional important functions. The first function is to provide data for the
VWF database. In other words, to store device parameters to the VWF database, you must
evaluate them using EXTRACT. The second function is when using the DeckBuild Optimizer to
tune parameters, use the EXTRACT statements as the optimization targets.

2.10.4 Functions In TonyPlot

The Edit/Functions Menu in TONYPLOT allows you to specify and plot functions of the
terminal characteristics in the Graph Function text fields. For example, you can calculate
transconductance using the following function:
dydx (drain current, gate voltage)
Current gain can be evaluated as:
collector current / base current
When creating functions, the key to correct syntax is that the name for any variable in a
function is the same as that in the Y Quantities list on the Display menu.

2.10.5 Solution Files

Solution files or structure files provide an image of the device at a particular bias point (DC
solution or transient solution point). This gives you the ability to view any evaluated quantity
within the device structure in question, from doping profiles and band parameters to electron
concentrations and electric fields. These files should be plotted using TONYPLOT.
The syntax used to generate these files is of two forms:
(1) SAVE OUTFILE=<filename>
(2) SOLVE .... OUTFILE=<filename>.sta MASTER [ONEFILEONLY]
Here in the second form, a file named <filename> will be saved with data from the
previously solved bias point.
In this case, a structure file will be saved at each bias point solved in the solve statement. The
last letter of the file name will be automatically incremented alphabetically (i.e., *.sta,
*.stb, *.stc..., and so on). If you need the solution for the last bias point, you can add the
onefileonly parameter to the command. The <filename>.sta file will be overwritten at
each solution point.
Structure files can be large (1 - 2 MB), depending on the mesh density and quantities saved. It
is recommended that unwanted structure files be deleted.
If many solution files have been written from a long simulation, it is often confusing to find
out which solution file belongs to which bias point or transient time. The solution files should
be plotted in TONYPLOT. In TONYPLOT, you can create 2D contour plots and 1D cutlines. To
find out the bias of any solution file in TONYPLOT, select the plot, and press B on the
keyboard.

93 ATLAS User’s Manual

Interpreting The Results Getting Started with ATLAS

Interpreting Contour Plots

Some quantities saved in the solution files are not evaluated at the node points during
solutions. They are evaluated at the center of the sides of each triangle in the mesh. Values of
quantities at each node are derived from averaging the values from the sides of triangles
connected to that node. The weighting method used to do the averaging can be selected with
options in the OUTPUT statement. It is possible that for some meshes, smoother contour plots
can be obtained by choosing a non-default averaging method.
When interpreting the contour plots, it’s important to remember that the solution file contains
values only at each node point. The color fills seen in TONYPLOT are simply interpolations
based on the node values. This may lead to occasional strange contour values. In these cases,
check the node values using the probe in TONYPLOT.
The primary solution variables (potential, carrier concentration, lattice temperature, and
carrier temperatures) are calculated on the nodes of the ATLAS mesh. Therefore, they are
always correct in TONYPLOT. But since ATLAS doesn’t use nodal values of quantities, such
as electric field and mobility, the actual values used in calculations cannot be determined
from TONYPLOT or the structure files. The PROBE statement allows you to directly probe
values at given locations in the structure. This provides the most accurate way to determine
the actual values used in ATLAS calculations.
Customizing Solution Files (OUTPUT Statement)
Several quantities are saved by default within a structure file. For example, doping, electron
concentration, and potential. You can also specify additional quantities (such as conduction
band potential) by using the OUTPUT statement. This must precede the SAVE statement in
question. For example, to save the conduction and valence band potentials, the following
command would be used at some point before the relevant SAVE.
OUTPUT CON.BAND VAL.BAND
Saving Quantities from the Structure at each Bias Point (PROBE statement)
Structure files provide all data from the structure at a single bias point. The log files provide
terminal characteristics for a set of bias points. Use the PROBE statement to combine these and
allow certain structural quantities to be saved at each bias point.
The PROBE statement allows you to specify quantities to be saved at given XY locations. You
can also save the maximum or minimum of certain quantities. The value from the PROBE at
each bias point in DC or timestep in transient mode is saved to the log file. The syntax:
PROBE NAME=mycarriers N.CONC X=1 Y=0.1
saves the electron concentration at (1, 0.1) for each solution in the log file. When TONYPLOT
displays the log file, the value will be labelled mycarriers. You can then plot mycarriers
versus terminal bias or current or other probed quantities.
Certain direction dependent quantities, such as electric field and mobility, can be probed. In
these cases, specify a direction for the vector quantity using the DIR parameter.
The PROBE statement provides the only way to extract the actual values of quantities, which
are calculated along the sides of each triangle in ATLAS. The PROBE statement actually stored
the triangle side value closest to the probed location, while taking into account the direction
for vector quantities.

94 ATLAS User’s Manual

Interpreting The Results Getting Started with ATLAS

Note: Specifying the probe location exactly at a material or region interface will often lead to erroneous results. It is
best to slightly offset the location of the probe inside the material or region of interest.

Re-initializing ATLAS at a Given Bias Point

Each SOLVE statement will begin with the device biased at the previous value solved. To
begin a solution at a previously solved bias point, re-load the structure file saved at that point.
This is accomplished in the following manner:
LOAD INFILE=<filename> MASTER
Information about that solution point will be displayed in the Output Window.
This command is useful for solving a set of I/V curves. For example, to solve a family of Id /
Vd (at various Vg), ramp the gate with zero drain bias. A structure file is then saved at each
desired value of Vg. These structure files can be reloaded in turn while a Vd sweep is
performed.

Note: An ATLAS input file cannot start with a LOAD statement. Before loading the structure file, use the MESH
statement to load the device mesh for the same structure. Also, the same MODELS, MATERIAL, and
CONTACT settings are required when the files are saved by ATLAS.

2.10.6 Technology Specific Issues in ATLAS

This chapter was designed to give an overview to the basic use of ATLAS without regard to
the details required for a given technology. Requirements for ATLAS simulation vary
considerably. The needs of the sub-micron MOS device engineer, the 1000V power device
engineer, and the III-V RF device engineer differ and cannot all be covered in one chapter.
SILVACO provides many references to individual technology problems using ATLAS. These
are:
• A library of standard examples that can be accessed on-line from DECKBUILD. Look at
these examples not only for their technology but also related ones. For example, different
aspects of high frequency analysis is covered in the MESFET and silicon bipolar example
sections.
• The chapters for each ATLAS module in this manual.
• The Simulation Standard, a newsletter distributed by SILVACO. To make sure you’re on
the mailing list, contact your local SILVACO office or go to www.silvaco.com.
• The SILVACO website also provides detailed information. It contains on-line versions of
the articles in our newsletter, on-line searchable index of the examples, links to other
TCAD web sites, and a section on solutions to known problems with all SILVACO
programs.
• For more information about suggested technology specific strategies, contact your local
SILVACO support engineer at [email protected].

95 ATLAS User’s Manual

Chapter 3
Physics
Basic Semiconductor Equations Physics

3.1 Basic Semiconductor Equations

Years of research into device physics have resulted in a mathematical model that operates on
any semiconductor device [214]. This model consists of a set of fundamental equations,
which link together the electrostatic potential and the carrier densities, within some
simulation domain. These equations, which are solved inside any general purpose device
simulator, have been derived from Maxwell’s laws and consist of Poisson’s Equation (see
Section 3.1.1 “Poisson’s Equation”), the continuity equations (see Section 3.1.2 “Carrier
Continuity Equations”) and the transport equations (see Section 3.1.3 “The Transport
Equations”). Poisson’s Equation relates variations in electrostatic potential to local charge
densities. The continuity and the transport equations describe the way that the electron and
hole densities evolve as a result of transport processes, generation processes, and
recombination processes.
This chapter describes the mathematical models implemented in ATLAS. Note that a
discretization of the equations is also performed so that they can be applied to the finite
element grid used to represent the simulation domain.

3.1.1 Poisson’s Equation

Poisson’s Equation relates the electrostatic potential to the space charge density:
div    = –  3-1
where  is the electrostatic potential,  is the local permittivity, and  is the local space charge
density. The reference potential can be defined in various ways. For ATLAS, this is always
the intrinsic Fermi potential i which is defined in the next section. The local space charge
density is the sum of contributions from all mobile and fixed charges, including electrons,
holes, and ionized impurities.
The electric field is obtained from the gradient of the potential (see Equation 3-2).

E = –  3-2

3.1.2 Carrier Continuity Equations

The continuity equations for electrons and holes are defined by equations:
n 1
------ = --- divJ n + G n – R n 3-3
t q

p 1
------ = – --- divJ p + G p – R p 3-4
t q
where n and p are the electron and hole concentration, J n and J p are the electron and hole
current densities, Gn and Gp are the generation rates for electrons and holes, Rn and Rp are the
recombination rates for electrons and holes, and q is the magnitude of the charge on an
electron.

97 ATLAS User’s Manual

Basic Semiconductor Equations Physics

By default ATLAS includes both Equations 3-3 and 3-4. In some circumstances, however, it
is sufficient to solve only one carrier continuity equation. The specification of which
continuity equations are to be solved is performed in the METHOD statement by turning off any
equation that is not to be solved. The syntax, ^ELECTRONS or ^HOLES, turns off the electron
continuity equation and the hole continuity equation respectively.

3.1.3 The Transport Equations

Equations 3-1, 3-3, and 3-4 provide the general framework for device simulation. But further
secondary equations are needed to specify particular physical models for: J n , J p , Gn, Rn, Gp
and Rp.
The current density equations, or charge transport models, are usually obtained by applying
approximations and simplifications to the Boltzmann Transport Equation. These assumptions
can result in a number of different transport models such as the drift-diffusion model, the
Energy Balance Transport Model or the hydrodynamic model. The choice of the charge
transport model will then have a major influence on the choice of generation and
recombination models.
The simplest model of charge transport that is useful is the Drift-Diffusion Model. This model
has the attractive feature that it does not introduce any independent variables in addition to ,
n and p. Until recently, the drift-diffusion model was adequate for nearly all devices that were
technologically feasible. The drift-diffusion approximation, however, becomes less accurate
for smaller feature sizes. More advanced energy balance and hydrodynamic models are
therefore becoming popular for simulating deep submicron devices. ATLAS supplies both
drift-diffusion and advanced transport models.
The charge transport models and the models for generation and recombination in ATLAS
make use of some concepts associated with carrier statistics. These concepts are summarized
in a further section of this chapter that deals with the carrier statistics.
Drift-Diffusion Transport Model
Derivations based upon the Boltzmann transport theory have shown that the current densities
in the continuity equations may be approximated by a drift-diffusion model [250]. In this
case, the current densities are expressed in terms of the quasi-Fermi levels n and p as:

J n = – q n n n 3-5

J p = – q p p p 3-6
where n and p are the electron and hole mobilities. The quasi-Fermi levels are then linked
to the carrier concentrations and the potential through the two Boltzmann approximations:
q   – n 
n = n ie exp ------------------------ 3-7
kT L

–q   – p 
p = n ie exp --------------------------- 3-8
kT L

98 ATLAS User’s Manual

Basic Semiconductor Equations Physics

where nie is the effective intrinsic concentration and TL is the lattice temperature. These two
equations may then be re-written to define the quasi-Fermi potentials:
kT L n
 n =  – --------- ln ------- 3-9
q n ie

kT L p
 p =  + --------- ln ------- 3-10
q n ie

By substituting these equations into the current density expressions, the following adapted
current relationships are obtained:

J n = qD n n – qn n  –  n n  kT L   ln n ie   3-11

J p = – q D p p – qp p  +  p p  kT L   ln n ie   3-12
The final term accounts for the gradient in the effective intrinsic carrier concentration, which
takes account of bandgap narrowing effects. Effective electric fields are normally defined
whereby:
kT L
E n = –   + --------- ln n ie 3-13
 q 

kT L
E p = –   – --------- ln n ie 3-14
 q 

which then allows the more conventional formulation of drift-diffusion equations to be

written (see Equations 3-15 and 3-16).

J n = qn n E n + qD n n 3-15

J p = qp p E p – q D p p 3-16
It should be noted that this derivation of the drift-diffusion model has tacitly assumed that the
Einstein relationship holds. In the case of Boltzmann statistics this corresponds to:
kT L
D n = ---------  n 3-17
q

kT L
D p = ---------  p 3-18
q

99 ATLAS User’s Manual

Basic Semiconductor Equations Physics

If Fermi-Dirac statistics are assumed for electrons, Equation 3-17 becomes:

kT
 --------L-   F  1 
---------   –  
 q n 1  2  kT L Fn C 
 
D n = ---------------------------------------------------------------------------- 3-19
 1 
F –1  2  ---------   Fn –  C  
 LkT 

where F  is the Fermi-Dirac integral of order  and Fn is given by -qn. An analogous
expression is used for holes with Fermi-Dirac statistics.

Note: See Section 3.2.9 “Bandgap Narrowing” for more information on the effects resulting from bandgap
narrowing and their implementation.

Energy Balance Transport Model

A higher order solution to the general Boltzmann Transport Equation consists of an additional
coupling of the current density to the carrier temperature, or energy. The current density
expressions from the drift-diffusion model are modified to include this additional physical
relationship. The electron current and energy flux densities are then expressed as:
T
J n = qD n n – q n n + qnD n T n 3-20

k n
S n = – K n T n –  -------- J n T n 3-21
 q 

T
J p = – q D p p – q p p – q pD p T p 3-22

k p
S p = – K p T p –  -------- J p T p 3-23
 q 

where Tn and Tp represent the electron and hole carrier temperatures and Sn and Sp are the
flux of energy (or heat) from the carrier to the lattice. The energy balance transport model
includes a number of very complex relationships and therefore a later section of this chapter
has been devoted to this model.

3.1.4 Displacement Current Equation

For time domain simulation, the displacement current is calculated and included in the
structure, log file and the run time output. The expression for displacement current is given
as:

E
J dis =   ------ 3-24
 t 

100 ATLAS User’s Manual

Basic Theory of Carrier Statistics Physics

3.2 Basic Theory of Carrier Statistics

3.2.1 Fermi-Dirac and Boltzmann Statistics
Electrons in thermal equilibrium at temperature TL with a semiconductor lattice obey Fermi-
Dirac statistics. That is the probability f() that an available electron state with energy  is
occupied by an electron is:
1
f    = ---------------------------------------- 3-25
 – EF
1 + exp  ---------------
 kT L 

where F is a spatially independent reference energy known as the Fermi level and k is
Boltzmann’s constant.
In the limit,  - EF >> kTL, Equation 3-25 can be approximated as:

EF – 
f    = exp  --------------- 3-26
 kT L 

Statistics based on the use of Equation 3-26 are referred to as Boltzmann statistics [326, 128].
The use of Boltzmann statistics instead of Fermi-Dirac statistics makes subsequent
calculations much simpler. The use of Boltzmann statistics is normally justified in
semiconductor device theory, but Fermi-Dirac statistics are necessary to account for certain
properties of very highly doped (degenerate) materials.
The Fermi-Dirac statistics have been implemented in ATLAS in a similar form to Boltzmann
statistics.
The remainder of this section outlines derivations and results for the simpler case of
Boltzmann statistics which are the default in ATLAS. You can have ATLAS use Fermi-Dirac
statistics by specifying the FERMIDIRAC parameter in the MODEL statement.

3.2.2 Effective Density of States

Integrating the Fermi-Dirac statistics over a parabolic density of states in the conduction and
valence bands, whose energy minimum is located at energies EC and EV respectively, yields
the following expressions for the electron and hole concentrations:
EF – EC
n = N C F 1  2  ------------------- 3-27
 kT L 

Ev – EF
p = N V F 1  2  ------------------ 3-28
 kT  L

where F1/2() is referred to as the Fermi-Dirac integral of order 1/2. If Equation 3-26 is a good
approximation, then Equations 3-27 and 3-28 can be simplified to
EF – EC
n = N C exp  ------------------- 3-29
 kT  L

101 ATLAS User’s Manual

Basic Theory of Carrier Statistics Physics

Ev – EF
p = N V exp  ------------------ 3-30
 kT  L

which are referred to as the Boltzmann approximation.

NC and NV are referred to as the effective density of states for electrons and holes and are
given by:
3
---
 2 m e* kT L 2 T L NC · F
N C  T L  = 2  -------------------------- M c =  ---------- NC300 3-31
 2
  300
h

3
* ---
 2 m h kT L 2 T L NV · F
N V  T L  = 2  --------------------------- =  ---------- NV300 3-32
 2
  300
h

where Mc is the number of equivalent conduction band minima. NC300 and NV300 are user-
definable on the MATERIAL statement as shown in Table 3-1.
In some circumstances, the lattice temperature, TL, is replaced by the electron temperature,
Tn, in Equation 3-31 and hole temperature, Tp, in Equation 3-32.

Table 3-1 User-Definable Parameters for the Density of States

Statement Parameter Default Units

MATERIAL NC300 2.81019 cm-3

MATERIAL NV300 1.041019 cm-3

MATERIAL NC.F 1.5

MATERIAL NV.F 1.5

3.2.3 Intrinsic Carrier Concentration

Multiplying Equations 3-29 and 3-30 yields:
2
np = n ie 3-33

where nie is the intrinsic carrier concentration and is given for Boltzmann statistics by:

 –Eg 
n ie = N C N V exp  ------------- 3-34
 2kT L

Eg = EC - EV is the band-gap energy.

For intrinsic (undoped) material, p = n. By equating Equations 3-29 and 3-30 and solving for
EF yields:

102 ATLAS User’s Manual

Basic Theory of Carrier Statistics Physics

EC + EV kT L  N v
E F = E i = – q i = --------------------- +  --------- 1n  ------ 3-35
2  2  N 
c

where i is the Fermi level for intrinsic doped silicon, and i is the intrinsic potential.
Equation 3-35 also defines the intrinsic potential under non-equilibrium conditions. As
indicated previously, for ATLAS the  used in Equation 3-1 is the intrinsic potential.
The electron and hole concentrations can be expressed in terms of the intrinsic carrier
concentration as:
q   – n 
n = n ie exp ------------------------ 3-36
kT L

–q   – p 
p = n ie exp --------------------------- 3-37
kT L

where is the intrinsic potential and  is the potential corresponding to the Fermi level (i.e.,
EF = q).
The expression for intrinsic carrier concentration, nie, can be generalized to Fermi-Dirac
statistics using Equations 3-27 and 3-28. Specifying the NI.FERMI parameter in the MODELS
statement will cause ATLAS to calculate nie using Fermi-Dirac statistics.

3.2.4 Evaluation of Fermi-Dirac Integrals

1
In addition to the Fermi-Dirac integral of order --- as used in Equations 3-27 and 3-28,
2
3 –1 3
ATLAS also needs to evaluate the Fermi-Dirac integrals of order –-----
-, ------, --- . There are simple,
2 2 2
quickly calculated, but relatively poor approximations available for these integrals. You can
also evaluate them to machine precision; however, the amount of computation required makes
the evaluations relatively slow. ATLAS uses a Rational Chebyshev approximation scheme,
which is efficient in terms of CPU use and also has accuracy close to the machine precision.
In the worst case, the approximation differs from the actual values by 1 part in 1010 and is
typically much better.

103 ATLAS User’s Manual

Basic Theory of Carrier Statistics Physics

3.2.5 Rules for Evaluation of Energy Bandgap

In the ATLAS simulator, there are several ways to evaluate the value of energy bandgap for a
given material. To uniquely define the approach used, there is a set of rules that are followed
in a hierarchical manner. The following is a description of this hierarchy:
1. For each material or material system, there may be a default approach. To determine the
default for a given material or material system, refer to the descriptions given in Section
5.4 “Material Dependent Physical Models”. If the material system is not listed in this
section, then look for default energy bandgap model constants in Appendix B “Material
Systems”.
2. If you set the EG300 parameter of the MATERIAL statement to a value greater than zero
and you do not set the PASSLER parameter of the MODELS statement, the universal energy
bandgap model described in Section 3.2.6 “The Universal Energy Bandgap Model” will
be used.
3. If you set the EG300 parameter of the MATERIAL statement to a value greater than zero
and you set the PASSLER parameter of the MODELS statement, then Passler's model for
temperature dependent energy bandgap described in Section 3.2.7 “Passler's Model for
Temperature Dependent Bandgap” will be used.
4. If you specify the K.P parameter of the MODELS statement for materials in the InAlGaN
system, then the bandgap is taken as the minimum transition energy calculated following
the approach given by the strained zincblende tow band model described in Section 3.9.8
“Strained Two-Band Zincblende Model for Gain and Radiative Recombination” or the
strained wurtzite 3 band model described in Section 3.9.10 “Strained Wurtzite Three-
Band Model for Gain and Radiative Recombination”.
5. If you assign the F.BANDCOMP parameter of the MATERIAL statement to a valid filename,
the C interpreter function for energy bandgap will be used.
6. If you assign the EG12BOW parameter to a non-zero value, the general terniary bowing
model described in Section 3.2.8 “General Ternary Bandgap Model with Bowing” will be
used.

3.2.6 The Universal Energy Bandgap Model

The temperature dependence of the bandgap energy is modeled in ATLAS as follows [268]:
2 2
EGALPHA T L  300
2 TL
E g  T L  = E g  0  – -------------------------------------------- = EG300 + EGALPHA ------------------------------------------ – -------------------------------------
- 3-38
T L + EGBETA 300 + EGBETA T L + EGBETA

You can specify EG300, EGALPHA and EGBETA parameters in the MATERIAL statement (see
Table 3-2).

Table 3-2 User-Specifiable Parameters for Equation 3-38

Statement Parameter Default Units

MATERIAL EG300 1.08 eV

MATERIAL EGALPHA 4.7310-4 eV/K

MATERIAL EGBETA 636 K

104 ATLAS User’s Manual

Basic Theory of Carrier Statistics Physics

Table 3-2 User-Specifiable Parameters for Equation 3-38

Statement Parameter Default Units

MATERIAL NC300 2.81019 cm-3

MATERIAL NV300 1.041019 cm-3

The default values are material dependent, which can be found in Appendix B “Material
Systems”. Table 3-2 displays the defaults for Silicon only.

3.2.7 Passler's Model for Temperature Dependent Bandgap

An alternative temperature dependent bandgap model by Passler [208] can be enabled by the
PASSLER parameter of the MODELS statement.
The bandgap is given by

A · PASSLER  T · PASSLER-
Eg  T  = Eg  0  – ---------------------------------------------------------------------- 3-39
2

 P · PASSLER 
  2T 1  P · PASSLER 

  --------------------------------- +1 – 1
 T · PASSLER 
 

where Eg(T) is the lattice temperature dependent bandgap, T is the lattice temperature,
A.PASSLER, T.PASSLER and P.PASSLER are user-specified parameters on the MATERIAL
statement and Eg(0) is given by

A · PASSLER  T · PASSLER
Eg  0  = EG300 – ----------------------------------------------------------------------
- 3-40
2

 P · PASSLER 
  2  300  1  P · PASSLER 
  --------------------------------- +1 – 1
 T · PASSLER 
 

where EG300 is a user specified parameter on the MATERIAL statement.

105 ATLAS User’s Manual

Basic Theory of Carrier Statistics Physics

3.2.8 General Ternary Bandgap Model with Bowing

For Ternary compound semiconductors, you can use a bandgap model that continuously and
non-linearly interpolates between the bandgaps of the two binary extremes in composition. To
enable this model, you must specify the EG12BOW, EG1300 and EG2300 parameters of the
MATERIAL statement. EG1300 is the 300K bandgap of the binary compound at a composition
fraction of x=0. EG2300 is the 300K bandgap of the binary compound at a composition
fraction of x=1. EG12BOW is the bowing factor. The bandgap as a function of composition is
given by Equation 3-41.
Eg = EG2300 x + EG1300  1 – x  – EG12BOW x  1 – x  3-41
where x is the composition fraction.
You can include the effects of temperature in the Bowing model by specifying the EG1ALPH,
EG1BETA, EG2ALPH and EG2BETA parameters of the MATERIAL statement. These parameters
are used in Equations 3-72 through 3-74 to calculate the bandgap.

2
300
2 TL
Eg 1 = EG1300 + EG1ALPH ------------------------------------------ – ---------------------------------------- 3-42
300 + EG1BETA T L + EG1BETA

2
300
2 TL
Eg2 = EG2300 + EG2ALPH ------------------------------------------ – ---------------------------------------- 3-43
300 + EG2BETA T L + EG2BETA

Eg = Eg2 x + Eg 1  1 – x  – EG12BOW x  1 – x  3-44

3.2.9 Bandgap Narrowing

In the presence of heavy doping, greater than 1018cm-3, experimental work has shown that the
pn product in silicon becomes doping dependent [260]. As the doping level increases, a
decrease in the bandgap separation occurs, where the conduction band is lowered by
approximately the same amount as the valence band is raised. In ATLAS this is simulated by
a spatially varying intrinsic concentration nie defined according to Equation 3-45.

Eg
n ie = n i exp  -----------
2 2
3-45
 kT 

Bandgap narrowing effects in ATLAS are enabled by specifying the BGN parameter of the
MODELS statement. These effects may be described by an analytic expression relating the
variation in bandgap, Eg, to the doping concentration, N. The expression used in ATLAS is
from Slotboom and de Graaf [261]:
1
 --- 
2
 
 N  N  2 
E g = BGN · E  ln ------------------ + ln ------------------ + BGN · C  3-46
 BGN · N   
BGN · N
 
 

106 ATLAS User’s Manual

Basic Theory of Carrier Statistics Physics

You can specify the BGN.E, BGN.N, and BGN.C parameters in the MATERIAL statement. The
default values from Slotboom [261] and Klaassen [141] are shown in Table 3-3. Specify BGN
or BGN.SLOTBOOM on the MODELS statement to choose the Slotboom set of defaults for use in
Equation 3-46. Alternatively, specify BGN2 or BGN.KLAASSEN on the MODELS statement to use
the Klaassen set of defaults in Equation 3-46. You can verify which values are being used
with the PRINT parameter on the MODELS statement.

Table 3-3 User-Definable Parameters of Bandgap Narrowing Model

Defaults Defaults
Statement Parameter Units
(Slotboom) (Klaassen)

MATERIAL BGN.E 9.010-3 6.9210-3 eV

MATERIAL BGN.N 1.01017 1.31017 cm-3

MATERIAL BGN.C 0.5 0.5

The variation in bandgap is introduced to the other physical models by subtracting the result
of Equation 3-47 from the bandgap, Eg. In addition an adjustment is also made to the electric
field terms in the transport models as described earlier. The adjustment takes the form:
kT L
E n = –   + --------- ln n ie 3-47
q

kT L
E p = –   – --------- ln n ie 3-48
q

The variation in the energy bandgap is also applied partially to the electron affinity,  . The
effective electron affinity,  eff given as follows:

 eff =  + E g  ASYMMETRY 3-49

where ASYMMETRY is a user-specifiable asymmetry factor. You can specify the value of the
asymmetry factor using the ASYMMETRY parameter of the MATERIAL statement.

Note: In addition to this in-built model for bandgap narrowing, ATLAS allows you to use its C-INTERPRETER
module. You can write an external file with C-code that defines an equation for bandgap narrowing. The
filename is specified with the F.BGN parameter in the MODEL statement. See Appendix A “C-Interpreter
Functions” for more information on C-INTERPRETER.

107 ATLAS User’s Manual

Basic Theory of Carrier Statistics Physics

3.2.10 The Universal Bandgap Narrowing Model

A universal bandgap narrowing model has been suggested [216]. This model given in
Equation 3-50 relates the change in bandgap to the material effective masses, mc and mv, and
the static dielectric constant s. The UBGN.C and UBGN.B parameters are fitting parameters to
define in the MATERIAL statement. To enable the model, specify UBGN on the MODELS
statement.
1
5 – ---
s mc + mv 2 s
4
E g = – UBGN · C -----  m 0 ------------------- + UBGN · BT ---- 3-50
N mc mv N

In Equation 3-50, N is the net doping concentration. The default values for UBGN.C and
UBGN.B are 3.910-5 eV cm3/4 and 3.101012 cm-3 K-2.

3.2.11 Bennett-Wilson and del Alamo Bandgap Models

These are variations of the Slotboom and de Graaf model. The Bennett-Wilson formula [25]
is
2
E g = – BGN · E  log  N  BGN · N   3-51

and the del-Alamo formula [5], [267] is

E g = – BGN · E  log  N  BGN · N   3-52

where N is the total doping concentration. In both cases, the formula is only used for
N>BGN.N, otherwise it is set at zero. Table 3-4 shows the default values of BGN.E and BGN.N
for Silicon.
.

Table 3-4 Silicon Default Values for Bennett-Wilson and del Alamo Bandgap models

Statement Parameter Units Default (Bennett) Default (Del Alamo)

MATERIAL BGN.E eV 6.8410-3 1.8710-2

MATERIAL BGN.N cm-3 3.1621018 7.01017

To enable the Bennett-Wilson model, use the BGN.BENNETT flag on the MODELS statement. To
enable the del Alamo model, use the BGN.ALAMO flag on the MODELS statement.

108 ATLAS User’s Manual

Basic Theory of Carrier Statistics Physics

3.2.12 Schenk Bandgap Narrowing Model

Some bandgap narrowing models are described in previous sections. In those models, the
bandgap narrowing depends only on the doping level and temperature. High doping levels are
required in these models to produce significant bandgap narrowing. It has been observed
experimentally that electron-hole plasma densities in regions with low doping levels can also
produce bandgap narrowing effects [193]. An analytical model of bandgap narrowing in
Silicon was subsequently developed by Schenk [244], which takes into account both the
contribution due to the electron-hole plasma and the contribution due to dopant-carrier
interactions. These are referred to as the exchange-correlation (xc) contributions and the ionic
xc i
contributions (i) respectively, and are denoted by  t and  t , where t can be either e for
electrons or h for holes. Schenk uses the excitonic Bohr radius as a length scale in his
formulae. This is simply the radius of the orbit of the ground state of an electron and hole
electrostatically bound to each other. It is obtained by multiplying the Bohr radius for the
hydrogen atom (5.29 nanometers) by the relative dielectric permittivity of Silicon and
dividing by the reduced mass of the excitonic system. The effective masses of electrons and
holes in this model can be set by the parameters BGN.SHNK.ME and BGN.SHNK.MH on the
MATERIAL statement. The reduced mass is given by

BGN.SHNK.ME  BGN.SHNK.MH
 = -------------------------------------------------------------------------------------- 3-53
 BGN.SHNK.ME + BGN.SHNK.MH 
and the excitonic Bohr radius in centimetres is calculated from
–9
5.29  10 BGN.SHNK.EPS
 ex = ----------------------------------------------------------------------- 3-54

The excitonic Rydberg energy is used as an energy scale. This is obtained by multiplying the
binding energy of the ground state of the Bohr hydrogen atom (13.6 eV) as follows
13.6
E ex = -------------------------------------------
- 3-55
2
BGN.SHNK.EPS
Default values of these quantities are given in Table 3-5.

Table 3-5 Parameters for the Schenk Band Gap Narrowing Model

Parameter Default Units

BGN.SHNK.ME 0.321

BGN.SHNK.MH 0.346

BGN.SHNK.EPS 11.7

BGN.SHNK.GE 12

BGN.SHNK.GH 4

 0.1665

ex 3.71910-7 cm
Eex 1.65510-2 eV

109 ATLAS User’s Manual

Basic Theory of Carrier Statistics Physics

xc
The scaled, dimensionless, carrier concentrations used in the expressions for  t are

3
n e = n ex 3-56

3
n h = p ex 3-57

ns = ne + nh 3-58

 
n w = ------------------------------------- n e + ------------------------------------- n h 3-59
BGN.SHNK.ME BGN.SHNK.MH
where n is the free electron concentration and p is the free hole concentration. Similarly, the
dimensionless thermal energy parameter is
KT
 = -------- 3-60
E ex

Then, the carrier concentration and temperature dependent energy band shifts are given by
13
3 2 48n t 52
 4  n s  ----------- + c t log  1 + d t n w  +  ---------- n t  + 8n s 
pt 8 2
xc  g t   g t m t
 t = – ------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 3-61
3 2 3 2 32
 4  n s +  + b t n s  + 40n s 

where t = [e,h] and ge= BGN.SHNK.GE, gh = BGN.SHNK.GH, me =BGN.SHNK.ME, mh=

BGN.SHNK.MH. The overall bandgap narrowing due to the electron-hole plasma is obtained
xc xc
from multiplying the sum  e +  h by the energy scale Eex. In order to obtain an analytical

expression which is accurate over a large range of dopant and carrier densities, Pad e
approximation theory was used by Schenk. This results in the parameters bt, ct, dt, and pt.
Because of the complexity of the fitting process, these are set to the values given in [244] and
you cannot modify them. Their values are shown in Table 3-6.
The other component of the bandgap narrowing is the ionic contribution, which we denote as
i
 t . This depends on the local dopant level, and the carrier densities which would occur
under charge neutrality conditions. Following the simplifications in [244], the following
parameters are defined
3
n d = N d  ex 3-62

3
n a = N a  ex 3-63

ns = nd + Na 3-64

 
n w = ------------------------------------- n d + ------------------------------------- n a 3-65
BGN.SHNK.ME BGN.SHNK.ME
where Nd is the donor density and Na is the acceptor density. The ionic contributions are

i – nd + na   1 +  
 t = ----------------------------------------------------------------------------------------------------------------------------------------- 3-66
34 qt
n s  2  1 + h t log  1 + n s     + j t n w  1 + k t n w 

110 ATLAS User’s Manual

Basic Theory of Carrier Statistics Physics

where
2
ns
 = ------
- 3-67
3


The Pad e fitting parameters ht, jt, kt, and qt have the same values used in [244] and you
cannot change them. Their values are given in Table 3-6.
The overall bandgap narrowing in electron Volts is obtained by multiplying the sum of
i i
 e +  h by the energy scaling factor Eex and adding the result to the plasma contributions to
the bandgap narrowing. A fraction of this bandgap narrowing is also applied to the electron
affinity. The default value of this fraction is 0.5. This can be changed by using the ASYMMETRY
parameter on the MATERIAL statement.
The model is enabled by specifying the parameter BGN.SCHENK on the MODELS statement.
The carrier density dependence of the bandgap and electron affinity calls for a more
complicated numerical implementation. It was found that convergence on the left hand side
norm could in some cases be achieved when the right hand side convergence had not fully
occurred. It is therefore recommended that you reduce the the left hand side convergence
criterion by using the CX.TOL parameter on the METHOD statement when using this model.
Alternatively, you can require right hand side convergence by specifying RHS.NORM on the
METHOD statement. In this latter case, it may be necessary to adjust the CR.TOL parameter on
the METHOD statement to achieve convergence.

Table 3-6 Fitting Parameters

Value Value Value Value Value Value Value Value

be 8 ce 1.3346 de 0.893 pe · he 3.91 je 2.8585 ke 0.012 qe 0.75

0.233 3

bh 1 ch 1.2365 dh 1.153 ph · hh 4.20 jh 2.9307 kh 0.19 qh 0.25

0.233 3

3.2.13 Lindefelt Bandgap Narrowing Model

A model for the doping induced bandgap narrowing in 3C-SiC, 4H-SiC, and 6H-SiC due to
Lindefelt [164] is available. You implement the model by specifying BGN.LIND on the
MODELS statement.
According to this model, the downward shift of the conduction band in n-doped material is
given by
+ 13 + 12
 Nd   Nd 
E c = BGN.LIND.ANC  ------------------------------------- + BGN.LIND.BNC  ------------------------------------- 3-68
 BGN.LIND.ND  BGN.LIND.ND

and the upward shift of the valence band is given by

+ 14 + 12
 Nd   Nd 
E v = BGN.LIND.ANV  ------------------------------------- + BGN.LIND.BNV  ------------------------------------- 3-69
 BGN.LIND.ND  BGN.LIND.ND

111 ATLAS User’s Manual

Basic Theory of Carrier Statistics Physics

+
where N d is the ionized donor concentration. The corresponding quantities in p-doped
material are
- 14 - 12
 Na   Na 
E c = BGN.LIND.APC  ------------------------------------- + BGN.LIND.BPC  ------------------------------------- 3-70
 BGN.LIND.NA  BGN.LIND.NA

and
- 13 - 12
 Na   Na 
E v = BGN.LIND.APV  ------------------------------------- + BGN.LIND.BPV  ------------------------------------- 3-71
 BGN.LIND.NA  BGN.LIND.NA

-
where N a is the ionized acceptor concentration. The bandgap narrowing is obtained from the
sum of E c and E v . The default values for these parameters are given in Table 3-7 for four
different materials.
Because the bandgap narrowing depends on the ionized impurity densities rather than the
physical density, it is recommended to take incomplete ionisation into account by specifying
INCOMPLETE on the MODELS statement when using this model. In SiC, polytypes dopant
ionization energies are relatively large and incomplete ionization can be significant [151].

Table 3-7 Default values of Parameters for Lindefelt Bandgap Narrowing Model.

Parameter Type Default (Si) Default (3C-SiC) Default (4H-SiC) Default (6H-SiC) Units

BGN.LIND.ANC Real 9.74 10-3 1.48 10-2 1.5 10-2 1.12 10-2 eV
BGN.LIND.BNC Real 1.39 10-3 3.06 10-3 2.93 10-3 1.01 10-3 eV
BGN.LIND.ANV Real 1.27 10-2 1.75 10-2 1.9 10-2 2.11 10-2 eV
BGN.LIND.BNV Real 1.4 10-3 6.85 10-3 8.74 10-3 1.73 10-3 eV
BGN.LIND.APC Real 1.14 10-2 1.5 10-2 1.57 10-2 1.74 10-2 eV
BGN.LIND.BPC Real 2.05 10-3 6.41 10-4 3.87 10-4 6.64 10-4 eV
BGN.LIND.APV Real 1.11 10-2 1.3 10-2 1.3 10-2 1.3 10-2 eV
BGN.LIND.BPV Real 2.06 10-3 1.43 10-3 1.15 10-3 1.14 10-3 eV

Although the model was developed for SiC polytypes, it has also been applied to Silicon. It is
therefore active in regions where the material is either Silicon, 3C-SiC, 4H-SiC, or 6H-SiC
when BGN.LIND is enabled on the MODELS statement. Other bandgap narrowing models can
be used in other regions so long as the REGION parameter is used on the MODELS statement to
localize each model.

112 ATLAS User’s Manual

Basic Theory of Carrier Statistics Physics

This model has also been fitted to the Slotboom expression, Equation 3-46, for 4H-SiC and
6H-SiC, by Lades [151]. The coefficients are different according to whether the material is n-
type or p-type. Thus, BGN.N is replaced by BGN.N.DON for n-type material and BGN.N.ACC
for p-type material. BGN.E is replaced by BGN.E.DON and BGN.E.ACC. BGN.C is fixed at 0.5
in all cases. If the BGN or BGN2 flags are specified on the MODELS statement, then the default
parameters for the Slotboom expression in 4H-SiC and 6H-SiC are as listed in Table 3-8.

Table 3-8 Default values of parameters for Slotboom Bandgap narrowing model for 4H-SiC and
6H-SiC.

Parameter Type Default (4H-SiC) Default (6H-SiC) Units

BGN.N.DON Real 1 1017 1 1017 cm-3

BGN.N.ACC Real 1 1017 1 1017 cm-3
BGN.E.DON Real 2 10-2 9 10-3 eV
BGN.E.ACC Real 9 10-3 9 10-3 eV

113 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

3.3 Space Charge from Incomplete Ionization, Traps, and Defects

Poisson’s Equation (Equation 3-1) including the carrier concentrations, the ionized donor and
acceptor impurity concentrations ND+ and NA-, charge due to traps and defects, QT, has the
form:
+ –
div    = q  n – p – N D + NA  –QT 3-72

In ATLAS the default is to assume full impurity ionization (i.e., ND+=ND,total and NA-
,=NA,total), and to set QT equal to zero.
ATLAS also provides the options of accounting for incomplete ionization of impurities and
accounting for additional charge associated with traps and defects.

3.3.1 Incomplete Ionization of Impurities

ATLAS can account for impurity freeze-out [126] with appropriate degeneracy factors GCB
and GVB for conduction and valence bands. The ionized donor and acceptor impurity
concentrations are then given by:
ND
ND + = ------------------------------------------------------------------------------- 3-73
  F –  E C – E DB 
 + GCB exp  -------------------------------------------
n
 kT L 
 

– NA
N A = --------------------------------------------------------------------------- 3-74
 E V + EAB –  F 
1 + GVB exp  --------------------------------------
p
 kT L 
 

where EDB and EAB are the dopant activation energies and ND and NA are net compensated n-
type and p-type doping, respectively. Net compensated doping is defined as follows:
If
Ntotal (ND,total - NA,total) > 0 3-75

then
ND = | Ntotal | and NA = 0 3-76

Otherwise
ND = 0 and NA = | Ntotal | ) 3-77

The INCOMPLETE parameter of the MODELS statement is used to select incomplete ionization
and the parameters.

114 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

Table 3-9 User-Specifiable Parameters for Equations 3-73 and 3-74

Statement Parameter Units

MATERIAL GCB

MATERIAL EDB eV
MATERIAL GVB

MATERIAL EAB eV

To properly handle incomplete ionization in silicon for high doping levels, a new incomplete
ionization model has been added.
The models that form incomplete ionization of impurities given by Equations 3-73 and 3-74
give good physical results for low to moderately doped semiconductors. For heavily (greater
than 31018/cm3) doped semiconductors, these models fail to predict experimental results of
complete ionization. For silicon, an optional model has been introduced that better matches
experimental results. This model is set by the IONIZ parameter of the MODELS statement.
In this model, the activation energies of the dopants in Equations 3-73 and 3-74 have been
modified for doping dependence as given in the following equations:
13
EDB eV  = A · EDB – B · EDBN D 3-78

13
EAB eV  = A · EAB – B · EABN A 3-79

At doping concentrations above 31018 cm-3, the model predicts complete ionization. At
doping concentrations between 1018 cm-3 and 31018 cm-3, the model predicts a linearly
interpolated value between the above expressions and complete ionization.
The default value for the parameters of Equations 3-78 and 3-79 were chosen to represent
reasonable values for silicon and are given in Table 3-10.

Table 3-10 User-Specifiable Parameters for Equations 3-78 and 3-79

Statement Parameter Units

MATERIAL A.EDB 0.044

MATERIAL B.EDB 3.6e-8
MATERIAL A.EAB 0.0438
MATERIAL B.EAB 3.037e-5

115 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

For general semiconductors, you can use Equations 3-78 and 3-79 without the transitions to
complete ionization. To enable this model, specify INC.ION in the MODEL statement.

3.3.2 Low Temperature Simulations

In conjunction with Fermi-Dirac statistics and impurity freeze-out, ATLAS simulates device
behavior under low operating temperatures. In general, simulations can be performed at
temperatures as low as 50K without loss of quadratic convergence. Below this temperature,
carrier and ionization statistics develop sharp transitions which cause slower convergence.
Since many more iterations will probably be required if temperatures below 50K are
specified, the ITLIMIT parameter of the METHOD statement should be increased.
Due to the limited exponent range on some machines, ATLAS can have trouble calculating
the quasi-Fermi level of minority carriers. As the temperature decreases, the intrinsic carrier
concentration also decreases. In quasi-neutral regions, the minority carrier concentration can
easily underflow. Such situations were handled in the past by setting these concentrations to
zero. This method does not allow an accurate subsequent calculation of minority carrier
quasi-Fermi levels. In order to accurately calculate quasi-Fermi levels, majority carrier
2 is used to obtain minority carrier concentrations in
concentration and the relation, np=n ie
case of an underflow. Despite these efforts, spurious glitches are occasionally observed at low
temperatures in the minority quasi-Fermi levels.

3.3.3 Traps and Defects

Semiconductor materials exhibit crystal flaws, which can be caused by dangling bonds at
interfaces or by the presence of impurities in the substrate. The presence of these defect
centers, or traps, in semiconductor substrates may significantly influence the electrical
characteristics of the device. Trap centers, whose associated energy lies in a forbidden gap,
exchange charge with the conduction and valence bands through the emission and capture of
electrons. The trap centers influence the density of space charge in semiconductor bulk and
the recombination statistics.
Device physics has established the existence of three different mechanisms, which add to the
space charge term in Poissons’s equation in addition to the ionized donor and acceptor
impurities. These are interface fixed charge, interface trap states and bulk trap states.
Interface fixed charge is modeled as a sheet of charge at the interface and therefore is
controlled by the interface boundary condition. Interface traps and bulk traps will add space
charge directly into the right hand side of Poisson’s equation. This section describes the
definition of bulk trap states and the implementation of these bulk trap states into ATLAS for
both steady state and transient conditions.
A donor-type trap can be either positive or neutral like a donor dopant. An acceptor-type trap
can be either negative or neutral like an acceptor dopant. A donor-like trap is positively
charged (ionized) when empty and neutral when filled (with an electron). An empty donor-
type trap, which is positive, can capture an electron or emit a hole. A filled donor-type trap,
which is neutral, can emit an electron or capture a hole. An acceptor-like trap is neutral when
empty and negatively charged (ionized) when filled (with an electron). A filled acceptor-like
trap can emit an electron or capture a hole. An empty acceptor-like trap can capture an
electron or emit a hole. Unlike donors, donor-like traps usually lie near the valence band.
Likewise, acceptor-like traps usually lie near the conduction band.

116 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

Figure 3-1 shows the terminology used within ATLAS to define the type of trap. The position
of the trap is defined relative to the conduction or valence bands using E.LEVEL so for
instance, an acceptor trap at 0.4eV would be 0.4eV below the conduction band.

EC
E.LEVEL for acceptor traps

EtA

EtD
E.LEVEL for donor traps

Figure 3-1: Definition of the trap energy level for acceptor and donor traps in reference to the conduction and
valence band edges.
Calculation of Trapped Charge in Poisson’s Equation
The total charge caused by the presence of traps is subtracted from the right hand side of
Poisson’s equation. The total charge value is defined by:
+ –
Q T = q  N tD – N tA  3-80

+ –
where N tD and N tA are the densities of ionized donor-like and acceptor-like traps
respectively. The ionized density depends upon the trap density, DENSITY, and its probability
of ionization, FtA and FtD. For donor-like and acceptor-like traps respectively, the ionized
densities are calculated by the equations:
+
Nt D = DENSITY  F tD 3-81

–
N tA = DENSITY  F tA 3-82

117 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

In the case where multiple traps at multiple trap energy levels are defined the total charge
becomes:
k m

  NtA
+ + – –
N tD = N tD  N tA = 3-83
=1 =1

where k is the number of donor-like traps and m is the number of acceptor-like traps.
The probability of ionization assumes that the capture cross sections are constant for all
energies in a given band and follows the analysis developed by Simmons and Taylor [259].
The probability of ionization is given by the following equations for acceptor and donor-like
traps.
v n SIGN n + e pA
F tA = ---------------------------------------------------------------------------------------
- 3-84
v n SIGN n + v p SIGP p + e nA + e pA

v p SIGP p + e nD
F tD = ------------------------------------------------------------------------------------------
- 3-85
v n SIGN n + v p SIGP p + e n D + e p D

where SIGN and SIGP are the carrier capture cross sections for electrons and holes
respectively, vn and vp are the thermal velocities for electrons and holes. For donor like traps,
the electron and hole emission rates, enD and epD, are defined by:

1 Et – Ei
e nD = --------------------------------- v n SIGN n i exp ---------------
- 3-86
DEGEN.FAC kT L

Ei – Et
e pD = DEGEN.FAC v p SIGP n i exp ---------------
- 3-87
kT L

where Ei is the intrinsic Fermi level position, Et is the trap energy level as defined by
E.LEVEL and DEGEN.FAC is the degeneracy factor of the trap center. The latter term takes into
account that spin degeneracy will exist, that is the “empty” and “filled” conditions of a defect
will normally have different spin and orbital degeneracy choices. For acceptor like traps, the
electron and hole emission rates, enA and epA, are defined by:

Et – Ei
e nA = DEGEN.FAC v n SIGN n i exp ---------------
- 3-88
kT L

1 Ei – Et
e pA = --------------------------------- v p SIGP n i exp ---------------
- 3-89
DEGEN.FAC kT L

118 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

Table 3-11 User-Specifiable Parameters for Equations 3-81 to 3-87

Statement Parameter Units

TRAP E.LEVEL eV
TRAP DENSITY cm-3
TRAP DEGEN.FAC

TRAP SIGN cm2

TRAP SIGP cm2

Trap Implementation into Recombination Models

To maintain self-consistency, you need to take into account that electrons are being emitted or
captured by the donor and acceptor-like traps. Therefore, the concentration of carriers will be
affected. This is accounted for by modifying the recombination rate in the carrier continuity
equations.
The standard SRH recombination term (see “Shockley-Read-Hall (SRH) Recombination” on
page 204) is modified as follows:
l m

R =  RD +  RA  
3-90
=1 =1

where l is the number of donor-like traps, m is the number of acceptor-like traps. For donor-
like traps, the function R is:
2
pn – n
ie
R
D
= ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 3-91
 E i – E t  E – E i
t
- n exp  ---------------------
1
TAUN p + DEGEN · FACn ie exp  ------------------ kT
- + TAUP n + ----------------------------------------------
ie  kT 
 L  DEGEN · FAC  L 

For acceptor like traps the function R is:

2
pn – n
ie
R
A
= ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 3-92
 E i – E t  E – E i
t
- n exp  ------------------- + TAUP n + DEGEN · FAC n exp  ---------------------
1
TAUN p + ---------------------------------------------- ie kT ie  kT 
DEGEN · FAC  L   L 

The electron and hole lifetimes TAUN and TAUP are related to the carrier capture cross sections
SIGN and SIGP through the equations:

1
TAUN = -------------------------------------------------- 3-93
SIGN v n DENSITY

1 3-94
TAUP = -------------------------------------------------
-
SIGP v p DENSITY

119 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

The thermal velocities vn and vp are calculated from the following electron and hole effective
masses.
3kT 1  2
v n =  -------------- 3-95
 mn m 
0

3kT 1  2
v p =  -------------- 3-96
 mp m 
0

To specify the effective masses directly, use the M.VTHN and M.VTHP parameters from the
MATERIAL statement. If M.VTHN or M.VTHP or both are not specified, the density of states
effective mass is extracted from the density of states (Nc or Nv) using Equations 3-31 and 3-
32. In the case of silicon if M.VTHN or M.VTHP are not specified, the effective masses are
calculated from:
–4
m n = 1.045 + 4.5  10 T 3-97

–3 –6 2
m p = 0.523 + 1.4  10 T – 1.48  10 T 3-98

where T is the lattice temperature.

Table 3-12 User-Specifiable Parameters for Equations 3-91-3-94

Statement Parameter Units

TRAP TAUN s
TRAP TAUP s

The TRAP statement activates the model and is used to:

• Specify the trap type DONOR or ACCEPTOR
• Specify the energy level E.LEVEL parameter
• Specify the density of the trap centers DENSITY
• Specify the degeneracy factor DEGEN.FAC
• Specify either the cross sections, SIGN and SIGP, or the electron and hole lifetimes TAUN
and TAUP.

120 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

Trap-Assisted Tunneling
Trap-Assisted Tunneling models the trap-to-band phonon-assisted tunneling effects for Dirac
wells. At high electric fields, tunneling of electrons from the valence band to the conduction
band through trap or defect states can have an important effect on the current.
Trap-assisted tunneling is modeled by including appropriate enhancement factors [115]
(nDIRAC and pDIRAC) in the trap lifetimes in Equation 3-91. These enhancement factors
modify the lifetimes so that they include the effects of phonon-assisted tunneling on the
emission of electrons and holes from a trap. This model is enabled by specifying
TRAP.TUNNEL in the MODELS statement.
For donor like traps, the recombination term for traps becomes:
2
pn – n
ie
R
D
= ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
TAUN  E i – E t TAUP 1  E t – E i
--------------------------------------- p + DEGEN · FAC n exp  ------------------- + --------------------------------------- n + ------------------------------------------ n exp  ------------------ 3-99
DIRAC ie  kT L  DIRAC DEGEN · FAC ie  kT L 
1+ 1+
n p

For acceptor like traps, the recombination term becomes:

2
pn – n
ie
R
A
= ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
- 3-100
E –E  i E –E
TAUN 1 t TAUP  t i
--------------------------------------- p + ------------------------------------------ n exp  ------------------- + --------------------------------------- n + DEGEN · FAC n exp  ------------------
DIRAC DEGEN · FAC ie  kT L  DIRAC ie  kT 
1+ 1+ L
n p

The field-effect enhancement term for electrons is given by:

1
En En 32
= ----------- exp  ----------- u – K n u  du

DIRAC
n 3-101
kT L  kT L 
0

while the field-effect enhancement term for hole is:

1
Ep En 32
= ----------- exp  ----------- u – K p u  du

DIRAC
p 3-102
kT L  kT L 
0

where u is the integration variable, En is the energy range where tunneling can occur for
electrons, Ep is the tunneling energy range for holes, and Kn and Kp are defined as:

3
2m 0 MASS · TUNNELEn
4
K n = --- --------------------------------------------------------------------- 3-103
3 qh E

3
2m 0 MASS · TUNNELEp
4
K p = --- ---------------------------------------------------------------------- 3-104
3 qh E

 is the reduced Planck's constant, m0 is the rest mass of an electron and MASS.TUNNEL is the
effective mass. You can specify MASS.TUNNEL by setting the MASS.TUNNEL parameter in the
MODELS statement.

121 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

Table 3-13 shows the user-specifiable parameter for Equations 3-103 and 3-104.

Table 3-13 User-Specifiable Parameters for Equations 3-103 and 3-104

Statement Parameter Default Units

MODELS MASS.TUNNEL 0.25

Non-local Trap Assisted Tunneling

The Trap-assisted-tunneling model in the previous section uses values of the field effect
DIRAC DIRAC
enhancement factors  n and  p which are approximate. They are evaluated using
an approximation for the tunneling probability. It is assumed that the value at each node point
depends only on the local field strength there, and that the field is assumed to be constant over
the range of the tunneling. This gives Airy function solutions for the tunneling probabilities.
A further approximation is also made by replacing the Airy function with the leading term of
its high field asymptotic behavior. The integration over allowed tunneling energies is also
DIRAC
evaluated approximately. A more accurate expression for  n is

E n
E  kT
e
DIRAC 1
n = ------- T  E  dE 3-105
KT
0

where T(E) is the probability that an electron with energy E tunnels elastically to the
conduction band from the trap position.
The exponential term gives the probability that the electron will gain energy (  E n – E ) from
phonon interactions before tunneling. In Equation 3-105, the reference of energy is the
conduction band edge and  E n is the maximum energy below this for which the tunneling is
possible, or the trap depth if less than this maximum (see Figure 3-2). There is a similar
expression for holes. The tunneling probability is evaluated using a Wentzel-Kramers-
Brillouin (WKB) method. The integral in Equation 3-105 is evaluated numerically in order to
DIRAC DIRAC
give the  n and  p terms.

The trap energy is at the Intrinsic Fermi energy in each material, unless you specify the
parameter TAT.NLDEPTH on the MODELS statement. TAT.NLDEPTH has units of energy in
Electron Volts. If you specify TAT.NLDEPTH and TAT.RELEI on the MODELS statement, then
the value of TAT.NLDEPTH will be added to the Intrinsic Fermi energy to give the trap energy
level. If TAT.NLDEPTH is specified without TAT.RELEI, then the trap energy is calculated by
subtracting TAT.NLDEPTH from the conduction band energy as follows:

E T = E c – TAT · NLDEPTH 3-106

To model Trap Assisted Tunneling through heterojunctions, you must use the TAT.RELEI
parameter if you want to use the TAT.NLDEPTH parameter.

122 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

Figure 3-2: Schematic of Phonon assisted electron tunneling from a trap to the conduction band
DIRAC DIRAC
The  n and  p terms are then used in Equations 3-99 and 3-100. The values of
wavevector used in the tunneling probability calculation are evaluated using the density of
states effective mass. To use different masses, then use the ME.TUNNEL and MH.TUNNEL
parameters on the MATERIAL statement. Using ME and MV on the MATERIAL statement will also
change the masses used but may affect the behavior of other models. For example, if you
specify both ME and ME.TUNNEL, the evaluation of the  factor will use the value of
ME.TUNNEL.
To use this model, specify an extra rectangular mesh encompassing the region where the 
terms are to be evaluated. To position the special mesh, use the QTX.MESH and QTY.MESH
statement. You must then set the required direction of the quantum tunneling using the
QTUNN.DIR parameter on the MODELS statement. Outside this mesh, the local model of the
previous section will be used to evaluate the  terms.
For example
qtx.mesh loc=0.0 spac=0.01
qtx.mesh loc=1.0 spac=0.01

qty.mesh loc=1.0 spac=1.0

qty.mesh loc=9.0 spac=1.0
will set up a mesh in the rectangle bounded by x=0.0, x=1.0, y=1.0 and y=9.0. The
tunneling will be assigned to be in the X direction using the QTUNN.DIR (=1) parameter on the

123 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

MODELS statement. All the required values are interpolated from the values on the ATLAS
mesh. In this case, they are interpolated onto each slice in the X direction.
You can also use one or more QTREGION statements to set up the special meshes required for
this model. The syntax is exactly the same as described in the “Non-local Band-to-Band
Tunneling” on page 246. This enables you to model non-local Trap-Assisted tunneling for
more than one p-n junction and for p-n junctions with complicated shapes.
To enable the model, specify TAT.NONLOCAL on the MODELS statement. You also need to
specify QTUNN.DIR on the MODELS statement to determine the direction of tunneling if you
have used QTX.MESH and QTY.MESH to set up the special mesh.
DIRAC DIRAC
To output the calculated values of n and  p to a structure file, specify
NLTAT.GAMMA on the OUTPUT statement. This allows you to visualize their values when
interpolated back onto the device mesh.

Table 3-14 Non-local Trap Assisted Tunneling Parameters for the MODELS Statements

Parameter Type Default

TAT.RELEI Logical False

TAT.NONLOCAL Logical False
TAT.NLDEPTH Real 0
QTUNN.DIR Real 0

Table 3-15 Non-local Trap Assisted Tunneling Parameters for the MATERIAL Statement

Parameter Type Default

ME.TUNNEL Real
MH.TUNNEL Real

Poole-Frenkel Barrier Lowering for Coulombic Wells

The Poole-Frenkel barrier lowering effect enhances the emission rate for trap-to-band
phonon-assisted tunneling as well as pure thermal emissions at low electric fields. The Poole-
Frenkel effect occurs when the Coulombic potential barrier is lowered due to the electric
field, and only occurs when there is a Coulomb interaction between the trap and the carrier.
The following interactions are Coulombic.
• Between an empty (positive) donor-type trap and an electron.
• Between a filled (negative) acceptor-type trap and a hole.
The following interactions are short-range (Dirac).
Between a filled (neutral) donor-type trap and a hole.
Between an empty (neutral) acceptor-type trap and an electron.

124 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

The Poole-Frenkel effect is modeled by including field-effect enhancement for Coulombic

wells (nCOUL and pCOUL) and thermal emission (F) [167] in the trap lifetimes in Equation
3-91. The trap-assisted tunneling effects for Dirac wells remains unmodified. To enable this
model, specify TRAP.COULOMBIC in the MODELS statement.
The recombination term for traps now becomes:
2
pn – n
ie
R = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
A
3-107
TAUN 1  E i – E t TAUP  E t – E i
------------------------------------------ p + ------------------------------------------n exp  ------------------- + --------------------------------------- n + DEGEN · FAC n exp  ------------------
ie  kT ie  kT L 
F +  nCOUL DEGEN · FAC L  1+
DIRAC
p

2
pn – n
ie
R = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
- 3-108
D
TAUN  E i – E t TAUP 1  E t – E i
--------------------------------------- p + DEGEN · FAC n exp  ------------------- + ------------------------------------------ n + ------------------------------------------ n exp  ------------------
DIRAC ie  kT L  DEGEN · FAC ie  kT L 
1+
n F +  pCOUL

where the Poole-Frenkel thermal emission enhancement factor, F, is given by:

  E fp
 F = A · TRAPCOULOMBIC exp  -------------- 3-109
kT  L

Efp is the barrier lowering term for a Coulombic well (see Equation 3-110).

3
B · TRAPCOULOMBICq E
 E fp = ---------------------------------------------------------------------- 3-110


The Coulombic field-enhancement terms, nCOUL and pCOUL, are defined as:
1
En En 32 E fp 5  3
 exp  ----------- u – k p u 1 –  -------------  du
COUL
n = ----------- 3-111
kT L  KT L  uE n 
E fp
-----------
E n

1
E p E p 32 E fp 5  3
 exp  ---------- u – k p u 1 –  -------------  du
COUL
p = ---------- 3-112
kT L  KT L  uE p 
E fp
-----------
E n

Table 3-16 User-Specifiable Parameters for Equations 3-109 and 3-110

Statement Parameter Default Units

MATERIAL A.TRAPCOULOMBIC 1.0

MATERIAL B.TRAPCOULOMBIC 1.0

125 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

Transient Traps
In the time domain the acceptor and donor traps do not reach equilibrium instantaneously but
require time for electrons to be emitted or captured. This is taken account of inside ATLAS by
solving an additional differential rate equation whenever a transient simulation is performed.
These rate equations for donor and acceptor traps are given in Equations 3-113 and 3-114
[307]:

+ Ei – Et
dN tD 
----------------- = DENSITY   p SIGP p  1 – F tD  – F tD n DEGEN.FAC exp ---------------- 3-113
dt  i kT
 1 – F tD n exp E – E 
i t i
–  n SIGN nF tD – ------------------------------------------ ---------------- 
DEGEN.FAC kT 

–
dN tA  Et – Ei
------------ = DENSITY   n SIGN n  1 – F tA  – DEGEN.FAC F tA n exp ---------------- 3-114
dt  i kT
 1 – F tA n i Ei – Et 
–  p SIGP pF tA – --------------------------------- exp ---------------- 
DEGEN.FAC kT 

A transient trap simulation using this model is more time consuming than using the static
model but gives a much more accurate description of the device physics. It may sometimes be
acceptable to perform transient calculations using the static trap distribution and assume that
traps reach equilibrium instantaneously. If this is the case, a flag (FAST) on the TRAP
statement will neglect the trap rate equation from the simulation.
As may be seen from Equations 3-113 and 3-114, each trap species has a net electron capture
rate and a net hole capture rate. The former arise from the terms proportional to nSIGN and
the latter from terms proportional to pSIGP. These capture rates enter the transient current
continuity Equations 3-3 and 3-4 as components of the Rn and Rp terms respectively. In some
circumstances, they will be generation rates rather than recombination rates, but by
convention they are still referred to as Rn and Rp with negative values. In steady state, the trap
occupation is constant and therefore the electron and hole capture rates are equal for each trap
species. By substituting Equations 3-84 to 3-89 into the right hand sides of Equations 3-113
and 3-114, you obtain the expressions for steady state recombination, namely Equations 3-91
and 3-92. Under transient simulation conditions Rn and Rp have different values. This
difference leads to the change over time of the trap occupation densities.
To view the values of Rn and Rp summed over the trap species present in a transient
simulation, ATLAS provides two options. To output the rates to a standard structure file for
SAVE statements, you specify U.TRANTRAP on the OUTPUT statement. You specify
E.TRANTRAP on the PROBE statement to obtain a probe of Rn and H.TRANTRAP on the PROBE
statement to obtain a probe of Rp.

126 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

Table 3-17 Output of Transient Recombination Rates

Statement Parameter Type Default

OUTPUT U.TRANTRAP Logical False

PROBE E.TRANTRAP Logical False
PROBE H.TRANTRAP Logical False

Hysteresis of Interface Traps

The default implementation of traps at a semiconductor-insulator interface assumes that the
traps are located exactly at the interface. Their rate of response to changes in the energy of the
Quasi-Fermi level in the semiconductor depends linearly on the values of the carrier capture
cross-sections, SIGN and SIGP, as Equations 3-113 and 3-114 show.
In the HEIMAN model, the traps are assumed to have a uniform distribution in depth in the
insulator [108]. They are further assumed to have values of SIGN and SIGP that depend on
their distance into the insulator, d, as
–e d
SIGN  d  = SIGN e 3-115

–h d
SIGP  d  = SIGP e 3-116

where

2 2m e  E c – E tA 
 e = --------------------------------------
- 3-117
2


and

2 2m h  E tD – E v 
 h = --------------------------------------- 3-118
2


are the evanescent wavevectors of the semiconductor electron and hole states as they tunnel
into the insulator. The capture cross-sections decrease with depth, d, in the insulator.
Consequently the spatially deeper the trap, the more slowly it will respond to changes of
carrier concentration at the semiconductor-insulator interface. For example, if the traps are
initially full and the device is biased to empty them, then after a certain time the traps up to a
certain depth will be empty and deeper traps will still be full. Heiman demonstrates that the
transition distance between the two cases is very narrow with a length scale of approximately
1/(2e) for electrons and 1/(2h) for holes [108].

127 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

To enable this model you specify HEIMAN on the INTTRAP statement. The trap parameters are
set up as usual for interface traps with the additional specification of the DEPTH parameter.
This determines the extent of the uniform trap density distribution into the insulator and is in
units of microns with a default value of 0.005 microns. The number of equally spaced
internal points describing the internal trap distribution for each interface node is the HPOINTS
parameter. This is optionally specified on the INTTRAPS statement and has a default value of
10. For each interface node, the trap distribution is calculated at each of the HPOINTS internal
points. The average occupancy of the interface charge at each interface point is obtained by
taking the arithmetic mean over these internal points.
These averages can be output by using the parameters FTACC.TRAP and FTDON.TRAP on the
PROBE statement. The PROBE must be positioned on the semiconductor-insulator interface
where hysteresis traps are present.
The HEIMAN model can be used with either steady-state bias ramps or as a fully transient
simulation. To calculate the effect of hysteresis on small-signal admittance values, you must
perform a steady-state simulation. To do this, you specify the total simulation timespan for
the bias ramp using the TIMESPAN parameter on the SOLVE statement. The assumed time to
sweep the bias through VSTEP volts is thus VSTEP*TIMESPAN/(VFINAL-V(initial))
seconds. To simulate a hysteresis cycle, you need two consecutive SOLVE statements–one to
ramp the voltage to the desired final value and the other to return it to the initial value. The
value of TIMESPAN can be different for the two SOLVE statements. Any steady-state SOLVE
statement executed without a TIMESPAN parameter will cause the trap occupation
probabilities to be set to their steady-state values consistent with the device bias state.

Note: Changes in the quasi-Fermi levels at the semiconductor interface cause trap occupation changes up to a
distance d into the insulator. Distance d depends on the natural logarithm of the measurement time
TIMESPAN divided by the trap capture lifetime. Traps that are situated deep in the insulator therefore
require a large value of TIMESPAN. The DEPTH parameter, therefore, should not be set to too high a
value.

For a transient simulation of trap hysteresis, you specify HEIMAN on the INTTRAPS statement
and use the transient SOLVE statement. You do not specify the TIMESPAN parameter in this
case because all necessary information is specified on the transient SOLVE statement. You can
intermix transient and steady-state hysteresis simulations.
The HEIMAN model can also be used to study Negative Bias Temperature Instability in p-
MOSFETS if the SR.HEIMAN option is used. This enhancement is based on the modification
of hole trapping and emission rates by Structural Relaxation (SR) of hole traps in SiO2 [118].
The carrier capture terms in Equations 3-113 and 3-114 (i.e., those that depend on free carrier

density) are multiplied by a factor P SR = exp –----------------------------

c SR · TRAP- , and the emission terms are
KT

multiplied by P SR = exp –----------------------------

e SR · EMIT . The effect of this is to give NBTI degradation a much
-
KT
stronger temperature dependence than is obtained purely from the tunneling factors and
introduce a tunable asymmetry between emission and capture processes. You set the values
of SR.TRAP and SR.EMIT on the INTTRAP statement, and the default values are 0.5 eV [118].

128 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

Table 3-18 Parameters for the HEIMAN Model

Statement Parameter Type Default Units

INTTRAP HEIMAN Logical False

INTTRAP DEPTH Real 5.0e-3 m

INTTRAP SR.HEIMAN Logical False

INTTRAP SR.TRAP Real 0.5 eV
INTTRAP SR.EMIT Real 0.5 eV
INTTRAP HPOINTS Real 10
SOLVE TIMESPAN Real seconds

3.3.4 Multistate Traps

The individual traps considered above can be in one of two internal states, that is they are
either occupied by a charge carrier or empty. The occupation probability gives the fraction of
traps which are in a given state, and the remaining fraction are in the other state.
It is possible to model traps that have two or more internal states by using the Multistate trap
model. In this model, which is currently restricted to interface traps, a trap can have up to 5
internal states.
This allows you to model structural changes of a trap as well as trapping of multiple carrier
types and ionic or atomic species. In this model, each state has its own occupation probability
with the sum over states of occupation probability being one. The time evolution of each
occupation probability depends on the transition rates between the different states. In the
multistate trap model, the steady state occupation probability is fixed, and the occupation
probabilities are only changed during a transient simulation. The model can be used with
electron and hole traps and it can also be coupled with the generic ion transport model (see
Section 3.16 “Generic Ion Transport Model”) and the hydrogen diffusion model, which forms
part of the Reaction-Diffusion degradation model (see Section 3.6.8 “Device Level
Reliability Modeling”).
For a trap with N-possible internal states, the equation for the time dependence of the
occupation probability of state i, fi, is given by
f
------i =
t   fj aji – fi aij  3-119
ji

where the term f j a ji represents transfer into state i from state j, and the term f i a ij represents
transfer from state i to state j. If we add together all N of these equations, then the right hand
sides cancel so we get
f i
 -----t- = 0 3-120
i=1

129 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

so that the sum of occupation probabilities of the state is constrained to be a constant. By

convention this constant is 1.0. In general a ij  a ji . In other words, the forward and reverse
transfer rates between two internal states are usually different. The transfer rates typically
depend on several simulation variables as well as various external parameters, and are usually
time dependent themselves. ATLAS solves the set of Equation 3-119 in the TR-BDF2 or
BDF1 transient schemes, taking into account the evolution of the interface charge and the
recombination/generation processes associated with the change of internal trap state
occupations. Any numerical truncation errors causing the Ni = 1 fi to be different than
exactly 1.0 are distributed among the individual probabilities in such a way as to maintain the
constraint.
Almost all parameters and flags for this model belong to the INTTRAP statement (see Table 3-
19). You specify the MSCTRAP flag to enable this model. You must also specify MSC.NSTATES
to give the number of internal states that the model has. This can be anywhere between 2 and
5. Details of each internal state can be specified using the MSCx.ELEC, MSCx.HOLE,
MSCx.SP1, MSCx.SP2, MSCx.SP3 parameters, where x can be any integer between 1 and 5.
The latter three parameters refer to generic ions. To use this functionality, you need to either
use the generic ion transport model or specify H.ATOM or H.MOLE or the MODELS statement to
include atomic and molecular hydrogen respectively. In this case, the parameters MSCx.H
(aliased to MSCx.SP1) refer to atomic hydrogen and MSCx.H2 (aliased to MSCx.SP2) to
molecular hydrogen. If you use hydrogen transport, the transport parameters and
dimerisation rates are specified on the MATERIAL statement and are described more fully in
“Reaction-Diffusion Degradation Model” on page 294. If instead you use the generic ion
model, the transport parameters are described in Section 3.16 “Generic Ion Transport Model”
and reaction parameters are described in Section 3.17 “Generic Ion Reaction Model”.
For example, you would indicate that internal state 3 contained a trapped electron by
specifying MSC3.ELEC=1 on the INTTRAP statement. If internal state 1 contained 2 ions of
ionic species 3, you would specify MSC1.SP3 = 2 on the INTTRAP statement. The charge of
each internal state must also be specified using the MSCx.CHARGE parameter, and changes in
the CHARGE parameter between internal states must be consistent with the changes in
individual components.
The CHARGE parameter is used in the contribution of the multistate traps to the interface
charge. ATLAS will automatically check the CHARGE parameters for consistency.
The initial occupation probabilities of the internal states can be initialized using the
MSCx.FT0 parameter. The sum of the MSCx.FT0 parameters must add to 1.0. ATLAS will also
check this automatically. Each internal state can also be assigned an energy level using the
MSCx.ELEVEL parameter. You specify that this energy is relative to the conduction band edge
by using the MSCx.ET.CB flag, and relative to the valence band edge using the MSCx.ET.VB
flag.
You specify the trap cross-sections, SIGN and SIGP, and the trap density, DENSITY in the usual
way. The parameters DONOR, ACCEPTOR, FAST, HEIMAN, E.LEVEL, MIDGAP, and
E.ACTIVATION are irrelevant to multistate traps and are ignored. The DEGEN parameter is
automatically set to 1 for multistate traps.

130 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

You must also specify the transition rates between internal states. They are usually
complicated functions of the solution variables and it is impossible to stipulate them directly
using parser parameters. Instead of this, you specify a filename for the F.MSCRATES
parameter, the file containing a C-Interpreter function. The C-Interpreter function is called
msctransitionrates, and the prototype is given in the file template.lib. Among other
parameters, it is supplied with two integers, i and j, which are never simultaneously the same
and which both run from 1 up to the number of internal states. This allows a transition rate to
be programmed for each combination of the start and ending states. The convention is that i
is the starting state and j the final state. Figure 3-3 shows the direction of the transitions for
rates aij in the case of a 3-state model. The function will give optimum performance if the first
derivatives of the rate with respect to the dependent variables are given. The file
template.lib gives an example of this for multistate traps with 3 internal states.

Figure 3-3: Multistate Traps with 3 Internal States

As the trap occupation probabilities evolve over time, they give rise to a recombination rate
for the type of carrier they contain. The electron recombination rate, for example, is obtained
from the following sum

N
f i
Rn =
 MSCi.ELEC-----t- 3-121
i=1

so that any states containing an electron will contribute. This is then optionally coupled to the
electron continuity equation, with analogous results for the other trapped species. In
simulations where the coupling to the transport equations can be ignored, you specify
^RG.MSCTRAP on the METHOD statement to remove the coupling and thereby improve
simulation speed.
Visualization of the trap internal states in a standard log file is obtained using the FT.MSC flag
on the PROBE statement. This also requires the particular internal state to be specified using
the MSC.STATE parameter of the PROBE statement. The position parameters of the statement
will localize it to the interface with multistate traps. The recombination rates of electrons and
holes can be visualized by using the E.TRANTRAP and H.TRANTRAP flags on the PROBE

131 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

statement respectively. Because the traps are transient, the electron and hole recombination
rates will in general differ from each other. For simplicity, any change in a state having a
trapped hole is modeled as a hole recombination rate (H.TRANTRAP), and for one containing
an electron as an electron recombination rate (E.TRANTAP). If the generic ion transport
model is also active, any recombination of ionic species into the multistate traps can also be
visualized by using the U.SP1TRAP, U.SP2TRAP, and U.SP3TRAP flags on the PROBE
statement. If the H.ATOM or H.MOLE flags are enabled, then U.SP1TRAP will output the atomic
hydrogen recombination rate, and U.SP2TRAP will output the molecular hydrogen
recombination rate. By convention, a recombination rate is given as a positive value and a
generation rate is given as a negative value. To visualize the state occupation probabilities in
a standard structure file, you specify TRAPS on the OUTPUT statement. Five occupation
probabilities per multistate configuration trap will then be output. The first MSC.NSTATES
values will contain valid data, the rest will be zero.
One possible use of the Multistate trap model is to simulate the degradation behavior of
passivated dangling bonds, as an alternative to the built-in DEVDEG.RD, DEVDEG.KN, and
DEVDEG.PL models (described in Section 3.6.8 “Device Level Reliability Modeling”). This is
best achieved using the hydrogen diffusion and reaction models enabled with the H.ATOM and
H.MOLE models. The multistate trap model has also been used as the basis for the Two-Stage
Negative Bias Temperature Instability models described in “Two Stage Negative Bias
Temperature Instability Models” on page 304. The most complex of these uses the maximum
of 5 internal states.

Table 3-19 Parameters for the Multistate Configuration Trap Model

Statement Parameter Type Default Units

INTTRAP F.MSCRATES Character

INTTRAP MSC.NSTATES Integer 0
INTTRAP MSC1.CHARGE Integer 0
INTTRAP MSC1.ELEC Integer 0
INTTRAP MSC1.ELEVEL Real 0 eV
INTTRAP MSC1.ET.CB Logical False
INTTRAP MSC1.ET.VB Logical False
INTTRAP MSC1.FT0 Real 0
INTTRAP MSC1.H1 Integer 0
INTTRAP MSC1.H2 Integer 0
INTTRAP MSC1.HOLE Integer 0
INTTRAP MSC1.SP1 Integer 0
INTTRAP MSC1.SP2 Integer 0
INTTRAP MSC1.SP3 Integer 0
INTTRAP MSC2.CHARGE Integer 0

132 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

Table 3-19 Parameters for the Multistate Configuration Trap Model

INTTRAP MSC2.ELEC Integer 0
INTTRAP MSC2.ELEVEL Real 0 eV
INTTRAP MSC2.ET.CB Logical False
INTTRAP MSC2.ET.VB Logical False
INTTRAP MSC2.FT0 Real 0
INTTRAP MSC2.HOLE Integer 0
INTTRAP MSC2.H1 Integer 0
INTTRAP MSC2.H2 Integer 0
INTTRAP MSC2.SP1 Integer 0
INTTRAP MSC2.SP2 Integer 0
INTTRAP MSC2.SP3 Integer 0
INTTRAP MSC3.CHARGE Integer 0
INTTRAP MSC3.ELEC Integer 0
INTTRAP MSC3.ELEVEL Real 0 eV
INTTRAP MSC3.ET.CB Logical False
INTTRAP MSC3.ET.VB Logical False
INTTRAP MSC3.FT0 Real 0
INTTRAP MSC3.H1 Integer 0
INTTRAP MSC3.H2 Integer 0
INTTRAP MSC3.HOLE Integer 0
INTTRAP MSC3.SP1 Integer 0
INTTRAP MSC3.SP2 Integer 0
INTTRAP MSC3.SP3 Integer 0
INTTRAP MSC4.CHARGE Integer 0
INTTRAP MSC4.ELEC Integer 0
INTTRAP MSC4.FT0 Real 0
INTTRAP MSC4.H1 Integer 0
INTTRAP MSC4.H2 Integer 0
INTTRAP MSC4.HOLE Integer 0

133 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

Table 3-19 Parameters for the Multistate Configuration Trap Model

INTTRAP MSC4.SP1 Integer 0
INTTRAP MSC4.SP2 Integer 0
INTTRAP MSC4.SP3 Integer 0
INTTRAP MSC5.CHARGE Integer 0
INTTRAP MSC5.ELEC Integer 0
INTTRAP MSC4.ELEVEL Real 0 eV
INTTRAP MSC4.ET.CB Logical False
INTTRAP MSC4.ET.VB Logical False
INTTRAP MSC5.ELEVEL Real 0 eV
INTTRAP MSC5.ET.CB Logical False
INTTRAP MSC5.ET.VB Logical False
INTTRAP MSC5.FT0 Real 0
INTTRAP MSC5.H1 Integer 0
INTTRAP MSC5.H2 Integer 0
INTTRAP MSC5.HOLE Integer 0
INTTRAP MSC5.SP1 Integer 0
INTTRAP MSC5.SP2 Integer 0
INTTRAP MSC5.SP3 Integer 0
INTTRAP MSCTRAP Logical False
METHOD RG.MSCTRAP Logical True
MODELS H.ATOM Logical False
MODELS H.MOLE Logical False
PROBE E.TRANTRAP Logical False
PROBE FT.MSC Logical False
PROBE H.TRANTRAP Logical False
PROBE MSC.STATE Integer 1
PROBE U.SP1TRAP Logical False
PROBE U.SP2TRAP Logical False
PROBE U.SP3TRAP Logical False

134 ATLAS User’s Manual

Space Charge from Incomplete Ionization, Traps, and Defects Physics

Examples
INTTRAP S.I [Localisation parameters] DENSITY=3.0E12 DEGEN=1 \
SIGN=1.0E-13 SIGP=1.0E-14 MSCTRAP MSC.NSTATES=3 MSC1.FT0=0.5 \
MSC2.FT0=0.3 MSC3.FT0=0.2 MSC1.CHARGE=1 MSC1.HOLE=1 \
MSC2.CHARGE=-1 MSC2.ELEC=1 MSC3.CHARGE=0 F.MSCRATES="example.lib"
This is a 3 state trap State 1 has a trapped hole and a charge of +1 [Q]. State 2 has a trapped
electron and a charge of -1 [Q]. State 3 has no trapped carriers and a charge of 0 [Q]. It is
initialized with 50% of the traps in state 1, 30% of the traps in state 2 and 20% of the traps in
state 3. The transition rate C-Interpreter function is in the file 'example.lib'.
To obtain the time dependence of the trap state occupancies in a log file, specify
PROBE FT.MSC MSC.STATE=1 X=1.5 Y=0
PROBE FT.MSC MSC.STATE=2 X=1.5 Y=0
PROBE FT.MSC MSC.STATE=3 X=1.5 Y=0
INTTRAP [Localisation parameters] DENSITY=1.0E11 SIGN=1.0E-13
SIGP=1.0E-14 MSCTRAP MSC.NSTATES=2 MSC1.FT0=1.0 MSC1.SP1=1
MSC1.CHARGE=1 MSC2.FT0=0.0 MSC2.ELEC=1 MSC2.CHARGE=-1
F.MSCRATES="example2.lib"
MODELS NSPECIES=1 SPECIES1.Z=1
The generic ion transport model is enabled on the MODELS statement, and there is one ionic
species with charge of +1 [Q] Coulombs.
This is a 2-state trap State 1 has a trapped ion of species1, which has a charge of +1 [Q]. State
2 has a trapped electron, which has a charge of -1 [Q]. The initial occupancy condition is that
all traps are in state 1. To look at the recombination rates of ionic species 1 and electrons into
the trap, specify
PROBE E.TRANTRAP [Localisation parameters]
PROBE U.SP1TRAP [Localisation parameters]
In this case, the electron and SP1 recombination rates at any point will be equal and opposite
because the rate of change of internal state 1 is equal and opposite to that of internal state 2
(because their sum must be constant).
INTTRAP [Localisation parameters] DENSITY=1.0e12 MSCTRAP
MSC.NSTATES=2 MSC1.CHARGE=0 MSC1.H1=1 MSC1.CHARGE=0 MSC2.HOLE=1
MSC2.CHARGE=1 F.MSCRATES="dbond.lib"
MODELS MOS H.ATOM H.MOLE
PROBE U.SP1TRAP [Localisation parameters]
In this case, the atomic hydrogen and molecular hydrogen transport and diffusion models are
enabled on the MODELS statement. State 1 is set to have one hydrogen atom per trap using the
MSC1.H1 parameter (aliased to MSC1.SP1). The charge of state 1 is set to 0 [Q], and the
charge of state 2 is set to 1 [Q] to be consistent with the fact that state 2 has a trapped hole.
The PROBE entry will probe the atomic hydrogen concentration in this case. The bulk
dimerisation rates can be set on the MATERIAL statement to control the creation of molecular
hydrogen.

135 ATLAS User’s Manual

The Energy Balance Transport Model Physics

3.4 The Energy Balance Transport Model

The conventional drift-diffusion model of charge transport neglects non-local transport
effects such as velocity overshoot, diffusion associated with the carrier temperature and the
dependence of impact ionization rates on carrier energy distributions. These phenomena can
have a significant effect on the terminal properties of submicron devices. As a result ATLAS
offers two non-local models of charge transport, the energy balance, and hydrodynamic
models.
The Energy Balance Transport Model follows the derivation by Stratton [264,265] which is
derived starting from the Boltzmann Transport Equation. By applying certain assumptions,
this model decomposes into the hydrodynamic model [184,12,13].
The Energy Balance Transport Model adds continuity equations for the carrier temperatures,
and treats mobilities and impact ionization coefficients as functions of the carrier
temperatures rather than functions of the local electric field.

3.4.1 The Energy Balance Equations

The Energy Balance Transport Model introduces two new independent variables Tn and Tp,
the carrier temperature for electrons and holes. The energy balance equations consist of an
energy balance equation with the associated equations for current density and energy flux
S n ,p .

For electrons the Energy Balance Transport Model consists of:

3k  
divS n = --- J  E – W n – ------   n nT n
1
3-122
q n 2 t 

T
J n = qD n n – q n n + qnD n T n 3-123

k n
S n = – K n T n –  --------- J n T n 3-124
 q 

and for holes:

3k  
divS p = --- J p  E – W p – ------   p pT p
1
3-125
q 2 t 

T
J p = – q D p  p – q p p  – q pD p T p 3-126

k p
S p = – K p T p –  --------- J p T p 3-127
q

where S n and S p are the energy flux densities associated with electrons and holes, and n
and p are the electron and hole mobilities.

136 ATLAS User’s Manual

The Energy Balance Transport Model Physics

The remaining terms, Dn and Dp, are the thermal diffusivities for electrons and holes as
defined in Equations 3-128 and 3-135 respectively. Wn and Wp are the energy density loss
rates for electrons and holes as defined in Equations 3-151 and 3-152 respectively. Kn and Kp
are the thermal conductivities of electrons and holes as defined in Equations 3-132 and 3-139
respectively.
 n kT n
D n = ---------------  n 3-128
q

F  1  2   n   F – c
- = F 1  2  ------
–1 n
 n = -------------------------------  n = ------------------
n
3-129
F –1  2  n  kT n  N c

D n =   2n – ---  n  n ---
T 3 k
3-130
 2 q

F + 3  2  n 
 2n =  n  5
--- +  n -----------------------------------
n
- 3-131
2  F + 1  2  n 
n

k 2
K n = q n  n  ---  n T n 3-132
q

F + 5  2  n  F + 3  2  n 
7 n
 n =  n   n + --- -----------------------------------
- –  n + 5
--- -----------------------------------
n
- 3-133
 2 F  + 3  2   n   2 F  + 1  2   n 
n n

 2n
 n = ---------- 3-134
n

Similar expressions for holes are as follows:

 p kT p
D p = ---------------  p 3-135
q

F  1  2   p  v – f
 p = -----------------p = F 1  2  -------
–1 p
 p = ------------------------------- 3-136
F –1  2  p  kT p  N V

D p =   2p – ---  p  p ---
T 3 k
3-137
 2 q

F + 3  2  p 
 2p =  p  5
--- +  p -----------------------------------
p
- 3-138
2  F + 1  2  p 
p

k 2
K p = q p  p  ---  p T p 3-139
 q

137 ATLAS User’s Manual

The Energy Balance Transport Model Physics

F + 5  2  p  F + 3  2  p 
7 p
 p =  p   p + --- -----------------------------------
- –  p + 5
--- -----------------------------------
p
- 3-140
2 F + 3  2  p   2 F  + 1  2   p 
p p

 2p
 p = ---------- 3-141
p

If Boltzmann statistics are used in preference to Fermi statistics, the above equations simplify
to:
 n =  p = 1 3-142

n = n =  5
--- +  n 3-143
2 

 p =  p =  --- +  p
5
3-144
2 

d  ln  n  T n  n
 n = ------------------- = ------ --------- 3-145
d  ln T n   n T n

d  ln  p  T p  p
 p = ------------------- = ------ --------- 3-146
d  ln T p   p T p

The parameters n and p are dependent on the carrier temperatures. Different assumptions
concerning n and p correspond to different non-local models. In the high-field saturated-
velocity limit, that corresponds to velocity saturation, the carrier mobilities are inversely
proportional to the carrier temperatures.
n = p = –1 3-147

and this corresponds to the Energy Balance Transport Model.

If instead the choice n=p=0 was chosen this would correspond to the simplified
Hydrodynamic Model. The parameters, n and p, can be specified using the KSN and KSP
parameters on the MODELS statement.
Boundary conditions for n, p, and  are the same as for the drift diffusion model. Energy
balance equations are solved only in the semiconductor region. Electron and hole
temperatures are set equal to the lattice temperature on the contacts. On the other part of the
boundary, the normal components of the energy fluxes vanish.
Hot carrier transport equations are activated by the MODELS statement parameters: HCTE.EL
(electron temperature), HCTE.HO (hole temperature), and HCTE (both carrier temperatures).

138 ATLAS User’s Manual

The Energy Balance Transport Model Physics

3.4.2 Density of States

The calculation of the effective density of states is modified for the Energy Balance Transport
Model. The electron and hole temperatures replace the lattice temperature in Equations 3-31
and 3-32. For example:
3 3
 --- ---
 2m e kT n 2 Tn 2
 ---------
N c =  -------------------------- = - NC  300  3-148
 2
  300
h

3 3
 --- ---
 2m n kT p 2 Tp 2
 ---------
N v =  -------------------------- = - NV  300  3-149
 2
  300
h

3.4.3 Energy Density Loss Rates

The energy density loss rates define physical mechanisms by which carriers exchange energy
with the surrounding lattice environment. These mechanisms include carrier heating with
increasing lattice temperature as well as energy exchange through recombination processes
(SRH and Auger) and generation processes (impact ionization). If the net generation-
recombination rate is written in the form:
A A
U = R srh + R n + R p – G n – G p 3-150

where Rsrh is the SRH recombination rate, RAn are RAp Auger recombination rates related to
electrons and holes, Gn and Gp are impact ionization rates, then the energy density loss rates
in Equations 3-122 and 3-125 can be written in the following form:

3 k  Tn – TL  3 A
W n = --- n ---------------------------------  n + --- kT n  n R SRH + E g  G n – R n  3-151
2 TAUREL EL 2
·

3 k  Tp – TL  3 A
W p = --- p ---------------------------------  p + --- k T p  p R SRH + E g  G p – R p  3-152
2 TAUREL HO 2
·

where
n = F 3  hn   F 1  hn  3-153
--- ---
2 2

p = F 3  hp   F 1  hp  3-154
--- ---
2 2

(and are equal to 1 for Boltzmann statistics), TAUREL.EL and TAUREL.HO are the electron and
hole energy relaxation times, Eg is the bandgap energy of the semiconductor. The relaxation
parameters are user-definable on the MATERIAL statement, which have their defaults shown in
Table 3-20.
The relaxation times are extremely important as they determine the time constant for the rate
of energy exchange and therefore precise values are required if the model is to be accurate.

139 ATLAS User’s Manual

The Energy Balance Transport Model Physics

But, this is an unmeasurable parameter and Monte Carlo analysis is the favored method
through which values can be extracted for the relaxation time.
It’s also important to take into consideration that different materials will have different values
for the energy relaxation time but within ATLAS, the relaxation time will always default to
the value for silicon.

Table 3-20 User-Specifiable Parameters for Equations 3-151 and 3-152

Statement Parameter Default Units

MATERIAL TAUREL.EL 410-13 s

MATERIAL TAUREL.HO 410-13 s

3.4.4 Temperature Dependence of Relaxation Times

ATLAS doesn’t provide an explicit default model for the temperature dependence of energy
relaxation times. Two methods exist, however, to make the relaxation time a function of
carrier energy.
For the first method, use a built-in model by specifying the TRE.T1, TRE.T2, TRE.T3,
TRE.W1, TRE.W2, and TRE.W3 parameters in the MATERIAL statement. Then, activate the
electron temperature (electron energy) dependent energy relaxation time by using
E.TAUR.VAR in the MODELS statement. Electron energy relaxation time will then be:


 TRE T1, W < TRE W1
 · ·
 e =  TRE T2, W = TRE W2 3-155
 · ·
 TRE · T3, W > TRE · W3


where:
3
W = --- kT n 3-156
2

For TRE.W1<W<TRE.W2 the energy relaxation time varies quadratically between TRE.T1 and
TRE.T2. For TRE.W2<W<TRE.W3 energy relaxation time varies quadratically between
TRE.T2 and TRE.T3. The corresponding parameter for hole energy relaxation time in the
MODELS statement is H.TAUR.VAR. Other parameters are listed in Table 3-21.

140 ATLAS User’s Manual

The Energy Balance Transport Model Physics

Table 3-21 User- Specifiable Parameters for Variable Energy Relaxation Time.

Statement Parameter Default Units

MATERIAL TRE.T1 810-13 s

MATERIAL TRE.T2 1.9210-12 s

MATERIAL TRE.T3 110-12 s

MATERIAL TRE.W1 0.06 eV

MATERIAL TRE.W2 0.3 eV
MATERIAL TRE.W3 0.45 eV
MATERIAL TRH.T1 110-12 s

MATERIAL TRH.T2 110-12 s

MATERIAL TRH.T3 110-12 s

MATERIAL TRH.W1 11010 eV

MATERIAL TRH.W2 11010 eV

MATERIAL TRH.W3 11010 eV

For the second method, use the C-INTERPRETER to apply a user-defined model for energy
relaxation time as a function of carrier energy. In the MODELS statement, assign the F.TAURN
and F.TAURP parameters with the names of external files that contain the user-defined C-
INTERPRETER function for the energy relaxation time. You should also specify E.TAUR.VAR
and H.TAUR.VAR flags in the MODELS statement when using C-INTERPRETER functions for the
energy relaxation times.

141 ATLAS User’s Manual

The Energy Balance Transport Model Physics

3.4.5 Energy Dependent Mobilities

The Energy Balance Transport Model requires the carrier mobility to be related to the carrier
energy. This has been achieved through the homogeneous steady state energy balance
relationship that pertains in the saturated velocity limit. This allows an effective electric field
to be calculated, which causes the carriers in a homogeneous sample to attain the same
temperature as at the node point in the device. The effective electric fields, Eeff,n and Eeff,p,
are calculated by solving the equations:

3 k  Tn – TL 
q n  E eff n E eff
2 --- ---------------------------------
 n = 2 TAUMOB.EL 3-157

3 k  Tp – TL 
q p  E eff p E eff
2 --- ---------------------------------
 p = 2 TAUMOB.HO 3-158

for Eeff,n and Eeff,p. These equations are derived from the energy balance equations by
stripping out all spatially varying terms. The effective electric fields are then introduced into
the relevant field dependent mobility model. A full description of the available models is
given in Section 3.6.1 “Mobility Modeling”.
ATLAS2D provides a general C-interpreter function allowing you to specify carrier mobility as
a function of Perpendicular field, carrier temperature, Lattice temperature, carrier
concentration and donor and acceptor concentrations.
For electron mobility, the parameter is F.ENMUN, which can be set in either the MATERIAL or
MOBILITY statement. The corresponding parameter for hole mobility is F.ENMUP, which can
be set in either the MATERIAL or MOBILITY statement. The respective C-functions are
endepmun() and endepmup(). The examples of these functions are provided in ATLAS.

142 ATLAS User’s Manual

Boundary Physics Physics

3.5 Boundary Physics

ATLAS supports several boundary conditions: Ohmic contacts, Schottky contacts, insulated
contacts, and Neumann (reflective) boundaries. Voltage boundary conditions are normally
specified at contacts. Current boundary conditions can also be specified. Additional boundary
conditions have been implemented to address the needs of specific applications. You can
connect lumped elements between applied biases and semiconductor device contacts. A true
distributed contact resistance is included to account for the finite resistivity of semiconductor
contacts.

3.5.1 Ohmic Contacts

Ohmic contacts are implemented as simple Dirichlet boundary conditions, where surface
potential, electron concentration, and hole concentrations (s, ns, ps) are fixed. Minority and
majority carrier quasi-Fermi potentials are equal to the applied bias of the electrode (i.e.,
n=p=Vapplied). The potential s is fixed at a value that is consistent with space charge
neutrality. For example:
+
ns + N A = ps + ND 3-159

Equation 3-159 can be solved for s, ns, and ps, since n and p are known. If Boltzmann
statistics are used, the substitution of Equations 3-36 and 3-37 into Equation 3-159 will yield:

2
– N A  +  – N A
1
n s = --2-  N D
+ N D+ 2
 + 4n ie 3-160

2
n ie
p s = ------
- 3-161
ns

kT L n kT L p
 s =  n + --------- ln ------s- =  p – --------- ln ------s- 3-162
q n ie
q n ie

Note: If you don’t specify a work function, the contacts will be Ohmic regardless of its material.

143 ATLAS User’s Manual

Boundary Physics Physics

3.5.2 Schottky Contacts

To specify a Schottky contact [232], specify a workfunction using the WORKFUN parameter of
the CONTACT statement. The surface potential of the Schottky contact is given by:
E kT L N
 s = AFFINITY + -----g- + --------- ln ------C- – WORKFUN + V applied 3-163
2q 2q N V

where AFFINITY is the electron affinity of the semiconductor material, Eg is the bandgap, NC
is the conduction band density of states, NV is the valence band density of states, and TL is the
ambient temperature. In practice, the workfunction is defined as:
WORKFUN = AFFINITY +  B 3-164

whereB is the barrier height at the metal-semiconductor interface in eV.

Table 3-22 User-Specifiable Parameters for Equation 3-164

Statement Parameter Units

CONTACT WORKFUN eV
MATERIAL AFFINITY eV

For example, if the Schottky contact were aluminum with a workfunction difference to the
silicon of 4.2 eV and a barrier height of 0.7eV, then you would define the Schottky contact
with the statement:
CONTACT NAME=GATE WORKFUN=4.9

Note: You can also define the workfunction using a C-Interpreter function. This function is referenced by the
F.WORKF parameter of the CONTACT statement. The function defines workfunction as a function of
electric field.

You can enable the thermionic emission model by specifying any of the following parameters
of the CONTACT statement: SURF.REC, E.TUNNEL, VSURFN, VSURFP, or BARRIERL. In this
case, the quasi-Fermi levels, n and p, are no longer equal to Vapplied. Instead, these
parameters are defined by a current boundary conditions at the surface [59]:
 b
J sn = qVSURFN  n s – n eq  exp  ---------  3-165
 kT 

 b
J sp = qVSURFP  p s – p eq  exp  --------- 3-166
 kT 

144 ATLAS User’s Manual

Boundary Physics Physics

Table 3-23 User-Specifiable Parameters for Equations 3-165 to 3-166

Statement Parameter Units

CONTACT VSURFN cm/s

CONTACT VSURFP cm/s

where Jsn and Jsp are the electron and hole currents at the contact, ns, is the surface electron
concentration and ps is the surface hole concentrations. The terms, neq and peq, are the
equilibrium electron and hole concentrations assuming infinite surface recombination
velocity (n = p = Vapplied). If VSURFN and VSURFP are not specified on the CONTACT
statement, their values will be calculated using:
2
ARICHN T L
VSURFN = --------------------------------
- 3-167
q NC

2
ARICHP T L
VSURNP = --------------------------------- 3-168
q NV

Table 3-24 User-Specifiable Parameters for Equations 3-167 to 3-168

Statement Parameter Default Units

MATERIAL ARICHN 110 A/cm2/K2

MATERIAL ARICHP 30 A/cm2/K2

Here, ARICHN and ARICHP are the effective Richardson constants for electrons and holes,
taking account of quantum mechanical reflections and tunneling, NC and NV are the
conduction and valence band density of states. The ARICHN and ARICHP parameters are user-
definable as shown in Table 3-24 and NC and NV are functions of the lattice temperature, TL,
according to Equations 3-31 and 3-32.
The schottky thermionic emission model also accounts for field-dependent barrier-lowering
mechanisms. These mechanisms are caused by image forces and possible static dipole layers
at the metal-semiconductor interface [268]. If the barrier height is defined as:
 bn = WORKFUN – AFFINITY 3-169

Eg
 bp = AFFINITY + ------ – WORKFUN 3-170
q

145 ATLAS User’s Manual

Boundary Physics Physics

With barrier lowering, the amount of energy by which these barrier heights are lowered is
defined by:
1
---
q 2 12 GAMMA
 b = BETA ------------ E + ALPHA  E 3-171
4 s

Table 3-25 User-Specifiable Parameters for Equation 3-171

Statement Parameter Default Units

CONTACT ALPHA 0.0 cm

CONTACT BETA 1.0
CONTACT GAMMA 1.0

Here, E is the magnitude of the electric field at the interface and ALPHA is the linear, dipole
barrier lowering coefficient. The Barrier Lowering Model can be turned on with the BARRIER
parameter in the CONTACT statement. Typical values of ALPHA may be found in [10]. Note that
the term with the square root dependence on electric field corresponds to the image force,
while the linear term corresponds to the Dipole Effect [268].
Thermionic emission is implemented on a triangle-by-triangle basis, which means using the
surface recombination velocity and geometrical data. A thermionic emission component is
then calculated for each triangle so that an element of interest is connected. Using the electric
field for each triangle, an adjusted recombination thermionic emission term can be computed
if barrier lowering is incorporated. This is in contrast to where a single field value for the
electrode node is used to compute total thermionic emission value [240].
You can take electron or hole tunneling through the barrier into account by specifying the
E.TUNNEL or H.TUNNEL parameter in the CONTACT statement. The electron tunneling current
(Jtn) is given by the Tsu-Esaki Model (Equation 3-172) and is applied to the contact as a
current boundary condition.
 bn
4  q ME.TUNNEL m 0 k T L
J tn = – --------------------------------------------------------------------------
h
3  P  E Z  N  E Z dEz 3-172
0
Here, bn is the barrier height, ME.TUNNEL is the relative effective mass for electrons, m0 is
the electron rest mass, h is Planck’s constant, k is Boltzmann’s constant, q is the electron
charge, and N(EZ) is given by:

  E F – E Z 
  -------------------- 
  kT L  
 1+e 
N  E Z  = ln  ---------------------------------------------------- 3-173
  E F – E Z – qV 
  --------------------------------- 
  kT L 
1 + e 

146 ATLAS User’s Manual

Boundary Physics Physics

The transmission probability P(EZ) is given by the WKB [268] approximation for a triangular
barrier and is expressed as:
1 3
 ---
2
---
2
 – 8  2ME.TUNNEL m*    bn – E Z  
P  E Z  = exp  ------------------------------------------------------------------------------------------------ 3-174
 3qhE 
 

The ME.TUNNEL parameter is user-definable in the CONTACT statement.

Similar expressions for Equations 3-172 through 3-174 exist for hole tunneling. You can
specify the MH.TUNNEL parameter on the CONTACT statement for the hole relative mass for
tunneling.
For example, to specify a thermoinic emission schottky contact with tunneling and barrier
lowering, the command would be:
CONTACT NAME=GATE SURF.REC E.TUNNEL BARRIER
Parabolic Field Emission Model
For wide bandgap materials such as SiC the parabolic field emission model is suggested
[58,73]. Equation 3-175 describes the current as a function of applied bias voltage V.
Equation 3-176 describes the tunneling probability T(E).
– E – En
1 + exp  --------------------
A T
  kB T 
kB 0 
J = ---------- T  E  ln ----------------------------------------------------------
– E – qV – E n
3-175
1 + exp  ---------------------------------
 kB T 

 2 m  Eb + Eb – E 
 exp  -----
- - – E
------- E ln  --------------------------------------- Eb – E  
TE =   hq N  E  b
 for  E  Eb  3-176
 
1  for  E  Eb 

Here A* is Richardson's constant, Eb is the barrier height on the metalside, N is the

semiconductor doping concentration, En=Ec-Ef=kT*ln(Nc/N) is the Fermi level energy
Vbi=(Eb-En)/q is the built in voltage, and Eb* = q(Vbi-V) = Eb-En-qV is the barrier height on
the semiconductor side.
2
Richardson's constant A* is given by 4  mqk where m is the effective mass.
----------------------
3
h

The effective mass is generally taken from the density of states. You can, however, specify a
mass for this tunneling calculation only by the ME.TUNNEL and MH.TUNNEL parameters of the
CONTACT statement. To enable the model, use the PARABOLIC parameter of the CONTACT
statement. To turn on the tunneling components independently for electrons and holes,
specify the E.TUNNEL and H.TUNNEL parameters of the CONTACT statement. To enable the
model, specify the NSURF.REC or PSURF.REC parameters of the CONTACT statement.

147 ATLAS User’s Manual

Boundary Physics Physics

The energy step size used in numeric integration of Equation 3-175 is controlled by the
DE.TUNNEL parameter of the CONTACT statement. This parameter secifies proportion of the
barrier height used in each step. In other words, the energy step is obtained by multiplying
DE.TUNNEL by the barrier height. The default value for DE.TUNNEL is 10.
In evaluation of the parabolic model Equation 3-175 is only used to evaluate the tunneling
part of the current. For pure thermionic emission, the expression in Equation 3-177 may be
used:
vs
J = q N c exp  – b  kT   V  KT + 1  3-177

The parameter THERM of the CONTACT statement controls whether the thermionic emission
component of Equation 3-177 is included. By default, the value of the THERM parameter is
TRUE.
You should notice that the parabolic Schottky model is essentially a voltage controlled current
source where the voltage applied to the contact, V, appears in the various expressions without
regard to the potentials applied anywhere else in the device. To promote flexibility, we have
added the ability to specify a reference electrode either by name using the REF.ELEC
parameter or by index using the REF.ENUM parameter of the CONTACT statement. When either
of these parameters are assigned, the parameter, V, in the above expressions can be interpreted
as the voltage of the contact minus the voltage at the reference electrode.
Universal Schottky Tunneling (UST) Model
In the Universal Schottky Tunneling (UST) Model [121, 178], the tunneling current is
represented by localized tunneling rates at grid locations near the Schottky contact. For
electrons, this is illustrated in Figure 3-4. The tunneling component of current can be
described by Equation 3-178.

Figure 3-4: Local tunneling generation rate representation of the universal Schottky tunneling model

148 ATLAS User’s Manual

Boundary Physics Physics


A T 1 + fs  E  
k 
J T = -------------L-   E   ln ----------------------------- dE 
1 + fm  E  
3-178
E
Here, JT is the tunneling current density, A* is the effective Richardson's coefficient, TL is the
lattice temperature, (E) is the tunneling probability, fs(E) and fm(E) are the Maxwell-
Boltzmann distribution functions in the semiconductor and metal and E is the carrier energy.
You can then apply the transformation [121] given in Equation 3-179 to obtain the localized
tunneling rates, GT.

1
G T = ---  J T 3-179
q

Applying the transformation given in Equations 3-178 and 3-179, you can obtain the
expression given in Equation 3-180.

A T L E 1 + n  n Nc
-   x  ln ---------------------------------------------------------------------------
G T = ------------------ 3-180
k 1 + exp  –  E c – E FM   kT L 

Here, E is the local electric field, n is the local electron concentration, Nc is the local
conduction band density of states, n is the local Fermi-Dirac factor, Ec is the local conduction
band edge energy and EFM is the Fermi level in the contact.
The tunneling probability (x) can be described by Equation 3-181.
x

0  Ec  x'  – Ec  x   dx'
– 2 2m
  x  = exp -------------------- 3-181


In Equation 3-181, m is the electron effective mass for tunneling, and Ec(x) is the conduction
band edge energy as a function of position.
Assuming linear variation of Ec around a grid location, Equation 3-181 can be reduced to
Equation 3-182.

4 2mx 12
  x  = exp –----------------------
-  E FM + q  b – E c  x   3-182
3

In Equation 3-182, b is the barrier height. Similar expressions to Equations 3-178 through 3-
182 exist for holes.
To enable the universal Schottky tunneling model you should specify UST on the MODELS
statement. You are also required to specify SURF.REC on the CONTACT statement for each
contact that you wish the model to apply. Once enabled, this model automatically applies to
both electrons and holes. For a given grid point, however, the model will only apply to one
carrier according to the relative direction of the local electric field.
Once you enable the model, for each electrode with thermionic emission, the model will be
applied to all grid points within a specified distance of the electrode, To specify this distance,
use D.TUNNEL parameter of the MATERIAL statement. The default value of D.TUNNEL is 10-6
in units of cm.

149 ATLAS User’s Manual

Boundary Physics Physics

Phonon-assisted Tunneling Model for GaN Schottky Diodes

It has been found that the reverse I-V characteristics of some Gallium Nitride (GaN) diodes
can best be explained by using a phonon-assisted electron tunneling model [215]. The
electrons are assumed to be emitted from local levels in the metal-semiconductor interface
and give a contribution to the current of

J = Q PIP · NT W 3-183

where PIP.NT gives the occupied state density near the interface in units of cm-2. The rate of
phonon-assisted tunneling, W, is given by

qF 2 12 12 2 –1  4
W = -----------------------------------------------------------   1 +   –  1 +  
8ME · TUNNELPIP · ET
12
 – 4  2ME TUNNEL  32 2 12 2 2 12  
 exp  ------ -------------------------------------------------
· - PIP · ET 1 +   –   1 +   + --- 
3 qFh 2 
3-184
where
12 2
 2ME · TUNNEL  
 = --------------------------------------------------------
- 3-185
12
8q  F  PIP ET 
·

F is the field strength at the interface and  is the energy width of the absorption band given
by

2 2
 = 8PIP · ACC  PIP · OMEGA   2n + 1  3-186

and n is the phonon density.

To enable the model, use the keyword PIPINYS on the CONTACT statement. The default value
of PIP.NT is 0 and so you must set a finite value for this quantity in order for the model to
have any effect. The other parameters may be specified on the MATERIAL statement. Although
the model was developed specifically for reverse bias current in GaN, ATLAS does not
restrict use of the model to that material. PIP.OMEGA is the phonon energy, PIP.ACC is the
electron-phonon interaction constant, and PIP.ET is the trap depth.

Table 3-26 User-Specifiable Parameters for the Pipinys model.

Statement Parameter Type Default Units

MATERIAL PIP.OMEGA Real 0.07 eV

MATERIAL PIP.ACC Real 2.0
MATERIAL PIP.ET Real 1.0 eV
MATERIAL PIP.NT Real 0.0 cm-2

150 ATLAS User’s Manual

Boundary Physics Physics

Scott-Malliaras Boundary Conditions

The Scott-Malliaras boundary conditions [247] define a reverse bias contact charge density of
–b 12
n = 4 N C exp  -------- exp f
2
3-187
 kT 

where the normalized electric field is

3
q E
f = ---------------------------- 3-188
2 2
4 0 k T

and the factor Ψ is

1 1 1 12 12
 = --- + ---------- – ---  1 + 2f  3-189
f f1  2 f

In forward bias, the contact charge density remains at the zero bias value, N C exp  –  b  kT  .

You can activate Scott-Malliaras boundary conditions by specifying SCOTT.MALLIARAS, in

addition to the zero bias work-function, on a CONTACT statement. For example
CONTACT NAME=ANODE WORKFUNCTION=4.47 SCOTT.MALLIARAS

3.5.3 Floating Contacts

A contact that isn’t connected to a current or voltage source and is totally insulated by
dielectric is called a floating contact. ATLAS accounts for floating contacts such as floating
gates in EPROM devices by using a distributed charge boundary condition, which is applied
to all nodes of the floating electrode (see Equation 3-190).

 D dS = QFG 3-190
s

where D is the electric displacement vector, s represents the external surface of the floating
gate, and QFG is the injected charge.
ATLAS performs an integration over the entire surface of the electrode and forces the
potential on the floating nodes to produce the correct total charge on the electrode. The total
charge when performing a simulation is by default zero but can be defined using the SOLVE
statement. The total charge may change if you activate a Charge Injection Model. For more
information about this model, see Section 3.6.7 “Gate Current Models”. To define a contact as
a floating contact, use:
CONTACT NAME=fgate FLOATING

Note: When specifying a floating contact, use the Newton scheme as the numerical technique. See Section 20.5.1
“Newton Iteration”.

151 ATLAS User’s Manual

Boundary Physics Physics

3.5.4 Current Boundary Conditions

In some devices, the terminal current is a multi-valued function of the applied voltage. This
means that for some voltage boundary conditions, the solution that is obtained depends on the
initial guess. An example of this is the CMOS latch-up trigger point. At the trigger point the I-
V curve changes from being flat to vertical and may exhibit a negative slope. The solution
will then have three different solutions of current for one applied bias. The particular solution
which the model finishes in will depend upon the initial conditions.
This trigger point is difficult to determine using a simple voltage boundary condition. In
addition, it is almost impossible to compute any solutions in the negative resistance regime
when using voltage boundary conditions. Some of these problems can be overcome using
current boundary conditions.
Calculation of current boundary conditions is activated by the CURRENT parameter in the
CONTACT statement.
The current boundary conditions are described by the Kirkhoff relation:
Js –   Jn + Jp  = 0 3-191

where Js is the applied current, Jn and Jp are given by Equations 3-11 and 3-12, and the
summation is taken over all grid points incident on the boundary. This relation is solved for
the boundary potential. The ohmic conditions in Equations 3-160 and 3-161 apply for
electron and hole concentrations.
The voltage boundary condition should be used in regions where dI/dV is small. The current
boundary condition may be preferable for operating regimes where dI/dV is large. It is
common for the negative resistance regime of a device to have a slope dI/dV very close to 0.
Such behavior should be considered when using a current source to trace out an entire I-V
curve. In these cases, use the CURVETRACE option in ATLAS.

Note: When a current boundary condition has been specified, choose the Newton numerical scheme on the
METHOD statement. You can perform ac small signal analysis with current boundary conditions.

3.5.5 Insulating Contacts

Insulating contacts are contacts that are completely surrounded by insulator. They may be
connected to a voltage source (tied contact) or they may be floating.
Insulating contacts that are connected to a voltage source generally have a work function that
dictates a value for s similar to that given by Equation 3-163. Electron and hole
concentrations within the insulator and at the insulating contact are forced to be zero (i.e.,
ns=ps=0).

152 ATLAS User’s Manual

Boundary Physics Physics

3.5.6 Neumann Boundaries

Along the outer (non-contact) edges of devices, homogeneous (reflecting) Neumann
boundary conditions are imposed so that current only flows out of the device through the
contacts. In the absence of surface charge along such edges, the normal electric field
component becomes zero. Current isn’t permitted to flow from the semiconductor into an
insulating region except via oxide tunneling models. At the interface between two different
materials, the difference between the normal components of the respective electric
displacements must be equal to any surface charge according to:
n̂   1  1 – n̂   2  2 =  s 3-192

where n is the unit normal vector, 1 and 2 are the permittivities of the materials on either
side of the interface and s is the sheet charge at the interface.

3.5.7 Lumped Element Boundaries

ATLAS supports some simple configurations of lumped elements. These are indicated in
Figure 3-5.

Figure 3-5: The Lumped elements supported by ATLAS

Lumped elements can be extremely useful when simulating CMOS or MOSFET structures.
For example, the p-tub contact in the CMOS cross-section might be tens or hundreds of
microns away from the active area of an embedded vertical npn bipolar transistor, which may
only be 10-20 µm on a side. If the whole structure were simulated, a tremendous number of
grid points (probably more than half) are essentially wasted in accounting for a purely
resistive region of the device.
In the case of a MOSFET substrate, you would not want to include grid points all the way to
the back side of a wafer. In either of these cases, a simple lumped resistance can be
substituted [225].

153 ATLAS User’s Manual

Boundary Physics Physics

Table 3-27 shows the user-specifiable parameters for Figure 3-5.

Table 3-27 User-Specifiable Parameters for Figure 3-2.

Symbol Statement Parameter Units

C CONTACT CAPACITANCE F/µm

R CONTACT RESISTANCE µm
L CONTACT INDUCTANCE Hµm

Resistance is specified in µm, capacitance is specified in F/µm, and inductance is specified

in Hµm.
The following combinations are possible: resistance only, resistance and capacitance,
inductance and resistance and inductance, capacitance and resistance. For the case of
inductance or capacitance only, ATLAS adds a small resistance as well. For more complicated
circuit configurations, use the MIXEDMODE module of ATLAS.
The boundary condition for a lumped resistance boundary is analogous to the current
boundary condition described in Section 3.5.4 “Current Boundary Conditions” and is given
by:
Va – Vs
------------------ –   J n + J p  = 0 3-193
R

where Va is the applied potential, R is the lumped resistance, and Jn and Jp are the electron
and hole currents given by Equations 3-11 and 3-12. Vs is the boundary potential or "internal
bias" solution. The summation is taken over all grid points incident on the boundary. The
ohmic conditions in Equations 3-160 and 3-161 apply for electron and hole concentrations.

Note: Capacitance increases with device width (into the Z plane) while resistance decreases. Except for the case
of extremely large resistances where the arrangement becomes similar to a pure current source, no
convergence degradation has been observed for a lumped element boundary in comparison to a simple
Ohmic contact. Transient simulation therefore becomes easier and is more well-defined.

You should use the simulator to calculate any resistance (or capacitance) components that
might be included as lumped elements. When performing CMOS simulations, you could
simulate just the p-tub with Ohmic contacts at either end. From the plot of terminal current (in
A/µm) versus voltage, resistance can be directly extracted from the slope. Be very careful to
consider any three-dimensional effects (e.g., current spreading) before using a resistance
value in further simulations.
When looking at the results of simulation with lumped elements it’s important to distinguish
between the applied voltage (Vapp) and the internal bias () in the log file produced by
ATLAS.
The lumped element boundary conditions can also be used with AC small signal analysis (see
Section 20.10.2 “Fourier Analysis Of Transient Responses”). In this case, the driving voltage
is a harmonic signal of small amplitude and the capacitance shown in Figure 3-5 is connected
to AC ground. Because some of the AC current goes to this ground, the measured admittances
may not always sum to zero.

154 ATLAS User’s Manual

Boundary Physics Physics

An alternative circuit arrangement is also available. If you specify the RESONANT parameter
on a CONTACT statement, then for that contact the capacitance will also connected to the
driving voltage.
In this case, the lumped elements form a parallel resonant circuit with the series resistor and
inductance in parallel with the capacitance. Also in this case, current continuity will be
explicitly satisfied.
The susceptance may be inductive rather than capacitive if inductors are present in the
lumped elements. This manifests itself as a negative capacitance (i.e., the conductance and
capacitance are of different signs). You can easily convert the output capacitance to an
inductance using the formula
2
L = 1   C 3-194

where  is the angular frequency, C is the capacitance in farads and L the inductance in
henrys.

3.5.8 Distributed Contact Resistance

Since contact materials have finite resistivities, the electrostatic potential isn’t always
uniform along the metal-semiconductor surface. To account for this effect, a distributed
contact resistance can be associated with any electrode. This is in contrast to the lumped
contact resistance described in the previous section.
ATLAS implements this distributed contact resistance by assigning a resistance value to each
surface element associated with the contact. If each surface element is a different length, a
different value for the contact resistance will be assigned. ATLAS calculates a resistance
value of Ri for each surface element from the value of CON.RESIST, as specified on the
CONTACT statement. The units of CON.RESIST are cm2 and Ri calculated as:

CON.RESIST-
R i = ------------------------------------ 3-195
d i WIDTH

where Ri is the resistance at node i, Pc is specified by the CON.RESIST parameter, di is the

length of the contact surface segment associated with node i and WIDTH is the width of the
device. The effect of the resistance, Ri is to add an extra equation to be satisfied to node i.
This equation is given by:
kT L
1
----- V –    --------- ln  N  n ie  –  I n + I p + I disp  = 0 3-196
R i applied  i q 

where Vapplied is the external applied voltage, i is the surface potential, N is the net doping,
ni is the intrinsic electron concentration, and In, Ip, Idisp are the electron, hole and
displacement currents at node i. This equation simply balances the current in and out of the
resistor added to each i node.

155 ATLAS User’s Manual

Boundary Physics Physics

As with the case for lumped elements, ATLAS can print out a value of contact resistance for
each contact in the run time output. Since the actual value depends on the length of each
surface segment for distributed contacts, ATLAS prints out the value of CON.RESIST/WIDTH
which is the same for all contact surface segments. This runtime output is enabled by adding
the PRINT option on the MODELS statement.

Table 3-28 User-Specifiable Parameters for Equation 3-195

Statement Parameter Units

CONTACT CON.RESIST -cm2

MESH WIDTH m

Note: AC small signal analysis cannot be performed when any distributed contact resistance has been specified.
Also, only the NEWTON numerical scheme should be chosen on the METHOD statement.

3.5.9 Energy Balance Boundary Conditions

When the Energy Balance Transport Model is applied, special boundary conditions are
applied for carrier temperatures. By default at the contacts, Dirichlet boundary conditions are
used for carrier temperatures:
Tn = Tp = TL 3-197

You can treat contacts as Neumann (reflective) boundaries with respect to carrier temperature
by specifying REFLECT on the CONTACT statement.
Elsewhere on the boundary, the normal components of the energy fluxes vanish. The
boundary conditions for (, n, p) are the same as for the drift-diffusion model.

3.5.10 Floating Semiconductor Regions

A semiconductor region that is totally surrounded by insulator can be modeled as a floating
contact as described in Section 3.5.3 “Floating Contacts”. The approximations made there are
that the floating region is an equipotential and that the value of the equipotential is derived
from a macroscopic application of Gauss's flux theorem to the whole region. The Poisson
equation is not solved in the interior of the region.
This approximation is appropriate for a metallic contact but highly doped polysilicon is
commonly used as a floating contact. In this case, it is more appropriate to make the
assumption that the potential is non-uniform and that the Fermi-level is constant in the
floating region. The electron and hole densities then vary across the region according to
Equations 3-36 and 3-37. The electron and hole quasi-Fermi potentials, n and p, in this
case are equal to the same constant value.

156 ATLAS User’s Manual

Boundary Physics Physics

The value of the Fermi level is obtained from the constraint that

V Q  r  dV = QFG 3-198

where the local charge density comprises the ionized dopant densities, along with the free
carrier densities. The integration is over the volume of the region and QFG is the total charge
associated with the region. Within the floating semiconductor region, the Poisson equation is
solved to give the potential distribution. The current-continuity equations are not solved there
because the assumption of constant quasi-Fermi level implies that there is no current flow
within the region.
It is convenient to refer to these floating semiconductor regions as contacts. The value of
QFG for the region can be set explicitly on the SOLVE statement as for the floating contacts in
Section 3.5.3 “Floating Contacts”. On transient SOLVE statements, if a gate charging model
is enabled, then the value of QFG can evolve with time. The value of QFG is sent to the runtime
output.
To define a floating region, you define an ELECTRODE and add the FLOATING parameter. You
must also set the ELECTRODE to be a semiconductor by using the MATERIAL parameter. For
example
ELECTRODE NAME=FGATE X.MIN=0.0 X.MAX=1.0 Y.MIN=-0.08 Y.MAX=-0.02
MATERIAL=POLYSILICON FLOATING
For a pre-existing electrode read in from a standard structure file, you use the MODIFY
parameter on the ELECTRODE statement along with the FLOATING parameter. For example
ELECTRODE MODIFY NAME="fgate" FLOATING

Table 3-29 Parameters for the FLOATING ELECTRODE Model

Statement Parameter Type Default Units

ELECTRODE FLOATING Logical False

ELECTRODE MODIFY Logical False

157 ATLAS User’s Manual

Boundary Physics Physics

3.5.11 High-K Effective Workfunction Model

ATLAS includes a model for effective metal workfunction in high-k gate insulators. This
model is based on the charge neutrality level model [322]. In this model, the effective
workfunction is given by:

 meff = E · HIGHK + S · HIGHK  WORKF – E · HIGHK  3-199

where meff is the effective gate workfunction, E.HIGHK is the charge neutrality level of the
high-k material, S.HIGHK is a slope parameter of the high-k material, and WORKF is the metal
workfunction in vacuum. The parameters WORKF, E.HIGHK, and S.HIGHK can all be specified
on the CONTACT statement.
The following default metal workfunctions can be used by specifying the associated
parameter name on the CONTACT statement.

Table 3-30 Parameterized Workfunction Defaults

WORKF
Parameter Name
(eV)

HIGHK.MG 3.66
HIGHK.TA 4.25
HIGHK.AL 4.28
HIGHK.TIN 4.53
HIGHK.W 4.63
HIGHK.MO 4.95
HIGHK.PT 5.65
HIGHK.NI 5.04
HIGHK.AU 5.39
HIGHK.HF 3.95

The following default high-k parameter values can be used by specifying the associated
parameter name on the CONTACT statement.

158 ATLAS User’s Manual

Boundary Physics Physics

Table 3-31 Parameterized Dielectric Parameter Defaults

E.HIGHK
Parameter Name S.HIGHK
(eV)

ON.SIO2 4.81 0.95

ON.AL2O3 3.43 0.69
ON.SI3N4 4.16 0.59
ON.HFO2 4.81 0.52
ON.ZRO2 4.53 0.52

Note: Here, the energy values are derived from the charge neutrality level values from [322] and the bandgap and
conduction band edge offset values from [236].

For example using Tables 3-30 and 3-31, you can specify an aluminum gate with a HfO2
dielectric by the following statement:
CONTACT NAME=gate HIGHK.AL ON.HFO2
This simple statement specifies that Equation 3-199 is used to calculate the effective gate
workfunction.
You can also modify the workfunction to include the effects of a second interfacial dielectric
layer in an anologous manner to the main dielectric layer. The interfacial layer is always
located between the high-k gate dielectric layer and the semiconductor.
In these cases, you can specify one of the logical parameters in Table 3-32. When you
include one of these parameters, the effective workfunction is calculated as in Equation 3-
199, except that the equation is applied recursively to the interfacial layer using the effective
workfunction from Equation 3-199 for WORKF and E.HIGHK and S.HIGHK from the interface
layer to obtain a composite effective workfunction.
The parameters used to enable interfacial layers are given in Table 3-32.

Table 3-32 Parameterized Dielectric Interface Layer Parameter Defaults

E.HIGHK
Parameter Name S.HIGHK
(eV)

IFL.SIO2 4.81 0.95

IFL.AL2O3 3.43 0.69
IFL.SI3N4 4.16 0.59
IFL.HFO2 4.81 0.52
IFL.ZRO2 4.53 0.52

159 ATLAS User’s Manual

Physical Models Physics

3.6 Physical Models

3.6.1 Mobility Modeling
Electrons and holes are accelerated by electric fields, but lose momentum as a result of
various scattering processes. These scattering mechanisms include lattice vibrations
(phonons), impurity ions, other carriers, surfaces, and other material imperfections. Since the
effects of all of these microscopic phenomena are lumped into the macroscopic mobilities
introduced by the transport equations these mobilities are therefore functions of the local
electric field, lattice temperature, doping concentration, and so on.
Mobility modeling is normally divided into: (i) low-field behavior, (ii) high field behavior,
(iii) bulk semiconductor regions and (iv) inversion layers.
The low electric field behavior has carriers almost in equilibrium with the lattice and the
mobility has a characteristic low-field value that is commonly denoted by the symbol n0,p0.
The value of this mobility is dependent upon phonon and impurity scattering. Both of which
act to decrease the low-field mobility.
The high electric field behavior shows that the carrier mobility declines with electric field
because the carriers that gain energy can take part in a wider range of scattering processes.
The mean drift velocity no longer increases linearly with increasing electric field, but rises
more slowly. Eventually, the velocity doesn’t increase any more with increasing field but
saturates at a constant velocity. This constant velocity is commonly denoted by the symbol
sat. Impurity scattering is relatively insignificant for energetic carriers, and so sat is
primarily a function of the lattice temperature.
Modeling mobility in bulk material involves: (i) characterizing n0 and p0 as a function of
doping and lattice temperature, (ii) characterizing sat as a function of lattice temperature, and
(iii) describing the transition between the low-field mobility and saturated velocity regions.
Modeling carrier mobilities in inversion layers introduces additional complications. Carriers
in inversion layers are subject to surface scattering, extreme carrier-carrier scattering, and
quantum mechanical size quantization effects. These effects must be accounted for in order to
perform accurate simulation of MOS devices. The transverse electric field is often used as a
parameter that indicates the strength of inversion layer phenomena.
You can define multiple non-conflicting mobility models simultaneously. You also need to
know which models are over-riding when conflicting models are defined.
Low-Field Mobility Models
The low-field carrier mobility can be defined in five different ways.
The first way is use the MUN and MUP parameters to set constant values for electron and hole
mobilities and optionally specify temperature dependence. The second way is by using a
look-up table model (CONMOB) to relate the low-field mobility at 300K to the impurity
concentration. The third way is by choosing the analytic low-field mobility models,
ANALYTIC, ARORA, or MASETTI, to relate the low-field carrier mobility to impurity
concentration and temperature.The fourth way is by choosing a carrier-carrier scattering
model (CCSMOB, CONWELL, or BROOKS) that relates the low-field mobility to the carrier
concentrations and temperature. The fifth way is to use a unified low-field mobility model
(KLAASSEN) that relates the low-field mobility to donor, acceptor, lattice, carrier-carrier
scattering, and temperature.

160 ATLAS User’s Manual

Physical Models Physics

Constant Low-Field Mobility Model

In ATLAS, the choice of mobility model is specified on the MODELS statement. The
parameters associated with mobility models are specified on a separate MOBILITY statement.
One or more mobility models should always be specified explicitly. The default is to use
constant low-field mobilities within each region of a device. This default model is
independent of doping concentration, carrier densities and electric field. It does account for
lattice scattering due to temperature according to:
T L – TMUN
 n0 = MUN  ---------- 3-200
 300

T L – TMUP
 p0 = MUP  ---------- 3-201
300

where T is the lattice temperature. The low-field mobility parameters: MUN, MUP, TMUN, and
TMUP can be specified in the MOBILITY statement with the defaults as shown in Table 3-33.

Table 3-33 User-Specifiable Parameters for the Constant Low-Field Mobility Model

Statement Parameter Default Units

MOBILITY MUN 1000 cm2/(V·s)

MOBILITY MUP 500 cm2/(V·s)
MOBILITY TMUN 1.5
MOBILITY TMUP 1.5

Concentration-Dependent Low-Field Mobility Tables

ATLAS provides empirical data for the doping dependent low-field mobilities of electrons
and holes in silicon at TL=300K only. This data is used if the CONMOB parameter is specified in
the MODELS statement. The data that is used is shown in Table 3-34.

Table 3-34 Mobility of Electrons and Holes in Silicon at T=300K

Concentration (cm-3) Mobility (cm2/Vs)

Electrons Holes

1.01014 1350.0 495.0

2.01014 1345.0 495.0
4.01014 1335.0 495.0
6.01014 1320.0 495.0
8.01014 1310.0 495.0

161 ATLAS User’s Manual

Physical Models Physics

Table 3-34 Mobility of Electrons and Holes in Silicon at T=300K

Concentration (cm-3) Mobility (cm2/Vs)

Electrons Holes

1.01015 1300.0 491.1

2.01015 1248.0 487.3
4.01015 1200.0 480.1
6.01015 1156.0 473.3
8.01015 1115.0 466.9
1.01016 1076.0 460.9
2.01016 960.0 434.8
4.01016 845.0 396.5
6.01016 760.0 369.2
8.01016 720.0 348.3
1.01017 675.0 331.5
2.01017 524.0 279.0
4.01017 385.0 229.8
6.01017 321.0 2103.8
8.01017 279.0 186.9
1.01018 252.0 178.0
2.01018 182.5 130.0
4.01018 140.6 90.0
6.01018 113.6 74.5
8.01018 99.5 66.6
1.01019 90.5 61.0
2.01019 86.9 55.0
4.01019 83.4 53.7
6.01019 78.8 52.9
8.01019 71.6 52.4
1.01020 67.8 52.0
2.01020 52.0 50.8

162 ATLAS User’s Manual

Physical Models Physics

Table 3-34 Mobility of Electrons and Holes in Silicon at T=300K

Concentration (cm-3) Mobility (cm2/Vs)

Electrons Holes

4.01020 35.5 49.6

6.01020 23.6 48.9
8.01020 19.0 48.4
1.01021 17.8 48.0

Analytic Low-Field Mobility Model

The following analytic function based upon the work of Caughey and Thomas [41, 251] can
be used to specify doping- and temperature-dependent low-field mobilities.

T L ALPHAN · CAUG
 n0 = MU1N CAUG   --------------
·  300K 3-202

T L BETAN · CAUG T L ALPHAN · CAUG

MU2N CAUG   -------------- – MU1N · CAUG   --------------
·  300K  300K
+ ------------------------------------------------------------------------------------------------------------------------------------------------------------------
T L GAMMAN · CAUG DELTAN · CAUG
1 +  --------------   ----------------------------------------
N
 300K  NCRITN CAUG
·

T L ALPHAP · CAUG
 p0 = MU1P · CAUG   -------------- 3-203
 300K

T L BETAP · CAUG TL ALPHAP · CAUG

MU2P · CAUG   -------------- – MU1P · CAUG   --------------
 300K  300K
+ -------------------------------------------------------------------------------------------------------------------------------------------------------------------
T L GAMMAP · CAUG DELTAP · CAUG
1 +  --------------   ----------------------------------------
N
 300K  NCRITP CAUG
·

where N is the local (total) impurity concentration in cm-3 and TL is the temperature in degrees
Kelvin.
This model is activated by specifying ANALYTIC in the MODELS statement. The parameters of
this model are specified in the MOBILITY statement. The default parameters are for silicon at
TL = 300K.

163 ATLAS User’s Manual

Physical Models Physics

Table 3-35 User-Specifiable Parameters for Equations 3-202 and 3-203

Statement Parameter Default Units

MOBILITY MU1N.CAUG 55.24 cm2/(V s)

MOBILITY MU1P.CAUG 49.7 cm2/(V s)
MOBILITY MU2N.CAUG 1429.23 cm2/(V s)
MOBILITY MU2P.CAUG 479.37 cm2/(V s)
MOBILITY ALPHAN.CAUG 0.0 arbitrary
MOBILITY ALPHAP.CAUG 0.0 arbitrary
MOBILITY BETAN. CAUG -2.3 arbitrary
MOBILITY BETAP. CAUG -2.2 arbitrary
MOBILITY GAMMAN.CAUG -3.8 arbitrary
MOBILITY GAMMAP.CAUG -3.7 arbitrary
MOBILITY DELTAN.CAUG 0.73 arbitrary
MOBILITY DELTAP.CAUG 0.70 arbitrary
MOBILITY NCRITN.CAUG 1.0721017 cm-3
MOBILITY NCRITP.CAUG 1.6061017 cm-3

Arora Model for Low-Field Mobility

Another analytic model for the doping and temperature dependence of the low-field mobility
is available in ATLAS. This model, which is due to Arora [16], has the following form:

T L BETAN.ARORA
MU2N.ARORA  ----------
TL ALPHAN.ARORA  300
 n0 = MU1N.ARORA  ---------- + ------------------------------------------------------------------------------------------------------------------------- 3-204
 300 N
1 + --------------------------------------------------------------------------------------------------------------
T L GAMMAN.ARORA
NCRITN.ARORA   ----------
 300

 TL 
BETAP.ARORA
TL ALPHAP.ARORA MU2P.ARORA  --------- -
 p0 = MU1P.ARORA  ----------
 300
+ ----------------------------------------------------------------------------------------------------------------------------------------------------- 3-205
 300 N
1 + ---------------------------------------------------------------------------------------------------------------------------------------- -
 TL 
GAMMAP.ARORA
NCRITP.ARORA   ----------
 300

where N is the total local dopant concentration.

164 ATLAS User’s Manual

Physical Models Physics

This model is activated by specifying ARORA in the MODELS statement. The parameters of the
model are specified in the MOBILITY statement. The default parameters are for silicon at
TL=300K.

Table 3-36 User-Specifiable Parameters for Equations 3-204 and 3-205

Statement Parameter Default Units

MOBILITY MU1N.ARORA 88.0 cm2/(V·s)

MOBILITY MU1P.ARORA 54.3 cm2/(V·s)
MOBILITY MU2N.ARORA 1252.0 cm2/(V·s)
MOBILITY MU2P.ARORA 407.0 cm2/(V·s)
MOBILITY ALPHAN.ARORA -0.57
MOBILITY ALPHAP.ARORA -0.57
MOBILITY BETAN.ARORA -2.33
MOBILITY BETAP.ARORA -2.33
MOBILITY GAMMAN.ARORA 2.546
MOBILITY GAMMAP.ARORA 2.546
MOBILITY NCRITN.ARORA 1.4321017 cm-3
MOBILITY NCRITP.ARORA 2.671017 cm-3

Masetti Model For Low-Field Mobility

Masetti et al. [176] modeled the dependence of mobility on carrier concentration over a range
of 8 orders of magnitude in carrier concentration from approximately 1013 cm-3 to 1021 cm-3.
They found that their model required different parameter sets for the electron mobility in
Arsenic and Phosphorous n-doped Silicon. The model is optimized for room temperature,
although it does model temperature dependence to some extent.
The functional form of the electron mobility is

 n = MTN · MIN1 exp  – MTN · PC + ----------------------------------------------------------------------

---------------------

MTN · MAX – MTN · MIN2 MTN · MU1
- – ------------------------------------------------------------------
- 3-206
N MTN · ALPHA
1 +  ---------------------
N MTN · CS MTN · BETA
 MTN CR 1 +  ---------------------
 N 
·

where N is the total or ionized doping level. The functional form of the hole mobility is

 p = MTP · MIN1 exp  – ---------------------

MTP · PC MTP · MAX – MTP · MIN2
+ ----------------------------------------------------------------------
MTP · MU1
- – ------------------------------------------------------------------
- 3-207
N  MTP · ALPHA
1 +  ---------------------
N MTP · CS MTP · BETA
 MTP CR 1 +  ---------------------
· N 

165 ATLAS User’s Manual

Physical Models Physics

where N is the total dopant concentration. MTN.MAX and MTP.MAX are given a lattice
temperature dependence as in Equations 3-200 and 3-201 respectively. These equations are
functionally equivalent to the bulk mobility term in the Lombardi CVT model (Equations 3-
253 and 3-254). To use it as a stand alone mobility model, use the MASETTI parameter on the
MODELS statement for both electron and hole mobility.
N.MASETTI on the MOBILITY statement enables it for electrons. P.MASETTI on the MOBILITY
statement enables it for holes.
For electron mobility, there is a choice of two default parameter sets. The set for arsenic
doping are used unless you set the MSTI.PHOS parameter on the MOBILITY statement to use
the set corresponding to Phosphorous. The following tables show default sets for electron
mobility and for Boron doping (p-type).

Table 3-37 Parameter Set for Electron Mobility with Arsenic Doping in Silicon.

Statement Parameter Default Units

MOBILITY MTN.MIN1 52.2 cm2/(Vs)

MOBILITY MTN.MIN2 52.2 cm2/(Vs)
MOBILITY MTN.MAX 1417 cm2/(Vs)
MOBILITY MTN.MU1 43.4 cm2/(Vs)
MOBILITY MTN.PC 0.0 cm-3
MOBILITY MTN.CR 9.681016 cm-3
MOBILITY MTN.CS 3.341020 cm-3
MOBILITY MTN.ALPHA 0.68
MOBILITY MTN.BETA 2.0

Table 3-38 Parameter Set for Electron Mobility with Phosphorous Doping in Silicon
(MSTI.PHOS set)

Statement Parameter Default Units

MOBILITY MTN.MIN1 68.5 cm2/(Vs)

MOBILITY MTN.MIN2 68.5 cm2/(Vs)
MOBILITY MTN.MAX 1414 cm2/(Vs)
MOBILITY MTN.MU1 56.1 cm2/(Vs)
MOBILITY MTN.PC 0.0 cm-3
MOBILITY MTN.CR 9.21016 cm-3

166 ATLAS User’s Manual

Physical Models Physics

Table 3-38 Parameter Set for Electron Mobility with Phosphorous Doping in Silicon
(MSTI.PHOS set)

Statement Parameter Default Units

MOBILITY MTN.CS 3.411020 cm-3

MOBILITY MTN.ALPHA 0. 711
MOBILITY MTN.BETA 1.98

Table 3-39 Parameter Set for Hole Mobility with Boron doping in Silicon.

Statement Parameter Default Units

MOBILITY MTP.MIN1 44.9 cm2/(Vs)

MOBILITY MTP.MIN2 0.0 cm2/(Vs)
MOBILITY MTP.MAX 470.5 cm2/(Vs)
MOBILITY MTP.MU1 29.0 cm2/(Vs)
MOBILITY MTP.PC 9.231016 cm-3
MOBILITY MTP.CR 2.231017 cm-3
MOBILITY MTP.CS 6.11020 cm-3
MOBILITY MTP.ALPHA 0.719
MOBILITY MTP.BETA 2.0

Carrier-Carrier Scattering Models For Low-Field Mobility

Dorkel and Leturcq Models
The Dorkel and Leturcq Model [67] for low-field mobility includes the dependence on
temperature, doping, and carrier-carrier scattering. This model is activated by specifying the
CCSMOB parameter of the MODELS statement. The parameters of the model are specified in the
MOBILITY statement. This model has the form:

 
 
 1.025 
 n0 p0 =  nL p  ---------------------------------------------------------------- – 0.025 3-208
   n p 
L 0.715

 1 + 2.126  --------------- - 
   n p  IC


where L is the lattice scattering, I is the ionized impurity scattering, and C is the carrier-carrier
scattering. Here,  nIC p is defined as:

1 1 –1
 nIC
p =
------ + ------------ 3-209
 C  n p
I
where:

167 ATLAS User’s Manual

Physical Models Physics

32
21 T L
1.04  10  ----------
C  300
 = ------------------------------------------------------------------------------------------------------ 3-210
2
13 T L –1  3
np ln 1 + 7.45  10  ----------  np 
 300

32 2
AN.CCS  T L  BN.CCS  T L 
 nI = ------------------------------------------ f ------------------------------------ 3-211
NT n+p

32 2
AP.CCS  T L  BP.CCS  T L 
 pI = ------------------------------------------ f ------------------------------------ 3-212
NT n+p

Here, NT is the total concentration, TL is the lattice temperature and n,p are the electron and
hole carrier concentrations respectively.

x –1
f x = ln  1 + x  – ------------ 3-213
1+x

The values of the lattice scattering terms, µN,PL are defined by Equations 3-200 and 3-201.

Table 3-40 User-Specifiable Parameters for Equations 3-211 and 3-212

Statement Parameter Default Units

MOBILITY AN.CCS 4.611017 cm-3

MOBILITY AP.CCS 1.01017 cm-3
MOBILITY BN.CCS 1.521015 cm-3
MOBILITY BP.CCS 6.251014 cm-3

Conwell-Weisskopf Model
This model adapts Conwell-Weisskopf theory to carrier-carrier scattering [47]. According to
this model, the carrier-carrier contribution to mobility is

T 32
D · CONWELL  ----------
 300
 ccs = ----------------------------------------------------------------------------------------- 3-214
  F · CONWELL  T  2 
pn  ln  1 + -------------------------------- - ----------  
 1  3   300
   pn  

where T is the Lattice temperature in Kelvin, n is the electron concentration, and p is the hole
concentration. This is then combined with other enabled low-field mobility models (if any)
using Matthiessens rule, giving overall low-field mobilities as follows

168 ATLAS User’s Manual

Physical Models Physics

1 1 1
------ = --------- + ---------- 3-215
n  n0  ccs

1 1 1
------ = --------- + ---------- 3-216
p  p0  ccs

To enable the model for both electrons and holes, specify CONWELL on the MODELS statement.
Alternatively, N.CONWELL on the MOBILITY statement enables it for electron mobility.
P.CONWELL on the MOBILITY statement enables it for hole mobility.

Table 3-41 MOBILITY Statement Parameters

Parameter Type Default Units

D.CONWELL Real 1.041021 (cmV·s)-1

F.CONWELL Real 7.4521013 cm-2
N.CONWELL Logical False
P.CONWELL Logical False

Brooks-Herring Model
Like the Conwell-Weisskopf model, this adds in the effects of carrier-carrier scattering to the
low-field mobility using Matthiesen's rule. The carrier-carrier contribution to mobility is
3
---
T 2
A · BROOKS  ----------
 300
 ccs = ------------------------------------------------ 3-217
pn    

where

    = log  1 +   – ------------- 3-218
1+

and

T 2
B · BROOKS  ----------
 300
  T  = -------------------------------------------------------------------------- 3-219
N c F – 1  2  ------ + N v F –1  2  ------
n p
 N c  N v

F-1/2 is the Fermi-Dirac function of order -1/2, Nc and Nv are the conduction band and valence
band effective densities of states, n is the electron concentration, p is the hole concentration,
and T is the Lattice temperature.
To enable the model for both electrons and holes, specify BROOKS on the MODELS statement.
Alternatively, N.BROOKS on the MOBILITY statement enables it for electron mobility.
P.BROOKS on the MOBILITY statement enables it for hole mobility.

169 ATLAS User’s Manual

Physical Models Physics

Table 3-42 MOBILITY Statement Parameters

Parameter Type Default Units

A.BROOKS Real 1.561021 (cmV·s)-1

B.BROOKS Real 7.631019 cm-3
N.BROOKS Logical False
P.BROOKS Logical False

Incomplete Ionization and Doping Dependent Mobility

The default in ATLAS is to use the total doping (NA+ND) in the formulae for the doping
concentration dependence of low-field mobility (e.g., ARORA model). As an alternative,
ATLAS can use the ionized dopant concentration.
The ionized doping concentration is obtained automatically and used in calculating the charge
density when you set the INCOMPLETE parameter on the MODELS statement. See Section 3.3.1
“Incomplete Ionization of Impurities” for more information.
ATLAS has a MOB.INCOMPL parameter on the MODELS statement. This allows you to use the
ionized dopant concentration to calculate doping dependent mobilities. Specifying this
parameter automatically sets the INCOMPLETE parameter. This ensures that the solution is
consistent with an ionized rather than total doping concentration.
For high doping levels, the difference between total doping and ionized doping
concentrations can be large and so this can result in quite different mobility values.
If it is required to use the ionized dopant concentration for mobility calculations and the total
dopant concentration for calculating the charge density in Poisson's equation, then specify
MOB.INCOMPL ^INCOMPLETE on the MODELS statement to explicitly clear this flag. This
combination of parameters is not recommended.
MOB.INCOMPL affects the ANALYTIC, ARORA, MASETTI, YAMAGUCHI, CVT, KLAASSEN,
ALBRECHT and tabulated Low-Field Mobility Models.

Klaassen’s Unified Low-Field Mobility Model

The model by D. B. M. Klaassen [139, 140], provides a unified description of majority and
minority carrier mobilities. In so doing, it includes the effects of lattice scattering, impurity
scattering (with screening from charged carriers), carrier-carrier scattering, and impurity
clustering effects at high concentration. The model shows excellent agreement between the
modeled and empirical data for:
• majority electron mobility as a function of donor concentration over the range of 1014
cm-3 to 1022 cm-3
• minority electron mobility as a function of acceptor concentration over the range of 1017
cm3 to 1020 cm-3
• minority hole mobility as a function of donor concentration from 1017 cm-3 to 1020 cm-3
• temperature dependence over the range of 70 K to 500 K

170 ATLAS User’s Manual

Physical Models Physics

The Klaassen Model accounts for a broader set of effects and has been calibrated over a wider
range of conditions than any other of the low-field bulk mobility models. This is the
recommended model for both MOS and bipolar simulation and is the default model for
silicon when you set MOS2 or BIPOLAR2 in the MODELS statement. You can enable or disable
the model by using the KLA parameter in the MODELS statement, or independently for electrons
and holes by the KLA.N and KLA.P parameters of the MOBILITY statement.
The total mobility can be described by its components using Matthiessen’s rule as:
–1 –1 –1
n 0 =  nL +  nDAP 3-220

–1 –1 –1
p 0 =  pL +  pDAP 3-221

n and p are the total low-field electron and hole mobilities, nL and pL are the electron
and hole mobilities due to lattice scattering, nDAP and pDAP are the electron and hole
mobilities due to donor (D), acceptor (A), screening (P) and carrier-carrier scattering.
The lattice scattering components, nL and pL are given as:

THETAN.KLA
 nL = MUMAXN.KLA  ----------
300
3-222
T 
L

THETAP.KLA
 pL = MUMAXP.KLA  ----------
300
3-223
T 
L

where TL is the temperature in degrees Kelvin. MUMAXN.KLA, MUMAXP.KLA, THETAN.KLA,

and THETAP.KLA are user-definable model parameters that can be specified as shown in Table
3-43.

Table 3-43 User-Specifiable Parameters for Equations 3-222 and 3-223

Statement Parameter Default Units

MOBILITY MUMAXN.KLA 1417.0 cm2/(Vs)

MOBILITY MUMAXP.KLA 470.5 cm2/(Vs)
MOBILITY THETAN.KLA 2.285
MOBILITY THETAP.KLA 2.247

The impurity-carrier scattering components of the total mobility are given by:
N nsc  NREF1N.KLA ALPHA1N.KLA
+  nc  ---------------------
n+p
 nDAP =  N, n --------------------
- ------------------------------------- 3-224
N nsc, eff  N nsc   N nsc, eff

171 ATLAS User’s Manual

Physical Models Physics

N psc  NREF1P.KLA ALPHA1P.KLA

+  pc  ---------------------
n+p
 pDAP =  N, p --------------------
- ------------------------------------- 3-225
N psc, eff  N psc
  N psc, eff


Table 3-44 User-Specifiable Parameters for Equations 3-224 and 3-225

Statement Parameter Default Units

MOBILITY ALPHA1N.KLA 0.68

MOBILITY ALPHA1P.KLA 0.719
MOBILITY NREF1N.KLA 9.681016 cm3
MOBILITY NREF1P.KLA 2.231017 cm3

The impurity scattering components, N,n and N,p, are given by:
2 T L  3  ALPHA1N.KLA – 1.5 
 N, n = ---------------------------------------------------------------------------------  ----------
MUMAXN.KLA
3-226
MUMAXN.KLA – MUMINN.KLA  300

2 T L  3  ALPHA1P.KLA – 1.5 
 N, p = ---------------------------------------------------------------------------------  ----------
MUMAXP.KLA
3-227
MUMAXP.KLA – MUMINP.KLA  300

where TL is the temperature in degrees Kelvin. MUMINN.KLA and MUMINP.KLA are user-
defined parameters shown in Table 3-45, and the other parameters are as described in Tables
3-43 and 3-44.

Table 3-45 User-Specifiable Parameters for Equations 3-226 and 3-227

Statement Parameter Default Units

MOBILITY MUMINN.KLA 52.2 cm2/(Vs)

MOBILITY MUMINP.KLA 44.9 cm2/(Vs)

The carrier-carrier scattering components, nc and pc, are given by:

MUMINN.KLA  M UMAXN.KLA 300 0.5

 nc = ------------------------------------------------------------------------------------  ---------- 3-228
MUMAXN.KLA – M UMINN.KLA  T L 

MUMINP.KLA  MUMAXP.KLA 300 0.5

 pc = ------------------------------------------------------------------------------------  ---------- 3-229
MUMAXP.KLA – M UMINP.KLA  T L 

The Nnsc and Npsc parameters of Equations 3-224 and 3-225 are given by:

N nsc = N D + N A + p 3-230

172 ATLAS User’s Manual

Physical Models Physics

N psc = N D + N A + n 3-231
where ND is the donor concentration in cm-3, NA is the acceptor concentration in cm-3, n is
the electron concentration in cm-3 and p is the hole concentration in cm-3.
The parameters of Equations 3-224 and 3-225 are given by:

N nsc eff = N D + G  P n N A +  ---------------

p
3-232
 F  P 
n

N psc eff = N A + G  P p N D +  ---------------

n
3-233
 F  P 
p

where ND is the donor concentration in cm-3, NA is the acceptor concentration in cm-3 and n
is the electron concentration in cm-3 and p is the hole concentration in cm-3. The two
functions, G(P) and F(P), are functions of the screening factors, Pn and Pp, for electrons and
holes. The function, G(P), in Equations 3-232 and 3-233 are given by:
S1.KLA S5.KLA
G  P  = 1 – -------------------------------------------------------------------------------------------------------------------------- + ------------------------------------------------------------------------------------------- 3-234
n

 L T  300  
S4.KLA S3.KLA  ME.KLA  S7.KLA
S6.KLA
S2.KLA + P n  ---------------------------- P  -----------------------
 ME.KLA 
n   T L  300 

S1.KLA S5.KLA
G  P p  = 1 – -------------------------------------------------------------------------------------------------------------------------- + ------------------------------------------------------------------------------------------- 3-235
  T L  300   MH.KLA  S7.KLA S6.KLA
S4.KLA S3.KLA
S2.KLA + P  ---------------------------- P  -----------------------
p  MH.KLA  p   T  300 
L

Here, TL is the temperature in degrees Kelvin, me and mh are the electron and hole masses
and the parameters S1.KLA through S7.KLA are user-specifiable model parameters as shown
in Table 3-46..

Table 3-46 User-Specifiable Parameters for Equations 3-234 and 3-235

Statement Parameter Default Units

MOBILITY S1.KLA 0.89233

MOBILITY S2.KLA 0.41372
MOBILITY S3.KLA 0.19778
MOBILITY S4.KLA 0.28227
MOBILITY S5.KLA 0.005978
MOBILITY S6.KLA 1.80618
MOBILITY S7.KLA 0.72169

173 ATLAS User’s Manual

Physical Models Physics

The functions, F(Pn) and F(Pp), in Equations 3-232 and 3-233 are given by:

ME.KLA
R1.KLA P nR6.KLA + R2.KLA + R3.KLA ----------------------
MH.KLA
F  P n  = -------------------------------------------------------------------------------------------------------------------------------------- 3-236
R6.KLA ME.KLA
Pn + R4.KLA + R5.KLA ----------------------
MH.KLA

MH.KLA
R1.KLA P pR6.KLA + R2.KLA + R3.KLA ------------
ME.KLA
F  P p  = -------------------------------------------------------------------------------------------------------------------------------------- 3-237
R6.KLA MH.KLA
Pp + R4.KLA + R5.KLA ------------
ME.KLA

where the parameters, R1.KLA through R6.KLA, are user-specifiable as shown in Table 3-47.

Table 3-47 User-Specifiable Parameters for Equations 3-236 and 3-237

Statement Parameter Default Units

MOBILITY ME.KLA 1.0

MOBILITY MH.KLA 1.258
MOBILITY R1.KLA 0.7643
MOBILITY R2.KLA 2.2999
MOBILITY R3.KLA 6.5502
MOBILITY R4.KLA 2.3670
MOBILITY R5.KLA -0.8552
MOBILITY R6.KLA 0.6478

The screening parameters, Pn and Pp, used in Equations 3-236 and 3-237 are given by:

–1
FCW.KLA FBH.KLA
P n = -------------------------- + -------------------------- 3-238
P CW n P BH n

–1
FCW.KLA FBH.KLA
P p = -------------------------- + -------------------------- 3-239
P CW p P BH p

Here, the FCW.KLA and FBH.KLA parameters are user-specifiable model parameters as shown
in Table 3-48.

174 ATLAS User’s Manual

Physical Models Physics

Table 3-48 User-Specifiable Parameters for Equations 3-238 and 3-239

Statement Parameter Default Units

MOBILITY FCW.KLA 2.459

MOBILITY FBH.KLA 3.828

The functions, PBH,n and PBH,p, PCW,n, and PCW,p are given by the following equations.

20 TL 2
1.36 10
P BH n = --------------------------  ME.KLA   ---------- 3-240
n  300

20 TL 2
1.36 10
P BH p = --------------------------  MH.KLA   ---------- 3-241
p  300

2
---
13  1 T L 3 3
P CW n = 3.97  10  ---------------  ----------  3-242
3  
 Z N 300 
n D
2
---
13  1 T L 3 3
P CW p = 3.97  10  --------------  ----------  3-243
3  
 Z N 300 
p A
where TL is the temperature in degrees Kelvin, me/m0 and mh/m0 are the normalized carrier
effective masses, and n and p are the electron and hole concentrations in cm-3.
Also here, ND and NA are the donor and acceptor concentrations in cm-3, TL is the temperature
in degrees Kelvin, and Zn and Zp are clustering functions given by:

1
Z n = 1 + ----------------------------------------------------------------------- 3-244
2
CD.KLA +  ---------------------------------
NREFD.KLA
ND

1
Z p = 1 + ----------------------------------------------------------------------- 3-245
2
CA.KLA +  ---------------------------------
NREFA.KLA
NA

where ND and NA are the donor and acceptor concentrations in cm-3 and CD.KLA, CA.KLA,
NREFD.KLA, and NREFA.KLA are user-definable parameters as given in Table 3-49.

175 ATLAS User’s Manual

Physical Models Physics

Table 3-49 User-Specifiable Parameters for Equations 3-244 and 3-245

Statement Parameter Default Units

MOBILITY CD.KLA 0.21

MOBILITY CA.KLA 0.50
MOBILITY NREFD.KLA 4.01020 cm3
MOBILITY NREFA.KLA 7.21020 cm3

Note: When the Klaassen low-field mobility is used, remember that it has been calibrated to work with Klaassen’s
models for bandgap narrowing, KLAAUG recombination, and KLASRH recombination. These models are
described in the Section 3.6.3 “Carrier Generation-Recombination Models”.

Uchida's Low-Field Model for Ultrathin SOI

The model by Uchida et.al. [277], provides a mobility limit in ultrathin-body MOSFET
transistors with SOI thickness less than 4 nm. This mobility limit is due to thickness
fluctuations in the nano scale SOI film. The model for electrons and holes is given by
Equations 3-246 and 3-247.
6
 nu = CN · UCHIDA  TN · UCHIDA 3-246

6
 p u = CP · UCHIDA  TP · UCHIDA 3-247

The parameters CN.UCHIDA, CP.UCHIDA are calibrated from the reference to a default value
of 0.78125 cm2/(Vsnm6). The parameters TN.UCHIDA and TP.UCHIDA represent the
thickness of the SOI in µm. This value should as close as possible to match the physical
thickness of the SOI layer in the simulated structure file. To enable this model for electrons
and holes, specify values for TN.UCHIDA and TP.UCHIDA, and specify the logical parameters
UCHIDA.N or UCHIDA.P or both on the MOBILITY statement. Table 3-50 lists the user
specifiable parameters.

Table 3-50 User-Specifiable Parameters for Equations 3-187 and 3-188

Statement Parameter Default Units

MOBILITY CN.UCHIDA 0.78125 cm2/(Vsnm6)

MOBILITY CP.UCHIDA 0.78125 cm2/(Vsnm6)
MOBILITY TN.UCHIDA 4.0 nm
MOBILITY TP.UCHIDA 4.0 nm

176 ATLAS User’s Manual

Physical Models Physics

Inversion Layer Mobility Models

To obtain accurate results for MOSFET simulations, you need to account for the mobility
degradation that occurs inside inversion layers. The degradation normally occurs as a result of
the substantially higher surface scattering near the semiconductor to insulator interface.
Lombardi CVT Model
The inversion layer model from Lombardi [166] is selected by setting CVT on the MODEL
statement. This model overrides any other mobility models which may be specified on the
MODELS statement. In the CVT model, the transverse field, doping dependent and temperature
dependent parts of the mobility are given by three components that are combined using
Matthiessen’s rule. These components are AC, sr and b and are combined using
Matthiessen’s rule as follows:
–1 –1 –1 –1
 T =  AC +  b +  sr 3-248

The first component, AC, is the surface mobility limited by scattering with acoustic phonons:

 N  TAUN.CVT
 ------
 CN.CVT  N  T
 AC n = ----------------------------------- +  ------------------------------------------------------------------- ------
BN.CVT 1 1
3-249
E  EN.CVT  E  DN.CVT
 -------
 TL
 -------  
 E1   E1 

 N  TAUP.CVT
 ------
 CP.CVT  N  T
- +  ------------------------------------------------------------------- ------
BP.CVT 1 1
 AC p = ---------------------------------- 3-250
E  EP.CVT  E  DP.CVT
 -------
 TL
 -------  
 E1   E1 

where TL is the temperature, E is the perpendicular electric field, and N is the total doping
concentration. In E is 1V/cm, N1 is 1cm-3, and T1 is 1K. You can define the parameters
BN.CVT, BP.CVT, CN.CVT, CP.CVT, DN.CVT, DP.CVT, TAUN.CVT, and TAUP.CVT in the
MOBILITY statement (see Table 3-51 for their defaults).
The second component, µsr, is the surface roughness factor and is given by:

KN · CVT
E E 3
 -------   -------
1  E1  E1
------------ = ---------------------------------- + ----------------------------- 3-251
 sr n DELN · CVT FELN · CVT

for electrons and

KP · CVT
E E 3
 -------  -------
1  E1  E1
------------ = ---------------------------------- + ----------------------------- 3-252
 sr p DELP · CVT FELP · CVT

for holes.

177 ATLAS User’s Manual

Physical Models Physics

The default values of FELN.CVT and FELP.CVT are set so high that the second terms are
negligible. The KN.CVT, KP.CVT, DELN.CVT, DELP.CVT, FELN.CVT and FELP.CVT parameters
are user-definable (see Table 3-51 for their defaults).
The third mobility component, b, is the mobility limited by scattering with optical
intervalley phonons. This component is given by:
T – G AMN.CVT
MUMAXN.CVT  ---------
L
-

– MU0N.CVT
– PCN.CVT 300
 b n = MU0N.CVT exp  ------------------------------------ + --------------------------------------------------------------------------------------------------------------------------------------------------------- 3-253
N ALPHN.CVT
1 +  --------------------------------
N
 CRN.CVT

MU1N.CVT
– ------------------------------------------------------------------------------------------
CSN.CVT BETAN.CVT
1 +  --------------------------------
 N 

T – G AMP.CVT
MUMAXP.CVT  ---------
L
-

– MU0P.CVT
– PCP.CVT 300
 b p = MU0P.CVT exp  ------------------------------------ + --------------------------------------------------------------------------------------------------------------------------------------------------------- 3-254
N ALPHP.CVT
1 +  --------------------------------
N
 CRP.CVT

MU1P.CVT
– ------------------------------------------------------------------------------------------
CSP.CVT BETAP.CVT
1 +  --------------------------------
 N 

Here, N is the total density of impurities and TL is the temperature in degrees Kelvin.
Because AC and sr are related to interaction with an interface, you can specify a typical
distance from the interface over which these components are significant [229]. If you specify
the N.LCRIT or P.LCRIT parameters or both then factors\

FE = exp  – l  N · LCRIT  3-255

and

FH = exp  – l  P · LCRIT  3-256

are obtained, where l is the shortest distance to the nearest interface from the point at which
the mobility will be calculated. These factors are then used to modify the Matthiesen's rule of
Equation 3-248 to
1 FE FE 1
------ = ---------- + ------- + ------ 3-257
T  AC  sr  b

for electrons and

1- = ---------
FH- + FH 1-
----- -------- + ----- 3-258
T  AC  sr  b

178 ATLAS User’s Manual

Physical Models Physics

for holes. Far from the interface, the mobility is the same as the bulk mobility.

Table 3-51 User-Specifiable Parameters for Equations to 3-254

Statement Parameter Default Units

MOBILITY ALPHN.CVT 0.680

MOBILITY ALPHP.CVT 0.71
MOBILITY BETAN.CVT 2.00
MOBILITY BETAP.CVT 2.00
MOBILITY BN.CVT 4.75107 cm2/(V·s)
MOBILITY BP.CVT 9.925106 cm2/(V·s)
MOBILITY CN.CVT 1.74105 cm2/(V·s)
MOBILITY CP.CVT 8.842105 cm2/(V·s)
MOBILITY CRN.CVT 9.681016 cm-3
MOBILITY CRP.CVT 2.231017 cm-3
MOBILITY CSN.CVT 3.431020 cm-3
MOBILITY CSP.CVT 6.101020 cm-3
MOBILITY DELN.CVT 5.821014 cm2/(V·s)
MOBILITY DELP.CVT 2.05461014 cm2/(V·s)
MOBILITY DN.CVT 0.333
MOBILITY DP.CVT 0.333
MOBILITY EN.CVT 1.0
MOBILITY EP.CVT 1.0
MOBILITY FELN.CVT 1.01050 cm2/(V·s)
MOBILITY FELP.CVT 1.01050 cm2/(V·s)
MOBILITY KN.CVT 2.0
MOBILITY KP.CVT 2.0
MOBILITY GAMN.CVT 2.5
MOBILITY GAMP.CVT 2.2
MOBILITY MU0N.CVT 52.2 cm2/(V·s)

179 ATLAS User’s Manual

Physical Models Physics

Table 3-51 User-Specifiable Parameters for Equations to 3-254

Statement Parameter Default Units

MOBILITY MU0P.CVT 44.9 cm2/(V·s)

MOBILITY MU1N.CVT 43.4 cm2/(V·s)
MOBILITY MU1P.CVT 29.0 cm2/(V·s)
MOBILITY MUMAXN.CVT 1417.0 cm2/(V·s)
MOBILITY MUMAXP.CVT 470.5 cm2/(V·s)
MOBILITY N.LCRIT 1.0 cm
MOBILITY P.LCRIT 1.0 cm
MOBILITY PCN.CVT 0.0 cm-3
MOBILITY PCP.CVT 9.231016 cm-3
MOBILITY TAUN.CVT 0.125
MOBILITY TAUP.CVT 0.0317

Note: The CVT model when activated will also, by default, apply the Parallel Electric Field Mobility Model which is
described in the “Parallel Electric Field-Dependent Mobility” on page 195. In this model, the low-field mobility
is supplied from the CVT model.

Darwish CVT Model

The CVT model already described has been successfully used in many device simulations.
Lately, an improved version of this model has been proposed by Darwish et al. [61] where
several modifications have been implemented. The first modification is that the bulk mobility
is calculated using Klaassen’s model (see “Klaassen’s Unified Low-Field Mobility Model” on
page 170) to take into account coulomb screening effects. The second modification is a new
expression for surface roughness is used.
The new model for surface roughness replaces Equations 3-251 and 3-252 with:
E  n E 3
 -------  -------
1  E2  E2
------------ = ----------------------------- + ----------------------------- 3-259
 sr n DELN · CVT FELN · CVT

E p E 3
 -------  -------
1  E2  E2
------------ = ----------------------------- + ----------------------------- 3-260
 sr p DELP · CVT FELP · CVT

180 ATLAS User’s Manual

Physical Models Physics

where:
ALN · CVT   n + p 
 n = AN · CVT + -------------------------------------------------- 3-261
N  ETAN · CVT
 ------
 N 2

ALP · CVT   n + p 
 p = AP · CVT + ------------------------------------------------- 3-262
N  ETAP · CVT
 ------
 N 2

Here, N is the total doping density (ND + NA), N2 is 1cm-3, and E2 is 1 V/cm.
Table 3-52 shows the default parameters for the new equations.

Table 3-52 Default Parameters for the Surface Roughness Components of New CVT Model

Parameter Statement Default Units

AN.CVT MOBILITY 2.58

AP.CVT MOBILITY 2.18
ALN.CVT MOBILITY 6.8510-21 cm3
ALP.CVT MOBILITY 7.8210-21 cm3
ETAN.CVT MOBILITY 0.0767
ETAP.CVT MOBILITY 0.123

Next, in the Darwish model, the expressions for acoustic phonon scattering are modified as
shown in Equations 3-263 and 3-264.

  N
TAUN.CVT
 CN.CVT  ------ N2  
 AC n = ----------------------------------- +  --------------------------------------------------------------------- -----
BN.CVT 1
3-263
E  EN.CVT  E DN.CVT  
 -------   ------

-  Tn
 E2   E2 

 N  TAUP.CVT
 ------
 CP.CVT  N  
 AC p = ----------------------------------- +  ------------------------------------------------------------------- -----
BP.CVT 2 1
3-264
E  EP.CVT  E  DP.CVT
 -------
 
 -------   Tp
 E2   E2 

where
 KAPPAN.CVT
T n =  T L  300  3-265

181 ATLAS User’s Manual

Physical Models Physics

 KAPPAP.CVT
T p =  T L  300  3-266

Table 3-53 shows the default values for KAPPAN.CVT and KAPPAP.CVT.

Table 3-53 Default Parameter Values for Equations 3-265 and 3-266

Statement Parameter Default Units

MOBILITY KAPPAN.CVT 1.7

MOBILITY KAPPAP.CVT 0.9

Finally, the default values for many of the parameters are changed from those listed Table 3-
51. Table 3-54 lists the defaults used for the Darwish model.

Table 3-54 Default Parameters for the Darwish model.

Statement Parameter Default Units

MOBILITY BN.CVT 3.61107 cm/s

MOBILITY BP.CVT 1.51107 cm/s

MOBILITY CN.CVT 1.7104

MOBILITY CP.CVT 4.18103
MOBILITY TAUN.CVT 0.0233
MOBILITY TAUP.CVT 0.0119
MOBILITY DELN.CVT 3.581018 cm2/(V·s)
MOBILITY DELP.CVT 4.11015 cm2/(V·s)

You can enable the Darwish model for electrons and holes by specifying NEWCVT.N or
NEWCVT.P on the MOBILITY statement.

182 ATLAS User’s Manual

Physical Models Physics

Yamaguchi Model
The Yamaguchi Model [317] is selected by setting YAMAGUCHI in the MODELS statement. This
model overrides any mobility model specifications other than the CVT model. The model
consists of calculating the low-field, doping dependent mobility. Surface degradation is then
accounted for based upon the transverse electric field before including the parallel electric
field dependence.
The low-field part of the Yamaguchi Model is given as follows:

– 1---
2
N
 n0 = MULN.YAMA 1 + ---------------------------------------------------------------------- 3-267
N
-------------------------- + NREFN.YAMA
SN.YAMA

–1
---
2
N
 p0 = MULP.YAMA 1 + ---------------------------------------------------------------------- 3-268
N
-------------------------- + NREFP.YAMA
SP.YAMA

where Ni is the total impurity concentration. The equation parameters: MULN.YAMA,

MULP.YAMA, SN.YAMA, SP.YAMA, NREFP.YAMA, and NREFP.YAMA are user-definable in the
MOBILITY statement (see Table 3-55 for their defaults).
The transverse electric field dependence is accounted for as follows:
1
– ---
2
 s, n =  n 0  1 + ASN.YAMA E   3-269

–1
---
2
 s, p =  p 0  1 + ASP.YAMA E   3-270

where  is the perpendicular electric field and the equation parameters, ASN.YAMA and
ASP.YAMA, are user-definable in the MOBILITY statement (see Table 3-55 for their defaults).
The final calculation of mobility takes into account the parallel electric field dependence
which takes the form:

–1
---
 s n E 2  s n E  – 1   s n E  2 2
 n =  s n 1 +  -----------------------------
-  GN.YAMA + -----------------------------
- + ------------------------------ 3-271
ULN.YAMA  ULN.YAMA  VSN.YAMA

– 1---
 s p E 2  s p E  – 1   s p E  2 2
 p =  s p 1 +  -----------------------------
-  GP.YAMA + -----------------------------
- + ------------------------------ 3-272
ULP.YAMA  ULP.YAMA  VSP.YAMA

where E is the parallel electric field and the equation parameters: ULN.YAMA, ULP.YAMA,
VSN.YAMA, VSP.YAMA, GN.YAMA, and GP.YAMA are user-definable in the MOBILITY
statement (see Table 3-55 for their defaults).

183 ATLAS User’s Manual

Physical Models Physics

Table 3-55 User-Specifiable Parameters for Equations to 3-272

Statement Parameter Default Units

MOBILITY SN.YAMA 350.0

MOBILITY SP.YAMA 81.0
MOBILITY NREFP.YAMA 3.01016 cm-3
MOBILITY NREFP.YAMA 4.01016 cm-3
MOBILITY MULN.YAMA 1400.0 cm2/(V·s)
MOBILITY MULP.YAMA 480.0 cm2/(V·s)
MOBILITY ASN.YAMA 1.5410-5 cm/V

MOBILITY ASP.YAMA 5.3510-5 cm/V

MOBILITY VSN.YAMA 1.036107 cm/s

MOBILITY VSP.YAMA 1.2107 cm/s

MOBILITY ULN.YAMA 4.9106 cm/s

MOBILITY ULP.YAMA 2.928106 cm/s

MOBILITY GN.YAMA 8.8

MOBILITY GP.YAMA 1.6

The Tasch Model

S-PISCES includes an improved local field-dependent mobility model. This model, which was
originally derived and published by Tasch et. al [255,256] has been designed explicitly for
MOSFETs. It defines the mobility as a function of the perpendicular and parallel electric
fields, the interface charge, the lattice temperature and the doping concentration. To activate
this model, is activated use the TASCH parameter on the MODELS statement. This mobility
model is given by the following expressions:

d n
 n =  n +  E perp – E 0  3-273
d E perp

d p
 p =  p +  E perp – E 0  3-274
d E perp

where Eperp is the transverse electric field and E0 is the transverse electric field at the edge of
the inversion layer. The functions n,p are defined as:

184 ATLAS User’s Manual

Physical Models Physics

 eff n
 n = --------------------------------------------------------------------------------------------------
- 3-275
 E  BETAN 1  BETAN
eff 
 1 +  --------------------------------
n 
  VSATN  

 eff p
 p = ---------------------------------------------------------------------------------------------
- 3-276
 BETAP 1  BETAP
eff p E  
 1 +  -------------------------- 
-
  VSATP  

The carrier mobilities eff,n and eff,p are defined by three components ph, sr,and c that are
combined by Mathiessen’s rule according to:
1 1 1 –1
 eff = -------- + ------- + ----- 3-277
 ph  sr  c

The term ph takes account of the degradation in mobility due to acoustic phonon scattering
through the expressions:
T L – T MUBN.TAS –1 T L 1  2 –1
 ph
–1 =  MUBN.TAS  ---------
-  +  Z  DN.TAS Y  ---------
-  3-278
n   300   n n  300 

T L – T MUBP.TAS –1 T L 1  2 –1
 ph
–1 =  MUBP.TAS  ---------
-  +  Z  DP.TAS Y  ---------
-  3-279
p   300   p p  300 

The function, Zn,p, is defined as:

T L –1 –1
---
  3
Z n = Z11N.TAS ---------- E eff n + Z22N.TAS E eff n 3-280
 300

1
T L –1 – ---
Z p = Z11P.TAS  ---------- E eff p + Z22P.TAS E eff p
3
3-281
 300

where:

 E perp + RN.TAS – 1   E 0 
E eff n = ------------------------------------------------------------------------
- 3-282
RP · TAS

 E perp + RN.TAS – 1   E 0 
E eff p = ------------------------------------------------------------------------
- 3-283
RP · TAS

Also, the function Yn,p, is defined as:

T L B1N.TAS –1
+ B2N.TAS  T L 
Y n = P1N.TAS  ---------- + P2N.TAS n ---------- N 3-284
 300  300 f

185 ATLAS User’s Manual

Physical Models Physics

T L B1P.TAS –1
+ B2P.TAS T L 
Y p = P1P.TAS  ---------- + P2P.TAS p ---------- N 3-285
 300  300 f

Mobility degradation due to surface roughness is accounted for by the term sr that is
calculated according to:

ESRN.TAS BETAN.TAS
 sr n =  ------------------------------ 3-286
 E 
eff n

ESRP.TAS BETAP.TAS
 sr p =  ------------------------------ 3-287
 E 
eff p

The final term, C, models Coulombic scattering with the expressions:

T L 1.5
N2N.TAS   ----------
 300
 C n = ------------------------------------------------------------------------------------- 3-288
 BH n
N A ln  1 +  BH n  – ------------------------------ -
 1 +  BH n 

T L 1.5
N2P.TAS   ----------
 300
 C p = ------------------------------------------------------------------------------------- 3-289
 BH p
N D ln  1 +  BH p  – -------------------------------
 1 +  BH p 

Here:

N1N.TAS T L
ALPHAN.TAS
 BH n = --------------------------   ---------- 3-290
n 300

N1P.TAS T L
ALPHAP.TAS
 BH p = --------------------------   ---------- 3-291
p  300

TL is the lattice temperature in degrees Kelvin, Nf is the fixed interface charge at the gate
dielectric-silicon interface (cm-2), NA is the channel acceptor doping concentration in cm-3,
ND is the channel donor doping concentration in cm-3, n and p are the electron and hole
concentrations per unit volume in the inversion layer (cm-3). The default parameters within
each equation is defined in the MOBILITY statement. The default parameters are shown in
Table 3-55.

186 ATLAS User’s Manual

Physical Models Physics

Table 3-56 Parameters for Equations 3-273 through 3-291

Statement Parameter Default Units

MOBILITY RN.TAS 2
MOBILITY RP.TAS 3
MOBILITY BETAN 2
MOBILITY BETAP 1
MOBILITY MUBN.TAS 1150
MOBILITY MUBP.TAS 270
MOBILITY TMUBN.TAS 2.5
MOBILITY TMUBP.TAS 1.4
MOBILITY DN.TAS 3.210-9
MOBILITY DP.TAS 2.3510-9
MOBILITY P1N.TAS 0.09
MOBILITY P1P.TAS 0.334
MOBILITY B1N.TAS 1.75
MOBILITY B1P.TAS 1.5
MOBILITY P2N.TAS 4.5310-8
MOBILITY PEP.TAS 3.1410-7
MOBILITY B2N.TAS -0.25
MOBILITY B2P.TAS -0.3
MOBILITY Z11N.TAS 0.0388
MOBILITY Z11P.TAS 0.039
MOBILITY Z22N.TAS 1.7310-5
MOBILITY Z22P.TAS 1.5110-5
MOBILITY ESRN.TAS 2.449107
MOBILITY ESRP.TAS 1.0108
MOBILITY BETAN.TAS 2
MOBILITY BETAP.TAS 1

187 ATLAS User’s Manual

Physical Models Physics

Table 3-56 Parameters for Equations 3-273 through 3-291

Statement Parameter Default Units

MOBILITY N2N.TAS 1.11021

MOBILITY N2P.TAS 1.41018
MOBILITY N1N.TAS 2.01019
MOBILITY N1P.TAS 8.41016
MOBILITY ALPHAN.TAS 2
MOBILITY ALPHAP.TAS 3.4

Shirahata’s Mobility Model

The Shirahata Mobility Model [257] is a general purpose MOS mobility model that accounts
for screening effects in the inversion layer and uses an improved perpendicular field
dependence for thin gate oxides. In the original paper, the authors present the model as a
combination of portions of Klaassen’s model for low-field mobility contributions and an
empirically fit expression for the perpendicular field dependent mobility in the inversion
layer. In this implementation, if the Klaassen Low-Field Model is used with the Shirahata
model, the lattice scattering term in the Klaassen model is omitted and Matthiesen’s rule is
used to combine the Klaassen and Shirahata mobilities. But for other low-field models, at any
given location, the lesser of the low-field and the Shirahata mobilities are used.
To enable the Shirahata Mobility Model, use the SHI parameter of the MODELS statement. You
can also enable it individually for electrons and holes using the SHI.N and SHI.P parameters
of the MOBILITY statement.
The Shirahata models for electrons and holes are given by:

T L –THETAN.SHI
MU0N.SHI  ----------
 300
 n = ------------------------------------------------------------------------------------------------------------------------------------- 3-292
E P1N.SHI E P2N.SHI
1 + ------------------------- - + --------------------------
E1N.SHI E2N.SHI

T L –THETAP.SHI
MU0P.SHI  ----------
300
 p = ------------------------------------------------------------------------------------------------------------------------------------- 3-293
E P1P.SHI E P2P.SHI
1 + -------------------------- + --------------------------
E1P.SHI E2P.SHI

where E is the perpendicular electric and the equation parameters: MU0N.SHI, MU0P.SHI,
E1N.SHI, E1P.SHI, E2N.SHI, E2P.SHI, P1N.SHI, P1P.SHI, P2N.SHI, P2P.SHI,
THETAN.SHI, and THETAP.SHI are user-definable in the MOBILITY statement (see Table 3-57
for their defaults).

188 ATLAS User’s Manual

Physical Models Physics

Table 3-57 User-Specifiable Parameters for Equations 3-292 and 3-293

Statement Parameter Default Units

MOBILITY MU0N.SHI 1430.0 cm2/(V·s)

MOBILITY MU0P.SHI 500.0 cm2/(V·s)
MOBILITY E1N.SHI 6.3103 V/cm

MOBILITY E1P.SHI 8.0103 V/cm

MOBILITY E2N.SHI 0.77106 V/cm

MOBILITY E2P.SHI 3.9105 V/cm

MOBILITY P1N.SHI 0.28

MOBILITY P1P.SHI 0.3
MOBILITY P2N.SHI 2.9
MOBILITY P2P.SHI 1.0
MOBILITY THETAN.SHI 2.285
MOBILITY THETAP.SHI 2.247

Note: If the maximum low-field mobility has been user-defined, then it’s important to also define this value inside
the Shirahata model with the MU0N.SHI and MU0P.SHI parameters in the MOBILITY statement.

Remote Coulomb Scattering Mobility Model

One problem with simulation of FET devices constructed with high-k gate dielectrics is the
observed mobility reduction in such devices. The principal contributions to this reduction in
mobility are believed to be due to remote Coulomb scattering [273] and remote phonon
scattering (discussed in the next section) [336].
The analytic forms of the remote coulomb scattering for electrons and holes are given in
Equations 3-294 and 3-295.

GAMMAN · RCS + DELTAN · RCS ln  ----------------------

N
 16
ALPHAN · RCS T BETAN · RCS  N s  3  10
 N
 rcsn = MUN RCS ----------------------  ----------  -----------
- 3-294
·  16  300  10 11
3  10

GAMMAP · RCS + DELTAP · RCS ln  ----------------------

N
 16
ALPHAP · RCS T BETAP · RCS  N s  3  10
 rcsp = MUP RCS  ----------------------  ----------
N
   300  -----------
- 3-295
· 16  10 11
3  10

where N is the doping concentration, T is the temperature in K, and Ns is the inversion layer
carrier density. The user-definable parameters are given in Table 3-58.

189 ATLAS User’s Manual

Physical Models Physics

Table 3-58 User Definable Parameters of the Remote Coulomb Scattering Mobility Model

Parameter Statement Default Units

MUN.RCS MOBILITY 240.0 cm2/(V*s)

MUP.RCS MOBILITY 240.0 cm2/(V*s)
ALPHAN.RCS MOBILITY -0.3
ALPHAP.RCS MOBILITY -0.3
BETAN.RCS MOBILITY 2.1
BETAP.RCS MOBILITY 2.1
GAMMAN.RCS MOBILITY 0.4
GAMMAP.RCS MOBILITY 0.4
DELTAN.RCS MOBILITY 0.035
DELTAP.RCS MOBILITY 0.035

Remote Phonon Scattering Mobility Model

Along with the remote Coulomb scattering mobility model, the remote phonon scattering
mobility model is suggested to help describe the mobility reduction observed in high-k gate
dielectric FET devices [336].
This model is enabled by specifying RPS.N for electrons and RPS.P for holes on the
MOBILITY statement. The analytic forms of the remote phonon scattering model are given in
Equations 3-296 and 3-297.

 E   ALPHAN · RPS T BETAN · RPS

 --------------
 rpsn = MUN · RPS  ------------------------ 3-296
 10 6 V  cm  300K

 E   ALPHAP · RPS T BETAP · RPS

 --------------
 rpsp = MUP · RPS  ------------------------ 3-297
 10 6 V  cm  300K

where N is the doping concentration, E  is the perpendicular field, and T is the temperature in
K. The user-definable parameters are given in Table 3-57.

190 ATLAS User’s Manual

Physical Models Physics

Table 3-59 User Definable Parameters of the Remote Coulomb Scattering Mobility Model

Default Alternate
Parameter Statement Units
(HfO2) (SiO2)

MUN.RPS MOBILITY 216.9 323.8 cm2/(V*s)

MUP.RPS MOBILITY 216.9 323.8 cm2/(V*s)
ALPHAN.RPS MOBILITY -0.415 -0.505
ALPHAP.RPS MOBILITY -0.415 -0.505
BETAN.RPS MOBILITY -0.898 -1.98
BETAP.RPS MOBILITY -0.898 -1.98

This table includes some alternate values for SiO2 that are suggested by the reference. In
general, the model parameters are material dependent and may require calibration.
Perpendicular Electric Field-Dependent Mobility
Perpendicular Field Mobility Model
Mobility degradation effects due to perpendicular fields can be accounted for using a simple
perpendicular field dependent mobility model. To enable this model described by Equations
3-298 and 3-299, specify PRPMOB on the MODELS statement.
 0n
 n = GSURFN  ------------------------------------ 3-298
E
1 + ---------------------
ECN · MU

 0n
 p = GSURFP  ------------------------------------ 3-299
E
1 + ---------------------
ECN · MP

The default values for the user specifiable parameters of Equations 3-298 and 3-299 are
described in Table 3-60.

Table 3-60 User Specifiable Parameters of the Perpendicular Field Mobility Model

Statement Parameter Default Units

MOBILITY GSURFN 1.0

MOBILITY GSURFP 1.0
MOBILITY ECN.MU 6.48104 V/cm
MOBILITY ECP.MU 1.87104 V/cm

191 ATLAS User’s Manual

Physical Models Physics

The Watt Model

A surface mobility model derived by J.T.Watt [299] is available in ATLAS. This mobility
model is activated when the parameter SURFMOB is specified on the MODELS statement. The
default model parameters are tuned to describe the measured mobilities in silicon at 300K.
You can modify model parameters by using the MOBILITY statement.
The Watt model takes into consideration the following primary scattering mechanisms in the
inversion layer:
1. Phonon scattering which results primarily from the interaction between two-dimensional
inversion layer carriers and bulk phonons.
2. Surface roughness scattering caused by the interaction between inversion layer carriers
and deviations from ideal planarity at the interface.
3. Charged impurity scattering caused by the interaction between inversion layer carriers
and ions located in the oxide, at the interface, or in the bulk
The phonon and surface roughness components are functions of effective electric field. The
charged impurity component is a function of the channel doping density.
The effective mobilities for electrons and holes are given by Equations 3-300 and 3-301.

1 1  10 6  AL1N.WATT
----------------- = -----------------------------------------  -------------- 3-300
 eff, n MREF1N.WATT  E eff n

1  10 6  AL2N.WATT
+ -----------------------------------------  --------------
MREF2N.WATT  E eff n
–1
1  10 18  10 12 AL3N.WATT
+ -----------------------------------------  ------------  ------------
MREF3N.WATT  N B   Ni 

1 1  10 6  AL1P.WATT
----------------- = -----------------------------------------  -------------- 3-301
 eff, p MREF1P.WATT  E eff p

1  10 6  AL2P.WATT
+ -----------------------------------------  --------------
MREF2P.WATT  E eff p

1  10 18 – 1  10 12 AL3N.WATT

+ -----------------------------------------  ------------  ------------
MREF3P.WATT  N B   N i 

Here, NB is the surface trapped charge density, Ni is the inversion layer charge density and
Eeff is the effective electric field given by:

E eff n = E  + ETAN.WATT  E 0 – E   3-302

E eff p = E  + ETAP.WATT  E 0 – E   3-303

Here, E is the electric field perpendicular to the current flow and E0 is the perpendicular
electric field at the insulator-semiconductor interface. The equation parameters and their
defaults are listed in Table 3-61.

192 ATLAS User’s Manual

Physical Models Physics

The terms on the right side of Equations 3-300 and 3-301, describe (in order) the three
scattering mechanisms previously discussed. Each component contains two constants: a pre-
exponential factor and the exponent of the principal independent parameter. The charge
impurity scattering component is assumed to be inversely proportional to doping density.
The expression for effective mobility contains a number of normalizing constants. These
normalizing constants are included to allow easy comparison of constants. The first two terms
in this mobility model are dependent on Eeff and represent the universal mobility-field
relationship. The third term accounts for deviation from the universal relationship resulting
from charged impurity scattering.

Table 3-61 User-Specifiable Parameters for Equations 3-300 - 3-303

Statement Parameter Default Units

MOBILITY ETAN.WATT 0.50

MOBILITY ETAP.WATT 0.33
MOBILITY MREF1N.WATT 481.0 cm2/(V·s)
MOBILITY MREF1P.WATT 92.8 cm2/(V·s)
MOBILITY MREF2N.WATT 591.0 cm2/(V·s)
MOBILITY MREF2P.WATT 124.0 cm2/(V·s)
MOBILITY MREF3N.WATT 1270.0 cm2/(V·s)
MOBILITY MREF3P.WATT 534.0 cm2/(V·s)
MOBILITY AL1N.WATT -0.16
MOBILITY AL1P.WATT -0.296
MOBILITY AL2N.WATT -2.17
MOBILITY AL2P.WATT -1.62
MOBILITY AL3N.WATT 1.07
MOBILITY AL3P.WATT 1.02

Modifications to the Watt’s Model

By default the Watt mobility model is a surface model that applies, only to those grid points
on the silicon/ oxide interface. A modification has been added that now applies the Watt
model to points below the interface. This extension to the Watt model is enabled using the
MOD.WATT.N and MOD.WATT.P parameters of the MOBILITY statement.
The distance over which the model is applied can be controlled by using the YMAXN.WATT and
the YMAXP.WATT parameters of the MOBILITY statement. These parameters specify the
maximum value of the Y coordinate over which the model is applies below the interface for
electrons and holes respectively.

193 ATLAS User’s Manual

Physical Models Physics

The XMINN.WATT, XMINP.WATT, XMAXN.WATT and XMAXP.WATT of the MOBILITY statement

can be used to limit the range of the model in the X direction, to prevent the model from
applying to the source and drain regions. The MIN.SURF parameter of the MODELS statement
can also be used for this purpose. When enabled, the MIN.SURF parameter will ensure that the
Watt model will only apply to minority regions.
The logical parameters EXP.WATT.N and EXP.WATT.P of the MOBILITY statement can also be
used to enable an additional modification to the Watt model. When these parameters are
enabled the effective normal electric field becomes a function of the depth beneath the
silicon/oxide interface according to:
–  y – y int 
E  n = E y exp ----------------------------------------
- 3-304
YCHARN.WATT

–  y – y int 
E  p = E y exp ----------------------------------------
- 3-305
YCHARP.WATT

where E is the perpendicular electric field, Ey is the perpendicular electric field at the
interface, y is the local Y coordinate and yint is the Y coordinate of the silicon/oxide interface.
The YCHARN.WATT and YCHARP.WATT parameters are user-definable in the MOBILITY
statement.

Table 3-62 User-Specifiable Parameters for Equations 3-194 and 3-195

Statement Parameter Default Units

MOBILITY XMINN.WATT -1.01032 microns

MOBILITY XMAXN.WATT 1.01032 microns

MOBILITY YMAXN.WATT -1.01032 microns

MOBILITY XMINP.WATT -1.01032 microns

MOBILITY XMAXP.WATT 1.01032 microns

MOBILITY YMAXP.WATT -1.01032 microns

MOBILITY YCHARN.WATT 1.01032 microns

MOBILITY YCHARP.WATT 1.01032 microns

194 ATLAS User’s Manual

Physical Models Physics

Parallel Electric Field-Dependent Mobility

Saturation Velocity Model
As carriers are accelerated in an electric field their velocity will begin to saturate when the
electric field magnitude becomes significant. This effect has to be accounted for by a
reduction of the effective mobility since the magnitude of the drift velocity is the product of
the mobility and the electric field component in the direction of the current flow. The
following Caughey and Thomas Expression [41] is used to implement a field-dependent
mobility. This provides a smooth transition between low-field and high field behavior where:
1
------------------
1 BETAN
 n  E  =  n0 ----------------------------------------------------- 3-306
 n0 E BETAN
1 +  ------------------
 VSATN

1 -
-----------------
1 BETAP
 p  E  =  p0 ----------------------------------------------------- 3-307
 p0 E BETAP
1 +  ------------------
 VSATP

Here, E is the parallel electric field and n0 and p0 are the low-field electron and hole
mobilities respectively. The low-field mobilities are either set explicitly in the MOBILITY
statement or calculated by one of the low-field mobility models. The BETAN and BETAP
parameters are user-definable in the MOBILITY statement (see Table 3-63 for their defaults).
The saturation velocities are calculated by default from the temperature-dependent models
[246]:
ALPHAN.FLD
VSATN = -------------------------------------------------------------------------------------------------
TL
1 + THETAN.FLD exp  ---------------------------------
 TNOMN.FLD 3-308

ALPHAP.FLD
VSATP = ------------------------------------------------------------------------------------------------- 3-309
TL
1 + THETAP.FLD exp  ---------------------------------
 TNOMP.FLD

But, you can set them to constant values on the MOBILITY statement using the VSATN and
VSATP parameters.
In this case, no temperature dependence is implemented. Specifying the FLDMOB parameter on
the MODELS statement invokes the field-dependent mobility. FLDMOB should always be
specified unless one of the inversion layer mobility models (which incorporate their own
dependence on the parallel field) are specified.
You can invoke a C-INTERPRETER function for the saturation velocities. The F.VSATN and
F.VSATP parameters in the MATERIAL statement can be set to provide the filenames of two
text files containing the particular functions. These functions allow you to include the
temperature dependence. See Appendix A “C-Interpreter Functions” for more details.

195 ATLAS User’s Manual

Physical Models Physics

Canali Modification
The Canali model [39] has been implemented as an alternative to using fixed values of BETAN
and BETAP in the Caughey-Thomas model. This uses the exponent BETA, which depends on
lattice temperature (TL), and will calculate the values of BETAN and BETAP as

 N · BETAEXP 
BETAN = N · BETA0  TL  300  3-310

 P · BETAEXP 
BETAP = P · BETA0  TL  300  3-311

The Canali model can be used for Silicon up to 430 K.

To enable the model, use the N.CANALI and P.CANALI parameters on the MOBILITY
statement. You should also set the FLDMOB flag on the MODELS statement. The EVSATMOD and
HVSATMOD should also have their default values of 0.

Driving Force for FLDMOB Model

It has been demonstrated that using the electric field as the driving force for field dependent
mobility models results in some inaccuracy in semiconductors with strongly inhomogeneous
fields. It has been suggested that using the gradient of quasi-fermi level in place of field can
lead to improved results. Additionally, it has been shown that using the square root of the
product of electric field and gradient of quasi-fermi level as driving force is an even better
approximation [329]. You have the option to select different quantities to replace E in
Equations 3-306 and 3-307 for the FLDMOB model and for the Canali extension of the FLDMOB
model. You specify VSAT.QFN on the MOBILITY statement to use a driver
 n 3-312

instead of E for the electron mobility, where  n is the Electron quasi-Fermi potential. You
specify VSAT.QFP on the MOBILITY statement to use a driver
 p 3-313

for the hole mobility, where  p is the Hole quasi-Fermi potential. You specify VSAT.ZAKN on
the MOBILITYstatement to use a driver

 n E 3-314

and VSAT.ZAKP on the MOBILITY statement to use a driver

 p E 3-315

196 ATLAS User’s Manual

Physical Models Physics

Table 3-63 User-Definable Parameters in the Field-Dependent Mobility Model

Statement Parameter Default Units

MOBILITY ALPHAN.FLD 2.4107 cm/s

MOBILITY ALPHAP.FLD 2.4107 cm/s

MOBILITY BETAN 2.0

MOBILITY BETAP 1.0
MOBILITY N.BETAEXP 0.66
MOBILITY N.BETA0 1.109
MOBILITY N.CANALI False
MOBILITY P.BETAEXP 0.17
MOBILITY P.BETA0 1.213
MOBILITY P.CANALI False
MOBILITY THETAN.FLD 0.8
MOBILITY THETAP.FLD 0.8
MOBILITY TNOMN.FLD 600.0 K
MOBILITY VSATN cm/s
MOBILITY VSATP cm/s
MOBILITY VSAT.QFN False
MOBILITY VSAT.QFP False
MOBILITY VSAT.ZAKN False
MOBILITY VSAT.ZAKP False
MOBILITY TNOMP.FLD 600.0 K

Note: Equation 3-309, which was derived for the drift-diffusion approximation, ensures that velocity overshoot
cannot occur. To model velocity overshoot in silicon, apply the Energy Balance Transport Model. This model
follows the above implementation but with the electric field term replaced by a new “effective” field calculated
from the carrier temperature see the following section for more details.

Note: BLAZE includes a different field dependent mobility model that does simulate velocity overshoot in GaAs.
See Chapter 5 “Blaze: Compound Material 2D Simulator” for more information.

197 ATLAS User’s Manual

Physical Models Physics

Carrier Temperature Dependent Mobility

Energy Balance Model
The Energy Balance Transport Model allows the carrier mobility to be related to the carrier
energy. This has been achieved through the homogeneous steady state energy balance
relationship that pertains in the saturated velocity limit. This allows an effective electric field
to be calculated, which is the uniform electric field value that causes the carriers in a
homogeneous sample to attain the same temperature as at the node point in the device. The
effective electric fields, Eeff,n and Eeff,p, are calculated by solving the equations:

3 k  Tn – TL 
q n  E eff n E eff
2
 n = --
- --------------------------------- 3-316
2 TAUMOB.EL

3 k  Tp – TL 
q p  E eff p E eff
2
 p = --
- --------------------------------- 3-317
2 TAUMOB.HO

for Eeff,n and Eeff,p. These equations are derived from the energy balance equations by
stripping out all spatially varying terms. The effective electric fields are then introduced into
the relevant field dependent mobility model. The TAUMOB.EL and TAUMOB.HO parameters can
be set on the MODELS statement and have the default values shown in Table 3-64.

Note: The TAUMOB.EL and TAUMOB.HO parameters are distinct from the energy balance relaxation time
parameters: TAUREL.EL and TAUREL.HO. But, if the E.TAUR.VAR and H.TAUR.VAR parameters
are set on the MODELS statement, then TAUREL.EL and TAUREL.HO are used in place of
TAUMOB.EL and TAUMOB.HO.

Table 3-64 User-Specifiable Parameters for Equations 3-316 and 3-317

Statement Parameter Default Units

MODELS TAUMOB.EL 2.510-13 s

MODELS TAUMOB.HO 2.510-13 s

Four different models have been implemented into the ATLAS Energy Balance Transport
Model which can be chosen by the parameter EVSATMOD on the MODELS statement. These
models shall be described next.
Setting EVSATMOD=0 implements, the default model for silicon based upon the Caughey-
Thomas field-dependent mobility model in Equation 3-306. The resultant relationship
between the carrier mobility and the carrier temperature is in the forms:
 n0
 n = --------------------------------------------------------
- 3-318
1 -
-----------------
BETAN
 1 + X BETAN
 n 

198 ATLAS User’s Manual

Physical Models Physics

 p0
 p = --------------------------------------------------------- 3-319
1
------------------
BETAP
 1 + X BETAP
 p 

BETAN 1 BETAN
Xn = ---   nBETAN  T n – TL  3-320
2

2 BETAN BETAN BETAN 

+  2 BETAN  T – T  – 4 T – T  
n n L n n L 

BETAP 1 BETAP
X = ---   ^ BETAP  T n – TL  3-321
p 2 p
2 BETAP BETAP BETAP
+  2 BETAP  T – TL  – 4 T – T 
p p p p L

k B  n0
n = 3
--- ------------------------------------------------------------------
- 3-322
2 qVSATN2  TAUREL EL 
·

3 k B  p0
 p = --- ----------------------------------------------------------------- 3-323
2 2
qVSATP  TAUREL · HO 

where n0 and p0 are the low-field carrier mobilities and VSATN and VSATP are the saturated
velocities for electrons and holes. The VSATN, VSATP, BETAN, and BETAP parameters are user-
definable in the MOBILITY statement. The terms, TAUREL.EL and TAUREL.HO, are the energy
relaxation times for electrons and holes and can be defined in the MATERIAL statement.
Setting EVSATMOD=0 with the additional parameter, MOBTEM.SIMPL, allows you to apply a
simplified form of the above model. This form is:
 n0
 n = ------------------------------------------------
- 3-324
2 2
1 + n  Tn – TL 

 p0
 p = ------------------------------------------------
- 3-325
2 2
1 + p  Tp – TL 

where n0 and p0 are again the low-field carrier mobilities and an,p are as defined above.
Setting EVSATMOD=1 implements the GaAs carrier temperature dependent mobility model.
See Chapter 5 “Blaze: Compound Material 2D Simulator” for more information about this
model.
Setting EVSATMOD=2 will apply the simple velocity limiting model based upon the electric
field. In other words, the temperature dependent mobility is turned off and the standard
electric field based mobility model is applied.

199 ATLAS User’s Manual

Physical Models Physics

Note: If the YAMAGUCHI or TASCH mobility models are chosen in the MODEL statement, then no energy
dependence is applied. No energy dependence is included in any perpendicular electric field model, such as
SHIRAHATA or SURFMOB.

Table 3-65 User-Specifiable Parameters for Equations 3-318– 3-325

Statement Parameter Units

MOBILITY MUN cm2/(Vs)

MOBILITY MUP cm2/(Vs)
MATERIAL VSATN cm/s
MATERIAL VSATP cm/s

Meinerzhagen-Engl Model
The Meinerzhagen-Engl model is another velocity saturation model in which the mobility
depends on the carrier temperature [185].
 n0
 n = -------------------------------------------------------
- 3-326
  1
 1 + n  Tn – TL  

 n0
 p = -------------------------------------------------------
- 3-327
  1
 1 + p  Tp – TL  

where n and p are the same as in Equations 3-322 and 3-323. If =2, then you have the
same result as in Equations 3-324 and 3-325. The exponent , however, may now depend on
lattice temperature (TL).
As in the Canali model, the values of BETAN and BETAP are calculated as

 N · BETAEXP 
BETAN = N · BETA0  TL  300  3-328

 P · BETAEXP 
BETAP = P · BETA0  TL  300  3-329

although the default values differ from the Canali model (see Table 3-66).
To activate the model, specify FLDMOB on the MODELS statement and additionally N.MEINR or
P.MEINR or both on the MOBILITY statement. You must at least solve for one of the electron
or hole temperatures. The exponent  is constant unless you enable Lattice temperature too.

200 ATLAS User’s Manual

Physical Models Physics

Table 3-66 User Specifiable Parameters for Meinerzhagen-Engl Model

Statement Parameter Default

MOBILITY N.MEINR False

MOBILITY P.MEINR False
MOBILITY N.BETA0 0.6
MOBILITY P.BETA0 0.6
MOBILITY N.BETAEXP 0.01
MOBILITY P.BETAEXP 0.01

Complete C-interpreter Functions for Mobility

ATLAS provides some C-Interpreter functions that allows you to completely specify the
mobility. These can be considered as stand-alone mobility specifications as they are not
combined with other mobility parameters.
First, there is a C-Interpreter function that allows the input of a mobility function, which
depends explicitly on position. The parameters supplied to the function include the (X, Y, Z)
coordinates of each node point, acceptor and donor concentrations, electric field, electrical
potential, lattice temperature and carrier temperature.
For electron mobility, use the F.LOCALMUN parameter on the MATERIAL or MOBILITY
statement. For hole mobility, use the F.LOCALMUP parameter on the MATERIAL or MOBILITY
statement.
There is also a C-Interpreter function that allows you to supply a mobility, which depends
explicitly on perpendicular electric field (as in a surface mobility model) and on parallel field
(as in a velocity saturation model). Other parameters that are supplied are dopant densities,
carrier densities, lattice temperature and composition fractions. For electron mobility, use the
F.TOFIMUN parameter on the MATERIAL or MOBILITY statement. For hole mobility, use the
F.TOFIMUP parameter on the MATERIAL or MOBILITY statement.
For details about C-Interpreter functions, see Appendix A “C-Interpreter Functions” and the
C-Interpreter templates, which come with your ATLAS distribution.

201 ATLAS User’s Manual

Physical Models Physics

3.6.2 Mobility Model Summary

Tables 3-67 and 3-68 shows a brief description of the mobility models.

Table 3-67 Mobility Models Summary

Model Syntax Notes

Concentration Dependent CONMOB Lookup table valid at 300K for Si and GaAs only.
Uses simple power law temperature dependence.
Concentration and ANALYTIC Caughey-Thomas formula. Tuned for 77-450K.
Temperature Dependent
Arora’s Model ARORA Alternative to ANALYTIC for Si.
Carrier-Carrier Scattering CCSMOB Dorkel-Leturq Model. Includes n, N and T
dependence. Important when carrier concentration is
high (e.g., forward bias power devices).
Parallel Electric Field FLDMOB Si and GaAs models. Required to model any type of
Dependence velocity saturation effect.
Tasch Model TASCH Includes transverse field dependence. Only for planar
devices. Needs very fine grid.
Watt Model WATT Transverse field model applied to surface nodes only.
Klaassen Model KLA Includes N, T, and n dependence. Applies separate
mobility to majority and minority carriers.
Recommended for bipolar devices
Shirahata Model SHI Includes N, E. An alternative surface mobility model
that can be combined with KLA.
Modified Watt MOD.WATT Extension of WATT model to non-surface nodes.
Applies constant Eeffects. Best model for planar
MOS devices
Lombardi (CVT) Model CVT Complete model including N, T, E// and E effects.
Good for non-planar devices.
Yamaguchi Model YAMAGUCHI Includes N, E// and E effects. Only for 300K.

202 ATLAS User’s Manual

Physical Models Physics

Table 3-68 Mobility Models Summary

CONMOB FLDMOB TFLDMB2 YAMAGUCHI CVT ARORA ANALYTIC CCSMOB SURFACE LATTICE H E.BALANCE

CONMOB [CM] — OK OK YA CV AR AN CC OK OK OK

FLDMOB [FM] OK — TF1 YA CV OK OK OK OK OK OK

TFLDMB2 [TF] OK TF1 — YA CV OK OK TF TF OK OK

YAMAGUCHI [YA] YA YA YA — CV YA YA YA YA NO NO

CVT [CV] CV CV CV CV — CV CV CV CV OK OK

ARORA [AR] AR OK OK YA CV — AR CC OK OK OK

ANALYTIC [AN] AN OK OK YA OK — CC OK OK OK

CCSMOB [CC] CC OK TF YA CV CC CC — OK OK OK

SURFMOB [SF] OK OK TF YA CV OK OK OK — OK OK

LATTICE H [LH] OK OK OK NO OK OK OK OK OK — OK

E.BALANCE [EB] OK OK OK NO OK OK OK OK OK OK 2

Key to Table Entries:

MODEL ABBREVIATION = The model that supersedes when a combination is specified. In some cases, but not all, a warning message is issued
when a model is ignored.
OK = This combination is allowed.
NO = This combination isn’t allowed.
NOTES:
1. Uses internal model similar to FLDMOB.
2. When models including a parallel electric field dependence are used with energy balance the electric field term is replaced by a function of
carrier temperature.

203 ATLAS User’s Manual

Physical Models Physics

3.6.3 Carrier Generation-Recombination Models

Carrier generation-recombination is the process through which the semiconductor material
attempts to return to equilibrium after being disturbed from it. If we consider a
homogeneously doped semiconductor with carrier concentrations n and p to the equilibrium
concentrations n0 an p0 then at equilibrium a steady state balance exists according to:
2
n0 p0 = ni 3-330

Semiconductors, however, are under continual excitation whereby n and p are disturbed from
their equilibrium states: n0 and p0. For instance, light shining on the surface of a p-type
semiconductor causes generation of electron-hole pairs, disturbing greatly the minority
carrier concentration. A net recombination results which attempts to return the semiconductor
to equilibrium. The processes responsible for generation-recombination are known to fall into
six main categories:
• phonon transitions
• photon transitions
• Auger transitions
• surface recombination
• impact ionization
• tunneling
The following sections describes the models implemented into ATLAS that attempts the
simulation of these six types of generation-recombination mechanisms.
Shockley-Read-Hall (SRH) Recombination
Phonon transitions occur in the presence of a trap (or defect) within the forbidden gap of the
semiconductor. This is essentially a two step process, the theory of which was first derived by
Shockley and Read [258] and then by Hall [96]. The Shockley-Read-Hall recombination is
modeled as follows:
2
pn – n ie
R SRH = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 3-331
TAUP0 n + n ie exp  ------------------ + TAUN0 p + n ie exp  ---------------------
ETRAP – E TRAP
 kT   kT 
L L

where ETRAP is the difference between the trap energy level and the intrinsic Fermi level, TL
is the lattice temperature in degrees Kelvin and TAUN0 and TAUP0 are the electron and hole
lifetimes. This model is activated by using the SRH parameter of the MODELS statement. The
electron and hole lifetime parameters, TAUN0 and TAUP0, are user-definable in the MATERIAL
statement. The default values for carrier lifetimes are shown in Table 3-69. Materials other
than silicon will have different defaults. A full description of these parameters are given in
Appendix B “Material Systems”.

204 ATLAS User’s Manual

Physical Models Physics

Table 3-69 User-Specifiable Parameters for Equation 3-331

Statement Parameter Default Units

MATERIAL ETRAP 0 eV
MATERIAL TAUN0 110-7 s

MATERIAL TAUP0 110-7 s

Note: This model only presumes one trap level which, by default, is ETRAP=0 and it corresponds to the most
efficient recombination centre. If the TRAP statement is used to define specific trap physics then separate
SRH statistics are implemented as described earlier in “Trap Implementation into Recombination Models” on
page 119.

SRH Concentration-Dependent Lifetime Model

The constant carrier lifetimes that are used in the SRH recombination model above can be
made a function of impurity concentration [238,159,78] using the following equation:
2
pn – n ie
R SRH = ------------------------------------------------------------------------------------------------------------------------------------------------- 3-332
 p n + n ie exp  ------------------ +  n p + n ie exp  ---------------------
ETRAP – E TRAP
kT L kT L

where:
TAUN0
 n = -------------------------------------------------------------------------------------------
- 3-333
Ntotal EN
AN + BN  ------------------- + CN  -------------------
Ntotal
 NSRHN  NSRHN

TAUP0
 p = -------------------------------------------------------------------------------------------
- 3-334
Ntotal EP
AP + BP  ------------------- + CP  -------------------
Ntotal
 NSRHP  NSRHP

Here, N is the local (total) impurity concentration. The TAUN0, TAUP0, NSRHN, and NSRHP
parameters can be defined on the MATERIAL statement (see Table 3-70 for their default
values). This model is activated with the CONSRH parameter of the MODELS statement.

205 ATLAS User’s Manual

Physical Models Physics

Table 3-70 User-Specifiable Parameters for Equations 3-332 to 3-334

Statement Parameter Default Units

MATERIAL TAUN0 1.010-7 s

MATERIAL NSRHN 5.01016 cm-3

MATERIAL TAUP0 1.010-7 s

MATERIAL NSRHP 5.01016 cm-3

MATERIAL AN 1.0

MATERIAL AP 1.0

MATERIAL BN 1.0

MATERIAL BP 1.0

MATERIAL CN 0.0

MATERIAL CP 0.0

MATERIAL EN 0.0

MATERIAL EP 0.0

You can choose an alternate set of default values for the concentration dependent SRH model
as suggested by Law et.al. [159]. Specifying CONSRH.LAW on the MODEL statement will select
the use of the alternate default values shown in Table 3-71.

Table 3-71 Alternate default values for Equations 3-332 to 3-334

Statement Parameter Default Units

MATERIAL TAUN0 3010-6 s

MATERIAL TAUP0 1010-6 s
MATERIAL NSRHN 11017 cm-3
MATERIAL NSRHP 11017 cm-3

206 ATLAS User’s Manual

Physical Models Physics

Klaassen’s Concentration Dependent Lifetime Model

The Klaassen Concentration and Temperature-dependent SRH Lifetime Model [140] is
enabled by setting the KLASRH logical parameter in the MODELS statement. The lifetimes for
electrons and holes in this model are given by the equations:

300 KSRHGN
+ KSRHCN  N   ----------
–1 –1
TAUN0 =  KSRHTN 3-335
T 
L

300 KSRHGP
+ KSRHCP  N   ----------
–1 –1
TAUP0 =  KSRHTP 3-336
T 
L

Here, N is the local (total) impurity concentration. You can define the KSRHTN, KSRHTP,
KSRHCN, KSRHCP, KSRHGN, and KSRHGP parameters in the MATERIAL statement. Their default
values are given in Table 3-72.

Table 3-72 User-Specifiable Parameters for Equations 3-335 to 3-336

Statement Parameter Default Units

MATERIAL KSRHTN 2.510-3 s

MATERIAL KSRHTP 2.510-3 s

MATERIAL KSRHCN 3.010-13 cm3/s

MATERIAL KSRHCP 11.7610-13 cm3/s
MATERIAL KSRHGN 1.77
MATERIAL KSRHGP 0.57

Scharfetter Concentration Dependent Lifetime Model

This is an another model for the dependence of recombination lifetimes on total doping level.
To enable this model, use the flag SCHSRH on the MODELS statement. The lifetimes are given
by

 N · SCH · MAX – N · SCH · MIN 

 n  N  = N · SCH · MIN + ---------------------------------------------------------------------------------------------------------- 3-337
N · SCH · GAMMA
1 +  N  N · SCH · NREF

and

 P · SCH · MAX – P · SCH · MIN 

 p  N  = P · SCH · MIN + ---------------------------------------------------------------------------------------------------------- 3-338
P · SCH · GAMMA
1 +  N  P · SCH · NREF 

where N is the Total doping density. Table 3-73 shows the MATERIAL statement parameters.

207 ATLAS User’s Manual

Physical Models Physics

Table 3-73 User Specifiable Parameters for Equations 3-337 and 3-338

Statement Parameter Units Default

MATERIAL N.SCH.MIN s 0.0

MATERIAL P.SCH.MIN s 0.0
MATERIAL N.SCH.MAX s 1.010-5
MATERIAL P.SCH.MAX s 3.010-6
MATERIAL N.SCH.GAMMA s 1.0
MATERIAL P.SCH.GAMMA s 1.0
MATERIAL N.SCH.NREF cm-3 1.01016
MATERIAL P.SCH.NREF cm-3 1.01016

Coupled Defect Level Recombination Model

This model is a modification of the SRH model to the situation where there is charge transfer
between two defect levels. This can lead to large excess currents in devices. The model has
explained some anomalous diode characteristics [241].
Figure 3-6 shows the assumptions about the trap levels.

Figure 3-6: Trap Level Assumptions

If there is no coupling between the two defect levels, then the recombination total
recombination rate is just the sum of two SRH terms, R1 + R2, where
2
np – n i
R 1 = -------------------- 3-339
r1

208 ATLAS User’s Manual

Physical Models Physics

and
2
np – n i
R 2 = -------------------- 3-340
r2

where the denominators are

r k =  nk  p + p k  +  pk  n + n k ?Colon?k = 1, 2 3-341

Assuming Boltzmann statistics apply, the quantities n1, n2, p1 and p2 are given by

Et 1 
n 1 = n i exp  --------- 
 kT  

–Et 1 

p 1 = n i exp ------------ 
 kT 
 3-342
Et 2 
n 2 = n i exp  --------- 
 kT 

–Et 2 
p 2 = n i exp  ------------ 
 kT  

If r12 is non-zero, then the total recombination rate is given by

2
R = R 1 + R 2 +  R 12 – S 12 – R 12  3-343

CDL · TN1  CDL · TP2  n + n 2   p + p 1  – CDL · TN2  C DL · TP1  n + n 1   p + p 2 

 -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
r1 r2

where

r1 r2
R 12 = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
- 3-344
2  CDL · COUPLING  CDL · TN1  CDL · TN2  C DL · TP1  CDL · TP2  1 –  
CDL · TN1 p + p 1  + CDL · TP2  n + n 2 
+ -----------------------------------------------------------------------------------------------------
2CDL · TN1  CDL · TP2  1 –  
  CDL · TN2  p + p 2  + CDL · TP1  n + n 1  
+ --------------------------------------------------------------------------------------------------------------
2CDL · TN2  CDL · TP1 1 –  

and

1  CDL · TN1  CDL · TP2  2

-  np – n i 
S 12 = -------------------------------------------------------------------------  1 – ---------------------------------------------------------- 3-345
CDL · TN1  CDL · TP2  1 –    CDL · TN2  CDL · TP1 

209 ATLAS User’s Manual

Physical Models Physics

The factor  is given by

– E –E 
 = exp  ----------------------------------
t2 t1 

3-346
KB T

and will generally be less than 1. It can be shown that as r12 becomes small, the total
recombination rate simplifies to R1 + R2 as required.
You can set all the relevant parameters in ATLAS. To set the rate r12, use CDL.COUPLING on
the MATERIAL statement. To set the carrier lifetimes, use CDL.TN1, CDL.TN2, CDL.TP1 and
CDL.TP2 on the MATERIAL statement. To set the energies of the defect levels relative to the
intrinsic level, use CDL.ETR1 and CDL.ETR2 on the MATERIAL statement.
With reference to the energies shown in Figure 3-6, CDL.ETR1 will be positive and CDL.ETR2
will be negative. Table 3-74 shows these quantities and the units and default values.
To enable the model, use either the CDL parameter or the CONCDL parameter on the MODELS
statement. If you set CDL, then the values used for the lifetimes are those set by CDL.TN1,
CDL.TN2, CDL.TP1 and CDL.TP2. If you set CONCDL, then the lifetimes used are doping
dependent and are derived from Equations 3-333 and 3-334 with the lifetimes being
CDL.TN1, CDL.TN2, CDL.TP1 and CDL.TP2, instead of TAUN0 and TAUP0. The default values
for the other parameters are shown in Table 3-70.
SRH and CDL are mutually exclusive parameters. If you enable CDL, the CDL recombination
rate is output to a structure file in the SRH slot and are added to the output of total
recombination rate.

Table 3-74 Material Parameters for Coupled Defect Level Model

Statement Parameter Type Units Default

MATERIAL CDL.ETR1 Real eV 0.0

MATERIAL CDL.ETR2 Real eV 0.0
MATERIAL CDL.TN1 Real s 1.0
MATERIAL CDL.TN2 Real s 1.0
MATERIAL CDL.TP1 Real s 1.0
MATERIAL CDL.TP2 Real s 1.0
MATERIAL CDL.COUPLING Real s-1 cm-3 1.0

The individual lifetimes cannot be output to a structure file.

210 ATLAS User’s Manual

Physical Models Physics

Trap Assisted Auger Recombination

This model adds in a dependence of recombination lifetime on carrier density, and will only
be significant at fairly high carrier densities [79].
The carrier lifetimes are reduced according to the following formula
n
 n = ------------------------------------------------------------
- 3-347
 1 + TAA · CN n + p  n 

p
 p = ------------------------------------------------------------
- 3-348
 1 + TAA · CP n + p  p 

where n is the electron density and p the hole density. To enable the model, specify
TRAP.AUGER on the MODELS statement. It will then apply to the SRH model if enabled or to
the CDL model if enabled.
You can set the parameters on the MATERIAL statement with the defaults as shown in Table 3-
75.
.

Table 3-75 Parameters for Trap assisted Auger model

Statement Parameter Units Default

MATERIAL TAA.CN cm3/s 1.010-12

MATERIAL TAA.CP cm3/s 1.010-12

Trap-Assisted Tunneling
In a strong electric field, electrons can tunnel through the bandgap via trap states. This trap-
assisted tunneling mechanism is enabled by specifying TRAP.TUNNEL on the MODELS
statement and is accounted for by modifying the Schockley-Read-Hall recombination model.
2
pn – n ie
R SRH = ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 3-349
----------------------------------- n + n ie exp  ------------------ + ----------------------------------- p + n ie exp  ---------------------
TAUP0 ETRAP TAUN0 – E TRAP
DIRAC  kT  DIRAC  kT 
1 + p L 1 + n L

Here,nDIRAC is the electron field-effect enhancement term for Dirac wells, and pDIRAC is
the hole field-effect enhancement term for Dirac wells. nDIRAC and pDIRAC are defined in
Equations 3-101 and 3-102.

211 ATLAS User’s Manual

Physical Models Physics

Schenk Model for Trap Assisted Tunneling [243]

The Schenk model gives the field-effect enhancement factors as an analytic function. Before
using this analytic function, however, ATLAS works out several intermediate quantities. The
first is an effective trap level Et given by
mc 3 3 14
E t = --- Eg  T  + --- kT ln  ------ + ETRAP –  32R c h  
1 3
3-350
2 4  m v

for electrons
mc 3 3 14
E t = --- Eg  T  – --- kT ln  ------ – E TRAP –  32R v h  
1 3
3-351
2 4  m v

for holes. The quantities Rc and Rv are the effective Rydberg energies given by

Z · SCHENK - 2
R c = m c  --------------------------------------------  13.6eV 3-352
 PERMITTIVITY

Z · SCHENK  2
R v = m v  --------------------------------------------
-  13.6eV 3-353
 PERMITTIVITY

The quantity  is known as the electro-optical frequency and is given as

 q F
2 2 1  3
 =  -----------------------------------------------
 2  m o ME · TUNNEL
3-354
 q F
2 2 1  3
 =  -----------------------------------------------
 2  m o MH · TUNNEL

where F is the electric field. If you do not specify ME.TUNNEL, then a value of 0.5*(MT1 +
MT2) will be used for silicon and the density of states effective electron mass for other
materials. If you do not specify MH.TUNNEL, then the density of states effective hole mass will
be used.
The next quantity is the lattice relaxation energy given by

 R = HUANG · RHYS  PHONON · ENERGY 3-355

From this is derived the quantity F given by

2
 2 R kT 
 F = ----------------------- 3-356
3


ATLAS then calculates the optimum horizontal transition energy, E0. This is the energy at
which the tunneling probability for carriers to tunnel from the trap level to the band edge has
a maximum. It is given by

E o = 2  F  F + E t +  R – 2 F –  R 3-357

212 ATLAS User’s Manual

Physical Models Physics

with the maximum permitted value of Eo being

2
 ET + R   E T +  R 
E 0 = E t – --------------------------  1 – ----------------------
- 3-358
4 F  2 F 

when E T +  R «  F and the minimum permitted value of Eo, when F is small, being

E 0 =  R exp  – --------------------------------------------
PHONON.ENERGY
3-359
 KT 

Equation 3-358 represents the low field limit and Equation 3-359 represents the high field
limit.
The field enhancement factor for electrons is then given by
32 –1  2 34 14
 Et – Eo   Et – Eo  32
 n = 1 + -------------------------------------------------------
- -  h
----------------------------------------------------- --------
2 Et Eo  kT 
E o PHONON · ENERGY
3-360
– Et – Eo  PHONON · ENERGY – kT-
 exp ------------------------------------------------ + -------------------------------------------------------------
PHONON · ENERGY 2PHONON · ENERGY

E t + kT  2 E o ln  E o   R 
+ ------------------------------------------------ ln  E t   R  – ------------------------------------------------
PHONON · ENERGY PHONON · ENERGY

Et – Eo 32
4 Et – Eo
 exp  -------------------- exp – ---  --------------------
 kT  3  kT 

where ET is calculated by using Equation 3-350. The field enhancement factor for holes is
obtained using Equation 3-360 but with Et obtained from Equation 3-351. The default effective
masses used in the above calculation are correct for Silicon with the electric field in the
<100> direction. They are also accurate for the <110> direction, but for the <111> direction a
value of
 1 1-
1 - + ------
–1
m c = 3  ---------
- + ---------  3-361
 m t 1 m t 2 m l

should be used for electrons. To enable the model, specify SCHENK.TUN on the MODELS
statement and the parameters shown in Table 3-76.

Table 3-76 User Definable Parameters for Schenk Trap Assisted Tunneling Model

Statement Parameter Units Default

MATERIAL PHONON.ENERGY eV 0.068

MATERIAL Z.SCHENK 0.0
MATERIAL HUANG.RHYS 3.5

213 ATLAS User’s Manual

Physical Models Physics

Poole-Frenkel Barrier Lowering for Coulombic Wells

The Poole-Frenkel effect can enhance the emission rate for Coulombic wells. I the
TRAP.COULOMBIC parameter is specified in the MODELS statement, the Schockley-Read-Hall
electron and hole recombination model becomes:
2
p n – n ie
R n, SRH = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ 3-362
   
 -----------------------------------
TAUP0 -
n + n ie exp  ------------------ +  ---------------------------------------- p + n ie exp  ---------------------
ETRAP TAUN0 – ETRAP
 DIRAC  kT   COUL   kT 
 1 + p  L  F + n  L

2
p n – n ie
R p, SRH = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ 3-363
   
 ----------------------------------------
TAUP0  n + n exp  ETRAP ------------------ +  ------------------------------------ p + n exp  –
TAUN0 ETRAP
---------------------
 COUL ie  kT   DIRAC  ie  kT 
 F + n  L  1 + p  L

The Poole-Frenkel thermal emission factor, F, is defined in Equation 3-109. The Coulombic
field-enhancement terms, nCOUL and PCOUL are defined in Equations 3-111 and 3-112.
JTAT Model
A Trap assisted Tunneling model, based on [43], has been implemented in ATLAS. It is an
alternative to the TRAP.TUNNEL model and cannot be used concurrently with the
TRAP.TUNNEL model. Instead of modifying the SRH lifetimes, as does the TRAP.TUNNEL
model, it calculates a Generation-Recombination rate dependent on the local value of electric
field. This is given by
3 2
Q m o m FM N t G t
3
 4 2m o m  E g – E t  
- exp  – ---------------------------------------------------
----------------------------------------------  3-364
3
8   E g – E t   3Q  F 

where
• F is the magnitude of the local electric field.
• m* is the effective mass.
• Eg is the semiconductor energy gap.
• Nt is the trap density.
• Et is the trap energy.
• Gt is the trap degeneracy.
The factor M2 is the square of the matrix element associated with the trap potential. The
quantity Eg-Et is assumed to correspond to the rate-limiting step of the two-stage tunnelling
process and will have a minimum possible value of 0.5 Eg. The model can be used with or
without traps being defined by the TRAPS or DEFECTS statements. To use it without TRAPS,
you specify the flag JTAT on the MODELS statement. The value of Eg-Et is obtained as
0.5E g + JTAT · LEVEL because the trap level will be relatively close in energy to the middle

of the energy gap. JTAT.LEVEL is the trap energy relative to the midgap energy. To set the
trap density, you use the JTAT.DENSITY parameter on the MODELS statement. The square of

214 ATLAS User’s Manual

Physical Models Physics

the matrix element is set by using the JTAT.M2 parameter on the MODELS statement. The value
of JTAT.M2 can also be set for specific materials using the JTAT.M2 parameter on the
MATERIAL statement. The factor Gt is assumed to be 1 for the JTAT model.

The units and default values of these quantities are shown in Table 3-77.

Table 3-77 Parameters for the JTAT Model

Statement Parameter Type Default Units

MODELS JTAT Logical False

MODELS JTAT.LEVEL Real 0.0 eV
MODELS JTAT.DENSITY Real 0.0 cm-3
MODELS JTAT.M2 Real 10-20 V2cm-3
MATERIAL JTAT.M2 Real 10-20 V2cm-3

TRAP.JTAT Model
The JTAT model can only be used for one tunneling level, and does not put the trap charge
density into the simulation. The TRAP.JTAT model is designed to be used along with traps as
specified on the TRAP statement (Section 3.3.3 “Traps and Defects”) or using the DEFECTS
statement (Chapter 14 “TFT: Thin-Film Transistor Simulator”). For traps defined by the
TRAPS (or DOPING) statements, the quantity Eg-Et is interpreted as being the maximum value
of Eg-E.LEVEL and E.LEVEL. The trap density is given by the DENSITY parameter and the
trap degeneracy Gt, by DEGEN.FAC. The square of the interaction matrix element, JTAT.M2, is
set on a material by material basis on the MATERIAL statement or globally on the MODELS
statement. With the DEFECTS statement, the density used for the TRAP.JTAT calculation is the
value obtained from Equations 14-2 and 14-5, evaluated at the energy JTAT.LEVEL. The
tunneling is assumed to occur through the defects at an energy level JTAT.LEVEL with Eg-Et
being the maximum value of Eg-JTAT.LEVEL and JTAT.LEVEL. The value of the interaction
matrix element is the one specified on the MATERIAL statement. The trap degeneracy Gt, is
assumed to be 1, and the parameter JTAT.M2 is used as before.
The extra parameters for the TRAP.JTAT model are shown in Table 3-78 .

Table 3-78 Parameters for the TRAP.JTAT Model

Statement Parameter Type Default Units

MODELS TRAP.JTAT Logical False

MATERIAL JTAT.M2 Real 10-20 V2cm-3

The generation rate given by Equation 3-364 is output to Silvaco Structure Files in the overall
recombination rate. You set the U.TRAP flag on the OUTPUT statement to additionally output it
as a trap recombination rate.

215 ATLAS User’s Manual

Physical Models Physics

Optical Generation/Radiative Recombination

The next physical mechanisms we have to consider for generation/recombination are photon
transition. This mechanism occurs primarily in one step and is therefore a direct generation/
recombination mechanism. There are two partial processes involved. For radiative
recombination, an electron loses energy on the order of the band gap and moves from the
conduction band to the valence band. For optical generation, an electron moves from the
valence band to the conduction. In silicon, band to band generation/recombination is
insignificant. This effect, however, is important for narrow gap semiconductors and
semiconductors whose specific band structure allows direct transitions. By assuming a
capture rate CcOPT and an emission rate CeOTP, the involved partial processes can be written
as
OPT OPT
R np = Cc np, 3-365

for recombination and

OPT OPT
G np = Ce 3-366

for generation.
These rates must be equal in thermal equilibrium so that
OPT OPT 2
C np = Cc n ie 3-367

The total band to band generation/recombination is the difference of the partial rates, which
equates to
OPT OPT 2
R np = Cc  np – n ie  . 3-368

In ATLAS, CcOPT and can be defined by COPT on the MATERIAL statement or implemented
using a C-Interpreter routine. To turn on the optical recombination/ generation model, define
the OPTR keyword on the MODELS statement.
Auger Recombination
Auger recombination occurs through a three particle transition whereby a mobile carrier is
either captured or emitted. The underlying physics for such processes is unclear and normally
a more qualitative understanding is sufficient [250].
Standard Auger Model
Auger Recombination is commonly modeled using the expression [68]:
2 2 2 2
R Auger = AUGN  pn – nn ie  + AUGP  np – pn ie  3-369

where the model parameters AUGN and AUGP are user-definable in the MATERIAL statement
(see Table 3-79 for its default value). You can activate this model with the AUGER parameter
from the MODELS statement.

216 ATLAS User’s Manual

Physical Models Physics

Table 3-79 User-Specifiable Parameters for Equation 3-369

Statement Parameter Default Units

MATERIAL AUGN 2.810-31 cm6/s

MATERIAL AUGP 9.910-32 cm6/s

Klaassen’s Temperature-Dependent Auger Model

The Klaassen Auger Recombination Model [141] is activated by specifying the KLAAUG
parameter of the MODELS statement. The form of this model is
2 2 2 2
R Auger = C n  pn – nn ie  + C p  np – pnie  3-370

where the Auger coefficients are temperature dependent according to:

T KAUGDN
= KAUGCN  ----------
L
C 3-371
n 300

T KAUGDP
= KAUGCP  ----------
L
C 3-372
p  300

Here, the KAUGCN, KAUGCP, KAUGDN, and KAUGDP parameters are user-definable in the
MATERIAL statement and have the defaults shown in Table 3-80.

Table 3-80 User-Specifiable Parameters for Equation 3-371 and 3-372

Statement Parameter Default Units

MATERIAL KAUGCN 1.8310-31 cm6/s

MATERIAL KAUGCP 2.7810-31 cm6/s
MATERIAL KAUGDN 1.18
MATERIAL KAUGDP 0.72

217 ATLAS User’s Manual

Physical Models Physics

Narrow Bandgap Auger Model

An alternative model for the Auger recombination coefficients that is more suitable for
modeling Auger processes in narrow bandgap semiconductors can be enabled by setting the
parameters AUGKN and AUGKP on the MATERIAL statement and by specifying AUGER on the
MODELS statement. The model in ATLAS is a simplification of that by Beattie [22] and takes
the form:
2 2 2 2
R Auger = C n  pn – nn ie  + C p  np – pnie  3-373

where the Auger coefficients are concentration dependent according to:

AUGN
C n = ---------------------------------- 3-374
1 + AUGKN n

AUGP
C p = ---------------------------------- 3-375
1 + AUGKP p

Here, n and p are the electron and hole carrier concentrations and the new parameters, AUGKN
and AUGKP, are user-definable on the MATERIALS statement.
Auger Recombination Model for High and Low Injection Conditions [55,135]
The Auger recombination rate given in Equation 3-373 can be modified to account for high
and low injection conditions. In this model, the Auger rate coefficients, Cn and Cp, are carrier
and doping concentration dependent and are given by:

 N D  AUG · CHI  p 
C n = AUG · CNL  ------------------- + ------------------------
- ------------------- 3-376
 N D + p 2  N + p
D

NA
C p = AUG · CPL  ----------------- + ------------------------
AUG · CHI-  ---------------- n -
3-377
 N A + n 2  N A + n

Here, you can specify the AUG.CNL, AUG.CPL, and AUG.CHI parameters on the MATERIAL
statement and have defaults described in Table 3-66.

Table 3-81 User Specifiable Parameters for Equations 3-330 and 3-331.

Statement Parameter Default Units

MATERIAL AUG.CNL 2.210-31 cm6/s

MATERIAL AUG.CPL 9.210-32 cm6/s
MATERIAL AUG.CHI 1.6610-30 cm6/s

To enable this model, specify PIC.AUG on the MODELS statement.

218 ATLAS User’s Manual

Physical Models Physics

Auger Recombination Model With Temperature and Concentration Dependent Coefficients

The rate of Auger recombination is usually given by
2 2 2 2
R Auger = C n  n p – nn ie  + C p  p n – pn ie  3-378

The coefficients are dependent on both lattice temperature and carrier density [113]. The
carrier density dependence is such that the coefficients are reduced in size as carrier density
increases. The physical explanation of this dependence is that carrier-carrier interactions
become more significant at higher values of carrier density and reduce the Auger
recombination rate [95].
The coefficients are given by

2
 HNS · AE + HNS · BE  T  300  + HNS · CE  T  300  
C n  T, n  = 3-379
  1 + HNS · HE  exp  – n  HNS · N0E   

2
 HNS · AH + HNS · BH T  300  + HNS · CH  T  300  
C p  T, p  = 3-380
  1 + HNS · HH  exp  – p  HNS · N0H   

where n is electron density, p is hole density, and T is lattice temperature.

To enable the model, set the flag HNSAUG on the MODELS statement. You can change all the
parameters by specifying them on the MATERIAL statement. Table 3-83 shows the default
values. Setting HNS.HE=0 and HNS.HH=0 will remove the carrier density dependence if
required.

Table 3-82 User-Specifiable Parameters for Equations 3-380 to 3-381

MATERIAL HNS.AE 6.710-32 cm6/s
MATERIAL HNS.AH 7.210-32 cm6/s
MATERIAL HNS.BE 2.4510-31 cm6/s
MATERIAL HNS.BH 4.510-33 cm6/s
MATERIAL HNS.CE -2.210-32 cm6/s
MATERIAL HNS.CH 2.6310-32 cm6/s
MATERIAL HNS.HE 3.4667
MATERIAL HNS.HE 8.25688
MATERIAL HNS.N0E 1.01018 cm-3
MATERIAL HNS.N0H 1.01018 cm-3

219 ATLAS User’s Manual

Physical Models Physics

Surface Recombination
In addition to generation-recombination within the bulk of the semiconductor, electrons or
holes may recombine or be generated at interfaces. The rate of surface recombination may be
even greater than within the bulk. The standard method is to model interface recombination in
a similar manner as the bulk generation-recombination rate [91] where:
2
pn – n ie
R surf = --------------------------------------------------------------------------------------------------------------------------------------------------- 3-381
 p n + n ie exp  ------------------ +  n p + n ie exp  ---------------------
eff ETRAP eff – E TRAP
 kT   kT 
L L

Here:
di
1 1
--------- = ----- + ----- S.N 3-382
eff i Ai
n n

and
di
1 1
--------- = ----- + ----- S.P 3-383
eff i Ai
p p

in is the bulk lifetime calculated at node i along the interface and which may be a function of
the impurity concentration as well. The di and Ai parameters are the length and area of the
interface for node i. The S.N and S.P parameters are the recombination velocities for
electrons and holes respectively, which are user-definable in the INTERFACE statement. The
X.MIN, X.MAX, Y.MIN, and Y.MAX parameters can also be set in the INTERFACE statement to
define the region, where the specified values of the surface recombination velocities apply.
This model is activated by the presence of the recombination velocities in the INTERFACE
statement.

Table 3-83 User-Specifiable Parameters for Equations 3-382 to 3-383

Statement Parameter Default Units

INTERFACE S.N 0 cm/s

INTERFACE S.P 0 cm/s

220 ATLAS User’s Manual

Physical Models Physics

Zamdmer Model for LT-GaAs

Low temperature grown GaAs (LT-GaAs) is GaAs grown at a temperature of 200 - 300°,
usually with a slight excess of Arsenic. The properties of LT-GaAs depend on the growth
temperature and depend on the degree of post-growth annealing [170]. It can be produced
with high resistivity and with sub-picosecond carrier lifetime, which makes it ideal for
ultrafast photodetectors that are based on displacement current. In that category of device, the
response time is limited by the recombination lifetime of photogenerated carriers. It has been
observed that operating these devices under high bias conditions tends to increase the
response time [330].
One possible explanation of this effect is the effective decrease in the capture cross-section
with field of the recombination centers as also observed in SiO2 [194]. According to [330],
this causes an increase in the electron lifetime with electric field, F, which can be modeled as
T e  F  1.5 r  0  3
  F  =   0   --------------  ----------- 3-384
 T L   r  F 

where   0  is the electron lifetime with no field present. The variable r is the radius of the
coulombic potential well caused by the trapping centers. With zero applied field, this radius is
isotropic but as a field is applied, it becomes anisotropic and you can define a miniumum and
maximum radius in the direction of the field. r(F) is the minimum radius defined as the
minimum radius at which the potential between two trapping centers is equal to the maximum
e T F
potential between them minus 2 KBT/q. The ratio -------------
- is the ratio of electron temperature to
TL
lattice temperature. The electron temperature is approximated as
Fe
T e = T L + ---------- 3-385
K B

where  is the optical phonon mean free path and KB is Boltzmanns constant.
The model is designed for discrete traps. These are created by either the TRAPS statement or
the TRAPS parameter on the DOPING statement. You enable the model by setting the ZAMDMER
parameter on the MODELS statement. You set the distance between the recombination centers
using the ZAMDMER.Z0 field on the MATERIAL statement, and set the value of  with the
IG.LRELE parameter on the MODELS statement.
The model evaluates the recombination lifetime at every point in the LT-GaAs according to
the value of electric field there. An example of the amount by which the electron lifetime is
enhanced as a function of field is given in Figure 3-7.

221 ATLAS User’s Manual

Physical Models Physics

Figure 3-7: Log-Log plot of lifetime enhancement factor as a function of electric field. The trap separation is 7
nm and the value of  is 30 nm for this example.
A summary of the parameters used for the ZAMDMER model are given in Table 3-84.

Table 3-84 Parameters for the ZAMDMER Model

Statement Parameter Type Default Units

MODELS ZAMDMER Logical False

MODELS IG.LRELE Real 2.010-6 cm
MATERIAL ZAMDMER.Z0 Real 7.010-7 cm

222 ATLAS User’s Manual

2
AP = AP 1 2 + CP2  T + DP2  T 3-398

where T is temperature and CN2, CP2, DN2, and DP2 are user-specifiable parameters on the
IMPACT statement. By default, the temperature model of Equations 3-393 and 3-394 are used
and the values of CN2, CP2, DN2 and DP2 are all zero. You can use the temperature dependence
models described in Equations 3-397 or 3-398 or both by specifying non-zero values for CN2,
CP2, DN2 and DP2.
The critical fields given by BN and BP may be modeled based on band gap and optical phonon
mean free paths using the following expressions:
Eg
BN = ---------- 3-399
0
q n

Eg
BP = ---------- 3-400
0
q p

0 0
where q  n and q  p are the optical phonon mean free paths for electrons and holes and Eg is
the local temperature dependent band gap. The free paths are modeled using the following
expressions:

o
tanh  qOPPHE  2kT L 
 n = LAMDAH -------------------------------------------------------------- 3-401
tanh  qOPPHE  2k300 

225 ATLAS User’s Manual

Physical Models Physics

o
tanh  qOPPHE  2kT L 
 n = LAMDAE -------------------------------------------------------------- 3-402
tanh  qOPPHE  2k300 

where T is the lattice temperature and LAMDAE, LAMDAH, OPPHE are user-specifiable
parameters listed in Table 3-87. To enable the models described by Equations 3-395, 3-396,
3-397, and 3-398, either specify BN1,2 or BP1,2 or both as zero.

Table 3-85 User-Definable Parameters in the Selberherr Impact Ionization Model

Statement Parameter Default

IMPACT AN1 7.03105 cm-1

IMPACT AN2 7.03105 cm-1
IMPACT AP1 6.71105 cm-1
IMPACT AP2 1.58106 cm-1
IMPACT BN1 1.231106V/cm
IMPACT BN2 1.231106 V/cm
IMPACT BP1 1.693106 V/cm
IMPACT BP2 2.036106 V/cm
IMPACT BETAN 1.0
IMPACT BETAP 1.0
IMPACT EGRAN 4105 V/cm

Table 3-86 Temperature Coefficient Parameters of the Selberherr Impact Ionization Model for
Silicon in Equations 3-393 to 3-396

Statement Parameter Default

IMPACT A.NT 0.588

IMPACT B.NT 0.248
IMPACT A.PT 0.588
IMPACT B.PT 0.248
IMPACT M.ANT 1.0
IMPACT M.BNT 1.0
IMPACT M.APT 1.0
IMPACT M.BPT 1.0

226 ATLAS User’s Manual

Physical Models Physics

Table 3-87 User specifiable parameters for the optical phonon mean free path model in Equations
3-401 and 3-402.

Statement Parameter Default Units

IMPACT LAMDAE 6.210-7 cm

IMPACT LAMDAH 3.810-7 cm
IMPACT OPPHE 0.063 eV

Selberherr Tabular Model

Given a tabulation of ionization rate versus field or reciprocal field, as it more commonly
appears, you can extract values of the A and B coefficients in Equations 3-390 and 3-391
assuming a value for BETA (in this case 1.0) for each consectutive pair of samples.
ATLAS provides an interface to automate this process whereby you will provide a table of
ionization rates versus fields or reciprocal fields and the Selberherr parameters are
automatically derived and used to reproduce a representation of continuous ionization rates
versus field to be used in the Selberherr model.
To use this feature, you must assign the values of IINOFF or IIPOFF or both to the name of a
file containing an ASCII table of the ionization rates as a function of field for electrons or
holes or both.
The table itself must first contain the integer number of samples (pairs) in the table. This is
followed by that number of pairs of either field - ionization rate or reciprocal field - ionization
rate values.
If reciprocal field is used, the designator flag "oe" must preceed the samples. If the "oe"
designator flag is missing the ordinate values will be interpreted as field values.
The units of fields are V/cm. Ionization rates units are 1/cm. Reciprocal field units are cm/V.
Comments may be embedded in the file following a "#" character in the first column any row
in the file.
The following is an example ionization file:
# Oguzman, I., Bellotti, E. and Brennan, K., "Theory of hole
initiated impact
# ionization in bulk zincblende and wurtzite GaN", J. Appl. Phys,
B. 81,
# 15 June 1997, pp. 7827-7834.
oe
5
2.465e-07 4.711e+04
2.840e-07 1.350e+04
3.335e-07 3.670e+03
4.011e-07 3.658e+02
5.012e-07 1.098e+01

227 ATLAS User’s Manual

Physical Models Physics

In this case, we have some comments that are useful for tracing down the origin of the data
followed by the reciprocal field designator flag.
Next, the number of samples is specified as 5. Finally, the five samples of reciprocal field
ionization rate pairs are specified.
For this case, you might introduce this impact ionization characteristic for a region using the
following statement:
MATERIAL REGION=3 IINOFF=mcclintockn.tbl
Using tabular ionization coefficients presents an issue when evaluating ionization coefficients
outside of the range of the table. We provide two extrapolation schemes. The default specified
by the parameter TBL.EXTRAP on the IMPACT statement implies that fields outside the range
of the table uses the Selberherr coefficients from the nearest pair of samples.
Extrapolation is disabled by specifying "^TBL.EXTRAP" on the IMPACT statement. Once
extrapolation is disabled, ionization coefficient values for fields larger than the peak field in
the table will be set to the value of the ionization coefficient at the peak field. If the fields are
less than the lowest field in the table, the ionization coefficient will be linearly interpolated
with field between zero at zero field and the value of the tabulated ionization rate at the
lowest tabulated field.
The difference between the extrapolation and non-extrapolation of tabulated ionization rates
is illustrated using the example table above in Figure 3-8.

Figure 3-8: Calculated ionization rates for the tabulated Selberherr model with and without extrapolation

228 ATLAS User’s Manual

Physical Models Physics

Van Overstraeten–de Man Impact Ionization Model

Based on the Chynoweth law [53], this model is very similar to the Selberherr model. The
differences, however, are that the exponents BETAN and BETAP are set equal to unity. The
lattice temperature dependence of the coefficients is also different. The functional forms of
the ionization rate are
–  n BN
 n =  n AN exp --------------- 3-403
E

–  p BP
 p =  p AP exp --------------- 3-404
E

where the  factors depend on Lattice temperature, or device temperature if GIGA is not
enabled. They are calculated as

tanh  ------------------------
N · VANHW
-
 2k  300  
 n = --------------------------------------------- 3-405
tanh  ------------------------
N · VANHW
-
2kT 

tanh  ------------------------
P · VANHW-
 2k300 
 p = --------------------------------------------- 3-406
tanh  ------------------------
P · VANHW-
 2kT 

and will be unity if the device is at a uniform 300 K. The same parameters are used as in the
Selberherr model but the BETAN and BETAP are fixed at unity. The other parameters are in
Table 3-88. To activate the model, use the VANOVERS parameter on the IMPACT statement.

Table 3-88 van Overstraeten–de Man model Parameters

Statement Parameter Default Units

IMPACT N.VANHW 0.063 eV

IMPACT P.VANHW 0.063 eV
IMPACT VANOVERS False --

229 ATLAS User’s Manual

Physical Models Physics

Valdinoci Impact Ionization Model

Valdinoci et al. [281] reported on a measurement technique to calibrate the temperature
dependence of impact ionization models and proposed a new model based on their study.
Their model of both electron and hole ionization coefficients was calibrated for silicon for
temperature ranging from 25 to 400°C. This model is based on the following:
E
 n, p = ---------------------------------------------------------------------------------------------------------------------------------------------- 3-407
a n, p  T L  + b n, p  T L  exp  d n, p  T L    E + c n, p  T L   

where E is the electric field along the current flow lines. The parameters in Equation 3-407
depend on temperature as follows:
VAL · AN2
 
a n  T L  = VAL · AN0 +  VAL · AN1  T L  3-408
 

b n  T L  = VAL · BN0 exp  VAL · BN1  T L 3-409

 

VAL · CN2
  2
c n  T L  = VAL · CN0 +  VAL · CN1  T L  + VAL · CN3  T L 3-410
 

2
d n  T L  = VAL · DN0 + VAL · DN1  T L + VAL · DN2  T L 3-411

VAL · AP2
 
a p  T L  = VAL · AP0 +  VAL · AP1  T L  3-412
 

b p  T L  = VAL · BP0 exp  VAL · BP1  T L 3-413

 

VAL · CP2
  2
c p  T L  = VAL · CP0 +  VAL · CP1  T L  + VAL · CP3  T L 3-414
 

2
d p  T L  = VAL · DP0 + VAL · DP1  T L + VAL · DP2  T L 3-415

230 ATLAS User’s Manual

Physical Models Physics

Our default parameters for silicon are shown in Table 3-89. These parameters are also taken
from Valdinoci et al. [281]. To enable this model, specify the logical parameter, VALDINOCI,
in the IMPACT statement.

Table 3-89 Default Parameters for Valdinoci Impact Ionization Model

Statement Parameter Type Default Units

IMPACT VAL.AN0 Real 4.3383

IMPACT VAL.AN1 Real -2.4210-12

IMPACT VAL.AN2 Real 4.1233

IMPACT VAL.BN0 Real 0.235

IMPACT VAL.BN1 Real 0.0

IMPACT VAL.CN0 Real 1.6831104

IMPACT VAL.CN1 Real 4.3796

IMPACT VAL.CN2 Real 1.0

IMPACT VAL.CN3 Real 0.13005

IMPACT VAL.DN0 Real 1.233735106

IMPACT VAL.DN1 Real 1.2039103

IMPACT VAL.DN2 Real 0.56703

IMPACT VAL.AP0 Real 2.376

IMPACT VAL.AP1 Real 0.01033

IMPACT VAL.AP2 Real 1.0

IMPACT VAL.BP0 Real 0.17714

IMPACT VAL.BP1 Real -0.002178

IMPACT VAL.CP0 Real 0.0

IMPACT VAL.CP1 Real 0.00947

IMPACT VAL.CP2 Real 2.4924

IMPACT VAL.CP3 Real 0.0

IMPACT VAL.DP0 Real 1.4043106

IMPACT VAL.DP1 Real 2.9744103

IMPACT VAL.DP2 Real 1.4829

231 ATLAS User’s Manual

Physical Models Physics

Zappa's Model for Ionization Rates in InP

A model for temperature and field dependent ionization rates in InP was proposed by Zappa
et. al. [331]. To enable this model, specify ZAPPA in the IMPACT statement. The model is
described for electrons and holes by Equations 3-416 through 3-421:

qE  
ZAP · AN  1.14   ZAP AN 1.14 2  ZAP AN 2 0.5 
 n = --------------------- exp  0.217  --------------------- –  0.217  ---------------------
·   +  ---------------------
·   3-416
ZAP · AN   En    En    qE n  

 n = ZAP · BN tanh  ---------------------

ZAP · CN
3-417
 2kT 

E n = ZAP · DN tanh  ---------------------

ZAP · EN
3-418
 2kT 

 1.14
ZAP · AP 1.14 2  ZAP 2 0.5 
 p = --------------------- exp  0.217  ---------------------
ZAP · AP
–  0.217  --------------------- · AP
qE
Ep    Ep  
+ ---------------------
 qE p   3-419
ZAP · AP  

 p = ZAP · BP tanh  ---------------------

ZAP · CP
3-420
 2kT 

E p = ZAP · DP tanh  ---------------------

ZAP · EP
3-421
 2kT 

where E is the local electric field and T is the local temperature. Table 3-90 describes the user-
specifiable parameters for Zappa's model.

Table 3-90 User-Specifiable Parameters for Zappa's Ionization Rate Model

Statement Parameter Type Default Units

IMPACT ZAP.AN Real 1.9 eV

IMPACT ZAP.BN Real 41.7 angstroms
IMPACT ZAP.CN Real 46.0 meV
IMPACT ZAP.DN Real 46.0 meV
IMPACT ZAP.EN Real 46.0 meV
IMPACT ZAP.AP Real 1.4 eV
IMPACT ZAP.BP Real 41.3 angstroms
IMPACT ZAP.CP Real 36.0 meV
IMPACT ZAP.DP Real 36.0 meV
IMPACT ZAP.EP Real 36.0 meV

232 ATLAS User’s Manual

Physical Models Physics

Grant’s Impact Ionization Model

The second ionization model has the same form as the Selberherr model but a simpler
implementation where:

 n = AN exp –  ------- 
BN
3-422
E

 p = AP exp –  ------- 
BP
3-423
E

This implementation has three key differences:

• The model has a low-field, an intermediate field and a high field region.
• The coefficients for silicon are different.
• There is no temperature dependence.
This model was developed after investigations by Baraff [20] suggested the existence of a
low, intermediate and high field response region for electron and hole ionization rates. The
coefficients implemented into this model match the experimental data of Grant [89], which
suggested that the three different regions existed.
This model is activated with the GRANT parameter of the IMPACT statement. The model
parameters: AN, AP, BN, and BP aren’t user-definable. Instead, the three electric field regions
have in-built values as follows:
1) Low Electric Field E < 2.4105 V/cm 3-424
AN = 2.610 6
AP = 2.010 6

BN = 1.43106 BN = 1.97106

2) Intermediate Electric Field 2.4105 > E > 5.3105 V/cm 3-425

AN = 6.2105 AP= 2.0106
BN = 1.08106 BP = 1.97106
3) High Electric Field E > 5.3105 V/cm 3-426
AN = 5.0105 AP = 5.6105
BN = 9.9106 BP = 1.32106
Crowell-Sze Impact Ionization Model
Crowell and Sze [60] have proposed a more physical relationship between the electric field
and the ionization rates. This model represents ionization coefficients as follows:
1 2
 n p = ---------- exp  C o  r  + C 1  r x + C 2  r x  3-427
 n p

2
C 0 = – 1.92 + 75.5r – 75.7r 3-428

–2 2
C 1  r  = – 1.75 10 – 11.9r + 46r 3-429

233 ATLAS User’s Manual

Physical Models Physics

–4 2
C 2  r  = 3.9 10 – 1.17r + 11.5r 3-430
where:

OPPHE Ei
r = ------------------ ; x = ------------------
- 3-431
Ei q  n p E

 1.1eV for electrons

Ei =  3-432
1.8eV for holes

o tanh  qOPPHE  2kT 

 n = LAMDAE -------------------------------------------------------------
L
- 3-433
tanh  qOPPHE  2k300 

o tanh  qE  2kT 
r L
 n = LAMDAH --------------------------------------------------
- 3-434
tanh  qE r  2k300 

The Crowell-Sze Model for impact ionization is selected by setting the CROWELL parameter of
the IMPACT statement.

Table 3-91 Crowell-Sze Impact Ionization Model Parameters.

Statement Parameter Default

IMPACT LAMDAE 6.210-7cm

IMPACT LAMDAH 3.810-7 cm

Okuto-Crowell Impact Ionization Model

This is another empirical model with temperature dependent coefficients where the ionization
coefficients take the following forms:

NGAMOKUTO
 n = ANOKUTO  1 + CNOKUTO  T – 300    F 3-435

BNOKUTO  1 + DNOKUTO  T – 300   NDELOKUTO

 exp –  --------------------------------------------------------------------------------------------
 F 

PGAMOKUTO
 p = APOKUTO  1 + CPOKUTO  T – 300    F 3-436

BPOKUTO  1 + DPOKUTO  T – 300   PDELOKUTO

 exp –  --------------------------------------------------------------------------------------------
F

where F is the effective electric field.

The parameters are optimized to fit in the Electric Field Range 105 to 106 V/cm but the same
ones are used for the whole range of field. It uses a relatively large set of parameters and so
you can easily adjust the default values (applicable to silicon) to fit other materials [198].

234 ATLAS User’s Manual

Physical Models Physics

To enable the model, use the OKUTO parameter on the IMPACT statement. The adjustable
model parameters are given in Table 3-92.

Table 3-92 Okuto-Crowell Model Parameters

Statement Parameter Default Units

IMPACT ANOKUTO 0.426 /V

IMPACT APOKUTO 0.243 /V
IMPACT BNOKUTO 4.81105 V/cm
IMPACT BPOKUTO 6.53105 V/cm
IMPACT CNOKUTO 3.0510-4 /K
IMPACT CPOKUTO 5.3510-4 /K
IMPACT DNOKUTO 6.8610-4 /K
IMPACT DPOKUTO 5.6710-4 /K
IMPACT NGAMOKUTO 1.0
IMPACT PGAMOKUTO 1.0
IMPACT NDELOKUTO 2.0
IMPACT PDELOKUTO 2.0

Note: If NGAMOKUTO or PGAMOKUTO is different from 1.0, the units of ANOKUTO or APOKUTO will be different
from /V.

Lackner Impact Ionization Model

This is another, more recent, impact ionization model based on Chynoweth’s law [150]. It is
similar to the van Overstraeten - de Man model but with an extra factor Z, dividing the
ionization rate. The factor Z and the ionization coefficients are

exp  – -----------------------------------------------------------
GAMMAN · BNLACKNER GAMMAN · BNLACKNER
Z = 1 + ----------------------------------------------------------- 3-437
F  F 

+ GAMMAP · BPLACKNER exp  – -----------------------------------------------------------

----------------------------------------------------------- GAMMAP · BPLACKNER
F  F 

 n = GAMMAN · ANLACKNER exp  – GAMMAN

----------------------------------------------------------- · BNLACKNER
----------------------------------------------------------- 3-438
Z  F 

 p = GAMMAP · APLACKNER exp  – GAMMAP

----------------------------------------------------------- · BPLACKNER
----------------------------------------------------------- 3-439
Z  F 

235 ATLAS User’s Manual

Physical Models Physics

tanh  ----------------------------
N · LACKHW
tanh  ----------------------------
- P · LACKHW
-
 2k300   2k300 
GAMMAN = ------------------------------------------------- GAMMAP = ------------------------------------------------- 3-440
tanh  ----------------------------
N · LACKHW
tanh  ----------------------------
- P · LACKHW
-
 2kT   2kT 

To enable the model, use the LACKNER parameter on the IMPACT statement. The adjustable
model parameters are given in Table 3-93.

Table 3-93 LACKNER Model Parameters

Statement Parameter Default Units

IMPACT ANLACKNER 1.316106 /cm

IMPACT APLACKNER 1.818106 /cm
IMPACT BNLACKNER 1.474106 V/cm
IMPACT BPLACKNER 2.036106 V/cm
IMPACT N.LACKHW 0.063 eV
IMPACT P.LACKHW 0.063 eV

Non-Local Carrier Energy Models for Impact Ionization

All local electric field based models will normally overestimate the rate of impact ionization.
This occurs because lucky electron theory inherently assumes that a carrier is traveling
through a constant electric field E. As a result it will predict a distance x=Ei/qE over which
the carrier will gain the ionization energy Ei. In real devices, however, the electric field is
never constant but is normally sharply peaked at metallurgical junctions. Therefore, as a
carrier passes through the peaked electric field the lucky electron model will predict the
ionization distance x to be too small. As a result the ionization rate is overestimated. The
effect of this is that all the simulated breakdown voltages will be underestimated and substrate
currents overestimated.
The Energy Balance Transport Model can be used to improve the simulation of impact
ionization by implementing ionization models based upon the carrier temperature instead of
the electric field. The carrier temperature is a more meaningful basis as the velocity-field
relationship is more closely modeled. This allows a non-local dependence on the electric field
within the impact ionization model. Energy Balance Transport Models will therefore result in
more accurate simulations of breakdown voltage and substrate current. Two different impact
ionization models have been implemented into ATLAS, the first is based upon the classical
Chynoweth relationship, modified to include carrier temperature, but the second is a more
advanced non-Maxwellian approach based upon carrier temperatures.
When the energy balance transport model is applied, only two impact ionization models are
available. These are the Toyabe model and the Concannon model.

236 ATLAS User’s Manual

Physical Models Physics

Toyabe Impact Ionization Model

The temperature dependent impact ionization model is founded on the Selberherr model and
is similar to that suggested by Toyabe [131]. The carrier temperature is used to calculate an
effective electric field based upon the homogeneous temperature-field relation. To maintain
self-consistency within the energy balance transport model this is the same relationship used
for the effective electric field within the carrier temperature dependent mobility. This model
is the default model for impact ionization with energy balance transport and is activated with
the TOYABE parameters on the IMPACT statement. The ionization rates now have the form:

 n = AN exp  ------------------
– BN
3-441
E eff, n

 p = AP exp  ------------------
– BP
3-442
E eff, p

where the model parameters: AN, AP, BN, and BP are user-definable in the IMPACT statement.
The AN, AP, BN, and BP parameters are set by the AN1, AP1, BP1, and BN1 parameters that
have the same default values as listed in Table 3-85. The parameters EGRAN, AN2, AP2, BN2,
and BP2 are not used in the Toyabe model.
The effective electric field is calculated according to:

3 kT n
E eff n = --- ----------------------------- 3-443
2 qLREL.EL

3 kT p
E eff p = --- ----------------------------- 3-444
2 qLREL.HO

where the energy relaxation lengths, LREL.EL and LREL.HO, can be explicitly defined in the
IMPACT statement, or can be calculated according to:

LREL.EL = VSATN* TAUSN 3-445

LREL.HO = VSATP* TAUSP 3-446

VSATN and VSATP are the saturation velocities for electrons and holes, and TAUSN and TAUSP
correspond to the electron energy relaxation times (TAUREL.EL and TAUREL.HO) in Equations
3-322 and 3-323. You must set the LENGTH.REL flag to use the values of LREL.EL and
LREL.HO specified in the IMPACT statement.

Table 3-94 User-Specifiable Parameters for Equations 3-443-3-446

Statement Parameter Units

IMPACT LREL.EL µm
IMPACT LREL.HO µm
MATERIAL VSAT cm/s

237 ATLAS User’s Manual

Physical Models Physics

Table 3-94 User-Specifiable Parameters for Equations 3-443-3-446

Statement Parameter Units

MATERIAL VSATN cm/s

MATERIAL VSATP cm/s
IMPACT TAUSN s
IMPACT TAUSP s

Note: As an added level of flexibility, the relaxation times used for the energy balance equation and those used in
the impact ionization model are separated into two user-definable parameters. In contrast to TAUREL.EL
and TAUREL.HO, which are used in different formulae, the TAUSN and TAUSP parameters are only
applicable in the impact ionization expression in Equations 3-443 and 3-444. By default,
TAUREL.EL=TAUSN and TAUREL.HO=TAUSP.

It can also be argued that the AN, AP, BN, and BP parameters should also be a function of the
carrier temperature. But, no clear theoretical basis for this has been proposed and accepted.
Instead, the C-INTERPRETER within ATLAS has been extended to include two C-INTERPRETER
functions. These functions are specified via the F.EDIIN and F.EDIIP parameters of the
IMPACT statement. These parameters specify the filename of a text file containing a C-
INTERPRETER function that describes the dependence of the model parameters AN, AP, BN and
BP as a function of the carrier temperatures. These values will then be used within Toyabe’s
energy dependent impact ionization model.
Concannon’s Impact Ionization Model
The previous non-local impact ionization model inherently assumes a Maxwellian shape to
the distribution of hot carriers. Recent work by Fiegna et. al. [75] using Monte Carlo
simulations suggests a non-Maxwellian high energy tail to the energy distribution function.
To accurately model these effects, a non-Maxwellian based model from Concannon [57] has
been implemented. Based upon this energy distribution the model calculates the probability
of a carrier having sufficient energy to cause impact ionization. The model results show good
agreement with measured results for a 0.9m flash EPROM device [57].
To enable the Concannon substrate current for the electron and hole continuity equations,
specify the N.CONCANNON and P.CONCANNON parameters in the IMPACT statement.
The generation rate is a function of the carrier temperature and concentration is given by:

G n  x , y  = CSUB.N  n
 F  , T n  x, y   d  3-447
ETH.N


G p  x, y  = CSUB.P  p
 F  , T p  x, y   d  3-448
ETH.P

238 ATLAS User’s Manual

Physical Models Physics

where n(x,y) and p(x,y) are the electron and hole carrier concentrations within the
semiconductor,  is energy, Tn(x,y) and Tp(x,y) are the electron and hole carrier temperatures
in the semiconductor, F is given in Equation 3-449, CSUB.N, CSUB.P, ETH.N, and ETH.P are
user-specifiable parameters as given in Table 3-95.

Table 3-95 User-Specifiable Parameters for Equations 3-447 and 3-448

Statement Parameter Default Units

IMPACT CSUB.N 2.01014

IMPACT CSUB.P 4.01014
IMPACT ETH.N 1.8 eV
IMPACT ETH.P 3.5 eV

The function, F(,T), in Equations 3-447 and 3-448 is given by the product of the density of
states function, g(), and the energy distribution function f() as:
g   f   
F = ---------------------------- 3-449



 g   f   

The density of states function is given by:

1.25
g =  3-450

The energy distribution functions for electrons and holes are:

 – CHIA  3  – CHIB  3
fn    = exp  ------------------------- + C0 exp  ------------------------- 3-451
 T n 1.5   T n 1.5 

 – CHI.HOLES  3
f p    = exp  -------------------------------------------- 3-452
1.5
 Tp 

239 ATLAS User’s Manual

Physical Models Physics

Here,  is energy, Tn,p are the carrier temperatures, CHIA, CHIB, and C0 are user-specifiable
parameters (see Table 3-96).

Table 3-96 User-Definable Parameters for the Energy Distribution Functions

Statement Parameter Default Units

IMPACT CHIA 3.0105

IMPACT CHIB 5.0104
IMPACT C0 2.510-10
IMPACT CHI.HOLES 4.6104

Two other parameters in the IMPACT statement are user-definable, which may affect the result
of the numeric integration. The ENERGY.STEP parameter specifies the energy step size in eV
used during the numeric integration. The default step size is 25 meV. The INFINITY parameter
sets the upper limit of the integration and specifies ratio of the increment added to the integral
divided by the current value of the integral. The default value of the INFINITY parameter is
0.001.

Note: To maintain self-consistent results, implement this model if the Concannon model is being used for the
simulation of gate current.

Other Hydrodynamic Models

You can also use the Lackner, Okuto-Crowell, and van Overstraeten - de Man models with the
hydrodynamic model. To use this model, set the command line flag -ISE instead of the
ATLAS default model (Equations 3-443 and 3-444). The ATLAS default model assumes the
effective field used to calculate impact ionization rate is directly proportional to carrier
Temperature.
A more physical relationship between effective field and carrier temperature is given by
1.5k B  T n – T L 
E eff, n = ------------------------------------------------------------------------------------------------
- 3-453
TAUREL · EL NISELAMBDA VSATN
+  n  E eff n N · UPSILON  E g  T L  + NISEDELTAk B T n 

1.5k B  T p – T L 
E eff p = ------------------------------------------------------------------------------------------------
- 3-454
TAUREL · HO PISELAMBDA VSATP
+  p  E eff p P · UPSILON  E g  T L  + PISEDELTAk B T p 

The parameters N.UPSILON and P.UPSILON can be either 0(unset) or 1(set). The default for
Silicon is 1. It is 0 for other materials. In the case where UPSILON is unity, the equation
becomes a transcendental one because it contains  as a function of Eeff. In this case, the
equation is solved iteratively in order to obtain . If no solution is found, then an error
message will output and you must set N.UPSILON=0 or P.UPSILON=0. The user adjustable
parameters are given in Table 3-97.

240 ATLAS User’s Manual

Physical Models Physics

Table 3-97 Hydrodynamic Model Parameters

Statement Parameter Default Units

IMPACT N.UPSILON True False

(in silicon) (otherwise)
IMPACT P.UPSILON True False
(in silicon) (otherwise)
IMPACT NISELAMBDA 1.0
IMPACT PISELAMBDA 1.0
IMPACT NISEDELTA 1.5
IMPACT PISEDELTA 1.5

Solutions for Wide Bandgap Semiconductors

In models of wide bandgap semiconductors, the drift-diffusion current may be so small that
the simulated impact ionization carrier generation rates remain so low that breakdown does
not occur at the expected bias state. In reality, impact ionization is a non-local process and is
caused by hot carriers in the non-Maxwellian tail of the energy distribution function.
The drift-diffusion model fails to correctly model these effects, although you can simulate the
impact ionization breakdown. This is done by putting a minimum value on the current
densities used in Equation 3-388 to calculate the impact generation rate. You set a minimum
current density by specifying the values of its directional components with the JNX.MIN,
JNY.MIN, JNZ.MIN, JPX.MIN, JPY.MIN, and JPZ.MIN parameters on the IMPACT statement.
The drift-diffusion current density along any side has the current density calculated from the
above parameters added to it before being used in Equation 3-388.
This minimum current density causes the impact ionization breakdown to be bootstrapped
when the ionization coefficients become sufficiently large. The simulated currents below the
breakdown will depend on the particular values chosen for the minimum current and will be
unphysical. As breakdown occurs, however, the generated current densities will significantly
exceed the imposed minimum currents, and the simulation of the breakdown should be
physically realistic.
This model is only available with the E.SIDE option at present. It is targeted at wide bandgap
semiconductors but can also be used in devices where the current densities are too noisy or
too low to reliably simulate breakdown by impact ionization.

Table 3-98 Parameters for Specifying Minimum Current Values

Statement Parameter Default Units

IMPACT JNX.MIN 0.0 A/cm2

IMPACT JNY.MIN 0.0 A/cm2
IMPACT JNZ.MIN 0.0 A/cm2

241 ATLAS User’s Manual

Physical Models Physics

Table 3-98 Parameters for Specifying Minimum Current Values

Statement Parameter Default Units

IMPACT JPX.MIN 0.0 A/cm2

IMPACT JPY.MIN 0.0 A/cm2
IMPACT JPZ.MIN 0.0 A/cm2

3.6.5 Geiger Mode Simulation

Geiger mode simulation is the calculation of the single photon, single electron, or single hole
probability of of avalanche breakdown usually in devices that are cooled sufficiently to
neglect thermal generation of carriers. These calculations are done using line integrals
suggested by McIntyre [182] of ionization rates along paths of steepest potential gradient.
These calculations are done as a post-processing step after convergence is obtained without
any impact ionization or thermal generation models being turned on.
To enable the post processing, you should specify GEIGER on the MODELS statement. The next
three different calculations can be done. First, you can probe the value of the probability of
avalanche due to introduction of an electron, hole, or electron-hole pair by specifying
PE.AVALANCE, PH.AVALANCHE, or PP.AVALANCE on the PROBE statement. This should be
accompanied by the location of the probe using the X and Y parameters of the PROBE
statement. Second, if you specify a value for the FILENAME parameter of the PROBE statement,
the line integral through the specified point will also be outputted to the specified file.
Finally, once GEIGER is specified on the MODELS statement, all structure files saved thereafter
will contain the spatial distribution of the probabilities of electron, hole, and pair generated
avalanche breakdown.

3.6.6 Band-to-Band Tunneling

Standard Models
If a sufficiently high electric field exists within a device local band bending may be sufficient
to allow electrons to tunnel, by internal field emission, from the valence band into the
conduction band. An additional electron is therefore generated in the conduction band and a
hole in the valence band. This generation mechanism is implemented into the right-hand side
of the continuity equations. The tunneling generation rate is [114,115,116,138] as:

exp  – --------------
BB.GAMMA BB.B
G BBT = D BB.A E 3-455
 E 

where E is the magnitude of the electric field, D is a statistical factor, and BB.A, BB.B, and
BB.GAMMA are user-definable parameters. In ATLAS, there are three different sets of values
that may be applied to the model parameters.
The model parameters can be set to the standard model [114] by specifying BBT.STD on the
MODELS statement. The parameter defaults for the standard model are as follows:

BB.A = 9.6615e18 cm-1 V-2 s-1 BB.B= 3.0e7 V/cm BB.GAMMA= 2.0

242 ATLAS User’s Manual

Physical Models Physics

The model parameters may also be set to the Klaassen Model [114,138,115] by specifying
BBT.KL on the MODELS statement. The parameter defaults for the Klaassen model are as
follows:

BB.A = 4.00e14 cm-1/2 V-5/2 s-1 BB.B = 1.9e7 V/cm BB.GAMMA= 2.5
In application, use the standard model with direct transitions while using the Klaassen model
with indirect transitions.
Another modification allows these model parameters to be calculated from the first principles
by specifying the AUTOBBT parameter in the MODELS statement. In this case, the parameters
are calculated according to
2
q  2  MASS.TUNNEL m 0 
BB.A = ------------------------------------------------------------------------------
- 3-456
2
h EG300

3
---
MASS.TUNNEL m
2 2
 EG300 --------------------------------------------------0
2
BB.B = ------------------------------------------------------------------------------------- 3-457
qh

BB.GAMMA = 2 3-458
where q is the electronic charge, h is Planck’s constant, Eg is the energy bandgap, m0 is the
rest mass of an electron and MASS.TUNNEL is the effective mass. The parameter
MASS.TUNNEL may be set on the MODELS statement and the bandgap at 300K, EG300, is
defined on the MATERIAL statement.
An alternative expression to Equation 3-529 for BB.B is used if the flag AUTOBBT2 is set on
the MODELS statement. This is
32
8 2m o MASS · TUNNEL EG300
BB · B = --------------------------------------------------------------------------------------------- 3-459
3qh

The value of BB.A is the same for the AUTOBBT parameter as with the AUTOBBT2 parameter.
The default value of the statistical factor D is 1. This factor can be calculated as suggested by
Hurkx et al [114]:
2
n ie – np
D = ------------------------------------------ 3-460
 n + n ie   p + n ie 

To enable this modification, specify BBT.HURKX in the MODELS statement. If BBT.DEHURKX is

specified in the MODELS statement, then D will be set to 0 if

0 if n  E  0 and p  E  0

 2
D =  n ie – np 3-461
------------------------------------------ otherwise
 n + n ie   p + n ie 

243 ATLAS User’s Manual

Physical Models Physics

If BBT.DJHURKX is specified in the MODELS statement, then D will be set to 1 if

 1
–3
j n  1  10 qn ie VSATN and

 –3
 j p  1  10 qn ie VSATP
D =  3-462
 2
 ------------------------------------------
n ie – np
 n + n p + n  otherwise
 ie ie

If the BBT.ALPHA parameter is specified in the MATERIAL statement, then the statistical factor
will be given by
n ie – np
D = ------------------------------------------  1 – BBT.ALPHA  – BBT.ALPHA 3-463
 n + n ie   p + n ie 

where
• BBT.ALPHA=0 corresponds to the original Hurkx model,
• BBT.ALPHA=1 corresponds to recombination only, and
• BBT.ALPHA=-1 corresponds to generation only.
If BBT.SHURKX is specified in the MODELS statement, then the statistical factor will be taken
from Schenk et al [245].

Table 3-99 User-Definable Parameters in the Band-to-Band Tunneling Model

Statement Parameter Default Units

MODELS BBT.ALPHA 0

MODELS BB.A 4.01014 cm-1/2 V-5/2 s-1

MODELS BB.B 1.9107 V/cm

MODELS BB.GAMMA 2.5

Schenk Band to Band Tunneling Model

A comprehensive study of band-to-band tunneling was carried out by Schenk [245]. A
rigorous theory was developed and then an approximate result suitable for use in device
simulations was derived. The result shows that phonon assisted band-to-band tunneling rate is
generally dominant compared to the band-to-band tunneling that doesn't involve a phonon
scattering event. The direct band-to-band tunneling is thus neglected. The model also
assumes that the electric field is constant over the tunneling length. Therefore, it is a local
model.
The recombination-generation rate is given by

 
 –3  2  A   –3  2  A  
 A  exp  ---------- A  exp  ---------- 
SCHENK 72   F   F  
GBBT = A · BBT · SCHENKF S  ------------------------------------------------------------------------- + ---------------------------------------------------------------------------- 3-464
  HW · BBT · SCHENK  – 1 – H W · BBT · SCHENK -
 exp ---------------------------------------------------- 1 – exp ------------------------------------------------------- 
 k T kT 

244 ATLAS User’s Manual

Physical Models Physics


where A =B.BBT.SCHENK(   HW.BBT.SCHENK)3/2, S is a statistical factor dependent on
carrier concentrations, and  is the energy of the transverse acoustic phonon. The upper sign
refers to tunneling generation of electron-hole pairs (reverse biased junction). The lower sign
refers to tunneling recombination of electron-hole pairs (forward bias).
You can set parameters from Table 3-100 on the MODELS or MATERIAL statement. To enable
the model, specify SCHENK.BBT on the MODELS statement.

Table 3-100 Schenk Band to Band Tunneling Model Parameters

Statement Parameter Type Default Units

MATERIAL/MODELS A.BBT.SCHENK Real 8.9771020 cm2s-1V-2

MATERIAL/MODELS B.BBT.SCHENK Real 2.14667107 eV(-3/2)Vcm-1
MATERIAL/MODELS HW.BBT.SCHENK Real 0.0186 eV

Kane Band-To-Band Tunneling Model

Another local field band-to-band tunneling model is based on the work of Kane [130]. In this
model, the tunneling generation rate is given by
3
 ---
2
D BBT · A KANE BBT GAMMA  E g 
G BBT = -------------------------------------------- F · exp  – BBT · B KANE ------ 3-465
Eg  F
 
where Eg is the position dependent bandgap and F is the magnitude of the electric field. It is
similar to the standard models, except that it automatically includes a dependence on
bandgap.
To enable this model specify BBT.KANE on the MODELS statement. The parameters are also
aliased as indicated in Table 3-101.

Table 3-101 Parameters (with aliases) for the BBT.KANE Model

Statement Parameter Alias Type Default Units

MODELS BBT.KANE BTBT Logical False

MODELS BBT.A_KANE A.BTBT Real 3.51021 eV1/2/cm-s-V2
MODELS BBT.B_KANE B.BTBT Real 2.25107 V/cm-eV3/2
MODELS BBT.GAMMA Real 2.5

245 ATLAS User’s Manual

Physical Models Physics

Non-local Band-to-Band Tunneling

The band-to-band tunneling models described so far in this section calculate a recombination-
generation rate at each point based solely on the field value local to that point. For this reason,
we refer to them as local models. To model the tunneling process more accurately, you need
to take into account the spatial variation of the energy bands. You also need to take into
account that the generation/recombination of opposite carrier types is not spatially coincident.
Figure 3-11 illustrates this for a reverse biased p-n junction where it is assumed that the
tunneling process is elastic. For degenerately doped p-n junctions, you can obtain tunneling
current at forward bias and consequently can obtain negative differential resistance in the
forward I-V curve.
The ATLAS Non-local Band-To-Band tunneling model, BBT.NONLOCAL, allows modeling of
the forward and reverse tunneling currents of degenerately doped p-n junctions.
It is less suitable for lightly doped p-n junctions. It is not suitable for application to the
unipolar, high-field regions of a device. It is currently only available in ATLAS2D.
The first step in using BBT.NONLOCAL is in defining the areas where it will be applied.
These areas must each contain a single p-n junction, which may be either planar or non-
planar. They have a mesh that interpolates data from the underlying device mesh and
performs the band-to-band tunneling calculations on the interpolated data.
BBT.NONLOCAL thus assumes that the tunneling takes place on a series of 1D slices through
the junction, each slice being locally perpendicular to the junction. The slices themselves will
be approximately parallel to their neighbors.
ATLAS has two methods of setting up these areas of slices. The first method is only
applicable to planar junctions parallel to the x or y-axes. It allows you to set up a rectangular
area using the QTX.MESH and QTY.MESH statements. You also specify the number of tunneling
slices and the number of mesh points along the slices.
For example:
QTX.MESH LOCATION=0.0 SPACING=0.25
QTX.MESH LOCATION=1.0 SPACING=0.25

QTY.MESH LOCATION=0.99 SPACING=0.0005

QTY.MESH LOCATION=1.01 SPACING=0.0005
will set up an area bounded by the sides X=0.0, X=1.0, Y=0.99, and Y=1.01. By additionally
setting the QTUNN.DIR parameter on the MODELS statement to be 0, you define the tunneling
direction to be in the y-direction. Therefore, we have 5 parallel tunneling slices at a uniform
separation of 0.25 microns, each with 0.02/0.0005 + 1 = 41 mesh points along them. This
region is consistent with a p-n junction parallel to the x-direction, and having a y-position of
close to 1.0 microns. Notice that the mesh is very fine in the direction of tunneling.
It is recommended to have as fine mesh as possible in the tunneling direction, both for the
physical mesh and this tunneling mesh, so that values can be accurately interpolated between
the two meshes. If the tunneling direction is the x-direction, then set QTUNN.DIR to 1 on the
MODELS statement.
The second method of setting up a mesh for tunneling can be applied to either planar or non-
planar junctions. It can be used instead of, or in addition to, the first method.

246 ATLAS User’s Manual

Physical Models Physics

To use the second method, create one or more QTREGION statements. These must be located
after any X.MESH, Y.MESH, ELECTRODE, and REGION statements and before any MODELS
statements that specify BBT.NONLOCAL. The parameters for QTREGION include the position
coordinates of the four corners of a general quadrilateral. These parameters are X1, Y1, X2, Y2,
X3, Y3, X4, Y4, and should describe the quadrilateral in a counter-clockwise sense as
illustrated in Figure 3-9. The pair (X1, Y1) describe the coordinates of point 1. The other pairs
(e.g., (X2, Y2)) describe the coordinates of the other points (e.g., 2).
The tunneling is calculated between points 1 and 4, and between points 2 and 3, and along all
the slices in between. The tunneling slices do not need to be parallel to one another as shown
in Figure 3-9 for the special case where the spacing between tunneling slices is regular along
side 3-4 and also along side 1-2. The tunneling direction changes gradually as the
quadrilateral is traversed (see Figure 3-9).

Figure 3-9: Quadrilateral region example arising from a QTREGION statement. PTS.NORMAL= 5 has been
specified, modeling tunneling along 5 rays, which are not quite parallel. The spacing between the lines is
regular along edge 1->2 and along edge 4->3. The tunneling direction is indicated by the arrows.
For tunneling through non-planar junctions, it is preferable to construct a complicated
polygonal region to use the BBT.NONLOCAL method.
You can achieve this by using several joined QTREGION's: all with a common value of the
NUMBER parameter. The simple rule that must be followed to join the quadrilaterals is to
ensure that point (X2, Y2) on the first matches (X1, Y1) on the other, and that point (X3, Y3) on
the first matches point (X4, Y4) on the next. If (X1, Y1) or (X4, Y4) or both are not specified,
then they default to the values of (X2, Y2) and (X3, Y3) respectively from the previous
QTREGION statement. The first QTREGION must have all 4 corner points described. In this way,
you can build up a tunneling mesh to cover quite complicated p-n junction shapes. An
example is given in Figure 3-10.

247 ATLAS User’s Manual

Physical Models Physics

Figure 3-10: Example of a polygonal region made by joining 4 quadrilaterals. The region encloses the p-n
junction, which is geometrically curved and ensures the BBT current calculation occurs along lines that are
approximately locally normal to the junction. Quadrilateral 1 must have (X1,Y1), (X2,Y2), (X3,Y3), and (X4,Y4)
specified. For quadrilaterals 2,3 and 4, specifying (X1,Y1) and (X4, Y4) is optional. If they are specified, they
must be equal to the values of (X2,Y2) and (X3,Y3) on the adjacent quadrilateral.
You must set up the number of mesh points you require in the tunneling direction and the
simplest way to do this is to use the PTS.TUNNEL parameter. The value of PTS.TUNNEL
applies to all slices in a region of one or more joined quadrilaterals. You can also set the
number of slices in each quadrilateral using the PTS.NORMAL parameter. You may have a
different value of PTS.NORMAL for each quadrilateral in the same region. If there is more than
one value of PTS.TUNNEL specified for a particular region, then the one specified by the last
QTREGION statement will be used.
To allow some flexibility, there are two alternatives for specifying non-uniform mesh spacing
in the tunneling direction. You can use the parameters STNL.BEG and STNL.END to give the
mesh spacing at the beginning and end of the slice. Otherwise, you can use the
F.RESOLUTION parameter to specify the name of an ASCII-format datafile. This data file
must contain a single column of ascending values, in the range [0,1]. The density of mesh
points in the tunneling direction is obtained from linearly mapping from the [0,1] range in the
datafile to the start and end position of each tunneling slice for each quadrilateral having the
same NUMBER parameter.

248 ATLAS User’s Manual

Physical Models Physics

If you wish to specify non-uniform separations of the tunnel slices in a quadrilateral, then you
can achieve this using the SNRM.BEG and SNRM.END parameters. The spacing at the left hand
side of the quadrilateral is given by SNRM.BEG and that at the end by SNRM.END. This
functionality is useful because it allows you to match the interpolated tunneling mesh to the
underlying device mesh. Each quadrilateral may have its own unique values of SNRM.BEG and
SNRM.END.
In summary, you can set up one or more quantum band-to-band tunneling regions, each
comprised of one or more QTREGION's. All QTREGION's having the same value of the NUMBER
parameter are treated as part of the same region, and they must be joined together.
Additionally, one region set up by QTX.MESH and QTY.MESH can also be set up. The number
of slices in each region, and the density of points along those slices can be set up in several
ways. Quantities from the underlying device mesh are obtained from interpolation, and it is
imperative that the device mesh resolves the rapid changes in potential near to a highly doped
p-n junction.
The resolution of the tunneling slices in the tunneling direction should also be fine enough to
capture these rapid changes. Each tunneling slice must include exactly one p-n junction for
the model to work. The start and end points of the tunnel slices should ideally be in the flat
band regions on either side of the junction.

Figure 3-11: Schematic of non-local band to band tunneling in reverse bias

249 ATLAS User’s Manual

Physical Models Physics

In order to explain how the tunneling current is calculated, let us consider the energy band
profile along each tunneling slice with reverse bias applied across the junction. The range of
valence band electron energies for which tunneling is permitted is shown in the schematic of
the energy band profile in Figure 3-11. The highest energy at which an electron can tunnel is
Eupper and the lowest is Elower. The tunneling can be thought of being either the transfer of
electrons or the transfer of holes across the junction. The rates for electrons and holes are
equal and opposite because the tunneling results in the generation or recombination of
electron-hole pairs.
Considering the tunneling process as a transfer of an electron across the junction the net
current per unit area for an electron with longitudinal energy E and transverse energy ET is

  T  E   fl  E + ET  – fr  E + ET   ET  dE dET
q
J  E  = ------ 3-466


where T(E) is the tunneling probability for an electron with longitudinal energy E. (ET) is the
2-dimensional density of states corresponding to the 2 transverse wavevector components and
–1
f l =  1 + exp   E + E T – E Fl   KT   3-467

is the Fermi-Dirac function using the quasi Fermi-level on the left hand side of the junction.
Similarly
–1
f r =  1 + exp   E + E T – E Fr   KT   3-468

uses the quasi-Fermi level on the right hand side of the junction. This assumes that the
transverse energy is conserved in the tunneling transition. Because we are using a 2-band
model to give the evanescent wavevector, the transverse electron effective mass and the
transverse hole effective mass are combined in the 2D density of states

 me mh 
  E T  = -------------------------
- 3-469
2
2 

By integrating over transverse carrier energies, we obtain the contribution to the current from
the longitudinal energy range E - E/2 to E + E/2 as follows:
qKT m e m h E max
J  E E = ------------------------------- T  E  log  1 + exp   E Fl – E – E T   KT   – log  1 + exp   E Fr – E – E T   KT   E 3-470
2 3 o
2 

where the upper limit of integration, Emax is smaller of E - Elower and Eupper - E. This is
consistent with the 2-band model of the tunneling and the restriction on total carrier energy
for tunneling Elower E + ET  Eupper. Putting the above integration limits into Equation 3-470
we obtain
qKT m e m h   1 + exp   E Fr – E   KT    1 + exp   E Fl – E – E max   KT  
J  E E = ------------------------------ T  E  log  ------------------------------------------------------------------------------------------------------------------------------------------------------- E 3-471
2 
2 3
  1 + exp   E Fl – E   KT    1 + exp   E Fr – E – E max   KT  

250 ATLAS User’s Manual

Physical Models Physics

The Fermi levels used in Equation 3-471 are the quasi-Fermi levels belonging to the majority
carrier at the relevant side of the junction. In Figure 3-11 for example, EFl would be the
Electron quasi-Fermi level and EFr would be the hole quasi-Fermi level. In equilibrium, EFl =
EFr and the current contributions are all zero. As also seen in Figure 3-11, the start and end
points of the tunneling paths, xstart and xend, depend on Energy. ATLAS calculates these start
and end points for each value of E and calculates the evanescent wavevector at points in
between as
ke kh
k  x  = --------------------
- 3-472
2 2
ke + kh

where
1
k e  x  = ----- 2m o m e  x   E – E c  x   3-473
i

and
1
k h  x  = ----- 2m o m h  x   E v  x  – E  3-474
i

This ensures that the energy dispersion relationship is electron-like near the conduction band
and hole-like near the valence band, and approximately mixed in between. The tunneling
probability, T(E), is then calculated using the WKB approximation

 x end

T  E  = exp  – 2
 x start

k  x  dx

3-475

This value is put into Equation 3-471 to give the tunneling current density at a given
perpendicular energy, E, and the resulting current is injected into the simulation at xstart and
xend. This is repeated for all values of E between Elower and Eupper and is done for every
tunneling slice in the tunneling regions.
To calibrate the BBT.NONLOCAL model, you can specify the values of effective mass used in
Equations 3-471, 3-473, and 3-474. Set the values of effective mass used in Equations 3-473
and 3-474 using ME.TUNNEL and MH.TUNNEL on the MATERIAL statement respectively. If
ME.TUNNEL (MH.TUNNEL) is not specified, then the value specified for MC (MV) on the
MATERIAL statement is used. For the transverse effective masses as used in Equation 3-471, a
suitable value will be calculated by ATLAS, unless MC or MV or both have been specified on
the MATERIAL statement, in which case these are used. The tunneling current is most sensitive
to the effective masses used in Equations 3-473 and 3-474 because the tunneling probability
depends exponentially on them.
Across very highly doped junctions, the Band-To-Band tunneling current can be very high
and depends very strongly on the band edge profile throughout the junction. In turn, the
injected tunnel current creates a charge dipole and strongly affects the potential, and
consequently the band energies, at and near the junction. This strong coupling can cause
convergence problems in some situations. ATLAS has two features to try to mitigate this
problem. First, you can specify BBT.NLDERIVS on the MODELS statement to include non-local
coupling. This puts the derivatives of J(E) with respect to all potential values along its path

251 ATLAS User’s Manual

Physical Models Physics

into the Jacobian matrix. Generally, this improves the convergence properties. But, this
increases the solution time per iteration because of the consequent increase in the size of the
Jacobian matrix.
Second, the position at which the tunnel current is injected into the simulation can affect
convergence. The default described above is the most physically accurate. As an alternative,
use the parameter BBT.FORWARD on the MODELS statement. This causes the tunnel current to
be injected at the start and end positions of each tunnel slice, further from the junction than
with the default model. The parameter BBT.REVERSE on the MODELS statement causes the
tunnel current to be injected in a locally charge neutral manner across the width of the
junction, thus avoiding the creation of a charge dipole. The injected currents can be viewed as
recombination rates in the structure file. They are automatically output whenever the
BBT.NONLOCAL model is used. They will be negative for a reverse biased junction and
positive for a forward biased junction.
Because the BBT.NONLOCAL model is unsuitable for all band-to-band tunneling, you can use
the local Band-To-Band tunneling models (e.g., BBT.STD and BBT.KLA) simultaneously with
BBT.NONLOCAL. These local models are not applied inside any of the Quantum Tunneling
regions set up by the QTREGION statement or QTX.MESH/QTY.MESH statements.

Table 3-102 Non-Local Band-to-Band Tunneling Parameters for the MODELS Statement

Parameter Type Default

BBT.NONLOCAL Logical False

BBT.NLDERIVS Logical False
BBT.FORWARD Logical False
BBT.REVERSE Logical False

Table 3-103 Non-Local Band-to-Band Tunneling Parameters for the MATERIAL Statement

Parameter Type Units

ME Real None
MV Real None
ME.TUNNEL Real None
MH.TUNNEL Real None

Table 3-104 Parameters for the QTREGION Statement

Statement Parameter Type Units

QTREGION F.RESOLUTION Character

QTREGION NUMBER Real
QTREGION PTS.NORMAL Real

252 ATLAS User’s Manual

Physical Models Physics

Table 3-104 Parameters for the QTREGION Statement

Statement Parameter Type Units

QTREGION PTS.TUNNEL Real

QTREGION SNRM.BEG Real Microns
QTREGION SNRM.END Real Microns
QTREGION STNL.BEG Real Microns
QTREGION STNL.END Real Microns
QTREGION X1 Real Microns
QTREGION X2 Real Microns
QTREGION X3 Real Microns
QTREGION X4 Real Microns
QTREGION Y1 Real Microns
QTREGION Y2 Real Microns
QTREGION Y3 Real Microns
QTREGION Y4 Real Microns

Note: QTREGION must be specified after any X.MESH,Y.MESH, ELECTRODE, and REGION statements, but
before any DOPING statements.

Note: If you want to fine tune the Band to Band tunneling, we recommend you use ME.TUNNEL and
MH.TUNNEL parameters because they will only modify the effective masses used in tunneling.

3.6.7 Gate Current Models

In devices that have a metal-insulator-semiconductor (MIS) formation, the conductance of the
insulating film would ideally be considered as zero. But, for the sub 0.5um generation of
MOS devices there is now considerable conductance being measured on the gate contacts.
This gate current has resulted in two major consequences, one negative and one positive.
On the negative side, the gate current is responsible for the degradation in device operating
characteristics with time. This reliability issue is important because the lifetime of electronic
parts has to be guaranteed. You can simulate reliability within the Silvaco suite of tools for
device level reliability which are described in a later chapter.
On the positive side, the existence of this gate current has caused the proliferation of the non-
volatile memory market. These devices use the existence of gate current to program and erase
the charge on a “floating” contact. This concept has resulted in a variety of different devices
such as FLASH, FLOTOX, and EEPROM. All such devices rely on the physics of the gate current
process for their existence.

253 ATLAS User’s Manual

Physical Models Physics

There are a variety of different conduction mechanisms within an insulating layer [268], but
in the case of nonvolatile memory, only two mechanism are relevant: Fowler-Nordheim
tunneling and hot carrier injection. Models for these two injection processes are described in
the following sections. In the case of hot electron injection, two models are available: the
lucky electron model and the Concannon gate current model.
Fowler-Nordheim Tunneling
If the electric field across an insulator is sufficiently high, then it may cause tunneling of
electrons from the semiconductor (or metal) Fermi level into the insulator conduction band.
This process is strongly dependent on the applied electric field but is independent of the
ambient temperature.
The Fowler-Nordheim Equation [132] expresses tunnel current density through the oxide as:

J FN = F.AE E exp  – --------------

2 F.BE
3-476
E

J FP = F.AH E exp  – --------------

2 F.BH 3-477
 E 

where E specifies the magnitude of the electric field in the oxide. The model parameters:
F.AE, F.AH, F.BE, and F.BH can be defined on the MODELS statement. The default values for
these parameters, obtained from Keeney, Piccini, and Morelli [132], are shown in Table 3-
105.

Table 3-105 User-Specifiable Parameters for Equations 3-476 and 3-477

Symbol Statement Parameter Default Values

AFN MODELS F.AE 1.8210-7

BFN MODELS F.BE 1.90108
AFH MODELS F.AH 1.8210-7
BFH MODELS F.BH 1.90108

The Fowler-Nordheim model in ATLAS has been implemented in two ways. In the post-
processing implementation, the tunneling current is calculated from the solution to the device
equations at each bias step. In the self-consistent implementation, the tunneling current is
included directly in the current-continuity equations. Thus, it is part of the solution to the
equations and gives accurate current continuity for the device.
To enable the post processing version, use the MODELS statement parameters FNPP for electron
current and FNHPP for hole current. To enable the self-consistent solution, use the MODELS
statement parameters FNORD for electrons and FNHOLES for holes. Setting the parameter
FN.CUR is the same as specifying FNORD and FNHOLES.
For either model, the implementation scheme is the same. Each electrode-insulator and
insulator-semiconductor interface is divided into discrete segments which are based upon the

254 ATLAS User’s Manual

Physical Models Physics

mesh. For each insulator-semiconductor segment, the Fowler-Nordheim current is calculated

as described above. This current is then added to a segment on the electrode-insulator
boundary. Two schemes have been implemented to find out to which segment this current
should be added.
The default model that calculates which electrode segment receives the Fowler-Nordheim
current follows the path of the electric field vector at the semiconductor-insulator interface.
The first electrode-insulator segment that is found along this trajectory, provided no other
semiconductors or metals are found along the trajectory, receives the Fowler-Nordheim
current.
A second model may be chosen using the NEARFLG parameter of the MODEL statement. In this
case, the electrode-insulator segment found closest to the semiconductor-insulator segment
receives the Fowler-Nordheim current.
The total current on the gate electrode is then the sum of the currents from all the individual
segments around the electrode boundary. To include the tunneling current in the logfile,
specify J.TUN on the LOG statement.
By default, the Fowler-Nordheim model, when enabled, is applied to all relevant interfaces in
the device structure. To localize the model, use one or more INTERFACE statements with one
or more of the parameters X.MIN, X.MAX, Y.MIN, and Y.MAX specified. This restricts the
application of Fowler-Nordheim tunneling to interfaces contained in the union of the
rectangular areas defined by the INTERFACE statements. The parameter S.I. must also be set
on the INTERFACE statement but is enabled by default.
In the special situation of an insulator layer between two semiconductor regions, the FN.SIS
parameter can be specified on the REGION statement. This precludes the Fowler-Nordheim
current generated on the interfaces of the region from being added to any terminal current.
The current is instead added to current continuity equations at one side of the region and
subracted on the other side. This simulates tunneling through an insulator placed between
two semiconductors.

Note: Since Fowler-Nordheim tunneling current is responsible for EPROM and EEPROM cell erasure, this model
should always be specified when performing erasure simulation. We also recommend that you model the
band-to-band tunneling model if you model Fowler-Norheim tunneling.

Note: When simulating EPROM erasure in a transient analysis with this model, the floating contact charge
becomes a function of the gate current. In this case, the total current flowing into the floating electrode is
multiplied by the time step to calculate the charge added to the electrode during that time step. The new
value of the charge is then used as the boundary condition for the next time step.

Lucky Electron Hot Carrier Injection Model

In the Lucky-Electron Hot Carrier Injection Model it is proposed that an electron is emitted
into the oxide by first gaining enough energy from the electric field in the channel to
surmount the insulator/semiconductor barrier. Once the required energy to surmount the
barrier has been obtained the electrons are redirected towards the insulator/semiconductor
interface by some form of phonon scattering. When these conditions are met, the carrier
travelling towards the interface will then have an additional probability that it will not suffer
any additional collision through which energy could be lost.

255 ATLAS User’s Manual

Physical Models Physics

The model implemented into ATLAS is a modified version of the model proposed by Tam
[272] and is activated by the parameters HEI and HHI, for electron and hole injection
respectively, on the MODELS statement. The gate electrode-insulator interface is subdivided
into a number of discrete segments which are defined by the mesh. For each segment the
lucky electron model is used to calculate the injected current into that segment. The total gate
current is then the sum of all of the discrete values.
If we consider a discrete point on the gate electrode-insulator boundary we can write a
mathematical formula for the current injected from the semiconductor. The formula calculates
the injected gate current contribution from every node point within the semiconductor
according to:

Iinj =
  Pn  x, y  Jn  x, y  dx dy +   Pp  x, y  Jp  x, y  dx dy 3-478

where Jn,p (x,y) are the electron and hole current densities at a point (x,y) within the
semiconductor, and Pn,p (x,y) are the probabilities that a fraction of this current reaches the
gate oxide and is injected across into the gate electrode. The total probability Pn,p (x,y) is
defined by:
P n  x, y  = P  B, n P 1, n P 2, n  IG.ELINR 3-479

P p  x, y  = P  B, p P 1, p P 2, p  IG.HLINR 3-480

where E is the electric field parallel to the current flow, IG.ELINR and IG.HLINR are the
electron and hole mean free path lengths between redirecting collisions. The three probability
factors will now be described.
The probability PB is the probability of a carrier gaining the energy B by moving in, and
parallel to, an electric field E, without suffering energy loss by optical phonon scattering and
is given by:
 B, n
= 0.25  ------------------------------------ exp  – ------------------------------------
E IG.ELINF
P 3-481
B, n   B, n   E IG.ELINF

 B p
= 0.25  ----------------------------------- exp  – -------------------------------------
E IG.HLINF
P 3-482
B p   B p   E IG.HLINF

where IG.ELINF and IG.HLINF are the mean free path lengths of electrons and holes for
scattering by optical phonons. The barrier heights Bn,p are defined according to:
23
 B, n = IG.EB0 – IG.EBETA E  – IG.EETAE –   x, y  3-483

23
 B, p = IG.HB0 – IG.HBETA E  – IG.HETAE –   x, y  3-484

256 ATLAS User’s Manual

Physical Models Physics

where  is the electric field perpendicular to the semiconductor-insulator interface. The

traditional barrier heights, IG.EB0 and IG.HB0, are reduced to take account of three effects.
The first effect is due to Schottky barrier lowering which depends on the perpendicular
electric field at the semiconductor-insulator interface. The second effect takes account of
tunneling through the gate oxide by reducing the barrier height. The third effect takes into
account that a potential difference exists between the semiconductor-insulator interface and
the starting position of the hot carrier. By default, this last effect is disabled. But you can
enable it by specifying the E.BENDING and H.BENDING parameters for electrons and holes
respectively.
The second probability P1 is the probability that no energy is lost by optical phonon scattering
as the hot carrier travels towards the semiconductor-insulator interface after being redirected,
and is given by:

P 1, n ~ exp  – ------------------------------
r
3-485
 IG.ELINF

P 1, p ~ exp  – ------------------------------
r
3-486
 IG.HLINF

where r is the distance from point of redirection to the semiconductor-insulator interface.

The final probability P2 accounts for the probability of scattering in the image force potential
well in the gate oxide and is given by:

q
 ----------------------------
 16 ox E ox
P 2 n = exp  – -------------------------------- for THETA.N 3-487
 PATH.N 
 

P 2 n = 0 for  < THETA.N 3-488

q
 ----------------------------
 16 ox E ox
P 2 p = exp  – -------------------------------- for  > THETA.P 3-489
 PATH.P 
 

P 2 p = 0 for  < THETA.P 3-490

Here, PATH.N and PATH.P are the electron and hole mean free path lengths within the oxide,
ox is the oxide permittivity and Eox is the electric field in the oxide. The angle  introduces
an angle dependence which is based upon the work of Wada [295]. His experiments indicate a
critical rejection angle, THETA.N and THETA.P, between the angle  formed between the
semiconductor-insulator interface and the electric field in the oxide. If the angle  is less than
the rejection angle then the electrons are repelled back to the substrate.
Table 3-106 lists the user-definable model parameters that can be set in the MODELS statement,
their default values, and their units.

257 ATLAS User’s Manual

Physical Models Physics

Table 3-106 User-Definable Parameters in Concannon’s Gate Current Model

Statement Parameter Default Units

MODELS IG.EB0 3.2 eV

MODELS IG.ELINR 6.1610-6 cm

MODELS IG.HLINR 6.1610-6 cm

MODELS IG.ELINF 9.210-7 cm

MODELS IG.HB0 4.0 eV

MODELS IG.HLINF 9.210-7 cm

MODELS IG.EBETA 2.5910-4 (Vcm)1/2

MODELS IG.HBETA 2.5910-4 (Vcm)1/2
MODELS IG.EETA 2.010-5 V1/3cm2/3
MODELS IG.HETA 2.010-5 V1/3cm2/3
MODELS IG.LRELE 3.010-6 [Q=1] cm

MODELS IG.LRELH 2.010-6 [Q=1] cm

MODELS PATH.N 3.410-7 cm

MODELS PATH.P 2.3810-7 cm

MODELS THETA.N 60 degrees

MODELS THETA.P 60 degrees

The implementation of this model is similar to that for Fowler-Nordheim tunneling. Each
electrode-insulator and insulator-semiconductor interface is divided into discrete segments,
which are based upon the mesh. For each insulator-semiconductor segment the Fowler-
Nordheim current is calculated as described above. This current will then be added to a
segment on the electrode-insulator boundary. Two schemes have been implemented to find
out to which segment this current should be added.
The default model that calculates which electrode segment receives the hot carrier injected
current follows the path of the electric field vector at the semiconductor-insulator interface.
The first electrode-insulator segment that is found along this trajectory, provided no other
semiconductors or metals are found along the trajectory, will receive the current.
A second model may be chosen using the NEARFLG parameter of the MODELS statement. In
this case, the electrode-insulator segment found closest to the semiconductor-insulator
segment will receive the hot carrier injected current.

258 ATLAS User’s Manual

Physical Models Physics

The total current on the gate electrode is then the sum of the currents from all the individual
segments around the electrode boundary.
The lucky electron hot carrier injection model can be used to include the electron or hole
carrier temperature in the solution because the carrier temperature does not directly enter the
equations.
The one exception is in ATLAS2D where the electric field parallel to the current flow is
calculated as
E = 1.5 KB Tn / IG.LRELE

for electrons if HCTE.EL is specified and

E = 1.5 KB Tp / IG.LRELH

for holes if HCTE.HO is specified. KB is Boltzmanns constant in units of eV/Kelvin. If

IG.LRELE is set to zero then E will be calculated the same way as if HCTE.EL is unspecified.
The same is true for IG.LRELH and HCTE.HO. In ATLAS3D, this model is unavailable and E is
calculated the same way regardless if HCTE is specified.

Note: When simulating EPROM programming with this model, the floating contact charging is simulated in the
transient mode. In this case, the total current flowing into the floating electrode is multiplied by the time step
to calculate the charge added to the electrode during that time step. The new value of charge is then used
as the boundary condition for the next time step.

The hot electron and hot hole currents can be output separately to a logfile by specifying the
parameters J.HEI and J.HHI on the LOG statement. They are also automatically included in
the electron and hole currents which are output to a logfile. The value of hot carrier current
density associated with each interface node can be output to any standard structure file by
specifying the HEI and HHI parameters on the OUTPUT statement.
Concannon’s Injection Model
The implicit assumption in the lucky electron approach is a Maxwellian shape for the energy
distribution of the hot carriers. Recent work by Fiegna [75] using Monte Carlo simulations
suggests a non-Maxwellian high energy tail to the distribution function. To accurately model
these effects, a non-Maxwellian based model from Concannon [57] has been implemented.
This model requires the solution to the energy balance equation for the carrier temperatures
but has been implemented in a similar manner to the lucky electron model. The Concannon
gate injection model may be specified with the parameters N.CONCANNON and P.CONCANNON
on the MODELS statement. This choice of parameters automatically activates the Energy
Balance Transport Model.
The Concannon injection model has a similar form to the lucky electron model. The injected
current is calculated according to:

Iinj =
  Pn  x, y  n  x, y  dx dy +   Pp  x, y  p  x, y  dx dy 3-491

where n(x,y) and p(x,y) are the carrier concentrations within the semiconductor. The
probability functions Pn(x,y) and Pp(x,y) are now defined by:
P n  x y  = – q CGATE.N P  P P 3-492
B, n 1 n 2 n

259 ATLAS User’s Manual

Physical Models Physics

P p  x, y  = q CGATE.P P  P P 3-493
B, p 1, p 2, p

where q is the electronic charge and the parameters CGATE.N and CGATE.P are user-definable
on the MODEL statement. The three probability functions in Equations 3-492 and 3-493 shall
now be described.
The probability that a carrier has sufficient energy to surmount the insulator-semiconductor
barrier of height B is now defined as a function of energy. The probability has the form:


P
B, n
=
B, n v   F   Tn  x y  d 3-494


P
B, p
=
B, p v   F   Tp  x y   d 3-495

Here, v() is the perpendicular velocity of a hot carrier and defines the probability of a hot
carrier with an energy  travelling in the direction of the insulator-semiconductor. The barrier
heights Bn,p are defined according to:
23
 B n = IG.EB0 – IG.EBETA E  – IG.EETA E  –   x, y  3-496

23
 B p = IG.HB0 – IG.HBETA E  – IG.HETA E  –   x, y  3-497

where E is the electric field perpendicular to the semiconductor-insulator interface. The

traditional barrier heights, IG.EB0 and IG.HB0, are reduced to take account of three effects.
The first effect is due to Schottky barrier lowering which depends on the perpendicular
electric field at the semiconductor-insulator interface. The second effect takes account of
tunneling through the gate oxide by reducing the barrier height. The third effect takes into
account that a potential difference exists between the semiconductor-insulator interface and
the starting position of the hot carrier. By default, this last effect is disabled. But you can
enable it by specifying the E.BENDING and H.BENDING parameters for electrons and holes
respectively.
The carrier velocity model follows the approach of Fiegna et. al. [75] where velocity is
proportional to energy according to:
0.25
v   3-498

The function, F(, Tn,p(x,y)), is determined by the density of states and the energy distribution
function according to:
g   f   
F   T n  x y    ---------------------------------- 3-499

 g   f    d

The density of states g() follows the analysis of Cassi [42] where:
1.25
g   3-500

260 ATLAS User’s Manual

Physical Models Physics

Finally, the energy distribution functions for electrons and holes are defined by:

 – CHIA  3  – CHIB  3
f n     exp  ------------------------- + C0 exp  ------------------------- 3-501
 T n 1.5   T n 1.5 

 – CHI.HOLES  3
f p     exp  -------------------------------------------- 3-502
1.5
 Tp 

where CHIA, CHIB, CHI.HOLES, and C0 are user-definable constants found from fitting to
measured data. The terms: Tn and Tp are the mean carrier temperatures for electrons and
holes, which are calculated from the Energy Balance Transport Model.
Normalization in all of the above equations is accounted for in the combined constants of
proportionality, CGATE.N and CGATE.P.
The second probability P1 is the probability that no energy is lost by optical phonon scattering
as the hot carrier travels towards the semiconductor-insulator interface after being redirected
and is given by:

P 1 n  exp  – ------------------------------
r
3-503
 IG.ELINF

P 1, p ~ exp  – ------------------------------
r 3-504
 IG.HLINF

where r is the distance from point of redirection to the semiconductor-insulator interface.

The final probability P2 accounts for the probability of scattering in the image force potential
well in the gate oxide and is given by:

q
 ----------------------------
 16 ox E ox
P 2 n = exp  – -------------------------------- for THETA.N 3-505
 PATH.N 
 

P 2 n = 0 for  > THETA.N 3-506

q
 ----------------------------
 16 ox E ox
P 2 p = exp  – -------------------------------
- for  > THETA.P 3-507
 PATH.P 
 

P 2 p = 0 for  < THETA.P 3-508

Here, PATH.N and PATH.P are the electron and hole mean free path lengths within the oxide,
ox is the oxide permittivity and Eox is the electric field in the oxide. The angle  introduces

261 ATLAS User’s Manual

Physical Models Physics

an angle dependence which is based upon the work of Wada [295]. His experiments indicate a
critical rejection angle, THETA.N and THETA.P between the angle  formed between the
semiconductor-insulator interface and the electric field in the oxide. If the angle  is less than
the rejection angle, then the electrons are repelled back to the substrate.

Note: The current implementation of the Concannon model for hot carrier injection is that only carriers along the
semiconductor-insulator interface are significant and as a result the probability P1 is assumed unity. This
also means that the integration is only applied to those node points along the semiconductor-insulator
interface.

Two other parameters of the MODELS statement that may affect the result of the numeric
integration are user-definable. The ENERGY.STEP parameter specifies the energy step size in
eV used during the numeric integration. The default step size is 25 meV. The INFINITY
parameter sets the upper limit of the integration and specifies ratio of the increment added to
the integral divided by the current value of the integral. The default value of the INFINITY
parameter is 0.001.
The implementation of this model is similar to that for Fowler-Nordheim tunneling. Each
electrode-insulator and insulator-semiconductor interface is divided into discrete segments
which are based upon the mesh. For each insulator-semiconductor segment the Fowler-
Nordheim current is calculated as described above. This current will then be added to a
segment on the electrode-insulator boundary. Two schemes have been implemented to find
out to which segment this current should be added.
The default model that calculates which electrode segment receives the hot carrier injected
current follows the path of the electric field vector at the semiconductor-insulator interface.
The first electrode-insulator segment that is found along this trajectory, provided no other
semiconductors or metals are found along the trajectory, will receive the current.
A second model may be chosen using the NEARFLG parameter of the MODELS statement. In
this case the electrode-insulator segment found closest to the semiconductor-insulator
segment will receive the hot carrier injected current.
The total current on the gate electrode is then the sum of the currents from all the individual
segments around the electrode boundary.

Note: To maintain self-consistent results, it’s important that this model is implemented if the Concannon model is
being used for the simulation of substrate current.

The Concannon hot electron and hole currents may be output separately to a logfile using the
parameters J.HEI and J.HHI on the LOG statement respectively. They are also automatically
included in the electron and hole currents that are output to a logfile. The value of Concannon
hot carrier current density associated with each interface node can be output to any standard
structure file by specifying the HEI and HHI parameters on the OUTPUT statement.

Direct Quantum Tunneling Model

262 ATLAS User’s Manual

Physical Models Physics

For deep submicron devices, the thickness of the insulating layers can be very small. For
example, gate oxide thicknesses in MOS devices can be as low as several nanometers. In this
case, the main assumptions of the Fowler-Nordheim approximation are generally invalid and
you need a more accurate expression for tunneling current. The one ATLAS uses is based on
a formula, which was introduced by Price and Radcliffe [223] and developed by later authors.
It formulates the Schrodinger equation in the effective mass approximation and solves it to
calculate the transmission probability, T(E), of an electron or hole through the potential
barrier formed by the oxide layer. The incident (perpendicular) energy of the charge carrier,
E, is a parameter. It is assumed that the tunneling process is elastic. After taking into account
carrier statistics and integrating over lateral energy, the formula

 1 + exp   E Fr – E   kT  

qkT
J = ---------------- m y m z T  E  ln  ---------------------------------------------------------- dE 3-509
 1 + exp   E Fl – E   kT  
2 3
2 h

is obtained, which gives the current density J (A/m2) though the barrier. The effective masses
my and mz are the effective masses in the lateral direction in the semiconductor. For example,
for a direct bandgap material, where the  valley is isotropic, both my and mz are the same as
the density of states effective mass. The logarithmic term includes the carrier statistics and Efl
and Efr are the quasi-Fermi levels on either side of the barrier (see Figure 3-12). The range of
integration is determined according to the band edge shape at any given contact bias.

Figure 3-12: Typical conduction and valence band profiles of a MOS capacitor. The right region
represents the MOS bulk, the left region represents the gate.
For indirect bandgap materials, Equation 3-510 is applied to each valley and the resulting sum
gives the tunneling current. For the conduction band of silicon, for example, summing over
the 6 valleys gives

263 ATLAS User’s Manual

Physical Models Physics

 1 + exp   E Fr – E   kT  

qkT
J = ----------------  2m t + 4 m l m t  T  E  ln  ---------------------------------------------------------- dE 3-510
 1 + exp   E Fl – E   kT  
2 3
2 h

where mt is the transverse effective mass and ml the longitudinal mass. For a valence band,
you need to sum the light hole and heavy hole contributions separately. For the valence band,
the light hole and heavy contributions are summed to give

 1 + exp   E Fr – E   kT  

qkT
J = ----------------  m lh + m hh  T  E  ln  ---------------------------------------------------------- dE 3-511
 1 + exp   E Fl – E   kT  
2 3
2 h

In equilibrium, EFl = EFr and the logarithmic term and consequently J is identically zero.
The tunneling probability T(E) is calculated by using a transfer matrix method to solve the
Schrodinger equation. The method uses a stepwise approximation of the potential.
Consequently, it is recommended to use a fine mesh in the tunnelling direction.
Tunneling Types
ATLAS can account for several types of tunneling. For clarity, we assume the material on the
left hand side of Figure 3-12 is polysilicon. Electron tunneling occurs when electrons tunnel
through the insulator energy barrier from one conduction band to the other. Hole tunneling
occurs when holes tunnel from one valence band to the other. If the bias applied to the contact
is sufficiently large, then the situation illustrated in Figure 3-13 can occur. In this case, an
electron in the silicon valence band tunnels through the insulator to the polysilicon
conduction band.
If the bias is reversed, then the opposite occurs in which an electron in the polysilicon valence
band tunnels to the Silicon conduction band. Both of these cases are referred to as band-to-
band Tunneling. These should not be confused, however, with the other band-to-band
tunneling models implemented in ATLAS, which apply to regions of high electric field within
a semiconductor. All these types of tunneling can be used as post-processing calculations or
self-consistent calculations. In the former case, the tunneling currents are calculated after
ATLAS has found a converged solution at a given bias. ATLAS then outputs the calculated
tunneling currents to the logfile. In the latter case, the tunneling currents are coupled with the
solution of the current continuity equations in the semiconductor. Therefore, the terminal
currents are self-consistent.
In situations where there is a strong quantum confinement in the semiconductor, ATLAS can
include charge quantization effects. To do this, use the Schrodinger-Poisson solver in the
Silicon, which causes the tunneling current to be calculated as a sum over eigenstates. The
relevant formulae are

 1 + exp   E Fl – E i   kT  
 vr  Ei T  Ei  ln  ---------------------------------------------------------------
qkT
1 + exp   E Fr – E i   kT  
J = ----------- m y m z - 3-512
2
 i

where the sum is over all bound state energies, Ei, for a particular band minimum.

For electrons in silicon, with 6 conduction band minima, we obtain

264 ATLAS User’s Manual

Physical Models Physics

 1 + exp   E Fl – E il   kT  

2qkT
J = --------------- m t v ril T  E il  ln  -----------------------------------------------------------------  3-513

2
 1 + exp   E Fr – E il   kT  
i
 1 + exp   E Fl – E it   kT  

4qkT
+ --------------- m t m l v rit T  E it  ln  ------------------------------------------------------------------ 

2
 1 + exp   E Fr – E it   kT  
i

where the sums over longitudinal bound eigenstates and transverse bound eigenstates are
done separately. For holes, the current is calculated using the equation

 1 + exp   E i lh – E Fl   kT  

qkT
J = ----------- m lh v TE  ln  -----------------------------------------------------------------------  3-514

2 ri lh i lh
 1 + exp   E i lh – E Fr   kT  
i
 1 + exp   E i hh – E Fl   kT  

qkT
+ ----------- m hh v TE  ln  ------------------------------------------------------------------------- 

2 ri hh i hh
 1 + exp   E i hh – E Fr   kT  
i

where the sums over light hole and heavy hole bound eigenstates are done separately.
The expression  is called the attempt frequency and is given by

k 2 d 21
v r = ---------   x r  + --------  x r  ------ 3-515
4m dx 2
k

It is evaluated at the semiconductor-oxide interface if quantum confinement occurs there.

This quantum confined modification of the tunneling will be applied when you set the
SCHRODINGER parameter and set quantum tunneling. This requires you to set CARRIERS=0 on
the METHOD statement. Therefore, this model cannot be used if it is required to solve the
current continuity equations.

265 ATLAS User’s Manual

Physical Models Physics

Figure 3-13: Schematic of band-to-band tunneling across a gate insulator with a polysilicon gate
Enabling the Models
To enable the quantum tunneling model for electrons, specify QTUNN.EL in the MODELS
statement. To enable the quantum tunneling model for holes, specify QTUNN.HO in the
MODELS statement. To enable Band-to-Band tunneling, specify QTUNN.BBT in the MODELS
statement. In order to enable all three options, specify QTUNN in the MODELS statement.

Table 3-107 MODELS Statement: Quantum Tunneling Parameters

Parameter Type Default

QTUNN Logical False

QTUNN.BBT Logical False
QTUNN.EL Logical False
QTUNN.HO Logical False

266 ATLAS User’s Manual

Physical Models Physics

QTUNN.EL and QTUNN.HO are available with the CARRIERS=0, CARRIERS=1, and
CARRIERS=2 options on the METHOD statement. With CARRIERS=0, the semi-classical and
quantum-confined variants are available. QTUNN.BBT is only available as part of a semi-
classical calculation with either CARRIERS=0, CARRIERS=1 or CARRIERS=2.
To enable the self-consistent versions of the quantum tunneling models, specify the
QTNLSC.EL and QTNLSC.HO parameters on the MODELS statement for electron and hole
tunneling respectively. The Band-to-Band tunneling option is enabled using QTNLSC.BBT.
You can enable all three models using the QTUNNSC parameter on the MODELS statement.

Table 3-108 MODELS Statement: Self-consistent Quantum Tunneling Parameters

Parameter Type Default

QTNL.DERIVS Logical false

QTNLSC.EL Logical false
QTNLSC.HO Logical false
QTNLSC.BBT Logical false
QTUNNSC Logical false

For the self-consistent implementations, convergence can be somewhat poor in some

circumstances, particularly for high values of tunneling current. In this case, you can set the
QTNL.DERIVS parameter on the MODELS statement. This will include extra terms in the
Jacobian matrix, which should improve convergence. It will also increase the time needed to
solve each iteration.
Mesh Considerations
The tunneling current model is quasi one-dimensional and the tunneling current is evaluated
on a series of parallel or nearly parallel slices through the insulator. Implementation details
therefore depend on the type of mesh. If the mesh was generated by ATLAS meshing
commands, then everything including the direction of tunneling is determined automatically.
If the mesh was created by another Silvaco product, for example ATHENA or DEVEDIT, then
there are two options.
The first option is to use the QTX.MESH and QTY.MESH commands to create a rectangular
mesh on which all quantum tunneling is calculated. Interpolation is used to couple between
this supplementary mesh and the device mesh. By default, the tunneling will be assumed to be
in the Y direction. You can change this by setting the QTUNN.DIR parameter on the MODELS
statement to 1 instead of its default value of 0. Place the QT mesh to enclose the insulator and
to not overextend into the conductor or semiconductor. The second option is to allow ATLAS
to automatically generate slices through the oxide layer. This is to be preferred if the oxide
geometry is non-planar.
There are a choice of two algorithms for determining the slices. The default is to calculate the
currents on slices constructed as being locally perpendicular to each semiconductor-oxide
segment on the interface. To enable the alternative algorithm, use the SHAPEOX parameter on
the MODELS statement. In this case, slices giving the shortest distance to the contact are
constructed from each node on the semiconductor-insulator interface.

267 ATLAS User’s Manual

Physical Models Physics

If you specified the Schrodinger equation using the NEW.SCHRODINGER parameter and solved
on a rectangular mesh specified by the SPX.MESH and SPY.MESH commands, then the
Quantum tunneling current must also be calculated on the rectangular mesh defined by the
QTX.MESH and QTY.MESH commands. Best results will be obtained if the mesh lines in the
direction of tunneling are roughly coincident and if the SP mesh ends at the semiconductor-
insulator interface.

Table 3-109 QTX.MESH and QTY.MESH Statements

Parameter Type Units Default

NODE Int -999

LOCATION Real microns -999
X Real microns -999
RATIO Real microns 1.0
SPACING Real microns -999

Table 3-110 MODELS Statement

Parameter Type Default

QTUNN.DIR Real 0
SHAPEOX Logical False

Contact Specifications
Carriers that tunnel through an oxide are added to the current of the electrode into or from
which they flow. If they tunnel into a polysilicon region with several contacts attached, then
the tunnel current is added to the electrode that is nearest to the segment of the oxide/
polysilicon interface across which the current is being calculated. The NEARFLG parameter in
the MODELS statement is automatically set for quantum tunneling. Therefore, the algorithm
used to obtain the nearest electrode is the same as when you set NEARFLG. To exclude any
electrode from this algorithm, set the EXCLUDE_NEAR flag in the CONTACTS statement.
You can also set the effective mass to use inside each contact for the tunneling current
calculation using the QTUNN.CMASS (electrons) and QTUNN.VMASS (holes) parameters in the
CONTACTS statement. If the contact is polysilicon, then the default effective mass is either the
conduction band or valence band density of states effective mass depending on its dopant
specification.

268 ATLAS User’s Manual

Physical Models Physics

Table 3-111 CONTACTS Statement

Parameter Type Units Default

EXCLUDE_NEAR Logical False

QTUNN.CMASS Real 1.0
QTUNN.VMASS Real 1.0

To calibrate the tunneling current, use the effective mass in the oxide region. You can set this
by using either the MC or ME.TUNNEL parameters on the MATERIAL statement for electrons and
the MV or MH.TUNNEL parameters on the MATERIAL statement for holes.
For example:
MATERIAL MATERIAL=OXIDE MC=0.6 MV=0.2
In addition to this direct tunneling model and Fowler Nordheim model, there are several
quantum tunneling models included in ATLAS for compatibility with other products. These
are mutually exclusive and the tunneling current outputs to a variable Itnl (A) or Jtnl (A/um)
regardless of model chosen.
The direct quantum tunneling current density associated with each interface node can be
output to any standard structure file using the parameters QTUNN.EL, QTUNN.HO, and
QTUNN.BBT on the OUTPUT statement. This applies to both post-processing and self-consistent
versions of the direct quantum tunneling model.
When using the direct quantum tunnelling model in conjunction with the Bohm Quantum
Potential or Density Gradient models, some correction is made for the quantum confinement
effects introduced by these models. In Equations 3-405 and 3-406, the lower limit of energy
integration is usually the band edge in the semiconductor at the interface with the insulator. It
is changed to be the ground state of the confining potential well formed by the band edge.
This is modeled as a triangular potential well. The ground state of the potential well is
approximated as
13
9 2  3   qF   2 
---   ------------------ 3-516
8  2m 0 m

relative to the minimum of the well. The quantity F is the local electric field at the interface.
If the potential is not confining, then the Field will be zero or negative, and the ground state is
set to zero. By integrating over a continuous energy range beginning at the ground state
energy, quantization effects in the channel are introduced into the gate tunnel current model
for a drift-diffusion solution.

269 ATLAS User’s Manual

Physical Models Physics

Schenk Oxide Tunneling model

Another approximate tunneling model is based on the Gundlach model [93] and includes the
effects of barrier lowering due to the image force potentials [242]. As such, it is especially
suited for tunneling through ultra-thin gate oxides. It involves mapping from the energy
barrier arising from the actual barrier profile plus the correction due to the image force to an
effective trapezoidal barrier. The exact Transmission Coefficient for a trapezoidal barrier can
then be obtained from
2
T  E  = ---------------------- 3-517
1 + g E

where

2 m k mc ks
 s c 2
g  E  = ------ ---------------  Bi' d Ai o – Ai' d Bi o  + ---------------  Bi d Ai' o – Ai d Bi' o 
2
3-518
2 m ks m k
c s c
2 2
mc ms 2
 o m ox k c k s 2
+ --------------------------------  Bi' Ai' – Ai' Bi'  + --------------------------------  Bi d Ai o – Ai d Bi o 
2 2 d o d o mc ms
 o m ox k c k s

where kc is the wavevector in the contact, ks is the wavevector in the semiconductor, mc is the
effective mass in the contact, mox is the effective mass in the oxide and ms is the effective
mass in the semiconductor.
Complication arises because the image potential is not of a simple functional form, it is given
by

2 k1 k2 2k 1 k 2
  k1 k2 
q n
E im  x  = ------------------  ----------------
- + ------------------------------
- + ---------------------
- 3-519
16  ox nd + x d  n + 1  – x d  n + 1 
n=0

with
 ox –  M  ox –  s
k 1 = ---------------------
- = – 1, k 2 = ------------------
- 3-520
 ox +  M  ox +  s

where  is the relative dielectric permittivity (of the material indicated by the subscript), d is
the oxide thickness and x is the position in the barrier. The sum can be evaluated numerically
and the potential added to the barrier potential. This allows one to evaluate the action of an
electron of incident energy E when moving through this barrier by numerically integrating the
barrier energy minus the electron energy as a function of distance between the classical
turning points of the motion.
This can be equated with the action of the carrier moving through a trapezoidal barrier with
the barrier height as a fitting parameter.
S eff  E  = S im + actual  E  3-521

270 ATLAS User’s Manual

Physical Models Physics

This results in an effective trapezoidal barrier height as a function of electron incident energy
E. This is evaluated at three different energies and these values are used to calculate the
effective barrier height as a function of electron energy. The interpolation formula used is
B  E2  – B  E0 
 B  E  =  B  E 0  + --------------------------------------------------
-  E – E0   E1 – E  3-522
 E2 – E0   E1 – E2 
B  E1  – B  E0 
– ---------------------------------------------------  E – E 0   E 2 – E 
 E1 – E0   E1 – E2 
For a given electron incident energy, you use Equation 3-522 to calculate the effective barrier
height and then use Equation 3-518 to calculate the Transmission Probability, T(E). This is
placed in an equation like Equation 3-509 and the integration over the range of tunneling
energies is carried out to give the tunneling current.
The model has been implemented as both a post-processing step and alternatively, as being
solved self-consistently with the current continuity equations. To enable the post-processing
option, specify SCHENK on the MODELS statement. To enable the self-consistent version, use
SCHKSC with the MODELS statement. You can enable them separately for electrons and holes if
required using the parameters SCHENK.EL, SHENCK.HO (post-processing) and SCHKSC.EL,
SCHKSC.HO (self-consistent) on the MODELS statement. The values of effective mass and
permittivity will affect the quantity of tunneling current as will the electron affinities and
work functions of the materials involved.

Table 3-112 Schenk Oxide Tunneling Flags

Parameter Type Default Units

SCHENK.EL Logical False

SCHENK.HO Logical False
SCHENK Logical False
SCHKSC.EL Logical False
SCHKSC.HO Logical False
SCHKSC Logical False
SCHENK.BBT Logical False

Gate Tunneling Models for SONOS Type Structures

One approach to obtaining Non-Volatile Memories with low operating Voltages is to use a
SONOS structure. The (S)emiconducting channel has a thin layer of tunnel (O)xide grown on
it, followed by a thin layer of silicon (N)itride, and then followed by a thicker blocking or
capping layer of (O)xide, and finally a (S)emiconducting polysilicon gate. The nitride has
trapping levels located within it and the nitride-oxide band offset allows charge to be
accumulated in the nitride layer.
The use of a charge-trapping Silicon nitride layer as a basis for a Non-Volatile memory device
goes back to 1967 [296]. There has recently been a lot of interest in SONOS devices because
of the continuing trend to smaller device dimensions.

271 ATLAS User’s Manual

Physical Models Physics

To achieve low-voltage, low-power operation, you need to have an ultrathin tunneling layer
so that charging by direct tunneling is possible. The presence of a defect in an ultrathin oxide
tunneling layer can cause the complete failure of a floating gate device because it will fully
discharge. In a SONOS device, the charge is stored in traps in the Silicon Nitride and
consequently a defect will have only a localized effect. Thus, the SONOS structure is
potentially more reliable [306].
ATLAS has three different SONOS models for attempting to model the behavior of these
devices.
The first is the FNONOS model that relies on using embedded floating gates in the Nitride
layer but can model capture efficiency and trap saturation.
The second is the SONOS model that assumes that the trapping of charge in the Silicon
Nitride occurs at the interface with the tunneling oxide.
The third, and most complete, is the DYNASONOS model. This model includes several
tunneling mechanisms, carrier transport and trap dynamics in the Silicon Nitride layer. It
assumes that the traps are evenly distributed throughout the Silicon Nitride layer. It is the
most complete model for SONOS devices in ATLAS.
All three models are described below.
FNONOS Model
To enable this model, specify FNONOS on the MODELS statement. To use the FNONOS model,
you must either set the whole Silicon Nitride layer as a floating contact or embed floating
contacts in the Silicon Nitride layer. Use the FLOATING flag on the CONTACT statement to to
make an electrode into a floating contact. For each point in the channel-oxide interface,
ATLAS calculates the distance to the nearest point in the Silicon Nitride layer. The tunneling
current for this point is then calculated as'
2
J n = F AE E  factor1 exp  – F · BE factor2  E  3-523
·

where
1 2
---
2
factor1 = 1 –  1 – DV  BH FNONOS  3-524
·

and
3
---
2
factor2 = 1 –  1 – DV  BH FNONOS  3-525
·

DV is the potential drop across the tunnel oxide layer and is calculated automatically by
ATLAS. The value BH.FNONOS is the Barrier height and if not specified directly ATLAS will
calculate it. This formula is calculated using WKB theory for the tunneling co-efficient
through a trapezoidal barrier.
If DV > BH.FNONOS, then factor1 and factor2 are both set to be unity. In this case, Equation 3-
523 is the same as the Fowler-Nordheim Expression (Equation 3-476).

272 ATLAS User’s Manual

Physical Models Physics

ETA.FNONOS times the tunneling current is added to the nearest floating electrode, where
ETA.FNONOS is the capture efficiency. A factor (1-ETA.FNONOS) times the tunneling current
is added to the next nearest (gate) electrode. The efficiency can depend on the Charge state of
the floating electrode itself. To enable this, set the NT.FNONOS parameter on the MODELS
statement to a positive value. The efficiency is modified by an extra term

1.0 – Q floating   qNT FNONOS  3-526

where Qfloating is the Floating gate charge density in C/micron. NT.FNONOS is the integrated
trap density /micron. If this term is negative, then zero is used instead. This allows you to
model the phenomenon of trap saturation.

Table 3-113 MODELS Statement

Parameter Type Default Units

FNONOS Logical False

ETA.FNONOS Real 1.0
BH.FNONOS Real 3.07 eV
NT.FNONOS Real 0.0 um-1

SONOS Model
This is a model of intermediate complexity for modeling SONOS devices. It assumes that the
trapped charge in the Silicon Nitride is at the interface of the Silicon Nitride and the
surrounding insulator materials. The traps can be charged up and discharged by direct
tunneling through the tunnel oxide, or by hot carrier injection (lucky electron or concannon
models).
To enable this model, you include an INTERFACE statement with the parameter N.I specified.
You must also explicitly suppress the DYNASONOS flag by specifying ^DYNASONOS. This is
because the DYNASONOS model (see below) is the default. The gate stack materials must
all be insulators. If they are changed to wide bandgap semiconductors, then the model will
not work correctly.
You can view the trapped carrier density as a sheet charge density under the Insulator
Charge field in TONYPLOT.
The tunneling current is calculated using the direct quantum tunneling Equation 3-509 with
the tunneling probability T(E) obtained using the WKB approximation. The quasi-Fermi
energy Efl in the Silicon Nitride is obtained from the energy level at the gate contact. This
value is then clipped to ensure that it lies within the Silicon Nitride bandgap. The value of Efr
is the usual value of quasi-Fermi level near the interface of the channel and the oxide. The
shortest paths between each point on the interface and the Silicon Nitride are used as the
tunneling paths.
The value of T(E) is strongly dependent on the values of effective mass in the tunnel oxide.
You can use the MC and MV parameters on the MATERIAL statement as fitting parameters.

273 ATLAS User’s Manual

Physical Models Physics

If a hot carrier gate current model is selected (HEI, HEI, N.CONCANNON, or P.CONCANNON on
the MODELS statement), then the hot carrier current charges points on the Silicon Nitride
interface in a way, which is consistent with the motion of the hot carriers in the local electric
field.
The limitations of this model are that:
• The trapped charge is constrained to be negative (holes can only erase).
• The trapped charge is localized at the Silicon Nitride boundary.
• Trap dynamics are not included.
• Charge transport in the Silicon Nitride is not included.
Its advantage is that it is a robust model and can model spatial variations in the trapped charge
density.
DYNASONOS Model
To address some of the limitations of the SONOS model, the DYNASONOS model was
developed by adapting features of various published models [224], [81], [51], [65], [82], [52].
To enable this model, you specify the N.I parameter on the INTERFACE statement. You can
also specify the DYNASONOS parameter, although this is enabled by default (it has no effect
unless you also specify N.I).
The N.I parameter causes every point on the interface between Nitride and other insulators to
be treated as a double point, allowing easier implementation of the thermionic and direct
tunneling models for these structures. This requires that the INTERFACE statement be placed
at the end of the structure specification group of statements. The current across this interface
is automatically modeled as a thermionic current.
You must also change all of the insulator materials in the gate stack to wide bandgap
semiconductors using the SEMICONDUCTOR parameter on the MATERIAL statement. You also
need to specify effective masses, mobilities, and effective densities of states in the gate stack
materials with the MATERIAL statement.
The properties of the traps in the Silicon Nitride must be set using the NITRIDECHARGE
statement. These parameters are now introduced. ATLAS models the trap states in the Nitride
as being either acceptor-like (for storing electrons) or donor-like (for storing holes), at a
single discrete energy level below the Nitride conduction band (acceptor-like) or above the
valence band (donor-like). The number of available trap states is assumed to be spatially
uniform, with the number density of acceptor-like traps being set by the NT.N parameter in
units of cm-3 and the number density of donor-like states being set by NT.P in the same units.
By default, the traps are initialized to be empty (uncharged) and are charged using a transient
SOLVE statement. It is possible, however, to set up an initial distribution of charged traps by
using the NIT.N and NIT.P parameters on the SOLVE statement. These set the trapped
electron and trapped hole densities respectively to the specified value within the limits
specified by the optional parameters: NIT.XMIN, NIT.XMAX, NIT.YMIN, and NIT.YMAX. The
default values of these parameters are the maximum extents of the Nitride region. The effect
of consecutive SOLVE statements with NIT.N or NIT.P is cumulative, allowing you to create a
complicated profile of trapped charge density. This is useful for studying known trapped
charge distributions.

274 ATLAS User’s Manual

Physical Models Physics

To simulate the charging of the Nitride layer, you use a transient SOLVE statement. The
Silicon Nitride layer can be charged by quantum mechanical tunneling or by hot carrier
injection. Two types of tunneling are modeled. The first is tunneling to the Silicon Nitride
conduction band and valence band. The second is direct tunneling to and from the trap levels.
In the former case, some of the free carriers are captured by the trapping centers. For
acceptor traps, the rate of capture of electron density is

SIGMAT · Nv th, n n  NT · N – n t  3-527

where SIGMAT.N is a capture cross-section in cm2, vth,n is the electron thermal velocity in cm/
s, n is the free electron density in cm-3 and nt is the trapped electron density in cm-3. Trapped
electrons can be re-emitted to the conduction band, the rate for this process being
nt en 3-528

where en is an emission rate in s-1. There are also rates for hole capture from the valence band

SIGMAN · Pv th, p pn t 3-529

where p is the free hole density in cm-3 and vth,p is the hole thermal velocity in cm/s. To get an
equilibrium solution, you need to include an hole emission term

 NT · N – n t e eq p 3-530

where the hole emission rate eeq,p is calculated from the other parameters and the equilibrium
carrier densities and cannot be independenly specified. The rate of direct trap charging due to
tunneling is given by

P  sNT · N – n t  3-531

where s is the Fermi-Dirac factor in the conduction band, and the tunneling rate per trap is P
in units of per second (see Equation 3-534).
Putting these terms together we obtain an expression for the rate of the change of the local
trapped electron density.

dn
-------t = SIGMAT · Nv th n n  NT · N – n t  – SIGMAN · Pv th p pn t +  NT · N – n t  3-532
dt
e eq p – n t e n + P  sNT · N – n t 

The equivalent expression for the rate of change of trapped hole density is
dp
-------t = SIGMAT · Pv th, p p  NT · P – p t  – SIGMAP · Nv th, n np t +  NT · P – p t  3-533
dt
e eq n – p t e p + P  sNT · P – p t 

275 ATLAS User’s Manual

Physical Models Physics

The tunneling terms are zero for any tunneling processes, which start or end in the band gap
of the channel material. We now consider electron trapping. To charge the traps, there must be
electrons in the Silicon Nitride and these are generated by direct quantum tunneling, shown
schematically as process A in part a) of Figure 3-14, where the tunnelling transitions are
allowed at energies greater than the conduction band energy in the channel. Any excess
electrons in the Silicon Nitride may tunnel to the contact as shown schematically as process
B. Clearly tunneling directly into the Silicon Nitride traps is not possible in this band lineup.
For a band lineup during a typical erase operation, as in part b) of Figure 3-14, direct trap-to-
channel tunneling is possible (process C). In process D, trapped electrons are emitted to the
Silicon Nitride conduction band, drift to the interface with the tunnel oxide and then tunnel
out of the Silicon Nitride in process E. Some electrons may be injected from the contact as
shown in process F. There are analogous process for hole traps.

276 ATLAS User’s Manual

Physical Models Physics

Figure 3-14: Schematic of band structure and processes for SONOS model
For trap-to-channel tunneling, the range of states in the channel corresponding to the
tunneling to or from a specific trap energy are such that the total energy must add up to the
energy at the trap level. This means that we do an integral over transverse wavevector k||,
with the constraint that the transverse and perpendicular energies sum to the value of the trap
energy. The electron band-to-trap tunneling rate in units of s-1 is
2
4  m nitride  Hk  2
2
m oxide

P = ---------------------------- ------------------ WKB  E k ||  -------------------------- k || dk ||
0 2 + k2  2 + 3 
2
3-534
1 

277 ATLAS User’s Manual

Physical Models Physics

2 nitride2m
- ELEC · DEPTH k  is the perpendicular wavevector at the channel side
where H = ----------------------
2

of the Channel/Oxide interface and 1 is the evanescent wavevector at the Oxide side of the
Channel/Oxide interface. Similarly, 2 and 3 are evanescent wavevectors evaluated at the
Nitride/Oxide interface and WKB(E, k||) is the product of the WKB tunneling probabilities
through the oxide and through part of the Nitride to the trap position. This is derived by
extending the result of Lundstörm et al. [168] to a realistic bandstructure.
The electron emission rate (for acceptor traps) and hole emission rate (for donor traps) are
important in determining the trapped carrier density after charging, and also the importance of
process D in Figure 3-14. These rates are either set using the TAU.N and TAU.P parameters on
the NITRIDECHARGE statement, or by enabling the Poole-Frenkel detrapping model by
specifying the PF.NITRIDE flag on the MODELS statement. In the former case, the emission
rates are constants en = 1/TAU.N and ep = 1/TAU.P. In the latter case, the emission rates depend
on the local value of field in the device. For electrons

ELEC DEPTH – q qF  - 

e n = PF · B exp  – -------------------------------------------------------------------
· 3-535
 K BT 

where ELEC.DEPTH is the depth of the electron traps below the conduction band, F is the
local electric field in V/m and  is the dielectric permittivity of the Silicon Nitride in F/m.
The overall rate is proportional to PF.B, which has a default value of 1013/s.
If you want to use the Poole-Frenkel model without modifying ELEC.DEPTH, then you can
use PF.BARRIER. If PF.BARRIER is specified, then it is used in place of ELEC.DEPTH in
Equation 3-535.
The parameters that are set on the NITRIDECHARGE statement are summarized in Table 3-114,
along with their default values. Parameters that are relevant to electron traps (acceptor-like)
are only used if NT.N is greater than zero, and parameters that are relevant to hole traps
(donor-like) are only used if NT.P is greater than zero.

Table 3-114 NITRIDECHARGE Statement

Parameter Type Default Units

NT.N Real 0.0 cm3

NT.P Real 0.0 cm3
TAU.N Real 1.0e300 s
TAU.P Real 1.0e300 s
ELEC.DEPTH Real -999.0 eV
HOLE.DEPTH Real -999.0 eV
SIGMAN.P Real 1.0e-15 cm2
SIGMAP.N Real 1.0e-14 cm2
SIGMAT.N Real 1.0e-16 cm2

278 ATLAS User’s Manual

Physical Models Physics

Table 3-114 NITRIDECHARGE Statement

Parameter Type Default Units

SIGMAT.P Real 1.0e-14 cm2

PF.BARRIER Real eV
PF.B Real 1.0e13 Hertz

In the Silicon Nitride, the fully transient current continuity equations are solved, self-
consistently along with the Equations 3-532 (if NT.N is non-zero) and 3-533 (if NT.P is non-
zero) for the trap occupancies. The Poisson Equation is also solved to self-consistently
include the effect of nitride trap charging.
The default tunneling path is determined as being the shortest distance between each node on
the Nitride–Tunneling Insulator interface and the channel, or the shortest distance between
each node on the Nitride–Blocking Insulator interface and the gate. The path chosen must be
outwardly oriented. In other words, it cannot cross any part of the Nitride layer itself.
For some geometries evaluating the tunnelling over only the shortest distance path is too
restrictive, as for example in Figure 3-15. ATLAS allows you to specify a range of angles for
tunneling path. These angles are relative to the shortest distance path. If you specify
NOS.ANGLE on the INTERFACE statement, then all valid tunneling paths within an angle of ±
NOS.ANGLE degrees will be considered for every point on the interface between the Nitride
and the tunneling oxide. For points on the interface between the Nitride and the Blocking
oxide, the same algorithm is used if the NOM.ANGLE parameter is specified on the MODELS
statement. In Figure 3-15, point A on the nitride/oxide interface has its shortest tunneling
path to point C. If NOS.ANGLE is specified, then it will also consider linear tunneling paths
between A and points like B and D, so long as the angle, , between these paths and A-C is
less than NOS.ANGLE.
To restrict the calculation to paths that have a significant contribution, the parameters
NOS.DIST and NOM.DIST are available. These reject any tunneling path longer than
NOS.DIST or NOM.DIST microns. Although the contribution of a tunneling path depends on
other factors as well as tunneling path length, applying this restriction eliminates much
unnecessary calculation. ATLAS prints out the average number of tunneling paths from the
nitride interface to both semiconductors and contacts. This information can be used to fine
tune the values of NOS.ANGLE, NOM.ANGLE, NOS.DIST, and NOM.DIST.

279 ATLAS User’s Manual

Physical Models Physics

Figure 3-15: Use of NOS.ANGLE parameter. All linear paths so that O1 <NOS.ANGLE or O2 <NOS.ANGLE are
considered.
In order to simulate hot carrier injection into the Silicon Nitride, you may enable the existing
hot carrier models HEI/HHI for drift diffusion simulations or N.CONCANNON/P.CONCANNON
with Energy balance simulations. In general, N.CONCANNON and P.CONCANNON should give
physically more reasonable results. An additional model is available to use in drift-diffusion
simulations. It is enabled by specifying N.HOTSONOS (for electrons) and P.HOTSONOS (for
holes) on the MODELS statement. It uses the parallel field along the channel/insulator interface
to obtain an effective carrier temperature using the formulae:

T n = F IG · LRELE   1.5K B 
3-536
T p = F IG · LRELH   1.5K B 

where F is the parallel field along the interface.

It then uses these effective carrier temperatures in the Concannon expressions for Hot carrier
gate current.

280 ATLAS User’s Manual

Physical Models Physics

The final distribution of trapped charge depends on the distribution of carrier generation and
recombination, the parameters of the trap equation and the simulation time. The trapped
carrier concentrations are output automatically to any structure file as Trapped Insulator e-
Concentration and Trapped Insulator h+ Concentration. The instantaneous rates of carrier
generation in the Silicon Nitride layer can be output to any structure file by specifying
SONOS.RATES on the OUTPUT statement. Both through barrier direct tunneling and trap-to-
channel tunneling are included and are given as a generation term in units of cm-3 s-1. The net
rate of electron capture from the conduction band and hole capture from the valence band are
calculated from Equations 3-532 and 3-533 and output as recombination rates in units of cm-3
s-1 if SONOS.RATES is specified.
You can obtain the net trapped charge density (electron density - hole density) by specifying
the SONOS.CHARGE statement on the PROBE statement. Furthermore, you can view the overall
tunneling current injected into the Nitride layer from the channel as a function of time by
specifying SONOS.CURR on the LOG statement. Choosing this option will also output the
tunnel current entering the external contact from the Nitride layer. The are refered to as the
SONOS Tunneling Insulator Current and SONOS Blocking Insulator Current respectively
and are given in units of A/m.
In steady state, the trapped charge in the Silicon Nitride layer remains fixed at its charged
value, allowing you to do threshold voltage shift calculations. If the gate stack materials
remain as wide bandgap semiconductors, the large stored fixed charge combined with weakly
coupled quasi-Fermi levels mean that SOLVE INIT may wrongly introduce a compensating
free carrier charge. To avoid this, either revert the gate stack materials to insulators or specify
the SONOS flag on the SOLVE statement when you also specify INITIAL.
The BESONOS model is an extension of the DYNASONOS model to SONOS devices with
Band-Engineered tunnel layers. It is enabled by setting the parameter BESONOS on the
INTERFACE statement together with the DYNASONOS parameter. The BESONOS model uses
the direct quantum tunneling model for calculating the tunneling current through the layered
insulator tunnel stack. This model is described in Section 3.6.7 “Gate Current Models” and
can calculate tunnel current through a stack of materials with different band offsets and
effective masses.
Some Band-Engineered tunnel layers may contain ultra-thin Nitride layers, which do not trap
charge. In order to specify which Silicon Nitride regions to charge in the BESONOS model,
you must use the TRAPPY parameter on the REGION statement. Only Silicon Nitride regions
that have the TRAPPY parameter set will be able to have trapped charge stored in them. The
TRAPPY parameter is only active for the BESONOS model.

Table 3-115 Parameters Relevant to the SONOS MODEL

Statement Parameter Type Default

INTERFACE BESONOS Logical False

INTERFACE N.I Logical False
INTERFACE DYNASONOS Logical True
LOG SONOS.CURR Logical False

281 ATLAS User’s Manual

Physical Models Physics

Table 3-115 Parameters Relevant to the SONOS MODEL

Statement Parameter Type Default

INTERFACE NOS.ANGLE Real 0.0

INTERFACE NOM.ANGLE Real 0.0
INTERFACE NOS.DIST Real 1.0
INTERFACE NOM.DIST Real 1.0
MODELS N.HOTSONOS Logical False
MODELS P.HOTSONOS Logical False
OUTPUT SONOS.RATES Logical False
PROBE SONOS.CHARGE Logical False
REGION TRAPPY Logical False
SOLVE NIT.N Real 0.0
SOLVE NIT.P Real 0.0
SOLVE NIT.XMAX Real 0.0
SOLVE NIT.XMIN Real 0.0
SOLVE NIT.YMAX Real 0.0
SOLVE NIT.YMIN Real 0.0
SOLVE SONOS Logical False

Floating Gate to Control Gate (FGCG) Current

In a conventional Flash memory device structure, it is usual to neglect the tunneling current
between the control gate and the floating gate. This is because the insulating layer separating
the floating gate from the channel (tunnel oxide) is much thinner than that separating the
floating gate from the control gate (barrier oxide). This will typically result in the dominance
of the channel to floating gate tunneling current, at least under normal operating conditions.
As flash memory technology develops, it may be necessary to quantify the floating gate to
control gate tunneling currents [65]. ATLAS incorporates this capability with a new model
called the FGCG model. This model calculates the tunneling currents between the floating
gate and the semiconducting channel and between the floating gate and the control gate. Both
of these currents can be taken into account when calculating the charge state of the floating
gate. The model also applies to more complicated geometries where there are several floating
gates and/or several control gates.
The tunneling currents are calculated using the direct tunneling formula, Equation 3-509,
with the Transmission co-efficient being obtained using the WKB approximation.
This formula includes the carrier statistics and automatically determines the sign of the
currents using the Fermi levels in the gates and quasi-Fermi in the channel.

282 ATLAS User’s Manual

Physical Models Physics

To enable the model for electron tunneling, use the parameter E.FGCGTUN on the MODELS
statement. To enable the model for hole tunneling, use the parameter H.FGCGTUN on the
MODELS statement.
It is also important to specify how the floating gate to control gate current is output and how it
affects the charge state of the floating gate. This is controlled by the CGTUNN parameter on the
CONTACT statement. This only has an effect on a floating gate. If it is set (default), then the
tunneling current attributed to the floating gate is obtained by calculating the current from
that floating gate to every control gate.
This is then subtracted from the channel to floating gate current to give a net current. This net
current is the one output. It is also the one used in the calculation of the amount of charging
during a transient simulation.
If CGTUNN is explicitly cleared for a floating gate, then only the channel to floating gate
current is used for the charging calculation and output.

Note: The tunneling current at a control gate is the sum of channel to control gate current, plus the floating gate to
control gate current from all floating gates in the device. This is not affected by the CGTUNN parameter
because the CGTUNN parameter is only effective for floating contacts.

A specific example is shown in Figure 3-16, where there are 3 control gates and 2 floating
gates. The following CONTACT statements are issued:
CONTACT NAME=fgate1 FLOATING WORKF=4.28 CGTUNN
CONTACT NAME=fgate2 FLOATING WORKF=4.28 ^CGTUNN
so that CGTUNN is set for fgate1 and explicitly cleared for fgate2. The currents attributed to
each electrode are calculated by ATLAS as follows:
jfgate1 = jchannel1 - jcg1fg1 - jcg2fg1 - jcg3fg1
jfgate2 = jchannel3
jcgate1 = jcg1fg1 + jcg1fg2
jcgate2 = jchannel2 + jcg2fg1 + jcg2fg2
jcgate3 = jcg3fg1 + jcg3fg2

283 ATLAS User’s Manual

Physical Models Physics

Figure 3-16: Example Structure that can be Modeled with FGCG Model, Showing the Current Components
The intention is that the model is used with CGTUNN set on every floating gate. You are,
however, given the ability to see the effect of leaving out the control gate components of the
floating gate current by clearing CGTUNN.
When used in transient mode, the floating gate tunneling currents are integrated over time to
give the charges on the floating gates. Because the floating gate charge affects the floating
gate voltage, which in turn affects the floating gate to control gate current, there is strong
coupling between floating gate to control gate current and floating gate charge. For large
timesteps this can cause numerical instability.
One approach to avoid this problem is to control the maximum timestep allowed using the
DT.MAX parameter on the METHOD statement. The recommended alternative, however, is to
specify FGCNTRL on the METHOD statement, which invokes an algorithm to control step size
based on the relative size of the charging current. There are two parameters that can be used
to control the behavior of this stabilization algorithm. One is QMAX.FGCNTRL, which is in
units of electronic charge. If the charge on the floating gate is less than QMAX.FGCNTRL, then
the algorithm will not be applied. The other parameter is RAT.FGCNTRL. This controls how
much the timestep is reduced by the algorithm. Reduce it from the default value of 1.0 to give
better control, but longer runtime.

284 ATLAS User’s Manual

Physical Models Physics

For example, the following statement will enable the stability algorithm for those floating
gateshaving an absolute charge of more than 8.0110-17 Coulombs, and give a high level of
stability due to the low value of RAT.FGCNTRL.
METHOD FGCNTRL RAT.FGCNTRL=0.05 QMAX.FGCNTRL=500
The parameters E.FGCGTUN and H.FGCGTUN enable the FGCG model. This will not include
the current tunneling from the semiconducting channel into the current continuity equations
in the semiconductor. The parameters E.SC.FGCGTUN and H.SC.FGCGTUN enable the FGCG
model. This will include this current in the current continuity equations. If you set also
CGTUNN is for every floating gate in the device, then overall current continuity for the device
will be attained when you set the E.SC.FGCGTUN and H.SC.FGCGTUN parameters.
Table 3-116 shows the parameters used in the FGCG model.

Table 3-116 FGCG Model Parameters

Statement Parameter Default Units

MODELS E.FGCGTUN False

MODELS H.FGCGTUN False
MODELS E.SC.FGCGTUN False
MODELS H.SC.FGCGTUN False
CONTACT CGTUNN True
METHOD FGCNTRL True
METHOD RAT.FGCNTRL 1.0
METHOD QMAX.FGCNTRL 100

Metal-Insulator-Metal Direct Tunneling

The simulation of some classes of device, for example Metal-Insulator-Metal tunnel diodes,
requires the calculation of the tunneling current between electrodes. ATLAS has the
MIMTUN model for this class of device. The MIMTUN model calculates the tunneling paths
between electrodes and obtains the tunnel currents using the Direct Quantum Tunneling
model of Equation 3-509.
The quasi-Fermi levels are defined by the biases on the electrodes. The tunneling current can
be calculated through a single or multi-layer insulator stack. The insulator materials can also
be modeled as wide bandgap semiconductors using the SEMICONDUCTOR parameter on the
MATERIAL statement. It is possible to model electrodes having different workfunctions by
using the WORKF parameter on the CONTACT statement.
To enable this MODEL, specify MIMTUN on the MODELS statement.
It is recommended that the effective masses of the insulator materials are explicitly assigned
using the MC and MV parameters on the MATERIAL statement.

285 ATLAS User’s Manual

Physical Models Physics

MIMTUN enables tunneling for both holes and electrons, including the band-to-band
component. To enable tunneling for electrons above the Fermi energy, specify MIMTUN.EL on
the MODELS statement. To enable tunneling for holes below the Fermi energy, specify
MIMTUN.HO on the MODELS statement. To model the tunneling of electrons from below the
Fermi-level on the more negatively biased electrode to hole states above the Fermi energy on
the more positively biased electrode, specify MIMTUN.BBT. This is usually the dominant
component.
Band structure data is interpolated onto the tunneling paths. The number of regularly spaced
points per tunneling path can be specified. Use the MIMSLICEPTS parameter of the MODELS
statement to change the number of points.

Table 3-117 Parameters for the MIMTUN Model

Statement Parameter Type Default

MODELS MIMTUN Logical False

MODELS MIMTUN.EL Logical False
MODELS MIMTUN.HO Logical False
MODELS MIMSLICEPTS Real 51

Ielmini Inelastic Trap Assisted Tunneling Model

A common problem with non-volatile memory devices is the increase in tunneling current
after a number of write-erase cycles. This Stress-Induced Leakage Current (SILC) is present
even at low biases and causes degradation of the device retention time. It is believed to be
mediated by traps that are created in the gate insulator by the repeated electrical stressing. The
current is therefore assumed to be a two-step tunneling process, where the trap levels form an
intermediate state for the tunneling process. Ielmini et al developed a model that can take into
account both transient [119] and steady state simulations [120]. This model does not simulate
the creation of oxide traps under stressing, rather it allows the simulation of the SILC current
for a predefined trap distribution. We consider a cross-section through a typical gate stack
with the gate at x = 0 and the channel at x = L. The gate stack may consist of one or more
regions of insulator or wide bandgap semiconductor.
For a trap with a discrete energy level Et(x) and a spatial distribution Nt(x) [cm-3], where x is
its spatial position within the gate insulator stack, four current contributions per carrier type
can be defined. These are capture from the gate, emission to the gate, capture from the
channel, and emission to the channel. The relationship between capture and emission
currents, allowing for inelastic tunneling is derived by Ielmini et al [119]. Using this
relationship for electrons, the current from the left hand side into an infinitesimal width x of
the traps at position x can be written as
J L  x  = QN t b L  E t E Fl x   f L  E t E Fl x  – f t  E t x  x 3-537

The effective capture rate from the left, bL, is obtained as

286 ATLAS User’s Manual

Physical Models Physics

4m K B T E max –  E – E Fl 
log  1 + exp --------------------------  W  E x  dE
3
h f L  E t E Fl x  E min

b L = --------------------------------------
 KB T 
3-538

where  is the trap cross-section, m* is the effective mass, KB, h are Boltzmann’s constant and
Planck’s constant respectively, and T is the device temperaure. The integrand consists of a
statistical factor depending on the left hand side quasi-Fermi level, Efl, and a tunneling
probability factor W(E, x) for tunneling to point x. Similarly, the expression for the current
from the right hand side into the same infinitesimal width x of the traps at position x is
J R  x  = QN t b R  E t E Fr x   f R  E t E Fr x  – f t  E t x  x 3-539

where bR is the effective capture rate from the right, given by

4m K B T E max –  E – E Fr 
log  1 + exp --------------------------  W  E L – x  dE
3
h f R  E t E Fr x  E min

b R = ---------------------------------------
 KB T 
3-540

where W  E L – x  is the tunneling probability factor from position L to position x. The

probabilities fL and fR are the probabilities of electrons being at energy Et at the left and right
edges of the oxide respectively. The probability of the trap being occupied by an electron is ft.
The minimum value of energy Emin in the evaluation of bL is the maximum of either the trap
energy at position x or the electron energy at the left hand side of the oxide. The maximum
energy is the greater of the conduction band energy at position x or the conduction band
energy at the left hand side of the oxide. The situation is analagous for the evaluation of bR.
The expressions for bL and bR differ from those in Ielmini et al [119] by the factors fL and fR
being included in the denominator in order to avoid underflow errors for trap energies a long
way outside the Silicon bandgap.
For the traps of infinitesimal width x about x and in the general time dependent case, the
charge continuity gives
f t  E t x 
QN t  x  x ---------------------- = J L + J R 3-541
t

In steady state, the trap occupancy is constant and so the condition reduces to JL + JR = 0,
giving the steady state trap occupancy at each position.
fL bL + fR bR
f t = ---------------------------
- 3-542
bL + bR

Thus, the total trap assisted tunneling current can be evaluated by using the value of ft(x) in
Equations 3-537 and 3-539 and summing the contributions from each length x at position xi

J ITAT =  JL  xi  = – JR  xi  3-543
i i

where the sum is across the oxide thickness.

There is an analagous expression for hole currents (working with hole occupation
probabilities and hole energy) and the models that consider electron and hole current
separately are refered to as the ITAT models. For traps that coincide with the bandgap of the

287 ATLAS User’s Manual

Physical Models Physics

channel material, there is a greater likelihood of the trap level bringing about recombination.
Therfore, the RTAT model can also be used. This takes into account both the electron and hole
currents from left and right (JLe, JLh, JRe, JRh, and so Equation 3-541 becomes
f t  E t x 
QN t  x  x ---------------------- = J Le – J Lh + J Re – J Rh 3-544
t

Thus, the probability of a trap containing an electron becomes

f Le b Le  x  + f Re b Re  x  + b Lh  x   1 – f Lh  + b Rh  x   1 – f Rh 
f t  x  = -----------------------------------------------------------------------------------------------------------------------------------------------
- 3-545
b Le  x  + b Re  x  + b Lh  x  + b Rh  x 

where fLe is the probability of finding an electron at energy Et(x) at the left hand contact and
fLh the probability of finding a hole at that energy. In general, it impossible to write fLe= 1 - fLh
because the quasi-Fermi levels for electrons and holes are in general different. The same
applies to fRh and fRe, and after substituting the value of ft(x) into the current equations we
obtain

J RTAT =
  JLe  xi  – JLh  xi   = –  JRe  xi  – JRh  xi   3-546
i i

The charge of the carrier type is correctly included when evaluating Equation 3-546 (i.e., an
equal flux of electrons and holes will result in zero current). The implementation of this
model in ATLAS requires an electrode on one side of the oxide and a semiconducting channel
on the other. It automatically detects the gate oxide and sets up a QTREGION automatically.
You do not need to use the QTREGION statement, but you can control the number of
interpolated mesh points on each slice through the oxide by using the TAT.SLICEPTS
parameter on the MODELS statement. The larger this number, the greater the resolution of the
integrand in Equations 3-543 and 3-546. This number should be chosen so that the calculated
currents are largely unchanged if one uses a slightly higher or lower value of TAT.SLICEPTS.
It is possible to treat the ITAT or RTAT current as a post-processing current, where it is not
coupled to the continuity equations. To do this, specify ITAT.PP.EL on the MODELS statement
for electron current, ITAT.PP.HO for hole current, or RTAT.PP for electron and hole current,
where the trap can act as a recombination center. To couple the tunneling current self-
consistently with the current continuity equations instead, specify ITAT.SC.EL on the
MODELS statement for electron current, ITAT.SC.HO for hole current, or RTAT.SC for both
electron and hole current with recombination. Selecting the ITAT model for both electrons
and holes is the same as selecting the RTAT model. Coupling both electron and hole tunneling
currents to a trap state means that recombination of electrons and holes must be taken into
account. This is exactly what the RTAT model achieves.
The trap densities and levels are set up in the usual way (see Section 3.3.3 “Traps and
Defects”) by using either the TRAPS statement or the DOPING statement, restricting the traps to
the required regions using the REGION statement. Additionally, the TAT.TRAP parameter must
be specified on the TRAPS or DOPING statement.
The charge density associated with the traps is incorporated into the Poisson equation by
default. For ACCEPTOR-like traps, the charge density is given by
–3
  x  = – Qft  x N t  x  C cm 3-547

288 ATLAS User’s Manual

Physical Models Physics

and for DONOR-like traps, it is

–3
  x  = Q  1 – f t  x  N t  x  C cm 3-548
It is possible to disable the inclusion of the charge in the Poisson equation by explicitly
clearing the flag TAT.POISSON on the MODELS statement (see the examples below). This flag
is enabled by default.
For a transient simulation, it is necessary to include the dynamic nature of the trap charging
and discharging. This is done by using Equation 3-544 to propagate the value of ft(x) with
time. It is therefore time dependent and consequently so are the currents and the trapped
charge. In this case, the particle currents obtained by the integrals in Equations 3-543 and 3-
546 are no longer equal. So the gate current and current injected into the channel are generally
different. Integrating the Poisson equation through the oxide from x=0 to x=L, gives
L
DL = D0 +
0   x  dx 3-549

where D(L) is the electric displacement. Differentiating this with respect to time, and using
Equations 3-544, 3-547, or 3-548, gives

D  0 - – Q N  x  f
L
D  L - = ---------------
---------------
t t 0
t  ------t dx
t
3-550

D  0 - –
= ---------------
t   JLe xi  – JLh  xi  + JRe  xi  – JRh  xi   3-551
i

where the sum is over the discrete points in the numerical integration. Therefore, the
imbalance in particle current through the oxide is exactly balanced by the displacement
current, and current continuity is assured so long as the charge due to the traps is included in
the Poisson equation. For this reason, it is not possible to disable the TAT.POISSON flag
during a transient simulation.
All of the above applies to a single discrete trap level. ATLAS allows an arbitrary number of
discrete TAT.TRAP levels for this model. A continuous trap distribution can be approximated
by a sum of discrete TAT.TRAP levels with different energies and concentrations. A
continuous trap distribution having Gaussian dependence on energy can also be set up using
the DEFECTS statement. You set the flag TAT.TRAP and specify either all of NGA, EGA, WGA for
acceptors or all of NGD, EGD, WGD for holes. The densities are calculated according to
Equations 14-4 and 14-5 with energy for acceptors taken relative to the conduction band and
energy for donors taken relative to the valence band. You specify the trap capture cross-
sections for acceptors by SIGGAE, SIGGAH and for donors by SIGGDE, SIGGDH. The number
of energy levels used in the integration is NUMA for acceptors and NUMD for DONORS. The
default energy range is the entire bandgap, but this can be reduced by specifying the lower
and upper limits using the LIMIT1 and LIMIT2 parameters on the DEFECTS statement.
Spatial variation of the TAT.TRAP density can be included by using the functionality of the
DOPING statement for including analytical or user-defined doping profiles.
The tunneling model assumes by default that tunneling is into a crystalline semiconductor. If
the semiconductor has localized energy states in the mobility gap, as is the case with

289 ATLAS User’s Manual

Physical Models Physics

amorphous semiconductors, you can enable the TAT.LOCAL flag on the MODELS statement to
include tunneling to localized energy states as defined on the DEFECTS statement.
The expression for bL for electrons within the mobility gap is
* E min
4m K B T
2 
b L = -----------------------------------------
h qfL  E t ,E Fl ,x  Etransition
v thn   donor  1 – f donor  E  g donor  E  3-552

+  acceptor f acceptor  E g acceptor  E  W  E x  dE

where the quantities f donor and f acceptor are the occupancy fractions of the localized states,
v thn is the thermal velocity in cm/s,  acceptor and  donor are capture cross-sections in cm2
and g donor and g acceptor are the overall densities of states distributions of donor states and
acceptor states respectively in units of cm-3eV-1. The value of trap occupancy in the channel,
f L  E t E Fl x  is evaluated at each tunnelling energy by getting the average of the defect
occupation probabilities at that energy. If there are no defects at a particular tunnelling
energy, the occupation probability is obtained from the Fermi-Dirac distribution function.
For hole tunneling, the same expression applies but with  1 – f donor  replaced by f donor and

f acceptor replaced by  1 – f acceptor  and with different values of capture cross-section and
thermal velocity. Above the mobility gap transition energy, E transition , the usual expression
*
4m K B T E int  E – E fl 
log  1 + exp ---------------------  W  E x  dE
3 
b L = ------------------------------------
h f L  E t E fl x  E transition
 KB T 
3-553

is used. The overall value of b L obtained is the sum of the individual values of b L from
Equations 3-25 and 3-26. If the tunneling from the opposite side of the barrier is from a
contact or a crystalline semiconductor, then the expression is unchanged from Equation 3-
540. Otherwise, it will be evaluated in a similar way to Equation 3-25 using the localized
defect densities and occupation fractions. Once these factors have been obtained, the
calculation of current proceeds as before. If the self-consistent versions of the models are
chosen, then the current is injected into the continuity equations in the semiconductor. This
affects the defect occupations and consequently the tunneling rate, resulting in strong
coupling, which can give poorer convergence of the non-linear solver. You can specify the
TAT.NLCURRS flag on the MODELS statement to improve non-local coupling in the system
matrix and convergence. TAT.NLCURRS may help to bring about convergence for tunneling
into crystalline semiconductors as well as amorphous ones.
The results of the model can be visualized in the run time output. Optionally, the tunneling
current associated with each contact is output to log files if the parameter J.TUN is specified
on the LOG statement. For the self-consistent version of the model, the current injected into
the channel is automatically output to Silvaco standard structure files. These are visualized
under Nonlocal e- tunneling rate and Nonlocal h+ tunnelling rate in TONYPLOT. Also, output
automatically to the structure files are the trap density and trap ionized density for each trap
separately. The trap occupation fraction can also be output for each trap by specifying
TRAPS.FT on the OUTPUT statement.

290 ATLAS User’s Manual

Physical Models Physics

The TAT.TRAP levels set up by the DEFECTS statement are output individually to a standard
structure file. If the CONTINUOUS parameter is also specified, then the energy integrated value
of density of states and ionized trap density are also output as an extra trap state. If the
TRAPS.FT parameter is specified on the OUTPUT statement, then the the average trap density,
defined as the ratio of integrated ionized density to integrated density of states, is also output.
For very high trap densities, the coupling between the Tunneling current and potential profile
across the insulator can be become so strong as to thwart convergence. One way to avoid this
is to remove the coupling by explicitly clearing the flag TAT.POISSON. A better alternative is
to use the TAT.NLDERIVS flag on the MODELS statement. This puts into the numerical solver
the couplings between the trap charge at every point with the electrostatic potential at every
other point on the same tunneling path. This can improve convergence properties at the minor
expense of taking longer for each iteration. Convergence can also be improved in the case of
high trap densities by reducing the value of the DVMAX parameter on the METHOD statement
from its default value.

Table 3-118 Parameters for the ITAT/RTAT Model

Statement Parameter Type Default

DOPING TAT.TRAP Logical False

DEFECTS TAT.TRAP Logical False
DEFECTS LIMIT1 Real
DEFECTS LIMIT2 Real
MODELS ITAT.SC.EL Logical False
MODELS ITAT.SC.HO Logical False
MODELS RTAT.SC Logical False
MODELS ITAT.PP.EL Logical False
MODELS ITAT.PP.HO Logical False
MODELS RTAT.PP Logical False
MODELS TAT.LOCAL Logical False
MODELS TAT.NLCURRS Logical False
MODELS TAT.NLDERIVS Logical False
MODELS TAT.POISSON Logical True
MODELS TAT.SLICEPTS Real 12
TRAPS TAT.TRAP Logical False

291 ATLAS User’s Manual

Physical Models Physics

Examples
// Setting up a uniform tat trap density for region number 2 via
doping statement
DOPING REGION=2 UNIFORM TAT.TRAP CONC=1E16 ACCEPTOR SIGN=1E-14
SIGP=1E-14 E.LEVEL=4.2

// Setting up the tat trap density for region number 2 via traps
statement
TRAPS REGION=2 TAT.TRAP DENSITY=1E16 ACCEPTOR SIGN=1E-14 SIGP=1E-14
E.LEVEL=4.2 DEGEN.FAC=1

//Setting up Self-consistent RTAT model, and disabling tat.poisson

flag.
// The number of meshpoints used in the integration of current over
the
// oxide thickness is increased to 21.
MODELS RTAT.SC TAT.SLICEPTS=21 ^TAT.POISSON
// Setting up an acceptor tat trap gaussian density profile with
DEFECTS statement
// The peak density is at energy of 2.0 eV below conduction band
edge and 10 (NUMA)
// traps are only created between 1.0 and 3.0 eV below the
conduction band.
// The CONTINUOUS flag means that output to structure file will
show an extra trap
// with integrated quantities.
DEFECTS NGA=1.0e22 EGA=2.0 WGA=0.5 NUMA=10 SIGGAE=1.0e-14
SIGGAH=1.0e-14
LIMIT1=1.0 LIMIT2=3.0 TAT.TRAP CONTINUOUS
Metal-Insulator-Metal Trap-Assisted-Tunneling
The physics of carrier tunneling through traps, as detailed in the previous section, is also
applicable to Metal-Insulator-Metal structures. In this case, there is no distinction between
self-consistent and post-processing versions because all current is injected directly into the
contacts. To enable the model for electron tunneling specify MIM.ITAT.EL on the MODELS
statement. To enable the model for hole tunneling, specify MIM.ITAT.HO on the MODELS
statement. To consider both electrons and holes that can recombine inside the insulator,
specify MIM.RTAT on the MODELS statement (the contact is considered as a zero bandgap
semiconductor). Depending on the dominant conduction species of the contact material, only
one of these models may be physically appropriate.
Both steady state and transient implementations of the models are available. The models can
be used concurrently with the MIMTUN models. The parameter to set the number of points in
the tunneling slices, MIMSLICEPTS, is common to both models. The trap distributions are set
up in exactly the same way as for the Ielmini model in the previous sections, using either

292 ATLAS User’s Manual

Physical Models Physics

TRAPS, DOPING, or DEFECTS statement. All other parameters have the same meaning as for
the Ielmini and MIMTUN models. Poor convergence in the case of high trap density can be
improved by using the TAT.NLDERIVS flag on the MODELS statement as described in the
previous section.

Table 3-119 Parameters for the MIM.TAT Models

Statement Parameter Type Default

MODELS MIM.ITAT.EL Logical False

MODELS MIM.ITAT.HO Logical False
MODELS MIM.RTAT Logical False
MODELS MIMSLICEPTS Real 51

3.6.8 Device Level Reliability Modeling

Hansch MOS Reliability Model
The Hansch Reliability Model [98,4,235] can be used to simulate MOS transistor degradation
under stress conditions. The causes of device characteristic degradation are the hot electron
(hole) injection into gate oxide, and the trapping of electron (hole) charge on the effective
interface acceptor (donor) like traps.
The model calculates hot electron (hole) injection current according to the lucky electron
model. The arbitrary position-dependent distributions of acceptor and donor-like traps are
specified on the oxide-semiconductor interface with corresponding capture cross sections.
The device degradation is calculated as a function of stress time by performing transient
calculations. The trap rate equation is solved on every time-step and thus the trapped electron
(hole) concentration is calculated. The rate of electron trapping can be described by the
equations:
dN n (x,t) SIGMAE
-------------------- = ----------------------  J inj, n (x,t)   NTA  x  – N (x,t)  3-554
dt q

dN (x,t) SIGMAH
----------------- = ----------------------  J inj, p (x,t)   NTD  x  – N (x,t)  3-555
dt q

where N(x,t) represents the trapped electron (hole) density, at the interface point x, at time=t
during a transient simulation. The NTA and NTD parameters represent the acceptor and donor-
like trap densities at time=0. The Jinj,n(x,t) and Jinj,p(x,t) parameters are the injected electron
and hole current densities, SIGMAE and SIGMAH are the capture cross section of electrons and
holes.
To activate this model, use the DEVDEG, DEVDEG.E, and DEVDEG.H parameters in the MODELS
statement (to account for both hot electron and hole injection, hot electron or hot hole
injection, respectively). The model parameters are user-definable on the DEGRADATION
statement.

293 ATLAS User’s Manual

Physical Models Physics

Table 3-120 User-Definable Parameters for Equations 3-554 and 3-555

Statement Parameter Units

DEGRADATION SIGMAE cm2

DEGRADATION SIGMAH cm2
DEGRADATION NTA/F.NTA cm2
DEGRADATION NTD/F.NTD cm2

The results of stress simulation can be used to calculate the characteristics of the degraded
device (the shift of the threshold voltage, transconductance degradation, and so on). You can
view the distribution of traps, hot electron (hole) current density, and trapped electron (hole)
distribution by using TONYPLOT.
The model parameters: NTA, NTD, SIGMAE, and SIGMAH can also be defined through the
C-INTERPRETER functions: F.NTA, F.NTD, F.SIGMAE, and F.SIGMAH. This allows you to
define these values as functions of their position (x,y) along the insulator-semiconductor
interface. These C-function libraries are also defined on the DEGRADATION statement. More
information on the C-INTERPRETER functions can be found in Appendix A “C-Interpreter
Functions”.
Reaction-Diffusion Degradation Model
The problem of the degradation of MOSFET devices with thin oxide layers under bias and
temperature stress conditions is generally accepted as being mainly associated with the
depassivation of silicon dangling bonds at the Si/SiO2 interface. These dangling bonds are
initially passivated at the end of the fabrication process by heating in an hydrogen or, more
rarely, a deuterium environment. The interface trap density is typically reduced by two orders
of magnitude by this passivation process, to around 1010cm-2 or even less [228].
The mechanisms for depassivation under device operating conditions have been the subject of
much investigation. It is found that devices passivated with deuterium show much improved
degradation properties [45] (and references therein), suggesting that the Si-D bond is much
stronger than the Si-H bond. As pointed out by van de Walle and Tuttle [288], the static
electronic bonding is the same for Si-H bonds and Si-D bonds. Therefore, the apparent
difference in bond strengths must arise from the different dynamic nature of the bonds caused
by the larger mass of the deuterium nucleus. It has been suggested that multiple excitation of
transverse vibrational states by inelastic carrier scattering can liberate the hydrogen or
deuterium atom from the bond. In the case of deuterium, the energy quanta of the excitation
are believed to couple strongly to a bulk phonon mode and thus the bending mode excitations
have a short relaxation time. In the case of hydrogen, however, the vibrational modes are only
weakly coupled to phonon states and a vibrational ladder is climbed until the bond dissociates
at a median energy of around 1.5 eV above the ground state [92]. Consequently, charge
carriers with relatively low energy may contribute to the depassivation process, and carriers
with sufficient energy to be injected into the gate oxide are not primarily responsible for this
generation of interface traps [45].

294 ATLAS User’s Manual

Physical Models Physics

Another suggested reaction for the depassivation of the Si-H bond is that the bond first
captures an inversion layer hole that reduces its binding energy [147]. Thereby, increasing the
rate of thermal dissocation. The model implemented here allows depassivation by one or a
combination of hot channel current, tunneling current, oxide field, or inversion layer hole
capture [171]. The mode of stressing determines which is the most significant mechanism.
An additional possibility, not considered here, is that a proton and an electron react with the
passivated bond to create a trap and an H2 molecule [205].
The DEVDEG.RD model assumes that atomic, neutral hydrogen is created by the depassivation
and that the trap becomes charged to ±Q Coulombs almost instantaneously. This is consistent
with the extensive evidence that the dangling bonds, also known as Pb centers, are the origin
of the interface charge [218]. The atomic hydrogen can diffuse and dimerize to H2 molecules.
The interface charge density Nit as a function of time is found by integrating the following
equation forward in time:
dN
---------it- = K f  RD · SIHTOT – N it  – RD · KR0 H N it 3-556
dt

where
• Kf is the forward reaction rate,
• RD.SIHTOT is the total available dangling bond density,
• RD.KR0 is a repassivation rate,
• H is the local density of atomic hydrogen.
This rate is a function of energy and includes contributions to bond-breaking from direct field
effects, channel hot carriers, and Fowler-Nordheim tunneling. The channel hot carrier current
is not the hot carrier current injected into the oxide, and we model it as the channel current
multiplied by the lucky electron expression for the fraction of carriers above the activation
energy E. The full expression for the forward reaction rate at a given energy, E, is

 E + E  RD · AESLOPE 
K f  E  = RD · KF0 exp  – ----------------------------------------------------------
-  3-557
 KT 

  E + RD E · OFFSET 
 1.0 + RD · E · HCCOFF exp  – ------------------------------------------------------
·
· - J n
  E  IG ELINF  
 E + RD · H · OFFSET 
+ RD · H · HCCOFF exp  – ------------------------------------------------------
- J p
  E  IG · HLINF  

+ RD · E · HCCOEFJn  FN  + RD · H · HCCOFFJ p  FN 

+ RD · INVHCOFF P 

The nominal (median) activation energy for the depassivation is given by the parameter
RD.AE. Kf(E) can either be evaluated at this discrete energy or integrated over the distribution
function

295 ATLAS User’s Manual

Physical Models Physics

1
G  E  = ------------------------------------------------------------------------------------------------ 3-558
  E – RD AE  
2RD · AEVAR  1 + cosh  ---------------------------- · - 
  RD · AEVAR 

where RD.AEVAR is the energy width parameter for the distribution of activation energies. If
RD.AEVAR = 0, then the forward reaction rate is evaluated only at RD.AE. In the above
expression, E is the field in the oxide perpendicular to the interface and E  is the field
along the semiconducting channel. Jn and Jp are the electron and hole channel currents and
J(FN)n and J(FP)p are the Fowler-Nordheim tunneling currents for electrons and holes
respectively. The hole density at the interface is given by P. The overall magnitudes of the
different terms can be controlled by setting the values of their prefactor coefficients, as well
as other relevant parameters. For example, the parameters IG.ELINF and IG.HLINF are the
optical phonon scattering mean free path lengths for electrons and holes respectively, as also
used by the HEI and HHI models (Section 3.6.6 “Band-to-Band Tunneling”). The
exponential prefactor is given a linear dependence on oxide field as more complicated
behavior seems quite contrived [92]. The overall depassivation rate is controlled by the
prefactor RD.KF0 and the repassivation rate (by released hydrogen) is controlled by RD.KR0.
An initial uniform density of depassivated bonds can be set using the RD.NIT0 parameter. All
of the parameters prefixed with RD are specified on the DEGRADATION statement, which must
follow the MODELS statements and have defaults values as below.

Table 3-121 DEGRADATION statement

Parameter Type Default Units

RD.SIHTOT Real 1012 /cm2

RD.NIT0 Real 0.0 /cm2
RD.KF0 Real 10-5 s-1
RD.KR0 Real 310-9 cm3s-1
RD.AE Real 1.5 eV
RD.AESLOPE Real -5.610-8 [Q=1]cm

RD.AEVAR Real 0.1 eV

RD.E.HCCOEF Real 0.0 cm2/A
RD.H.HCCOEF Real 0.0 cm2/A
RD.E.FNCOEF Real 0.0 cm2/A
RD.H.FNCOEF Real 0.0 cm2/A
RD.INVHCOEF Real 0.0 cm3

296 ATLAS User’s Manual

Physical Models Physics

Table 3-121 DEGRADATION statement

RD.E.OFFSET Real 0.0 eV
RD.H.OFFSET Real 0.0 eV
RD.COUPLED Logical True

This form of the forward depassivation rate can therefore model the effects of Channel Hot
Carrier stress, Fowler-Nordheim stress, and Negative Bias Temperature Instability (NBTI)
stress. In this model, each depassivation event releases an atom of hydrogen and creates an
amphoteric interface trap. The trap is immediately charged and the released hydrogen does
one of three things. It can diffuse away as atomic hydrogen in any direction, it can react with
another hydrogen to create molecular hydrogen, or it can be captured by another depassivated
bond, thereby repassivating it. The repassivation rate is controlled by the parameter RD.KR0
as seen in Equation 3-556. Atomic Hydrogen (H) and Molecular Hydrogen (H2) are allowed
to react and diffuse according to the equations:
2
H  H-
------- = D0 · H1 exp  – -----------------
EA · H1 --------- 2
- – H1TOH2RATE H + H2TOH1RATE H 2 3-559
t  KT  x 2

2
H  H2
---------2- = D0 · H2 exp  – -----------------
EA · H2 ------------ - – 0.5H2TOH1RATE H 2 + 0.5H1TOH2RATE H
2
- 3-560
t KT x
2

On the Si/SiO2 interface, Equation 3-559 has an extra hydrogen generation term given by the
rate of trap creation and transformed into an volume rather than an area generation rate. The
diffusion coefficients and reaction rates can be set on a material by material basis through the
MATERIAL statement. The material dependence of diffusivity was found to be necessary to
explain some observed results of NBTI degradation [104, 146]. The hydrogen diffusivities are

given by D0 · H1 exp  – EA · H1- and D0 · H2 exp  – EA

----------------- · H2- for atomic and molecular hydrogen
 -----------------
KT  KT 
respectively. The reaction rates are H1TOH2RATE H -> H2 and H2TOH1RATE H2 -> H, and the
surface recombination velocities at the device exteriors are H1SRV and H2SRV respectively.
These Hydrogen parameters are specified on the MATERIAL statement and the default values
used when material specific ones are absent are given below:

Table 3-122 MATERIAL statement

Parameter Type Default Units

D0.H1 Real 1.7510-4 cm2/s

D0.H2 Real 1.7510-4 cm2/s

EA.H1 Real 0.1685 eV

EA.H2 Real 0.1685 eV

297 ATLAS User’s Manual

Physical Models Physics

Table 3-122 MATERIAL statement

Parameter Type Default Units

H1TOH2RATE Real 1.0E-2 cm-3/s

H2TOH1RATE Real 100.0 s-1
H1SRV Real 0.0 cm/s
H2SRV Real 0.0 cm/s

and have the following material dependent defaults (same units as in previous table):

Material D0.H1 D0.H2 EA.H1 EA.H2 H1TOH2RATE H2TOH1RATE H1SRV H2SRV

Silicon [337, 338] 110-4 2.0 1.0 0.83 1.410-3 95.4 0 0

SiO2 [276] 8.110-3 8.110-5 0.2 0.2 1.410-3 95.4 0 0

Aluminum [323] 1.7510-4 1.7510-4 0.1685 0.1685 1.0e-2 100 0 0

PolySi [112] 110-4 9.410-3 1.0 0.48 1.410-3 95.4 0 0

SiN [105, 324] 1.0 1.0 1.0 1.0 1.410-3 95.4 0 0

It is a common assumption in interface depassivation studies to neglect hydrogen diffusion in

the Silicon channel [35]. Hydrogen does diffuse in Silicon, however, and so the DEVDEG.RD
model includes it [288] with the following caveat. The diffusion of hydrogen in silicon
depends strongly on a number of factors, including dopant species, concentration and whether
the silicon is crystalline or amorphous. Other charge states H+, H- are also present [112],
leading to field dependent transport. The default values for crystalline and polycrystalline
silicon should not be regarded as definitive and possible alternatives can be found in [112].
The hydrogen equations are solved in every part of the device. The boundary conditions are
applied at all exterior boundaries of the device. The default is to apply dirichlet boundary
conditions on a contact (infinite surface recombination velocity) and zero normal gradient
boundary conditions elsewhere. This may not be realistic and therefore it is possible to
specify a surface recombination velocity for each material. To do this, you use the H1SRV and
H2SRV parameters on the MATERIAL statement. Then, the normal gradient of hydrogen flux at
an exterior surface is given by

H
------- = ------------------ exp  – -----------------
H1SRV EA · H1
- H 3-561
x D0 · H1  KT 

and

H- = H2SRV
------------------ exp  – -----------------
EA · H2- H 3-562
------
x D0 · H2  KT  2

for atomic Hydrogen and molecular hydrogen respectively. For an electrode, it is possible to
specify these parameters on the CONTACT statement.

298 ATLAS User’s Manual

Physical Models Physics

Table 3-123 CONTACT statement

Parameter Type Default Units

H1SRV Real  cm/s

H2SRV Real  cm/s

where the default of  is equivalent to Dirichlet boundary conditions (H=0, H2 =0) on the
exterior edges of contacts. The values of the hydrogen density are output to a standard
structure file. They can also be obtained using the PROBE statement. Specify H.ATOM for
atomic hydrogen and specify H.MOLE for molecular hydrogen. To enable the model, specify
DEVDEG.RD on the MODELS statement. To charge created traps with electrons specify
DEVDEG.E, and with holes specify DEVDEG.H on the MODELS statement. Specify DEVDEG.A for
amphoteric traps, where electron current creates negatively charged traps and hole current
creates positively charged traps. The default, if none of DEGDEG.H, DEVDEG.A, or DEVDEG.E
are specified, is to charge the traps with electrons. The total interface electron and hole
charges in units of C/cm as a function of time can be saved to a LOG file by specifying
DEVDEG on the LOG statement.
Faster run times may be obtained by setting the the flag RD.COUPLED equal to false. This is
not recommended when the repassivation coefficient is significant because the solution will
be less accurate. It is recommended, however, to use the LTE2STEP flag on the METHOD
statement when using the DEVDEG.RD model and to use a small value of TOL.TIME when the
repassivation rate is significant. In this case, the net repassivation rate is the difference of a
large depassivation rate and a large repassivation rate and a small step size is required to
dN
accurately solve the equation. The maximum net depassivation rate, ---------it- in units of cm-2s-1,
dt
and the position at which it occurs are automatically part of the run time output.
With a large repassivation coefficient, RD.KR0, the net depassivation rate will become small
relative to the forward and reverse passivation rates. In this case, a small stepsize is necessary
as are higher precision solutions. The optimum parameters for stepsize control should be
determined pragmatically. To achieve a small stepsize, you adjust the value of the TOL.TIME
parameter on the METHOD statement and reduce the value of CX.TOL parameter on the METHOD
statement. This results in the solution for carrier and hydrogen concentrations being
determined more precisely. It also is recommended that you also specify the flag XNORM on
the METHOD statement. In some cases, it may also be necessary to use a version of ATLAS
with higher floating point precision.

299 ATLAS User’s Manual

Physical Models Physics

Power-Law Degradation Model

In this model, interface charge arising from degradation is given an explicit dependence on
simulated time. The time factor used for evolution of electron traps is

PL · E · ALPHA
 e =  PL · E · NU0 exp  – ------------------------
PL · E · NA- t  3-563
  K T  simulated
B

where PL.E.NU0 is a rate prefactor, PL.E.EA is an activation energy for the rate constant, KB
is Boltzmann constant, and T is device temperature. The exponent PL.E.ALPHA gives the
power law dependence of the time factor, and the elapsed simulated time is tsimulated. There is
an entirely analagous expression for hole traps,

PL · H · ALPHA
 h =  PL · H · NU0 exp  – ------------------------
PL · H · NA- t  3-564
  K T  simulated
B

To enable electron trap modeling, you specify DEVDEG.E and DEVDEG.PL on the MODELS
statement. The trap charge density along the interface is then given by
Q deg  x  = – QNTA  x  e   1 +  e  3-565

where NTA(x) is the Si-H bond density and can be set by either the NTA or F.NTA parameter
on the DEGRADATION statement. The units of charge are Coulomb/cm2.
Similarly for hole traps with DEVDEG.H and DEVDEG.PL selected, the charge is
Q deg  x  = + QNTD  x  h   1 +  h  3-566

At large simulation times, the interface charge saturates at the value of the Si-H bond density.
It is recommended to use the C-Interpreter functions F.NTA and F.NTD to define position
dependent trap densities when using this model.

Table 3-124 Parameters for Power Law Degradation Model

Statement Parameter Type Default Units

DEGRADATION PL.E.NU0 Real 1.0 /s

DEGRADATION PL.H.NU0 Real 1.0 /s
DEGRADATION PL.E.EA Real 0.25 eV
DEGRADATION PL.H.EA Real 0.25 eV
DEGRADATION PL.E.ALPHA Real 0.5
DEGRADATION PL.H.ALPHA Real 0.5

The integrated charge density can be output to a log file by setting the DEVDEG parameter on
the LOG statement.

300 ATLAS User’s Manual

Physical Models Physics

Kinetic Degradation Model

An alternative model of MOSFET degradation based on [213] has been implemented in
ATLAS. Like the DEVDEG.RD model, it assumes that the depassivation of interface Si-H
bonds is the main mechanism for device degradation. The interface equation is assumed to be
dN hb
- = – k f N hb +   N – N hb 
----------- 3-567
dt

where N hb is the number of passivated Si-H bonds and N is the total number of passivated
and unpassivated bonds and k f and are rates. It is more convenient to write this in terms of
the number of depassivated bonds, Nit where
o o
N = N hb + N it = N hb + N it 3-568

where the superscript denotes the concentrations at the start of the simulation. The equation
then becomes
dN
---------it- = k f  N – N it  – k r  ------ +   N it – N it  N it
H o
3-569
dt H 
o

H
and the term  has been replaced with a more detailed expression. The term ------ is only
Ho
included if hydrogen diffusion is also enabled, otherwise it is one. The model parameters are
all set on the DEGRADATION statement, with the following symbolic identifications
dN it
----------- = A KC.KF0  KC.SIHTOT-N it  – KC.KR0 H  KC.AMBIENTH + KC.VOLUME  N it – KC.NIT0  N it 3-570
dt

The initial value of the depassivated traps is KC.NIT0, a depassivation rate constant is
KC.KF0, a repassivation rate constant is KC.KR0, and the maximum dangling bond density is
KC.SIHTOT. If the hydrogen diffusion model is enabled, then KC.AMBIENTH is a scaling
factor for the repassivation of bonds due to excess hydrogen. You enable the hydrogen
diffusion model by specifying the KC.HYDROGEN flag on the MODELS statement. If hydrogen
diffusion is not enabled, then H/KC.AMBIENT = 1.
A non-zero value of the KC.VOLUME parameter enables a model for repassivation that doesn't
require detailed modelling of the hydrogen diffusion. It assumes that the hydrogen density
present at the interface is equal to the difference between the present and initial interface
charge densities, Nit - KC.NIT0.
The parameter A in Equation 3-570 is quite complicated. It can be written as

KC.EA0 KC.EA0+E
A = exp  ------------------------ – ----------------------------------a- J eff 3-571
 K KC.T0 T 
B

where
KC.RHOPARL
 T = K B T + KC.DELTPARL F || 3-572

and
KC.RHOPERP
E a = – KC.DELTAPERP F  + 1  1 +    T  log  1 + N it  – log  1 + KC.NIT0   3-573

301 ATLAS User’s Manual

Physical Models Physics

and
 = – KC.BETA+KC.DELTAPERP F + KC.BETAPARL F || 3-574

The field at the interface is decomposed into parallel and perpendicular components, F || and
F  respectively. The multiplicative factor Jeff is given by

KC.E.RHOHC KC.H.RHOHC
J eff =  1 + KC.E.HCCOEF J hei + KC.H.HCCOEF J hhi  3-575
KC.E.RHOFN KC.H.RHOFN
  1 + KC.E.FNCOEF J tun + KC.H.FNCOEF J tun 
e h

The hot electron or hot hole flags, HEI or HHI, must be specified to obtain values for hot
electron current, Jhei, and hot hole current, Jhhi, respectively. Similarly, a tunneling model
such as FNORD or QTUNN must be enabled to obtain values for the gate tunnel currents for
electrons, J tun , and holes J tun .
e h

As with the DEVDEG.RD model, it is assumed that the dangling bonds become charged very
soon after they are created.
The DEVDEG.E and DEVDEG.H parameters on the MODELS statement make the interface charge
negative or positive respectively. The dangling bonds created at the Si/SiO2 interface are
commonly identified as Pb centers.
These are thought to act as amphoteric traps [36, 218] meaning that they can have either
negative or positive charge. The DEVDEG.A parameter on the MODELS statement charges the
traps according to the dominant carrier type at each point of the interface. If none of
DEVDEG.A, DEVDEG.E, or DEVDEG.H are specified then the traps will be charged with
electrons.
Specify the DEVDEG.KN flag on the MODELS statement to enable this model. The coupling
between the interface state density and other parameters can be omitted from the Newton
iterations by clearing the KC.COUPLED flag on the DEGRADATION statement. This can improve
runtime but may give lower quality results.
A summary of the parameters pertinent to the DEVDEG.KN model is given in Table 3-125.

Table 3-125 Parameters for the Kinetic Degradation Model

Statement Parameter Type Default Units

DEGRADATION KC.AE0 Real 1.5 eV

DEGRADATION KC.KF0 Real 10-5 s
DEGRADATION KC.KR0 Real 3.010-9 s
DEGRADATION KC.SIHTOT Real 1012 cm-2
DEGRADATION KC.NIT0 Real 0.0 cm-2
DEGRADATION KC.AMBIENTH Real 1.0 cm-3
DEGRADATION KC.VOLUME Real 0.0 cm3
DEGRADATION KC.T0 Real 300.0 Kelvin

302 ATLAS User’s Manual

Physical Models Physics

Table 3-125 Parameters for the Kinetic Degradation Model

DEGRADATION KC.BETA Real 0.0
DEGRADATION KC.BETAPERP Real 0.0 cm/V
DEGRADATION KC.BETAPARL Real 0.0 cm/V
DEGRADATION KC.DELTPARL Real 0.0 [Q=1] cm
DEGRADATION KC.DELTPERP Real 0.0 [Q=1] cm
DEGRADATION KC.RHOPARL Real 1.0
DEGRADATION KC.RHOPERP Real 1.0
DEGRADATION KC.E.HCCOEF Real 0.0 cm2/A
DEGRADATION KC.H.HCCOEF Real 0.0 cm2/A
DEGRADATION KC.E.FNCOEF Real 0.0 cm2/A
DEGRADATION KC.H.FNCOEF Real 0.0 cm2/A
DEGRADATION KC.E.RHOHC Real 1.0
DEGRADATION KC.H.RHOHC Real 1.0
DEGRADATION KC.E.RHOFN Real 1.0
DEGRADATION KC.H.RHOFN Real 1.0
MODELS DEVDEG.KN Logical False
MODELS KC.HYDROGEN Logical False
MODELS KC.COUPLED Logical False

MODELS and DEGRADATION Statement Examples

MODELS MOS SRH DEVDEG.KN DEVDEG.A FNPP HEI
DEGRADATION RD.SIHTOT=1.0E12 RD.KF0=1.0E-4 RD.KR0=1.0E-5
KC.VOLUME=1.0
RD.NIT0=1.0E6 KC.BETA=0.1 KC.BETAPERP=1.0E-11 KC.BETAPARL=-5.0E-8
KC.DELTPARL=1.0E-7 KC.DELTPERP=1.0E-8 KC.RHOPERP=1.0
KC.RHOPARL=1.0
KC.AMBIENTH=1.0 KC.T0=300.0 ^KC.COUPLED KC.E.HCCOEF=1.0E4
KC.H.HCCOEF=0.0
KC.E.FNCOEF=1.0E10 KC.H.FNCOEF=0.0

303 ATLAS User’s Manual

p  – GRA.EC.HOLE  K B T  23 n  – GRA.EC.ELEC  K B T  23
a
23
= pv GRA.SIGPe
th
S + nv GRA.SIGNe
p th
S
n
3-579

and the energy level of the switching trap states is GRA.ET2, relative to silicon valence band
edge, and the statistical factors are

   E qfp – Ev   KB T 
23 e if GRA.ET2  0
S
p
=  3-580
 e –   GRA.ET2+ Ev – Eqfp   K B T  if GRA.ET2  0


and

1 if GRA.ET2-E g  0
23 
S
n
=    E – GRA.ET2   K T  3-581
e g B
if GRA.ET2-E g  0


The neutral switching trap can become positively charged again by the processes of hole
capture or electron emission. The rate is

p  – GRA.EC.HOLE  K B T  32 n  – GRA.EC.ELEC  K B T  32
a
32
= pv GRA.SIGPe
th
S + nv GRA.SIGNe
p th
S
n
3-582

and the statistical factors are

32   GRA.ET2  K B T  if GRA.ET2  0
sp =  e 3-583
1 if GRA.ET2  0

for hole capture and

   GRA.ET2+ Ev – E qfn   K B T 
32 e if GRA.ET2-E g  0
S
n
=  3-584
 e   E c – E qfn   K B T  if GRA.ET2-E g  0


for electron emission. The neutral state of the switching trap can revert back to the precursor
state. The rate for this transition is
–  GRA.EA   K B T  
a 31 = GRA.NUe 3-585

where GRA.NU is an attempt frequency and GRA.EA a thermal activation energy. The trap
emission and capture rates are assumed to be independent of their depth in the interface.

306 ATLAS User’s Manual

Physical Models Physics

To enable the 3-state model, you specify either GRA.3.DET or GRA.3.STO on the INTTRAPS
statement. If you specify GRA.3.DET, then the parameters used for the capture rates are
precisely as specified above. If you specify GRA.3.STO, then the parameters used for the
capture rate are obtained from a uniform statistical distribution, in the range
GRA.X-GRA.X.SD< X <GRA.X+GRA.X.SD 3-586
where X refers to any of the model parameters and the GRA.X.SD parameter gives the range
about the average value of GRA.X. For example, the precursor energy would be GRA.ET1 in
the deterministic model and chosen from the range from GRA.ET1–GRA.ET1.SD to
GRA.ET1+GRA.ET1.SD in the stochastic model. The full range of deterministic parameters
and their related variance parameters are given in Table 3-126. When using the GRA.3.STO
model, the multistate trap associated with each interface point can be divided into
GRA.SAMPLES subtraps. The trap density is divided equally among the subtraps, and each
subtrap will generally have different rate parameters from all other subtraps. The default
value of GRA.SAMPLES is 10 and a practical maximum is 1000. GRA.SAMPLES has no effect
for the GRA.3.DET model.
The model is automatically initialized so that all multistate traps are in state 1. You set up a
transient SOLVE statement and ATLAS solves the equation system to give the time evolution
of the occupancy of all the states. The changes in interface state charge and the transient
recombination rates are included self-consistently in the simulation. The occupation fraction
ft of each trap can be saved in a log file by using the PROBE statement with FT.MSC and
MSC.STATE parameters as described in Section 3.3.4 “Multistate Traps”. In the case of the
GRA.3.STO model, a value averaged over the subtraps is logged. The occupation probability
of each multistate trap can be written to a Standard Structure File by using the TRAPS flag on
the OUTPUT statement followed by a SAVE statement. This is also described in Section 3.3.4
“Multistate Traps”. In the case of the GRA.3.STO model, the occupation fractions of each
subtrap are output (up to a maximum of 1000 per point). This then allows the simulation to
be reinitialized with existing data, making it easier to separate the stress and recovery phases
of a simulation.
4-State Model
The positively charged switching trap described in the previous section can be modeled as
being part of a complex that also contains an unpassivated silicon dangling bond. Passivated
silicon dangling bonds will typically already exist at the interface. It is claimed in [90] that it
is energetically favorable for the hydrogen atoms associated with the passivated bonds to
migrate into the insulator and passivate the dangling bonds created by the negative bias stress.
The residual dangling bonds on the interface thus form a Pb center, which will be positively
charged for most NBTI scenarios. In the 4-state model, this extra state couples to state 2 of
the multistate trap model. The transition rates are as follows
–  GRA.ED-GRA.ET2-GRA.GAMMA F   K B T
a 24 = GRA.NUe 3-587

–  GRA.ED-GRA.ET4+GRA.GAMMA F   K B T
a 42 = GRA.NUe 3-588

where F is the field normal to the interface. If the rate a24 is much greater than a42 under
recovery bias conditions, then state 4 will correspond to a non-recoverable part of the NBTI
degradation.

307 ATLAS User’s Manual

Physical Models Physics

To enable this model, specify either GRA.4.DET or GRA.4.STO on the INTTRAPS statement.
In addition to the parameters pertaining to the 3-state model, you optionally specify values for
GRA.ED, GRA.GAMMA, and GRA.ET4. For the stochastic model, you can specify GRA.ED.SD,
GRA.ET4.SD, and GRA.GAMMA.SD. The parameter values are obtained from a uniform
distribution, as described above, except in the case of GRA.ED. For that energy, the stochastic
values are chosen from a gaussian probability distribution with GRA.ED as the mean and
GRA.ED.SD as the standard deviation. In addition the energy, GRA.ET4, of state 4 is restricted
to be in the lower half of the silicon bandgap. This is because the Pb center is thought to be
amphoteric and becomes negatively charged if it is in the upper half of the silcon bandgap.
This possibility is not modeled here, although neutralization of positively charged Pb centers
is considered in the 5-state model. The trap state occupation probabilities can be visualized in
the same way as for the 3-state model.
5-State Model
In the 4-state model it is assumed that state 4 is always positively charged, an assumption
which will only be true if the Fermi-level in the channel is below the trap state. The trap
quickly responds to changes in the position of the Fermi-level and can become neutral by
electron capture and hole emission. If the Fermi-level decreases again it can capture a hole or
emit an electron to regain its positive charge. This is modeled by adding a 5th state to the
multistate traps, which is a neutral Pb center. The transition rates are
n   E i – E t K B T  p
a 45 = nGRA.SIGNv th + n i e GRA.SIGPv th 3-589

p   E i – E t K B T  n
a 54 = pGRA.SIGPv th + n i e GRA.SIGNv th 3-590

where Ei is the intrinsic energy and Et is the trap energy GRA.ET4 + Ev. The intrinsic
concentration is given by ni. To enable this model, specify either GRA.5.DET for the
deterministic model or GRA.5.STO for the stochastic model. It requires no extra parameters
compared to the 4-state model. The trap state occupation probabilities can be visualized in
the same way as for the 4-state model.
The parameters and flags for all these models are given in Table 3-126. The default values of
these are given, but you should use your own suitable values for best results. A schematic
diagram of the states and transitions utilized in this model are shown in Figure 3-17. The
parameters for the PROBE statement that are useful for this model are given in Table 3-19 in
Section 3.3.4 “Multistate Traps”.

Table 3-126 Parameters for 2-Stage NBTI Degradation Multistate Trap Model

Statement Parameter Type Default Units

INTTRAP GRA.3.DET Logical False

INTTRAP GRA.3.STO Logical False
INTTRAP GRA.4.DET Logical False
INTTRAP GRA.4.STO Logical False
INTTRAP GRA.5.DET Logical False

308 ATLAS User’s Manual

Physical Models Physics

INTTRAP GRA.5.STO Logical False

INTTRAP GRA.EA Real 1.0 eV
INTTRAP GRA.EA.SD Real 0.025 eV
INTTRAP GRA.EB.SD Real 0.025 eV
INTTRAP GRA.EB.ELEC Real 0.054 eV
INTTRAP GRA.EB.HOLE Real 0.054 eV
INTTRAP GRA.EC.ELEC Real 0.02 eV
INTTRAP GRA.EC.HOLE Real 0.02 eV
INTTRAP GRA.EC.SD Real 0.025 eV
INTTRAP GRA.ED Real 0.1 eV
INTTRAP GRA.ED.SD Real 0.1, GRA.ED eV
INTTRAP GRA.ET1 Real -0.6 eV
INTTRAP GRA.ET2 Real 0.2 eV
INTTRAP GRA.ET4 Real 0.25 eV
INTTRAP GRA.ET1.SD Real 0.025 eV
INTTRAP GRA.ET2.SD Real 0.025 eV
INTTRAP GRA.ET4.SD Real 0.025 eV
INTTRAP GRA.FC Real 1.0107 V/cm

INTTRAP GRA.FC.SD Real 0.1, GRA.FC V/cm

INTTRAP GRA.GAMMA Real 0.0 [Q] cm
INTTRAP GRA.GAMMA.SD Real 0.0 [Q] cm
INTTRAP GRA.NU Real 1.01013 Hz

INTTRAP GRA.NU.SD Real 0.1, GRA.NU Hz

INTTRAP GRA.SAMPLES Integer 10
INTTRAP GRA.SIGN Real 3.010-14 cm2

INTTRAP GRA.SIGN.SD Real 0.0 eV

INTTRAP GRA.SIGP Real 3.010-14 cm2

INTTRAP GRA.SIGP.SD Real 0.0 eV

309 ATLAS User’s Manual

Physical Models Physics

Note: GRA.ET1, GRA.ET2, and GRA.ET4 are automatically referenced to valence band energy, with a
positive value meaning that the energy level is above the valence band.

Note: GRA.SAMPLES must be limited to 1000 if you wish to save and reinitialize a simulation from a standard
structure file.

It is claimed in [90] that the 2-stage model is applicable to a range of technologies and not
only Si/SiO2 systems.
Examples
INTTRAP GRA.4.DET DENSITY=3.0e12 GRA.SIGN=1.23e-15
GRA.SIGP=1.1e-13 X.MIN=0.65 X.MAX=0.85
GRA.EA=0.77 GRA.NU=1.47e13 GRA.EB.ELEC=0.015
GRA.EB.HOLE=0.016
GRA.EC.ELEC=0.056 GRA.EC.HOLE=0.057 GRA.ET1=-1.0
GRA.ET2=0.19 GRA.ET4=0.1 GRA.GAMMA=1.0e-8 GRA.ED=0.112
GRA.FC=1.1e7
This sets up a 4-state deterministic model with an areal density of 3.01012cm-2, located on
the interface between x=0.65 and x=0.85 microns.
INTTRAP GRA.3.STO GRA.SAMPLES=20 DENSITY=1.0e12 GRA.SIGN=1.0e-14
GRA.SIGP=1.1e-13 X.MIN=0.0 X.MAX=0.6 GRA.EA=0.6
GRA.NU=1.0e13 GRA.EB.ELEC=0.015 GRA.EB.HOLE=0.015
GRA.EC.ELEC=0.056 GRA.EC.HOLE=0.056 GRA.ET1=-1.0
GRA.ET2=0.19 GRA.FC=1.1e7 GRA.SIGN.SD=1.0e-15
GRA.SIGP.SD=1.0e-14 GRA.ET1.SD=0.04 GRA.ET2.SD=0.02
GRA.EB.SD=0.0 GRA.NU.SD=2.0e12 GRA.EC.SD=0.0
GRA.EA.SD=0.001 GRA.FC.SD=1.0e6
This sets up a 3-state stochastic model with an areal density of 1.01012cm-2, located on the
interface between x=0.0 and x=0.6 microns. There are 20 traps simulated per interface point,
each with the relevant parameters chosen from a probability distribution.
PROBE FT.MSC MSC.STATE=1 [Position params]
PROBE FT.MSC MSC.STATE=2 [Position params]
PROBE FT.MSC MSC.STATE=3 [Position params]
PROBE FT.MSC MSC.STATE=4 [Position params]
PROBE FT.MSC MSC.STATE=5 [Position params]
This puts the value of occupation probabilities of each of 5 states at the specified interface
location into a log file. If a stochastic version of the model has been chosen, the values
returned will be sample averaged values.

310 ATLAS User’s Manual

Physical Models Physics

Figure 3-17: Schematic of transitions and states in 5-state, two-stage NBTI degradation model.

3.6.9 The Ferroelectric Permittivity Model

Ferroelectric materials exhibit high dielectric constants, polarization and hysteresis. Such
materials are finding more and more applications in integrated memory devices. To simulate
these effects, a modified version of the ferroelectric model from Miller [187] has been
implemented.
To enable the Ferroelectric Model, set the FERRO parameter in the MODELS statement. In this
model the permittivity used in Poisson’s Equation (Equation 3-1) is given the following
functional form:
FERRO.PS 2 E – FERRO.EC
  E  = FERRO.EPSF + ------------------------------  sech --------------------------------------- 3-591
2 2

where FERRO.EPSF is the permittivity, E is the electric field and  is given as follows:

1 + FERRO.PR  FERRO.PS – 1
 = FERRO.EC log ------------------------------------------------------------------------- 3-592
1 – FERRO.PR  FERRO.PS

The FERRO.EPSF, FERRO.PS, FERRO.PR, and FERRO.EC parameters can be modified in the
MATERIAL statement (see Table 3-127).
The permittivity in Equation 3-591 can be replaced with a user-defined expression with the
C-INTERPRETER. The F.FERRO parameter of the MATERIAL statement (see Section 21.32
“MATERIAL”) defines the file that contains the C-function. This function allows the
permittivity to be position and field dependent.
For more information about C-Interpreter, see Appendix A “C-Interpreter Functions”.

311 ATLAS User’s Manual

Physical Models Physics

The derivative of the dipole polarization with respect to electric field is given by:
dP d dP sat
---------- =  -------------- 3-593
dE dE

where Pd is the position dependent dipole polarization. A numeric integration of this function
is carried out in ATLAS to determine the position dependent dipole polarization.
For saturated loop polarization, the  function is equal to unity, which corresponds to the
default model. If you specify the UNSAT.FERRO parameter in the MODELS statement, the 
function will take on a more general form suitable for simulation of unsaturated loops. In this
case, the  function is given by:
P d – P sat 1  2
 = 1 – tanh  ------------------------ 3-594
P s – P d

where  =1 for increasing fields and  = –1 for decreasing fields.

Table 3-127 User-Specifiable Parameters for Equations 3-591 to 3-592

Statement Parameter Default Units

MATERIAL FERRO.EC 0.0 V/cm

MATERIAL FERRO.EPS 1.0
MATERIAL FERRO.PS 0.0 C/sqcm
MATERIAL FERRO.PR 0.0 C/sqcm

3.6.10 Epitaxial Strain Tensor Calculations in Zincblende

The strain tensor in epitaxial layers is used in the calculation of the zincblende gain and
spontaneous recombination models discussed in Section 3.9.8 “Strained Two-Band
Zincblende Model for Gain and Radiative Recombination”. In epitaxial layers, the strain
tensor can be represented by xx, yy, zz, xy, yz and zx. The relationship between the various
components of the strain tensor are given as follows:
as – a0
 xx =  yy = -----------------
- 3-595
a0

C 12
 zz = – 2 ---------  3-596
C 11 xx

 xy =  yz =  zx = 0 3-597

where C12 and C11 are the elastic constants, which can be specified by the parameters C12 and
C11 on the MATERIAL statement. The default values for C12 and C11 are given for various
binary zincblende materials in Table B-29. Values for terniary and quaternary materials are
linearly interpolated from the binary values.

312 ATLAS User’s Manual

Physical Models Physics

The principal value of strain, xx, can be specified or calculated. To specify the value of xx,
assign the desired value to the STRAIN parameter of the REGION statement.
The parameter a0 is the lattice constant in the “substrate”.
In Equation 3-595, as is the lattice constant in the layer in question. You can specify the lattice
constant in the given layer by the ALATTICE parameter of the MATERIAL statement.
Default values for ALATTICE are taken from Table B-29. Temperature coefficients are
included in the default calculation.
There are several ways to un-ambiguously specify the substrate or the substrate lattice
constant. You can specify a region as the substrate for all strain calculations by specifying the
logical parameter SUBSTRATE on the REGION statement of the substrate. Alternatively, you
can directly specify the substrate lattice constant “local” REGION statement using the ASUB
parameter.
If the substrate lattice constant is not otherwise specified through the use of the ASUB or
SUBSTRATE parameters, the substrate lattice constant is taken as the average of the lattice
constants of the two adjacent epitaxial layers (regions above and below the local region). If
there is only one adjacent region, the lattice constant of that region is used as the substrate
lattice constant.

3.6.11 Epitaxial Strain Tensor Calculation in Wurtzite

The strain tensor in epitaxial layers is used to calculate piezoelectric polarization (Section
3.6.12 “Polarization in Wurtzite Materials”) or in gain modeling (Section 3.9.10 “Strained
Wurtzite Three-Band Model for Gain and Radiative Recombination”) or both. In epitaxial
layers, the strain tensor can be represented by xx, yy, zz, xy, yz and zx. The relationship
between the various components of the strain tensor are given as follows:
as – a0
 xx =  yy = -----------------
- 3-598
a0

C 13
 zz = – 2 ---------  3-599
C 33 xx

 xy =  yz =  zx = 0 3-600

where C13 and C33 are elastic constants, which can be specified by the parameters C13 and C33
on the MATERIAL statement. The default values for C13 and C33 are given for the GaN system
in Section B.8 “Material Defaults for GaN/InN/AlN System”.
The principal value of strain, xx, can be specified or calculated. To specify xx, assign the
desired value to the STRAIN parameter of the REGION statement.
The parameter a0 is the lattice constant in the “substrate”.
In Equation 3-598, as is the lattice constant in the layer in question. You can specify the lattice
constant in the given layer by the ALATTICE parameter of the MATERIAL statement.
Default values for ALATTICE can be found for the GaN/AlN/InN system in Section B.8
“Material Defaults for GaN/InN/AlN System”.

313 ATLAS User’s Manual

Physical Models Physics

The substrate is more ambiguously defined so there are several ways to specify the substrate
or the substrate lattice constant. First, you can specify a region as the substrate for strain
calculations by specifying the logical parameter SUBSTRATE on the associated REGION
statement. You can then specify the lattice constant for that region using the ALATTICE
parameter of the corresponding MATERIAL statement. Alternatively, you can directly specify
the substrate lattice constant in the REGION statement, which make the strain calculations,
using the ASUB parameter.
If the substrate lattice constant is not otherwise specified through the ASUB or SUBSTRATE
parameters, the substrate lattice constant is taken as the average of the lattice constants of the
two adjacent epitaxial layers (region above and below the region in question). If there is only
one adjacent region, the lattice constant of that region is used as the substrate lattice constant.

3.6.12 Polarization in Wurtzite Materials

Polarization modeling is critical for GaN based devices. Automatic calculation of polarization
is enabled by the POLARIZ parameter of the MODELS statement. STET in wurtzite materials is
characterized by two components, spontaneous polarization, Psp, and piezoelectric
polarization, Ppi. Therefore, the total polarization, Pt, is given by:
P t = PSP + P pi 3-601

where PSP is specified on the MATERIAL statement and specifies the total spontaneous
polarization, Psp, for the given material(s). The piezoelectric polarization, Ppi, is given by:

as – a0
P pi = 2 ------------------  E31 – ----------E33
C13
3-602
a0 C33

where E31 and E33 are piezoelectric constants, and C13 and C33 are elastic constants all
specified in the MATERIAL statement. The a0 parameter is the lattice constant of the material
layer in question, which can be specified by the ALATTICE parameter of the MATERIAL
statement. The as parameter is the average value of the lattice constants of the layers directly
above and below the layer in question.
To enable the polarization model, specify POLARIZATION in the REGION or MATERIAL
statement for the region for which you wish to characterize polarization effects. Typically,
this will be a quantum well layer or active layer.
The polarization enters into the simulation as a positive and negative fixed charges appearing
at the top (most negative Y coordinate) and bottom (most positive Y coordinate) of the layer
in question. By default, the positive charge is added at the bottom and the negative charge is
added at the top. You can modify the sign and magnitude of this charge by specifying
POLAR.SCALE in the REGION or MATERIAL statement. This parameter is multiplied by the
polarization determined by Equation 3-601 to obtain the applied charge. The default value for
POLAR.SCALE is 1.0.
In some cases, the introduction of polarization charges may introduce difficulties with
convergence due to problems with initial guess. If these problems arise, you can use the
PIEZSCALE parameter of the SOLVE statement to gradually introduce the effects of
polarization. This parameter defaults to 1.0 and is multiplied by the net charge given by the
product of the results of Equation 3-602 and the POLAR.SCALE parameter.

314 ATLAS User’s Manual

Physical Models Physics

Table 3-128 shows the parameters of the wurtzite polarization model.

Table 3-128 User Specifiable Parameters of the Wurtzite Polarization Model

Statement Parameter Default Units

MATERIAL PSP see Tables B-17-B-23 cm-2

MATERIAL ALATTICE see Tables B-17-B-23 Å
MATERIAL E13 see Tables B-17-B-23 cm-2
MATERIAL E33 see Tables B-17-B-23 cm-2
MATERIAL C13 see Tables B-17-B-23 GPa
MATERIAL C33 see Tables B-17-B-23 GPa

In 3D simulation, you may choose to apply this uniaxial model along layers arranged in the Z
axis material. To accomplish this, you should specify Z.EPI on the MESH statement.
If the layer directly above or below a layer that is being polarized is composed of an insulator
or non-crystaline material or if there is no layer above or below the layer in question by
default, then no polarization charges are added to the interface in question. When the
PCH.INS parameter of the MODELS statement is specified, the polarization charge is added at
such interfaces.
The parameter IF.CHAR of the MATERIAL statement can be used to spread the distribution of
polarizaiton charges spatially in a direction normal to the interface analogous to Equation 21-
3.
The parameter IF.CHAR specifies the characteristic distance from the interface for an
equivalent Gaussian distribution of polarization charge that is applied instead of the 2D
charge produced by polarization. The units of IF.CHAR are microns.
You can also calculate polarization charge exactly as in Equations 3-595 and 3-367 but also
includes contributions by external mechanical strain. These strains are loaded from the
simulator (e.g., ATHENA) and contain strain fields induced by mechanical means. The
POLARIZ model accounts for axial strain due to lattice mismatch. All of the polarization scale
factors POLAR.SCALE and PSP.SCALE apply to external strain but the mechanically induced
polarization can be scaled separately by using the TENSOR.SCALE parameter of the MATERIAL
statement.

315 ATLAS User’s Manual

Physical Models Physics

3.6.13 Stress Effects on Bandgap in Si

Mechanical stress causes change in the band edges in silicon. These band edge shifts are
given by the deformation potential theory [29]. The shifts for the conduction band edges are
given by:
i
E c = D · DEFPOT   xx +  yy +  zz  + U · DEFPOT  ii 3-603

i
where E c is the shift in the band edge of the ith ellipsoidal conduction band minima. The
parameters U.DEFPOT and D.DEFPOT are the user-definable dialation and shear deformation
potentials for the conduction band.
The xx, yy, and zz parameters are the diagonal components of the strain tensor.
The shifts in the valence band edges are calculated by:
 hl 
E v = A · DEFPOT   xx +  yy +  zz    3-604

 hl 
where E v are the band edge shifts in the light and heavy hole valence band maxima. The
 parameter is given by:

2
B · DEFPOT  2 2  2 2 2 2 2
 = -------------------------------
-    xx –  yy  +   yy –  zz  +   zz –  xx   + C · DEFPOT   xy +  yz +  zx  3-605
2  

where xy, yz, and zx are the off diagonal components of the strain tensor.
The strain components are calculated from the stress components stored in the input structure
file if included (typically imported from ATHENA). The conversion between stress and strain
is given by Equations 3-606, 3-607, and 3-608.
 xx =  xx s 11 +  yy s 12 3-606

 yy =  xx s 12 +  yy s 11 3-607

 zz = 2 zz s 44 3-608

Here, xx, yy, and zz are the diagonal components of the stress tensor and s11, s12, and s44 are
the material compliance coefficients.
The compliance coefficients are given by:
c 11 + c 12
s 11 = ------------------------------------------------
- 3-609
2 2
c 11 + c 11 c 12 – 2c 12

– c 12
s 12 = ------------------------------------------------
- 3-610
2 2
c 11 + c 11 c 12 – 2c 12

316 ATLAS User’s Manual

Physical Models Physics

1
s 44 = -------- 3-611
c 44

where c11, c12, and c44 are the elastic stiffness coefficients. The stiffness coefficients for
silicon and germanium are given by [122]:
c 11 = 163.8 – T  0.0128 

c 12 = 59.2 – T  0.0048 Si 3-612

c 44 = 81.7 – T  0.0059 

c 11 = 126.0 

c 12 = 44.0 Ge 3-613

c 44 = 67.7 

where T is the temperature. The stiffness coefficients for SiGe at composition are given by
linear interpolation.
If the stress tensor is not loaded from a structure file, you can specify the values of the strain
tensor as described in Table 3-129. If the stress tensor is not loaded from a structure file and
the strain tensor is not specified, then the strain tensor is calculated as [172]:

 1 +   a SiGe – a Si
 xx =  yy = ----------------- --------------------------- 3-614
 1 –   a SiGe

where  is Poisson’s ratio, aSiGe is the lattice constant of SiGe, and aSi is the lattice constant of
Ge. The lattice constants and Poisson’s ratio for Si and Ge are given by:
–5
a Si = 5.43102 + 1.41  10  T – 300  3-615

–5
a Ge = 5.6579 + 3.34  10  T – 300  3-616

 Si = 0.28 3-617

 Ge = 0.273 3-618

The lattice constant and Poisson’s ratio for Si1-xGex is calculated by linear interpolation.
In Equations 3-603 through 3-604, A.DEFPOT, B.DEFPOT, and C.DEFPOT are user-definable
valence band deformation potential constants.
The user-definable parameters are shown in Table 3-129.

317 ATLAS User’s Manual

Physical Models Physics

Table 3-129 User Definable Parameters for Strained Silicon Band Gap

Statement Parameter Default Units

MATERIAL A.DEFPOT 2.1 eV

MATERIAL B.DEFPOT -2.33 eV
MATERIAL C.DEFPOT -4.75 eV
MATERIAL D.DEFPOT 1.1 eV
MATERIAL U.DEFPOT 10.5 eV
MATERIAL EPS11 *
MATERIAL EPS22 *
MATERIAL EPS33 *
MATERIAL EPS12 *
MATERIAL EPS13 *
MATERIAL EPS23 *

Note: * If unspecified, Equations 3-606, 3-607, and 3-608 calculate these parameters.

The net changes in the band edges, under Boltzman's statistics, are given by Equations 3-619
and 3-620.

 E  i 
exp  – ---------------
c
3  kT 
 
 E c = kT ln  ----------------------------------
3
3-619
=1

1 h
r  E v  1  E v 
 E v = k T ln ------------ exp  – --------------- + ------------ exp  – --------------- 3-620
1+r  kT  1 + r  kT 

Here, the parameter r is given by Equation 3-621.

32
r =  ml  mh  3-621

In Equation 3-621, ml and mh are the effective masses of light and heavy holes as described in
other places in this manual.
To enable the model for stress dependent band gap in silicon, specify the STRESS parameter
of the MODELS statement.

318 ATLAS User’s Manual

Physical Models Physics

3.6.14 Low-Field Mobility in Strained Silicon

For Boltzman's statistics, the following expressions can be used for strain dependent electron
and hole low-field mobilities in silicon [69]:

  E – E i  
 1 – ML  MT1 
 n =  n 0  1 + ---------------------------------------------- exp  --------------------------- – 1
c c
   3-622
 1 + 2  ML  MT1  kT 
 
 

  E l – E v h 
  MLH  MHH 
1.5

 p =  p0  1 +  EGLEY · R – 1  --------------------------------------------------------- exp  ------------------------------- – 1
v
1.5  kT   3-623
 1 +  MLH  MHH    
 
where  n 0 and  p0 are the concentration dependent low field mobilities for electrons and
i l h
holes, and E c , E c , E v , and E v are given by Equations 3-603, 3-619, and 3-604
respectively. Table 3-130 shows user-definable parameters for this model.

Table 3-130 User Definable Parameters for the Strained Silicon Low-Field Mobility Model

Statement Parameter Default Units

MATERIAL ML 0.916
MATERIAL MT1 0.191
MATERIAL MLH 0.16
MATERIAL MHH 0.49
MOBILITY EGLEY.R 2.79

To enable the strained silicon low-field mobility model, specify EGLEY.N for electrons and
EGLEY.P for holes on the MOBILITY statement.

319 ATLAS User’s Manual

Physical Models Physics

3.6.15 Light Absorption in Strained Silicon

The effects of strain on the optical absorption in silicon can be modeled by modifications
[211] made to a model suggested by Rajkanan [226]. The Rajkanan model can be described
by:
2 2
   – E gj  T  + E pi     – E gj  T  + E pi  12
 T =
 C i A j -----------------------------------------------------
 exp  E pi  kT  – 1 
– -----------------------------------------------------
 1 – exp  E pi  kT  
+ A d    – E gd  T   3-624
i = 1 2
j = 1 2

where:
• (T) is the temperature dependent absorption coefficient.
•  is the optical frequency.
• Egi(T) are the indirect energy band gaps given by Equations 3-625 and 3-626.
• Egd(T) is the direct energy band gap given by Equation 3-627.
• T is the temperature in Kelvin.
• Ad, Aj, Ci, and Epi are empirical constants given in Table 3-131.
The temperature dependent Eg in Equation 3-624 are given by Equations 3-625 through 3-
627.
2
E g1  T  = EG1.RAJ –  BETA · RAJ  T   T + GAMMA · RAJ   3-625

2
E g2  T  = EG2.RAJ –  BETA · RAJ  T   T + GAMMA · RAJ   3-626

2
E gD  T  = EGD · RAJ –  BETA · RAJ  T   T + GAMMA · RAJ   3-627

To enable the model given in Equation 3-624, enable the RAJKANAN parameter of the MODELS
statement.

Table 3-131 User Modifiable Parameters of Equation 3-624.

Symbol Parameter Statement Default Units

Ep1 EP1.RAJ MATERIAL 1.82710-2

Ep2 EP2.RAJ MATERIAL 5.77310-2
C1 C1.RAJ MATERIAL 5.5
C2 C2.RAJ MATERIAL 4.0
A1 A1.RAJ MATERIAL 3.231102
A2 A2.RAJ MATERIAL 7.237103
Ad AD.RAJ MATERIAL 1.052106

320 ATLAS User’s Manual

Physical Models Physics

Table 3-131 User Modifiable Parameters of Equation 3-624.

Symbol Parameter Statement Default Units

BETA.RAJ MATERIAL 7.02110-4 eV/K

GAMMA.RAJ MATERIAL 1108.0 K
EG1.RAJ MATERIAL 1.1557 eV
EG2.RAJ MATERIAL 2.5 eV
EGD.RAJ MATERIAL 3.2 eV

To account for strain effects, the expressions in Equations 3-625 through 3-627 are modified
by the changes in band edges induced by strain as given in Equations 3-603 and 3-604. To
enable these modifications, set the PATRIN parameter of the MODELS statement.

321 ATLAS User’s Manual

The standard gain model [197] is enabled by specifying G.STANDARD in the MODELS
statement. The standard gain model is given by

 – E E c – E fn + GAMMA   – E g 
g  x, y  = GAIN0 -------------------g- f  -------------------------------------------------------------------------- 3-633
kT kT
E v – E fp –  1 – GAMMA    – E g 
– f  -----------------------------------------------------------------------------------------
kT
where:
•  is Planck's constant,
• Eg is the energy bandgap,
• k is Boltzman's constant,
• T is the lattice temperature,
• Efn and Efp are the electron and hole quasi-Fermi energies,
•  is the emission frequency,
• Ev and Ec are the valence and conduction band edge energies,
• the function f is defined in Equation 3-632,
• GAMMA and GAIN0 user-definable parameters specified on the MATERIAL statement.
Table 3-132 describes the user defined parameters of Equation 3-633.

Table 3-132 User-Specifiable Parameters for Equation 3-633

POLARIZATION DRIFT-DIFFUSION GAIN AND SPONTANEOUS

POISSON EMISSION

radiative recomb
photo-generation stimulated recomb

LUMINOUS LASER LED

VCSEL

Figure 3-18: Simulation Flow For Physically Based Optoelectronic Models

• mr is a reduced electron-hole effective mass.

• nr is a material refractive index.

• M is a polarization- dependent bulk momentum matrix element
b
• The sum is taken over all valence bands.
Expressions for gain and spontaneous emssion for quantum wells are similar and given in
Chapter 13 “Quantum: Quantum Effect Simulator”.
Spontaneous Recombination
In order to compute total radiative recombination rate and include it into drift-diffusion
equations, specify the SPONTANEOUS parameter on the MODELS statement. Recombination
rate is calculated by performing averaging over polarizations and integration over the whole
emission spectrum:

  2 r TE  h  + r TM  h 
 spon spon 
R sp  x y z  =  ------------------------------------------------------------------------ d h  
3
3-648
0

The integration over the spectrum is done using Gauss-Laguere quadrature rule. The number
of energy points is set by the WELL.NERSP parameter (default is 32) on the METHOD statement.
If the parameter is set to zero (not recommended), a trapezoidal integration will be used with
a spacing set by WELL.DENERGY parameter on the MODELS or MATERIAL statement.
Spectrum of spontaneous recombination and gain for each bulk LED region can be stored in a
separate file using SPECTRUM=’filename’ parameter on the SAVE statement. The spectra are
computed at the geometrical center of the LED region.

332 ATLAS User’s Manual

Optoelectronic Models Physics

3.9.8 Strained Two-Band Zincblende Model for Gain and Radiative Recombination
The zincblende two-band model is derived from the 4x4 k·p model assuming parabolic bands.
The model accounts for light and heavy holes and the effects of strain and is enabled by
specifying ZB.TWO on the MODELS statement.
In the zincblende two band model we first calculate P and Q as given by Equations 3-649
and 3-650.
P  = – a v   xx +  yy +  zz  3-649

b
Q  = – ---   xx +  yy +  zz  3-650
2

Here, av and b are the valence band hydrostatic deformation potentials and xx, yy, and zz are
the strain tensor as calculated in Section 3.6.10 “Epitaxial Strain Tensor Calculations in
Zincblende”. The default values of av and b for various binary materials are given in Table B-
B-29. Values for terniary and quaternary materials are linearly interpolated from the binary
values.
Next, we can calculate the conduction, heavy hole and light hole band edge energies from
Equations 3-651, 3-652, and 3-653.
E c = E v + E g + a c   xx +  yy +  zz  3-651

2
E hh = E v – P  – sgn  Q   Q  3-652

2
E lh = E v – P  + sgn  Q   Q  3-653

Here, ac is the conduction band hydrostatic deformation potential, Ec, Ev, and Eg are the
unstrained conduction band edge energy, valence band edge energy, and band gap. These are
calculated from the material specific affinity and bandgap.
The default values of ac for various binary materials are given in Table B-29. Values for
terniary and quaternary materials are linearly interpolated from the binary values.
The function sgn() is the “sign” function (i.e., +1 for positive arguments and -1 for negative
arguments).
Next, we calculate the effective masses for the various bands using Equations 3-654 through
3-657.
z
m hh = m 0    1 – 2 2  3-654

z
m lh = m 0    1 + 2 2  3-655

t
m hh = m 0    1 +  2  3-656

333 ATLAS User’s Manual

Optoelectronic Models Physics

t
m lh = m 0    1 –  2  3-657

Here, m0 is the rest mass of an electron, and 1 and 2 are the Luttinger parameters. The
default values of the luttinger paramters for various binary materials are given in Table B-30.
Values for terniary and quaternary materials are linearly interpolated from the binary values.
Finally, the band edges and effective masses are placed into the gain and spontaneous
recombination models described in Section 3.9.7 “Unstrained Zincblende Models for Gain
and Radiative Recombination”.

3.9.9 Strained Three-Band Zincblende Model for Gain and Radiative Recombination
The zincblende three-band model is derived from the 6x6 k·p model assuming parabolic
bands. The model accounts for light, heavy and split-off holes and the effects of strain and is
enabled by specifying ZB.THREE on the MODELS statement.
The conduction, heavy hole, light hole and split-off hole band edge energies are given by the
following expressions.
E c = E v + E g + a c   xx +  yy +  zz  3-658

E hh = E v – P  – Q  3-659

1 2 2
E lh = E v – P  + ---  Q  –  0 +  0 + 9Q  + 2Q   0  3-660
2

1 2 2
E so = E v – P  + ---  Q  –  0 –  0 + 9Q  + 2Q   0  3-661
2
where P and Q are given by Equations 3-649 and 3-650 above and 0 is the split-off energy.
The effective masses for the various bands are given by the following expressions:
z
m hh = m 0    1 – 2 2  3-662

z
m lh = m 0    1 + 2 2 f +  3-663

z
m so = m 0    1 + 2 2 f -  3-664

t
m hh = m 0    1 +  2  3-665

t
m lh = m 0    1 –  2 f +  3-666

334 ATLAS User’s Manual

Optoelectronic Models Physics

t
m so = m 0    1 –  2 f –  3-667

employing the strain factor

2s  1 + 1.5  s – 1  1 + 2s + 9s 2   + 6s 2
f  = --------------------------------------------------------------------------------------------------------------------------------------------------- 3-668
2
0.75  s – 1  1 + 2s + 9s 2  + s – 1  1 + 2s + 9s 2 – 3s 2

with the strain parameter s = Q    0 . Without strain, s = 0 and f + = 1 , the effective masses
for heavy and light holes are identical to those of the two-band model.

3.9.10 Strained Wurtzite Three-Band Model for Gain and Radiative Recombination
The strained wurtzite three-band model [48, 49, 148] is derived from the k·p method for three
valence bands in wurtzite crystalline structure. Given the assumptions of parabolic bands, no
valence band mixing and momentum approaching zero, the following approach can be used.
First, we calculate the parameters from Equations 3-669 and 3-670.
  = D 3  zz + D 4   xx +  yy  3-669

  = D 1  zz + D 2   xx +  yy  3-670

Here, D1, D2, D3, and D4 are shear deformation potentials and xx, yy, and zz are taken from
the strain tensor calculations described in Section 3.6.11 “Epitaxial Strain Tensor Calculation
in Wurtzite”. Next, we can calculate the valence band energies from Equations 3-671 through
3-673.
0 0
E hh = E v +  1 +  2 +   +   3-671

0 1 – 2 +  1 – 2 +  2
- +   +  ------------------------------- + 2 3
0 2
E lh = E v + ------------------------------ 3-672
2  2 

0 1 – 2 +  1 – 2 +  2
- +   –  ------------------------------- + 2 3
0 2
E ch = E v + ------------------------------ 3-673
2  2 

Here,1,2 and 3 are split energies and Ev is the valence band reference level. Next, we can
calculate the hydrostatic energy shift from Equation 3-674.
P c = a cz  zz + a ct   xx +  yy  3-674

Here, acz and act are hydrostatic deformation potentials. From which we can calculate the
conduction band energy as given in Equation 3-675.
0 0
E c = E v +  1 +  2 + E g + P c 3-675

335 ATLAS User’s Manual

Optoelectronic Models Physics

Here, Eg is the energy bandgap. The default bandgaps for the InGaN system given in Section
B.8 “Material Defaults for GaN/InN/AlN System”. You can override the default models for
bandgap by specifying the EG1300, EG2300, and EG12BOW parameters of the MATERIAL
statement.
Next, we can calculate the effective masses in the various bands using the expressions in
Equations 3-676 through 3-681.
z –1
m hh = – m 0  A 1 + A 3  3-676

t –1
m hh = – m 0  A 2 + A 4  3-677

0 –1
z  E lh –   
m lh = – m 0 A 1 +  ---------------------- A 3 3-678
 E 0lh – E 0ch

0 –1
t  E lh –   
m lh = – m 0 A 2 +  ---------------------- A 4 3-679
 E 0lh – E 0ch

0 –1
z  E ch –   
m ch = – m 0 A 1 +  ---------------------- A 3 3-680
 E 0ch – E 0lh

0 –1
t  E lh –   
m lh = – m 0 A 2 +  ---------------------- A 4 3-681
 E 0lh – E 0ch

z t z t z t
Here, m is the free space mass of an electron, m hh , m hh , m lh , m lh , m ch , and m ch are the
effective masses for heavy holes, light holes and crystal split off holes in the axial and
transverse directions and A1, A2, A3 and A4 are hole effective mass parameters.
The momentum matrix elements for the various transitions can be calculated by Equations 3-
682 through 3-687.
M hh 11 = 0 3-682

2 m0
M hh 11 = b  ------- E pz 3-683
 2 

2 m0
M ch 11 = a  ------- E pz 3-684
 2 

m0
M hh  = ------- E px 3-685
4

336 ATLAS User’s Manual

Optoelectronic Models Physics

2 m0
M lh  = a  ------- E px 3-686
4

2 m0
M ch  = b  ------- E px 3-687
 4

Here, Epx and Epz are given by Equations 3-688 and 3-689, a2 and b2 are given by Equations 3-
2 2
692 and 3-693. The values of P 1 and P 2 are given by Equations 3-690 and 3-691.

2 2
E px =  2m 0  h P 2 3-688

2 2
E pz =  2m 0  h P 1 3-689

2
2 h m
2   E g +  2 +  2   E g + 2 2  – 2 3
P1 = ----------  ------0- – 1 --------------------------------------------------------------------------------- 3-690
2m 0  m z   E g + 2 2 
e

2
2 h m
2  E g   E g +  1 +  2   E g + 2 2  – 2 3 
P2 = ----------  ------0- – 1 --------------------------------------------------------------------------------------------- 3-691
2m 0  m 2
t   E +  +    E +   –
g 1 2 g 2 3
e

0
2  E lh –   
a =  ---------------------- 3-692
 E 0lh – E 0ch

0
2  E ch –   
b =  ---------------------- 3-693
 E 0ch – E 0lh

t z
In Equations 3-688 through 3-693, m e and m e are the transverse and axial conduction band
effective masses.
The default values for the parameters of Equations 3-669 through 3-693 are shown in Section
B.8 “Material Defaults for GaN/InN/AlN System”.

337 ATLAS User’s Manual

Optoelectronic Models Physics

3.9.11 Lorentzian Gain Broadening

Gain broadening due to intra-band scattering can be introduced by specifying LORENTZ in the
MODEL statement. The following equation describes gain broadening when it’s applied.


gE =
 g  E L  E – E  dE 3-694
E ij

The Lorentzian shape function is given by Equation 3-695.

1 WELL · GAMMA0
L  E – E  = ---  ------------------------------------------------------------------------
- 3-695
  E – E  2 + WELL GAMMA0 2
·

where WELL.GAMMA0 is user-specifiable in the MATERIAL statement.

3.9.12 Ishikawa's Strain Effects Model

The strained layer InGaAsP and InGaAlAs quantum well material models from [123] are
implemented in the ATLAS simulator. To enable these models, specify the ISHIKAWA
parameter in the MODELS statement. Strain percentages are specified by the STRAIN parameter
in the MQW statement. STRAIN between a substrate material with lattice constant, ds, and an
epitaxial material without STRAIN, with lattice constant, de, is (de -ds)|ds.
When you enable this model, the band edge parameters for materials in the InGaAsP and
InGaAlAs systems are calculated using Equations 3-696, 3-697,and 3-698.
InGaAs
2
E c = 1.040 – 0.0474STRAIN + 0.003303STRAIN
2
E v hh = 0.3331 + 0.05503STRAIN – 0.002212STRAIN 3-696
2
E v lh = 0.331 – 0.01503 – 0.003695STRAIN

InGaAsP
a STRAIN < 0
E c =  0.6958 + 0.4836E g  – 0.03031STRAIN
E v, lh =  0.5766 – 0.3439E g  – 0.3031STRAIN
3-697
b STRAIN > 0
E c =  0.6958 + 0.4836E g  + 0.003382STRAIN
E v hh =  0.6958 – 0.5164E g  + 0.003382STRAIN

338 ATLAS User’s Manual

Optoelectronic Models Physics

InGaAlAs
a STRAIN  0
E c =  0.5766 + 0.6561E g  – 0.02307STRAIN

E v lh =  0.5766 – 0.3439E g  – 0.2307STRAIN

3-698
b STRAIN  0
E c =  0.5766 + 0.06561E g  + 0.01888STRAIN
E v hh =  0.5766 – 0.3439E g  + 0.01888STRAIN

A STRAIN parameter has also been added to the REGION statement to account for strain in the
bulk materials. Note that for the InGaAs material system, the equations ignore composition
fraction and variation in the parameters is accounted strictly through strain (see Equation 3-
696).
These band edge parameters are used through all subsequent calculations. Most importantly,
the band edges are used in solving the Schrodinger's Equation (See Equations 13-2 and 13-3)
to obtain the bound state energies in multiple quantum wells.
If you enable the Ishikawa Model, this would include the effects of strain in the valence and
conduction band effective masses as described in [123]. The effects of strain are introduced in
Equation 3-699.
1 2
--------- = ASTR + BSTRSTRAIN + CSTRSTRAIN
m lh
3-699
1 2
---------- = DSTR + ESTRSTRAIN + FSTRSTRAIN
m hh

The ASTR, BSTR, CSTR, DSTR, ESTR, and FSTR parameters are user-definable in the MATERIAL
statement. You can also choose the appropriate values for these parameters from Tables 3-136
or 3-137.

Table 3-136 In-Plane Effective Mass of InGaAsP Barrier (1.2µm)/InGaAs(P) Well and
InGaAlAs Barrier (1.2µm)/InGa(Al)As Well System for 1.55µm Operation [123]

2.0<Strain(%)<0.5 0<Strain(%)<2.0

ASTR BSTR CSTR DSTR ESTR FSTR

Lattice-matched InGaAsP 4.238 6.913 1.687 4.260 2.253 0.584

barrier/InGaAs well

InGaAsP (1.6-µm) well 0.936 4.825 1.754 2.986 5.505 1.736

Strain-compensated 7.761 13.601 5.100 4.166 1.933 0.558

InGaAsP barrier/InGaAs
well

339 ATLAS User’s Manual

Optoelectronic Models Physics

Table 3-136 In-Plane Effective Mass of InGaAsP Barrier (1.2µm)/InGaAs(P) Well and
InGaAlAs Barrier (1.2µm)/InGa(Al)As Well System for 1.55µm Operation [123]

2.0<Strain(%)<0.5 0<Strain(%)<2.0

Lattice-matched InGaAlAs 3.469 6.133 1.481 3.9134 2.336 0.633

barrier/InGaAs well

InGaAlAs(1.6-µm) well 1.297 5.598 2.104 2.725 6.317 1.766

Strain-compensated 5.889 9.467 2.730 4.193 1.075 -0.155

InGaAlAs barrier/InGaAs
well

Table 3-137 In-Plane Effective Mass of InGaAsP Barrier (1.1µm)/InGaAs(P) Well and
InGaAlAs Barrier (1.1µm)/InGa(Al)As Well System for 1.30µm Operation [123]

2.0<Strain(%)<0.5 0.5<Strain(%)<2.0

ASTR BSTR CSTR DSTR ESTR FSTR

Lattice-matched InGaAsP 9.558 8.634 1.847 4.631 1.249 0.313

barrier/InGaAs well

InGaAsP (1.4-µm) well 2.453 4.432 1.222 3.327 3.425 1.542

Strain-compensated 4.720 0.421 0.014

InGaAsP barrier/InGaAs
well

Lattice-matched InGaAlAs 8.332 8.582 2.031 3.931 1.760 0.543

barrier/InGaAs well

InGaAlAs(1.4-µm) well 3.269 5.559 1.594 3.164 4.065 1.321

Strain-compensated 4.078 0.516 0.0621

InGaAlAs barrier/InGaAs
well

To choose these values, use the STABLE parameter in the MATERIAL statement. You can set
this parameter to the row number in the tables. The rows are numbered sequentially. For
example, the first row in Table 3-136 is selected by specifying STABLE=1. The first row of
Table 3-137 is selected by specifying STABLE=7.
You can also choose to specify the effective masses directly. The conduction band effective
mass is specified by the MC parameter in the MATERIAL statement. There are two ways to
specify the valence band. One way, is to use the MV parameter to specify a net effective mass.
The other way, is to use the MLH and MHH parameters to specify the light and heavy hole
effective masses individually.
If you don’t specify the effective masses by using one of these methods, then the masses will
be calculated from the default density of states for the given material as described in Section
3.4.2 “Density of States”.
The effective masses described are also used to solve the Schrodinger Equation (see
Equations 13-2 and 13-3).

340 ATLAS User’s Manual

Optical Index Models Physics

3.10 Optical Index Models

The index of refraction of materials are used in several places in the ATLAS simulation
environment. For LUMINOUS and LUMINOUS3D, the index is used to calculate reflections
during raytrace. For LASER and VCSEL, the index is key to determining optical confinement.
For LED, the index is used to calculate reflections for reverse raytrace.
In the ATLAS environment there are a variety of ways the optical index of refraction can be
introduced as well as reasonable energy dependent defaults for many material systems. For
some these defaults are tabular (see Table B-37). For others there exist analytic functions.
Here we describe some of those analytic models.

3.10.1 The Sellmeier Dispersion Model

To enable the Sellmeier dispersion model, specify NDX.SELLMEIER on the MATERIAL
statement. The Sellmeier dispersion model is given in Equation 3-700.
2 2
S1SELL  S2SELL 
nr    = S0SELL + ------------------------------------- + ------------------------------------- 3-700
2 2 2 2
 – L1SELL  – L2SELL

In this equation,  is the wavelength in microns, and S0SELL, S1SELL, S2SELL, L1SELL and
L2SELL are all user-definable parameters on the MATERIAL statement. Default values for
these parameters of various materials are given in Table B-38.

3.10.2 Adachi's Dispersion Model

To enable a model by Adachi [1], specify NDX.ADACHI on the MATERIAL statement. This
model is given by Equations 3-389, 3-390, and 3-391.
Eg 1.5
n r  w  = AADACHI f  x 1  + 0.5  -------------------------------------- f  x 2  + BADACHI
2
3-701
 E g + DADACHI

1
f  x 1  = -----  2 – 1 + x 1 – 1 – x 1 

 x 1 = ------- 3-702
2 Eg
x1

1
f  x 2  = -----  2 – 1 + x 2 – 1 – x 2  
 x 2 = -------------------------------------- 3-703
2 E g + DADACHI
x2

In these equations, Eg is the band gap,  is plank's constant,  is the optical frequency, and
AADACHI, BADACHI and DADACHI are user-specifiable parameters on the MATERIAL statement.
Default values for these parameters for various binary III-V materials are given in Table B-
39. The compositionally dependent default values for the ternary combinations listed in Table
B-39 are linearly interpolated from the binary values listed in the same table.
For nitride compounds, you can use a modified version of this model, which is described in
Section 5.4.8 “GaN, InN, AlN, Al(x)Ga(1-x)N, In(x)Ga(1-x)N, Al(x)In(1-x)N, and
Al(x)In(y)Ga(1-x-y)N”.

341 ATLAS User’s Manual

Optical Index Models Physics

3.10.3 Tauc-Lorentz Dielectric Function with Optional Urbach Tail Model for Complex
Index of Refraction
A parameterization of the Tauc-Lorentz dielectric function was proposed by Foldyna et.al.
[77], which includes Urbach tails. The imaginary part of the dielectric function is given by
Equation 3-704.
2
 1 TLU · A TLU · E0 TLU · C  E – TLU · EG 
 --
- --------------------------------------------------------------------------------------------------------- E  TLU · EC
E 2 2 2 2 2
2  E  =   E – TLU · E0  + TLU · C E 3-704
 Au
 ------ exp  ------
E
E  TLU · EC
E  E u

Here, for energies greater than TLU.EC, the standard Tauc-Lorentz form is used while for
energies less than TLU.EC, Urbach tails are parameterized as a function of energy and two
matching parameters Au and Eu that are calculated by Equations 3-705 and 3-706 to give a
continuous function and first derivative at E=TLU.EC.

 TLU · EC – TLU · EG 
E u = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
- 3-705
2 2 2
TLU · EC + 2  TLU · EC – TLU · E0 
2 – 2TLU · EC  TLU · EC – TLU · EG  -----------------------------------------------------------------------------------------------------------------------
2 2 2 2 2
TLU · C TLU · EC +  TLU · EC – TLU · E0 

2
TLU · EC TLU · A TLU · E0 TLU · C  TLU · EC – TLU · EG 
A u = exp  --------------------- --------------------------------------------------------------------------------------------------------------------------- 3-706
 Eu  2 2 2 2 2
 TLU · EC – TLU · E0  + TLU · C TLU · EC

In our implementation, if you do not specify TLU.EC, then the standard form of the Tauc-
Lorentz dielectric function is used. For energies less than TLU.EG, the imaginary part is zero.
The real part of the dielectric function is obtained from the imaginary part using the Kramers-
Krönig relation as given in Equation 3-707:
    


2 2
 1  E  = TLU · EPS + ---  C· P·  ------------------ d 3-707
 2
0 –E
2

where (C.P.) denotes the Cauchy principal value of the integral. The integral itself is
resolved using some analytic forms the details of which are given in [77].
The complex index of refraction is calculated from the complex dielectric function using
Equations 3-708 and 3-709.

2 2
1 + 2 + 1
n = --------------------------------
- 3-708
2

2 2
1 + 2 – 1
k = --------------------------------
- 3-709
2

342 ATLAS User’s Manual

Optical Index Models Physics

You can enable the model by specifying TLU.INDEX on the MODELS statement. Table 3-138
gives a summary of the relevant model parameters.

Table 3-138 Parameters for the Tauc-Lorentz-Urbach Dielectric Model

Parameter Statement Default Units

TLU.A MATERIAL eV
TLU.C MATERIAL eV
TLU.E0 MATERIAL eV
TLU.EG MATERIAL eV
TLU.EC MATERIAL eV
TLU.EPS MATERIAL 1.0

343 ATLAS User’s Manual

Carrier Transport in an Applied Magnetic Field Physics

3.11 Carrier Transport in an Applied Magnetic Field

Magnetic field effects have been used in semiconductor characterization measurements and
exploited in semiconductor device applications. ATLAS can model the effect of applied
magnetic fields on the behavior of a semiconductor device.
The basic property, which is measured is the Hall coefficient RH. For a uniformly doped
device with a current flowing in the X direction and a magnetic field in the Z direction, an
electric field (the Hall field) is set up in the Y direction. The Hall coefficient is then obtained
by combining these measured quantities in the ratio:
Ey
R H = ----------
- 3-710
Jx Bz

The Hall co-efficient also has a theoretical value:

2
rp – s n
R H = -------------------------- 3-711
2
q  p + sn 

where s is the electron mobility divided by hole mobility. The Hall scattering factor, r, can be
obtained from the mean free time between carrier collisions and is typically not very different
from unity.
Another quantity that is measured experimentally is the Hall mobility:
 = R H  3-712

where  is the electrical conductivity. We will consider the electron and hole currents
separately. Therefore, taking the limits (n >> p) and then (p >> n) in Equation 3-712, we
obtain the Hall mobilities for electrons and holes respectively as:
 n = r n
3-713
 p = r p

where n and p are the carrier mobilities. The Hall scattering factor for electrons can be
different to that for holes in general. Published values for silicon are 1.2 and 1.2 [7] and 1.1
and 0.7 [234] for electrons and holes respectively. The default in ATLAS is the latter. You can
set the values on the MODELS statement using the R.ELEC and R.HOLE parameters.
The effect of the magnetic field on a carrier travelling with velocity v is to add a term called
the Lorentz force:
q  vxB  3-714

to the force it feels. The magnetic field density B is a vector (Bx, By, Bz) and is in units of Tesla
(V.s/m2) in ATLAS.

344 ATLAS User’s Manual

Carrier Transport in an Applied Magnetic Field Physics

The consequence of this extra force in a semiconductor can be calculated by including it in a

relaxation time based transport equation and making a low-field approximation [7]. You can
relate the current density vector obtained with a magnetic field applied to that obtained with
no magnetic field using a linear equation:
JB = M J0 3-715
˜

The Matrix M can be written as:

2
1 + a ab – c ca + b
1  
M = ---------------------------------------  c + ab 1+b
2
bc – a 3-716
˜ 1+a +b +c  2 2 2

 ca – b a + bc 1+c
2

where a = Bx, b = By, c = Bz. The matrix takes this form for both electrons and holes.
Using this form, you can include the magnetic field effects in ATLAS. The model is derived
from an expansion in powers of Magnetic field and is only accurate if the weak field
condition:
* B «1 3-717

is satisfied. Therefore, for example, a field of 1 Tesla in Silicon with a typical mobility of 0.1
m2/(V.s), the product is 0.1 so that the condition is satisfied.
There are several published methods that have been considered for discretizing Equation 3-
715 on a triangular or prismatic mesh. Some of them are based on the Box method extension
of Allegretto et al [7]. It can be proven that, except in the case of uniform carrier density, this
method has the considerable drawback that it does not conserve current at the level of
individual Voronoi cells. Alternative means are used by Allegretto to obtain continuity of
terminal currents, but this is not acceptable for ATLAS.

Another method is to obtain a least squares value for the current density vector J 0 , for each
triangle [210]. The components of these vectors are then multiplied by M and the Box
discretization obtained by projecting the resulting current density vector J B along each side
of the triangle. It was found that this method, along with several variations of this method,
could result in carrier densities which exhibit unphysical spatial oscillations and also result in
negative carrier densities. This instability for methods based on a least squares current has
previously been observed [233]. The least-squares based method is accessible in
MAGNETIC2D for comparative purposes only. You specify MAG2DLSQ on the METHOD
statement to enable it.
To avoid the problems associated with the aforementioned methods, we developed two
discretizations that are less problematic. The MAGNETIC3D module accomodates both of
these discretizations. For the first one, we start by expressing the integral of the current flux
over a Voronoi cell,

 JB  dS =  MJ0  dS 3-718

345 ATLAS User’s Manual

Carrier Transport in an Applied Magnetic Field Physics

We can rewrite the right hand side as


1
--------------------------------------------  N + T J 0  dS 3-719
2 2 2
1 + a + b + c 

where the normal matrix N=

2
1 + a 0 0
 
0 1+b
2
0 3-720
 
0 0 1+c 
2

and the tangential matrix T =

 0 ab – c ca + b
 c + ab 0 bc – a 3-721

 ca – b a + bc 0 
using the notation for the matrix entries as defined above. So long as the mesh elements are
right-angled prisms with non-diagonal faces in the coordinate directions (a rectangular mesh),
the normal part results in the usual edge based discretization with only a scalar modification.
Also under this restriction, the normals to the Voronoi edges become after multiplication by
TT.
 1 0 0    0 ab – c ca + b 
 0 1 0    c + ab 0 bc – a  3-722
 0 0 1    c – a b a + bc 0 

If we take into account the direction of the outward normals (i.e.,   1 0 0  ,  0  1 0  , and
 0 0  1  ), then you can understand the currents as band integrals around the Voronoi cell in
a direction tangential to the magnetic direction. In a prismatic mesh, there are three different
configurations to consider. Each requires the current components parallel to the Voronoi cell
boundaries to be approximated. This is done by interpolating solution variables to edge and
face midpoints, and then using the Scharfetter-Gummel method to obtain the discretised
current. By multiplying the interpolated current components by the appropriate magnetic field
terms, the contributions to the assembly of the continuity equation over the Voronoi cell may
be evaluated. This method apparently gives good convergence behavior and stable solutions.
It is enabled for the electron continuity equation by specifying MAG3DINT.N on the METHOD
statement, and for the hole continuity equation, by specifying MAG3DINT.P on the METHOD
statement.
The restriction of a rectangular mesh is unacceptable for many devices, and for this reason
SILVACO developed a new discretization for both the MAGNETIC2D and MAGNETIC3D
modules. The stability of Finite Element discretizations of the current continuity equations
was discussed by He [106]. SILVACO indirectly extended this approach to the Box
discretization method. The analysis is lengthy, so only the result is given.
The current along any edge in the device is obtained using the Scharfetter-Gummel
discretization but with two modifications. The first is that there is a prefactor equal to (1 + a2
+ b2 + c2 )-1, and the second is that the argument to the Bernoulli function has the following
term added to it

346 ATLAS User’s Manual

Carrier Transport in an Applied Magnetic Field Physics

  a 2     2  
------ +  ab – c  ------ +  ca + b  ------ x +   c + ab  ------ + b ------ +  bc – a  ------ y  K B T 3-723
  x y z   x x z  
  2 
+   ca – b  ------ +  a + bc  ------ + c ------ z  K B T
 x y z 

where (x, y, z) is the spatial vector describing the edge, and KB is the Boltzmann constant.
The quantity  is the electron quasi-Fermi level for electron current and the hole quasi-Fermi
level for hole current. This discretization is the default in MAGNETIC3D and the only one
available in MAGNETIC2D.
One possible weakness of this method is that it uses the quasi-Fermi levels, and for minority
carriers these are sometimes only weakly coupled to the boundaries. This can cause
instabilities in the solution. To counteract this, you use the parameters MAGMIN.N and
MAGMIN.P on the MODELS statement. The effect of these parameters is to modify the value of
electron concentration and hole concentration used in the calculation of the respective quasi-
Fermi levels to

n eff = n + MAGMIN · N  N c 3-724

and

p eff = p + MAGMIN · P  N v 3-725

The values of MAGMIN.N and MAGMIN.P are typically in the range 0 to 1. The larger the value,
the more insensitive the gradients of the quasi-Fermi levels are to the actual carrier densities.
Consequently, they become like the electric field at large values of MAGMIN.N and MAGMIN.P.
If MAGMIN.N or MAGMIN.P are set to be a negative value, then the respective addition to the
Bernoulli function argument for electrons or holes Equation 3-723 is set to zero.
The boundary conditions on the electrostatic potential are modified in the presence of a
magnetic field. On all external surfaces and internal interfaces with insulating materials, the
normal gradient of the electrostatic potential is set to be
 =  An  B  Jn  + Ap  B  Jp    N 3-726
-------
N sc

where N is an outward normal to the surface, J n and J p are the current densities near the
surface. Also
n Rn
A n = ------------ 3-727


and
p Rp
A p = ------------ 3-728


and where Rn,p are given explicitly by

347 ATLAS User’s Manual

Carrier Transport in an Applied Magnetic Field Physics

–n rn
R n = ------------------------
- 3-729
n n + p p

p rp
R p = ------------------------
- 3-730
n n + p p

The quantities Rn and Rp simplify to the respective unipolar Hall coefficients in the limits n
>> p and p >> n. The boundary conditions can be viewed as a local Hall field.
To invoke this model in ATLAS2D, simply supply a value for BZ on the MODELS statement.
Nonzero values of BX and BY are not permitted because they would generally result in current
in the Z direction, thereby voiding the assumption of two-dimensionality. In ATLAS3D, any
combination of BX, BY or BZ can be specified as long as condition (Equation 3-717) is
satisfied.
Alternatively, a position dependent magnetic field vector can be specified by using a C-
Interpreter function. Use the parameter F.BFIELD on the MODELS statement to specify the
name of the file containing an implementation of the C-Interpreter function. The magnetic
field should satisfy the constraint imposed by Maxwells equations that the divergence of the
magnetic field vector is everywhere zero.
Table 3-139 shows the parameters used for the carrier transport in a magnetic field.

Table 3-139 Parameters for Galvanic Transport Model

Statement Parameter Type Default Units

MODELS BX Real 0.0 Tesla

MODELS BY Real 0.0 Tesla
MODELS BZ Real 0.0 Tesla
MODELS F.BFIELD Character
METHOD MAG3DINT.N Logical False
METHOD MAG3DINT.P Logical False
METHOD MAG2DLSQ Logical False
MODELS MAGMIN.N Real 0.0
MODELS MAGMIN.P Real 0.0
MODELS R.ELEC Real 1.1
MODELS R.HOLE Real 0.7

348 ATLAS User’s Manual

Anisotropic Relative Dielectric Permittivity Physics

3.12 Anisotropic Relative Dielectric Permittivity

In the general case, the relationship between the electric displacement vector and the electric
field is given by:

D = 0  E 3-731
˜

where , the relative dielectric permittivity, is a symmetric second order tensor represented by
a 33 matrix. Since it is symmetric, it can be written as:

  xx x y  x z
 
 =   x y  yy y x 

3-732
˜  
 x z  yz  zz 

In the isotropic case, the off-diagonal terms are zero and the diagonal terms all have the same
value as each other. In the anisotropic case, you can always set the off-diagonal elements to
zero by a suitable transformation of coordinates. This coordinate system may not match the
coordinate system used by ATLAS for the simulation. Therefore, it is necessary to allow
ATLAS to model anisotropic permittivity for the relationship Equation 3-732 where the off-
diagonal elements are non-zero.
The discretization of the flux term in the Poisson equation:

0     E  3-733
˜

requires a more general treatment in the anisotropic case. For example, the x-component of
the electric displacement vector will now contain a contribution from all the components of
the electric field:
D x =  xx E x +  xy E y +  xz E z 3-734

and so the discretization is more complicated and uses more CPU time. A simplification can
be made when:
• the permittivity matrix is diagonal.
• the mesh used in modeling the device is rectangular.
In this case, discretization will occur only in the directions of the coordinate axes and simply
using a scalar permittivity for each direction is correct. This retains the simplicity and speed
of the discretization for isotropic materials.
ATLAS allows you to specify a value of anisotropic relative dielectric permittivity using the
PERM.ANISO parameter on the MATERIAL statement. By default, this is the value of (yy) if the
simulation is in 2D and (zz) if it is in 3D. The off-diagonal elements are assumed to be zero.
In the case of a 3D simulation, you can apply the value of PERM.ANISO to the Y direction
instead of the Z direction by specifying YDIR.ANISO or ^ZDIR.ANISO on the MATERIAL
statement. You can specify the value of (xx) by using the PERMITTIVITY parameter on the

349 ATLAS User’s Manual

Anisotropic Relative Dielectric Permittivity Physics

MATERIAL statement. The simpler version of discretization will be used in this case by
default.
If the coordinate system in which the permittivity tensor is diagonal and the ATLAS
coordinate system are non-coincident, then the coordinate transformation can be specified as
a set of vectors
X = (X.X, X.Y, X.Z)
Y = (Y.X, Y.Y, Y.Z)
Z = (Z.X, Z.Y, Z.Z)
where the vector components can be specified on the MATERIAL statement.
The default is that
X = (1.0, 0.0, 0.0)
Y = (0.0, 1.0, 0.0)
Z = (0.0, 0.0, 1.0)
and you should specify all necessary components. You do not need to normalize the vectors to
unit length. ATLAS will normalize them and transform the relative dielectric tensor according
to the usual rules. If you specify any component of X, Y or Z, then the more complete form of
discretization will be used. You can also enable the more complete form of electric
displacement vector discretization by using the E.FULL.ANISO parameter on the MATERIAL
statement.
Table 3-140 shows the parameters used in the Anisotropic Relative Dielectric Permittivity.

Table 3-140 MATERIAL Statement Parameters

Parameter Type Default

PERM.ANISO Real -999.0

YDIR.ANISO Logical False
ZDIR.ANISO Logical True
E.FULL.ANISO Logical False
X.X Real 1.0
X.Y Real 0.0
X.Z Real 0.0
Y.X Real 0.0
Y.Y Real 1.0
Y.Z Real 0.0

350 ATLAS User’s Manual

Anisotropic Relative Dielectric Permittivity Physics

Table 3-140 MATERIAL Statement Parameters

Parameter Type Default

Z.X Real 0.0

Z.Y Real 0.0
Z.Z Real 1.0

351 ATLAS User’s Manual

Single Event Upset Simulation Physics

3.13 Single Event Upset Simulation

The capability of single event upset/photogeneration transient simulation is included in 2D
and 3D using the SINGLEEVENTUPSET statement (see Section 21.55
“SINGLEEVENTUPSET” for more information). It allows you to specify the radial, length,
and time dependence of generated charge along tracks. There can be a single particle strike or
multiple strikes. Each track is specified by an Entry Point Location (x0,y0,z0) and an Exit
Point Location (x1,y1,z1). This is assumed to be a cylinder with the radius defined with the
RADIUS parameter. In 2D, z0 and z1 should be neglected.
The entry and exit points are specified by the ENTRYPOINT and EXITPOINT parameters of the
SINGLEEVENTUPSET statement. These are character parameters that represent the ordered
triplet coordinates of the entry and exit points of the particle track.
The electron/hole pairs generated at any point is a function of the radial distance, r, from the
center of the track to the point, the distance l along the track and the time, t. The
implementation into ATLAS allows you to define the generation rate as the number of
electron-hole pairs per cm3 along the track according to the equation:

G  r l t  =  D ENSITY L1  l  + S B · DENSITY L2  l   R  r  T  t  3-735

where DENSITY and B.DENSITY are defined as the number of generated electron/hole pairs
per cm3.
In radiation studies, the ionizing particle is typically described by the linear charge deposition
(LCD) value, which defines the actual charge deposited in units of pC/m. You can use this
definition within ATLAS by specifying the PCUNITS parameter in the SINGLEEVENTUPSET
statement. If the user- defined parameter, PCUNITS, is set in the SINGLEEVENTUPSET
statement, then B.DENSITY is the generated charge, in pC/m, and the scaling factor S is:
1
S = ---------------------------------
- 3-736
2
q RADIUS

where RADIUS is defined on the SINGLEEVENTUPSET statement. If the PCUNITS parameter

isn’t set, then B.DENSITY is the number of generated electron/hole pairs in cm-3 and the
scaling parameter, S, is unity.
Another common measure of the loss of energy of the SEU particle, as it suffers collisions in
a material, is the linear energy transfer (LET) value, which is given in units of MeV/mg/cm2
(or MeV-cm2/mg). In silicon, you can convert energy to charge by considering you need
approximately 3.6 eV of energy to generate an electron-hole pair. The conversion factor from
the LET value to the LCD value is then approximately 0.01 for silicon. So for instance, a LET
value of 25 MeV-cm2/mg is equivalent to 0.25 pC/m, which can then be defined with the
B.DENSITY and PCUNITS parameters.
The factors, L1 and L2, in Equation 3-735 define the variation of charge or carrier generation
along the path of the SEU track. These variations are defined by the equations:
L1  l  = A1 + A2  l + A3 exp  A4  l  3-737

352 ATLAS User’s Manual

Single Event Upset Simulation Physics

B4
L2  l  = B1  B2 + l  B3  3-738

where the A1, A2, A3, A4, B1, B2, B3, and B4 parameters are user-definable as shown in Table
3-141.

Table 3-141 User-Specifiable Parameters for Equations 3-737 and 3-738

Statement Parameter Default Units

SINGLEEVENTUPSET A1 1
SINGLEEVENTUPSET A2 0 cm-1
SINGLEEVENTUPSET A3 0
SINGLEEVENTUPSET A4 0 cm-1
SINGLEEVENTUPSET B1 1
SINGLEEVENTUPSET B2 1
SINGLEEVENTUPSET B3 0 cm-1
SINGLEEVENTUPSET B4 0

Note: The default parameters in Table 3-141 were chosen to result in constant carrier or charge generation as a
function of distance along the particle track.

The factor R(r) is the radial parameter, which is defined by one of two equations. The default
is:

R  r  = exp  – ------r------ 3-739

 RADIUS

where r is the radial distance from the center of the track to the point and RADIUS is a user-
definable parameter as shown in Table 3-142. You can choose an alternative expression if you
specify the RADIALGAUSS parameter on the SINGLEEVENTUPSET statement. In this case,
R(r) is given by:
2
R  r  = exp  – ------------
r
3-740
 RADIUS

The time dependency of the charge generation T(t) is controlled with the TC parameter
through two functions.
For TC=0:
T  t  = deltafunction  t – T0  3-741

353 ATLAS User’s Manual

Single Event Upset Simulation Physics

For TC>0:

t – TO 2
2e –  ---------------
 TC 
T  t  = ------------------------------------------- 3-742
TC  erfc  ----------
– TO
 TC 

where TO and TC are parameters of the SINGLEEVENTUPSET statement.

Table 3-142 User-Specifiable Parameters for Equations 3-735, 3-736, 3-741, and 3-742

Statement Parameter Default Units

SINGLEEVENTUPSET DENSITY 0.0 cm-3

SINGLEEVENTUPSET B.DENSITY 0.0 cm-3 or pC/m
SINGLEEVENTUPSET TO 0.0 s
SINGLEEVENTUPSET TC 0.0 s
SINGLEEVENTUPSET RADIUS 0.05 m

The following example shows a particle strike that is perpendicular to the surface along the Z
plane. Therefore, only the Z coordinates change has a radius of 0.1 m and a LET value of 20
MeV-cm2/mg (B.DENSITY=0.2). The strike has a delay time of 60 ps and the Gaussian
profile has a characteristic time of 10 ps. The generation track for this striking particle is from
z=0 to z=10 m and carrier generation occurs along its entire length.
single entry="5.5,0.0,0.0" exit="5.5,0.0,10" radius=0.1 \
pcunits b.density=0.2 t0=60.e-12 tc=10.e-12
User-Defined SEU
In addition to the model described above, you can use the C-Interpreter to specify an arbitrary
generation profile as a function of position and time. This is specified using the F.SEU
parameter of the SINGLEEVENTUPSET statement.
SINGLEEVENTUPSET F.SEU=myseu.c
.
SOLVE DT=1e-14 TSTOP=1e-7
The F.SEU parameter indicates an external C-language sub-routine conforming to the
template supplied. The file, myseu.c, returns a time and position dependent value of carrier
generation. For more information about the C-Interpreter, see Appendix A “C-Interpreter
Functions”.

354 ATLAS User’s Manual

Field Emission from Semiconductor Surfaces Physics

3.14 Field Emission from Semiconductor Surfaces

Vacuum microelectronic devices often utilize field emission from the sharp tips of emitters
made from semiconductor materials [66][191]. The current emitted can be modeled using a
tunneling model, usually a variant of the Fowler-Nordheim model. ATLAS implements one
such model, which we refer to as the Zaidman model [328]. In this model, the tunneling
current is calculated at each segment of a semiconductor/vacuum interface. The current
density in A/cm2 is given by
2 –7 2 32
– F · BE  0.95 – 1.43641  10 F    
- exp  ----------------------------------------------------------------------------------------------------------
F · AEF
J = ------------------------------------------- - 3-743
TSQ · ZAIDMAN  F 

Here,  is the semiconductor-vacuum barrier height in eV. F is the electric field strength
normal to the surface in units of V/cm and TSQ.ZAIDMAN is a fitting parameter. The
parameters F.AE and F.BE are also fitting parameters.
The current produced in each segment then follows the field lines until it either reaches a
contact or exits the device boundary. To enable the model, use the ZAIDMAN parameter on the
MODELS statement. It should not be used concurrently with the FNORD model.
It presently omits self-consistent evaluation with respect to the carrier continuity equations. It
also neglects any modification of the current in the vacuum arising from the space charge
present there.

Table 3-143 Zaidmann Model Parameters

Statement Parameter Default Units

MODELS ZAIDMAN False

MODELS TSQ.ZAIDMAN 1.1
MODELS F.AE 1.541410-6 Q2/Vs
MODELS F.BE 6.8308107 cm-1 V-1/2Q-3/2

355 ATLAS User’s Manual

Conduction in Silicon Nitride Physics

3.15 Conduction in Silicon Nitride

It is known that at high fields and for temperatures above 325 Kelvin, the conduction in
Silicon Nitride is dominated by Poole-Frenkel emission. ATLAS can model this by adding an
electron generation term to the electron continuity equation. To use this model, you define a
trap density, NT.N, using the NITRIDECHARGE statement. You also change the Silicon Nitride
layer into a semiconductor using the SEMICONDUCTOR parameter on the MATERIAL statement.
The electron generation term is proportional to

PF · A exp  – ------------------------------------
PF BARRIER
- + PF · B exp  – ------------------------------------
PF BARRIER Beta F
· · - + -------------------- 3-744
 KB TL   KB TL KB TL 

where PF.A, PF.B, and PF.BARRIER can be specified on the NITRIDECHARGE statement. The
first term is an ohmic term that will dominate at low fields. The local electric field is F, the
lattice temperature is TL, KB is the Boltzmann constant, and Beta is a factor obtained from the
equation:

Beta = Q Q     3-745

where Q is the electonic charge, and  is the dielectric permittivity of Silicon Nitride. This
model only applies to steady state simulations. The traps should first be charged using the
NIT.N parameter on the SOLVE statement. The value MUN on the MOBILITY statement can
also be used to calibrate the resulting I-V curves.

356 ATLAS User’s Manual

Generic Ion Transport Model Physics

3.16 Generic Ion Transport Model

The ability to model the transport of ionic species in semiconductor devices is important for
studies of performance degradation and for investigating the behavior of novel devices [282].
ATLAS is able to simulate the transport of up to three different ionic species within the drift-
diffusion framework. Denoting the concentrations of the ionic species as C1 , C2, and C3 , and
their fluxes as Fc1, Fc2, and Fc3 respectively, the continuity equation for species i is
C i  r t 
- +   F ci  r t  = G i  r t  – R i  r t 
-------------------- 3-746
t

where G i  r t  and R i  r t  are the generation and recombination rates of the ions.

All quantities are functions of position r, and time t. The general equation for the flux is
F Ci =  i C i  r t  – D i C i  r t  3-747

where  i is the velocity and D i the diffusion co-efficient. The velocity due to the
electrostatic potential,  is given by
Zi
 i = – -------  i   r t  3-748
Zi

where Zi is an integer denoting the number of electronic charges on the ion. For a positively
charged species, Zi is positive, and is negative for a negatively charged species. Instead of
working with the ionic mobility i, we use the generalised Einstein relationship
Di Zi
 i = -------------
- 3-749
KB T

to give a flux equation

Zi
F Ci = – D i  ---------- C i  r t    r t  + C i  r t  3-750
 KB T 

where K B is the Boltzmann constant and T is the device temperature. For a neutral species
Zi = 0, and Equation 3-750 simplifies to Fick's law. For non-zero values of Zi, the
electrostatic field adds a contribution to the flux according to the sign and magnitude of Zi.
The boundary conditions used are that no ions can enter or leave the device, either through its
external surfaces or its contacts. In practice some transport out of the contacts may occur.
But, a steady state flow of ions through the device is not expected and thus cannot be
simulated. These conditions of zero external flux mean that the total ion concentrations are
preserved and that the net Generation-Recombination term for each ionic species must be
zero. This can be proven by integrating Equation 3-746 over the device and applying Gauss's
flux theorem to give



t V S

---- C i  r t  dV + F ci  n̂ dS =
V

 G i – R i  dV 3-751

The integral of the flux over the device surface is identically zero and so for a steady state
solution the integral of generation-recombination terms must be zero. This implies that the
total number of ions in the device remains constant.

357 ATLAS User’s Manual

Generic Ion Transport Model Physics

To simulate one or more ionic species you first specify an initial distribution by using the
DOPING statement. One of the parameters SPECIES1, SPECIES2, or SPECIES3 must be
specified on a DOPING statement to set an initial ionic species distribution. The species
distribution can be set in insulator regions as well as semiconductor regions. In the absence
of inter-species reactions, the initial total dose of each simulated species will be preserved
throughout the simulation. SOLVE INIT does not calculate equilibrium values of the SPECIES
concentrations. This is because the timescale to achieve equilibrium may be much longer
than a typical observation time of a device.
You select the number of species to simulate using the NSPECIES parameter on the MODELS
statement. This can be 1, 2, or 3. If GIGA is enabled this number can only be 1 or 2.
NSPECIES=1 causes the simulation of SPECIES1. NSPECIES=2 simulates SPECIES1 and
SPECIES2. NSPECIES=3 simulates SPECIES1, SPECIES2, and SPECIES3. You utilize the
SPECIES1.Z, SPECIES2.Z, and SPECIES3.Z parameters on the MODELS statement to assign
charges to the ionic species.
You set the diffusion coefficient Di for each species on a material by material basis with the
MATERIAL statement. It is given by
2
D 1 = SPECIES1.HOP SPECIES1.AF EXP  – SPECIES1.EA/KT 
2
D 2 = SPECIES2.HOP SPECIES2.AF EXP  – SPECIES2.EA/KT  3-752
2
D 3 = SPECIES3.HOP SPECIES3.AF EXP  – SPECIES3.EA/KT 

where all the parameters are set on the MATERIAL statement. The units of Di are cm2 s-1.
Different ionic species with the same diffusion co-efficient but different values of Zi will have
different transport properties. This can be seen from consideration of Equation 3-750.
For ions with a very small diffusion coefficient, the steady state solution may not be
appropriate as it corresponds to an infinite simulation time. On a practical timescale, the
species distribution may not have changed by a significant amount. In this case, you specify
the FREEZESPECIES parameter on a steady state SOLVE, followed by a transient SOLVE
without the FREEZESPECIES parameter. The FREEZESPECIES parameter gives the steady
state solution with the species concentration held fixed. The transient solve then gives the
dynamic development of the species distribution. It is also possible to perform small signal
AC simulations in addition to both steady state and transient simulations.
The multistate traps described in Section 3.3.4 “Multistate Traps” can emit and trap ionic
species. Because ionic species can be modeled in all materials, the INTTRAP [MSCTRAP]
statement can be made to apply to interfaces between insulators and electrodes, insulators and
conducting regions, and different insulating regions. The respective flags are I.M, I.C, and
I.I. This is in addition to the other defined interface types.
The default is that the transport of the ionic species is simulated in the whole device. You can
restrict the ionic species to insulator regions by using the SPECIES.INS parameter on the
METHOD statement. This applies the interface condition that there is no ionic flow across
semiconductor-insulator boundaries. In addition, the maximum relative update of the species
concentrations in a non-linear solver iteration can be limited to SPECIES.MAXX, where
SPECIES.MAXX is specified on the METHOD statement.
This can help improve convergence, as can limiting the maximum update for other variables.

358 ATLAS User’s Manual

Generic Ion Transport Model Physics

The densities of the ionic species are automatically output to any Standard Structure file that
is saved. You can use the parameters SPECIES1, SPECIES2, or SPECIES3 on the PROBE
statement to save ionic species quantities to a log file.
The input deck parameters that you use for the generic ionic species transport model are
summarized in Table 3-144 and some example statements are given below.
How to include reactions between the species is described in Section 3.17 “Generic Ion
Reaction Model”.

Table 3-144 Parameters for the Generic Ionic Species Transport Model

Statement Parameter Type Default Units

DOPING SPECIES1 Logical False

DOPING SPECIES2 Logical False
DOPING SPECIES3 Logical False
MODELS NSPECIES Integer 0
MODELS SPECIES1.Z Integer 0
MODELS SPECIES2.Z Integer 0
MODELS SPECIES3.Z Integer 0
PROBE SPECIES1 Logical False
PROBE SPECIES2 Logical False
PROBE SPECIES3 Logical False
MATERIAL SPECIES1.HOP Real 1.0e-6 cm
MATERIAL SPECIES1.AF Real 1.0e15 Hz
MATERIAL SPECIES1.EA Real 0.25 eV
MATERIAL SPECIES2.HOP Real 1.0e-6 cm
MATERIAL SPECIES2.AF Real 1.0e15 Hz
MATERIAL SPECIES2.EA Real 0.25 eV
MATERIAL SPECIES3.HOP Real 1.0e-6 cm
MATERIAL SPECIES3.AF Real 1.0e15 Hz
MATERIAL SPECIES3.EA Real 0.25 eV
METHOD SPECIES.INS Logical False
METHOD SPECIES.MAXX Real 1000.0
SOLVE FREEZESPECIES Logical False

359 ATLAS User’s Manual

Generic Ion Transport Model Physics

Examples
//Set up initial profile for SPECIES1
DOPING REGION=1 SPECIES1 GAUSSIAN CONC=1e15 CHAR=0.06 PEAK=0.25

//Enable simulation of 3 species with charges +1,-2 and -1

respectively.
MODELS NSPECIES=3 SPECIES1.Z=1 SPECIES2.Z=-2 SPECIES3.Z=-1

// Get the value of species1 integrated over whole device

PROBE SPECIES1 INTEGRATE

// Restrict simulation to insulator regions only, and limit maximum

relative update of all species in non-linear iteration to be 100.
METHOD SPECIES.MAXX=100 SPECIES.INS

360 ATLAS User’s Manual

Generic Ion Reaction Model Physics

3.17 Generic Ion Reaction Model

The generic ion transport model of the previous section is augmented by the capability to
simulate general reactions between the ionic species. The examples are
• Capture of free carriers to change the ionic charge state.
• Emission of ionic charges to change the ionic charge state.
• Dimerisation reactions.
• Cation-Anion reactions.
• Vacancy-Interstitial pair generation and recombination. (This is not available yet due to
licensing issues).
You specify the reactions required on one or more REACTION statements. The syntax to do
this requires separating each reaction into reactants and products, or more colloquially into
before and after. The reactions are potentially reversible so this is only a semantic distinction.
Each equation is written as
- +
BEFORE.N e + BEFORE.P h + BEFORE.SP1 SPECIES1 + BEFORE.SP2 SPECIES2 + BEFORE.SP3 SPECIES3
 3-753
- +
AFTER.N e + AFTER.P h + AFTER.SP1 SPECIES1 + AFTER.SP2 SPECIES2 + AFTER.SP3 SPECIES3

For example, to simulate two charge states of a species, X+ and X++, you need to take into
account the possible reactions with free carriers
++ - +
X +e X
3-754
++ + ++
X +h X

Assume that X+ has been set up as SPECIES1 and X++ as SPECIES2, using SPECIES1.Z=1 and
SPECIES2.Z=2 on the MODELS statement. Then the first equation is set up with the statement
REACTION BEFORE.N=1 BEFORE.SP2=1 AFTER.SP1=1
and the second reaction by
REACTION BEFORE.P=1 BEFORE.SP1=1 AFTER.SP2=1
Modeling a double electron capture/emission
3+ - - +
X +e +e X 3-755
3+ +
with X as SPECIES1 and X as SPECIES2 requires the statement
REACTION BEFORE.N=2 BEFORE.SP1=1 AFTER.SP2=1
To model a dimerisation reaction 2H  H2 with H as SPECIES1 and H2 as SPECIES2
requires the statement
REACTION BEFORE.SP1=2 AFTER.SP2=1
The format allows a wide range of reactions to be specified. ATLAS checks that charge
conservation is obeyed in each REACTION statement and exits with an error if charge
conservation is violated.

361 ATLAS User’s Manual

Generic Ion Reaction Model Physics

You also need to model the reaction rates. This is calculated by multiplying together all the
densities of the reactant components present in the reaction, raised to the power of their
BEFORE coefficients. Thus, a SPECIES1 concentration is raised to the power of BEFORE.SP1
and the result multiplies the forward rate coefficient, similarly for the other SPECIES
concentrations and electron and hole concentrations. This is then multipled by
 – FORWARD.EA/KT 
FORWARD.RATEe to give the forward reaction rate. The same process is
 – REVERSE.EA/KT 
applied to the product species, and this is multiplied by REVERSE.RATEe .
If a BEFORE or AFTER coefficient is zero, then the corresponding concentration will not appear
in the rate equation.
For the reaction described above in Equation 3-755, the net rate is
 – FORWARD.EA/KT  3+ -2  – REVERSE.EA/KT  +
FORWARD.RATEe X  e – REVERSE.RATEe X  3-756

where the square brackets denote the local concentration of the bracketed quantity. Because
the reaction consumes two electrons for every ion, the electron concentration is raised to the
power of 2, which is the value of BEFORE.N.
You can add a constant value to the reaction rate using the EQUIRATE parameter on the
REACTION statement. This is useful if you have a constant product of species in equilibrium,
0 0
say a thermally generated vacancy-interstitial pair with reaction VX + IX   NULL  , then
the reaction rate would be
 – FORWARD.EA/KT  0 0
FORWARD.RATEe  VX   IX  – EQUIRATE 3-757

because there are no terms on the AFTER side of the reaction.

The reaction rates are included in the Gi - Ri terms of the continuity equations of the free
carriers and ionic species. Terms on the BEFORE side of the equation are included as
recombination rates and those on the AFTER side as generation rates. The rates are multiplied
by the relevant BEFORE or AFTER coefficient. This leads to a conservation law for the sums of
the volume integrals of the species, so long as one takes into account the composition
numbers of the species. This conservation law is

V  SPECIES1.A SPECIES1 r  + SPECIES2.A SPECIES2 r  + SPECIES3.A SPECIES3 r   dV = const 3-758

where SPECIES1.A, SPECIES2.A, and SPECIES3.A are the composition numbers. ATLAS
attempts to work these out automatically, but if this is impossible then you set them on the
MODELS statement. The following example clarifies this.
We consider two possible reactions between three species. SPECIES1 is an electrically
neutral atom, H. SPECIES2 is the neutral molecule, H2. SPECIES3 is the negatively charged
ion, H-. You set these up on the MODELS statement
MODELS NSPECIES=3 SPECIES1.Z=0 SPECIES2.Z=0
SPECIES3.Z=-1 [SPECIES1.A=1 SPECIES2.A=2 SPECIES3.A=1]
where the composition number specification is optional. The reactions considered are a
dimerisation

362 ATLAS User’s Manual

Generic Ion Reaction Model Physics

2H  H 2 3-759

with a net reaction rate Rdimer and an electron capture reaction

- -
H+e H 3-760
with a net reaction rate Relec and a hole capture reaction
- +
H +h H 3-761
and a net reaction rate Rhole. These reactions are setup using the following three REACTION
statements (rate parameters are omitted for clarity)
REACTION BEFORE.SP1=2 AFTER.SP2=1
REACTION BEFORE.SP1=1 BEFORE.N=1 AFTER.SP3=1
REACTION BEFORE.SP3=1 BEFORE.P=1 AFTER.SP1=1
The continuity equations for the three species become
dH
------- +   F H = – 2 dimer – R elec + R hole
dt
dH
---------2- +   F H = R dimer 3-762
dt 2

-
dH
--------- + F - = R elec – R hole
dt H

If two times the middle equation is added to the other two equations, then the right hand side
of the equation identically vanishes. The resulting equation is integrated over the device
volume. The divergence theorem is applied to the flux terms, which will then vanish because
of the zero flux boundary condition. The equation is integrated over time to give

V  H + 2H2 + H  dV = const
-
3-763

which is the same as obtained from Equation 3-758 after putting in the composition numbers.
If ATLAS cannot automatically determine the composition numbers, then you must specify
them on the MODELS statement using the parameters SPECIES1.A, SPECIES2.A, and
SPECIES3.A. ATLAS will check that your values are correct.
The above set of reactions also result in a recombination rate of Relec going into the electron
continuity equation and Rhole going into the hole continuity equation. These carriers are
supplied in the usual way.
The net reaction rates of the ionic species and electrons and holes are output to any Standard
Structure file, which is saved if you set the parameter U.SPECIES on the OUTPUT statement.
You use the parameters REACTION.ELEC, REACTION.HOLE, REACTION.SP1, REACTION.SP2,
or REACTION.SP3 on the PROBE statement to output the net reaction rates for each carrier or
species to a log file.
The input deck parameters that you use for the generic species reaction model are
summarised in Table 3-145. The associated generic transport model is described in the
previous section. All REACTION statements must occur after the MODELS statement which
defines the SPECIES.

363 ATLAS User’s Manual

Generic Ion Reaction Model Physics

Table 3-145 Parameters for the Generic Ionic Species Reaction Model

Statement Parameter Type Default Units

REACTION BEFORE.N Integer 0

REACTION BEFORE.P Integer 0
REACTION BEFORE.SP1 Integer 0
REACTION BEFORE.SP2 Integer 0
REACTION BEFORE.SP3 Integer 0
REACTION AFTER.N Integer 0
REACTION AFTER.P Integer 0
REACTION AFTER.SP1 Integer 0
REACTION AFTER.SP2 Integer 0
REACTION AFTER.SP3 Integer 0
REACTION FORWARD.RATE Real 0.0 Note 1
REACTION REVERSE.RATE Real 0.0 Note 2
REACTION EQUIRATE Real 0.0 cm-3s-1
REACTION FORWARD.EA Real 0.0 eV
REACTION REVERSE.EA Real 0.0 eV
MODELS SPECIES1.A Integer 1
MODELS SPECIES2.A Integer 1
MODELS SPECIES3.A Integer 1
PROBE REACTION.ELEC Logical False
PROBE REACTION.HOLE Logical False
PROBE REACTION.SP1 Logical False
PROBE REACTION.SP2 Logical False
PROBE REACTION.SP3 Logical False
OUTPUT U.SPECIES Logical False

1. Should give dimensions of forward rate as cm-3s-1

2. Should give dimensions of reverse rate as cm-3s-1

364 ATLAS User’s Manual

Chapter 4
S-Pisces: Silicon Based 2D
Simulator
Overview S-Pisces: Silicon Based 2D Simulator

4.1 Overview
S-PISCES is a powerful and accurate two-dimensional device modeling program that
simulates the electrical characteristics of silicon based semiconductor devices including
MOS, bipolar, SOI, EEPROM, and power device technologies.
S-PISCES calculates the internal distributions of physical parameters and predicts the
electrical behavior of devices under either steady-state, transient, or small signal AC
conditions. This is performed by solving Poisson’s Equation (see Section 3.1.1 “Poisson’s
Equation”) and the electron and hole carrier continuity equations in two dimensions (see
Section 3.1.2 “Carrier Continuity Equations”).
S-PISCES solves basic semiconductor equations on non-uniform triangular grids. The
structure of the simulated device can be completely arbitrary. Doping profiles and the
structure of the device may be obtained from analytical functions, experimentally measured
data, or from process modeling programs SSUPREM3 and ATHENA.
For more information on semiconductor equations, see Section 3.1 “Basic Semiconductor
Equations”.
You should be familiar with ATLAS before reading this chapter any further. If not, read
Chapter 2 “Getting Started with ATLAS”.

366 ATLAS User’s Manual

Simulating Silicon Devices Using S-Pisces S-Pisces: Silicon Based 2D Simulator

4.2 Simulating Silicon Devices Using S-PISCES

4.2.1 Simulating MOS Technologies
Physical Models for MOSFETs
S-PISCES provides special physical models tuned for simulating MOS devices. Most of these
models are accessed from the MODEL statement. The MOS parameter of the MODEL statement
can be specified to turn on a default set of physical models that are most useful for MOS
simulation. The MOS parameter enables Shockley-Read-Hall (SRH), Fermi Statistics
(FERMI), and the Lombardi Mobility model (CVT) for transverse field dependence. To set the
default MOS simulation models , use:
MODEL MOS PRINT
The transverse field dependent mobility models are of particular importance for simulating
MOS devices. S-PISCES currently supports several different transverse field dependent
mobility models. The CVT parameter selects the Lombardi CVT model. The YAMA parameter
selects the Yamaguchi model. The TASCH parameter selects the Tasch model. The WATT
parameter selects the Watt Surface Model, which can be operated in a more accurate mode
with the extra parameter, MOD.WATT, on the MOBILITY statement.
You will find that the MOBILITY statement can be used to modify some of the parameters of
the various models, to apply different models to different regions, or to apply different models
to electrons and holes.
Meshing for MOS Devices
In device simulation of MOS devices, the key areas for a tight mesh are:
• very small vertical mesh spacing in the channel under the gate. The exact size of mesh
required depends on the transverse field or surface mobility model chosen. See Figure 4-1
for an example of the effect of mesh size on drain current.
• lateral mesh spacing along the length of the channel for deep sub-micron devices. This is
required to get the correct source-drain resistance and to resolve the channel pinch-off
point.
• lateral mesh at the drain/channel junction for breakdown simulations. This is required to
resolve the peak of electric field, carrier temperature and impact ionization.
• several vertical grid spacings inside the gate oxide when simulating gate field effects such
as gate induced drain leakage (GIDL) or using any hot electron or tunneling gate current
models
Figures 4-1 and 4-2 show the effect of mesh size in the MOS channel on IV curves. In Figure
4-1, the mesh density in the vertical direction is increased. As the mesh density increases, the
resolution of the electric field and carrier concentration is improved. This example uses the
CVT mobility model. Improvements in transverse electric field resolution lead to a reduced
mobility in the channel and a stronger IV roll-off.
But Figure 4-2 shows the effect of surface channel mesh in MOSFETs is model dependent.
This result shows the current at Vds=3.0V and Vgs=0.1V versus the size of the first grid
division into the silicon. Results vary for each model but note that for all models, a very fine
grid is required to reduce the grid dependence to acceptable levels.

367 ATLAS User’s Manual

Simulating Silicon Devices Using S-Pisces S-Pisces: Silicon Based 2D Simulator

Figure 4-1: Effect on MOS IV curve of progressive refinement of the vertical mesh spacing at the surface of
the MOS channel

Figure 4-2: Effect of surface mesh spacing on simulated current for several MOS Mobility Models

368 ATLAS User’s Manual

Simulating Silicon Devices Using S-Pisces S-Pisces: Silicon Based 2D Simulator

MOS Electrode Naming

For MOS simulation, S-PISCES allows you to use standard electrode names to reduce
confusion with the use of electrode indices. These names include: source, drain, gate, and
substrate. Electrode names can be defined in ATHENA or DEVEDIT or in the ELECTRODE
statement in ATLAS. These names can be used in the SOLVE statements for setting bias
voltages such as:
SOLVE VGATE=1.0 VSTEP=1.0 VFINAL=5.0 NAME=GATE
Gate Workfunction
In MOS simulations, the workfunction of the gate is an important parameter. This must be set
in each input deck using the WORK parameter of the CONTACT statement. For example:
CONTACT NAME=GATE WORK=4.17
would set the workfunction on the gate at 4.17 eV.
Certain material names can also be used to set the workfunction of common gate materials.
For example:
CONTACT NAME=GATE N.POLY
would set the workfunction on the gate to that of n type polysilicon.

Note: The gate workfunction should be set on a CONTACT statement even though the material or workfunction
might be set from ATHENA or DEVEDIT.

Interface Charge
For accurate simulation of MOS devices, the interface charge at the oxide, specify
semiconductor interface. To do this, set the QF parameters for the INTERFACE statement.
Typically, a value of 31010 cm-2 is representative for the interface charge found in silicon
MOS devices. The proper syntax for setting the value for this interface fixed charge is:
INTERFACE QF=3e10
You can also try to model the fixed charge more directly by using interface traps to simulate
the surface states. To do this, use the INTTRAP statement. But this is rarely done in practice.
Single Carrier Solutions
Frequently for MOS simulation, you can choose to simulate only the majority carrier. This
will significantly speed up simulations where minority carrier effects can be neglected. This
can be done by turning off the minority carrier. To do this, use the ATLAS negation character,
^ , and one of the carrier parameters (ELECTRONS or HOLES) in the METHOD statement. For
example, to simulate electrons only, you can specify one of the following:
METHOD CARRIERS=1 ELECTRONS
METHOD ^HOLES
Single carrier solutions should not be performed where impact ionization, any recombination
mechanism, lumped element boundaries, or small signal AC analysis are involved.

369 ATLAS User’s Manual

Simulating Silicon Devices Using S-Pisces S-Pisces: Silicon Based 2D Simulator

Energy Balance Solutions

As MOS devices become smaller and smaller, non-local carrier heating effects becomes
important. You can perform accurate simulation of non-local carrier heating effects using the
energy balance model (EBM). As a general rule, you can use the gate length as a gauge to
predict when non-local effects are important. Generally, for drain currents, energy balance
should be applied for gate lengths less than 0.5 microns. For substrate currents, energy
balance should be applied for gate lengths less than 1.0 micron.
To enable energy balance for electrons/holes, set either the HCTE.EL or HCTE.HO parameters
on the MODEL statement. For example:
MODEL HCTE.EL
enables the energy balance model for electrons.
ESD Simulation
In some cases, lattice heating may be important to MOS simulation. This typically occurs in
cases with very high currents, just like the case with ESD simulation. In these cases, GIGA
should be used to simulate the heat-flow in the device. To enable heat flow simulation, set the
LAT.TEMP parameter of the MODEL statement (a license for GIGA is required). For example,
the statement:
MODEL LAT.TEMP
enables heat-flow simulation.

4.2.2 Simulating Silicon Bipolar Devices

Physical Models for BJTs
S-PISCES provides special physical models for bipolar device simulation. You can select these
models using the MODEL statement. The BIPOLAR parameter of the MODEL statement enables a
reasonable default set of bipolar models. These include: concentration dependent mobility
(CONMOB), field dependent mobility (FLDMOB), bandgap narrowing (BGN), concentration-
dependent lifetime (CONSRH) and Auger recombination (AUGER). If LAT.TEMP is also
specified on the MODELS statement, or the TEMPERATURE parameter differs from 300 Kelvin
by more than 10 Kelvin, then the ANALYTIC model is used instead of CONMOB. This is because
of the narrow valid temperature range for CONMOB.
For the most accurate bipolar simulations, the recommended mobility model is the Klaassen
model (KLA). This includes doping, temperature and carrier dependence. It applies separate
mobility expressions for majority and minority carriers. You should also use this model with
Klaassen’s Auger model (KLAAUG) and Klaassen’s concentration dependent SRH model
(KLASRH). Combine the mobility model with FLDMOB to model velocity saturation. For
surface (or lateral) bipolar devices, you can use the Shirahata model (SHI) to extend the
Klaassen model with a transverse electric field dependence. The most accurate and
appropriate model statement for bipolar devices is therefore:
MODELS KLA FLDMOB KLASRH KLAAUG BGN FERMI PRINT
You can choose this set of models by using the BIPOLAR2 parameter from the MODEL
statement.

370 ATLAS User’s Manual

Simulating Silicon Devices Using S-Pisces S-Pisces: Silicon Based 2D Simulator

Note: For a complete syntax including description of models and method for simulating polysilicon emitter bipolar
devices, see the BJT directory in the on-line examples.

Meshing Issues for BJTs

The most important areas to resolve in bipolar transistors are the emitter/base and base/
collector junctions. Typically, a very fine mesh throughout the base region is required. The
gain of the bipolar device is determined primarily by the recombination at the emitter/base
junction or inside the emitter. Therefore, these regions need to be resolved with a fine mesh.
BJT Electrode Naming
S-PISCES also provides special electrode names for bipolar simulation that can be used to ease
confusion over electrode indices. These electrode names include: “emitter”, “base”,
“collector”, “anode”, and “cathode”. Electrode names can be defined in ATHENA or
DEVEDIT or in the ELECTRODE statement in ATLAS. The electrode names are used on SOLVE
statement. For example:
SOLVE VBASE=1.0 VSTEP=1.0 VFINAL=5.0 NAME=BASE
Dual Base BJTs
It is possible in S-PISCES to tie two or more contacts together so that voltages on both contacts
are equal. This is useful for many technologies for example dual base bipolar transistors.
There are several methods for achieving this depending on how the structure was initially
defined.
If the structure is defined using ATLAS syntax, it is possible to have multiple ELECTRODE
statements with the same NAME parameter defining separate locations within the device
structure. In this case, the areas defined to be electrodes will be considered as having the same
applied voltage. A single current will appear combining the current through both ELECTRODE
areas.
Similarly, if two separate metal regions in ATHENA are defined using the ATHENA
ELECTRODE statement to have the same name, then in ATLAS these two electrodes will be
considered as shorted together.
If the electrodes are defined with different names, the following syntax can be used to link the
voltages applied to the two electrodes.
CONTACT NAME=base1 COMMON=base
.
SOLVE VBASE=0.1
Here, the electrode, base1, will be linked to the electrode, base. Then, the applied 0.1V on
base will also appear on base1. But ATLAS will calculate and store separate currents for
both base and base1. This can be a useful feature. In some cases, however, such as where
functions of the currents are required in EXTRACT or TONYPLOT, it is undesirable. You can
add the SHORT parameter to the CONTACT statement above to specify that only a single base
current will appear combining the currents from base and base1.
When loading a structure from ATHENA or DEVEDIT where two defined electrode regions
are touching, ATLAS will automatically short these and use the first defined electrode name.

371 ATLAS User’s Manual

Simulating Silicon Devices Using S-Pisces S-Pisces: Silicon Based 2D Simulator

Creating an Open Circuit Electrode

It is often required to perform a simulation with an open circuit, such as for an open-base
breakdown voltage simulation, on one of the defined electrodes. There are three different
methods that will accomplish this. The first method is to delete an electrode from the structure
file. The second method is to add an extremely large lumped resistance, for example 1020,
onto the contact to be made open circuited. The third method is to first switch the boundary
conditions on the contact to be made open circuited from voltage controlled to current
controlled. Then, specify a very small current through that electrode.
Each of these methods are feasible. But if a floating region is created within the structure
while using these methods, then numerical convergence may be affected. As a result, it is
normally recommended to use the second method to ensure that no floating region is created.
Solution Techniques for BJTs
To obtain bipolar solutions, you almost always need to simulate using two carriers. This is
due to the importance of minority carriers to device operation.
In certain cases, non-local carrier heating may be important to accurately simulate bipolar
devices. In these cases, use the energy balance model. To model non-local carrier heating for
electrons/holes, set the HCTE.EL and HCTE.HO parameters in the MODEL statement. For
example, the statement:
MODEL HCTE.EL
invokes the carrier heating equation for electrons.

4.2.3 Simulating Non-Volatile Memory Technologies (EEPROMs, FLASH Memories)

To simulate non-volatile memory devices, you must become familiar with the basics of
MOSFET simulation (see Section 4.2.1 “Simulating MOS Technologies”).
Defining Floating Gates
To simulate non-volatile memory technologies, such as EEPROMs or FLASH EEPROMs,
specify one or more electrodes as floating gates. To do this, set the FLOATING parameter of
the CONTACT statement. For example:
CONTACT NAME=fgate FLOATING
This specifies that the electrode named fgate is simulated as a floating gate. This means that
the charge on the floating gate is calculated as the integral of the gate current at that gate
during a transient simulation.
Modeling the correct coupling capacitance ratio between the floating gate and control gate
often requires adding an extra lumped capacitor from the floating gate to the control gate or
one of the other device terminals. This is often required since S-PISCES is performing a 2D
simulation whereas the coupling of the gates is often determined by their 3D geometry.
Parameters on the CONTACT statement are used to apply these extra lumped capacitances. For
example, to add a capacitor of 1fF/mm between the control and floating gates the syntax is:
CONTACT NAME=fgate FLOATING FG1.CAP=1.0e-15 EL1.CAP=cgate

372 ATLAS User’s Manual

Simulating Silicon Devices Using S-Pisces S-Pisces: Silicon Based 2D Simulator

Gate Current Models

You can supply the gate currents for the floating gate structure by one of three sources: hot
electron injection (HEI or N.CONCAN), hot hole injection (HHI or P.CONCAN) and Fowler-
Nordheim tunneling current (FNORD).
These currents are of importance, depending on whether electrons are being moved onto the
gate or off the floating gate. In the case of placing electrons on the floating gate, use hot
electron injection and Fowler-Nordheim tunneling. In the case of removing electrons from the
floating gate, set hot hole injection and Fowler-Nordheim tunneling.
In drift diffusion simulations, perform hot electron injection by setting the HEI parameter of
the MODELS statement. To simulate hot hole injection, use the HHI parameter of the MODELS
statement. To enable Fowler-Nordheim tunneling, set the FNORD parameter of the MODELS
statement. The following example demonstrated the proper setting for Flash EPROM
programming.
MODELS MOS HEI PRINT
This next example is appropriate for EEPROM erasure.
MODELS MOS HHI FNORD PRINT
With energy balance simulations, use the Concannon Models for EPROM programming and
erasing. For more information about these models, see “Concannon’s Injection Model” on
page 259.

Note: Writing and erasure of floating gate devices should be done using transient simulation.

Gate Current Assignment (NEARFLG)

The actual calculation of floating gate current magnitude is done at the silicon-oxide
interface. The question of distribution of oxide currents to the various electrodes near the
interface is resolved using one of two models. The actual flow of carriers in oxides is not well
known. Accepted physical models of carrier flow in oxides are still under research. As such,
S-PISCES provides two heuristic models to choose from. The default is to distribute currents
calculated at points along the interface to the electrode in the direction of highest contributing
field. This model is somewhat analogous to a purely drift model of oxide carrier transport.
The alternative is to set the NEARFLG parameter of the MODEL statement. In this case, the
currents calculated at points along the interface are distributed to the geometrically closest
electrode. This model is analogous to a purely diffusion model of carrier transport in oxide.

4.2.4 Simulating SOI Technologies

Silicon substrates are now being produced that contain an oxide layer buried below the
surface of the silicon at some predefined depth. The existence of this buried oxide layer has
resulted in a change not only in the fabrication process used to manufacture a device in the
surface silicon but also in the challenges facing device simulation.
All of the issues raised previously about MOS device simulation should be considered with
some extra SOI specific problems.
The most common device technology that uses these SOI substrates is the SOI MOSFET.
This section summarizes the simulation requirements for SOI using this particular technology
as a reference.

373 ATLAS User’s Manual

Simulating Silicon Devices Using S-Pisces S-Pisces: Silicon Based 2D Simulator

Meshing in SOI devices

The mesh requirements for SOI MOSFETs is very similar to that described in the previous
section for bulk MOS transistors. In addition to these requirements, there are some additional
points to meshing these devices. These are:
• Two channel regions may exist: one underneath the top (front) gate oxide and the other
above the buried (back gate) oxide.
• Inside the buried oxide layer, the mesh constraints can be relaxed considerably compared
with the top gate oxide.
• The active silicon underneath the top gate can act as the base region of a bipolar transistor
and as such may require a finer mesh when compared to bulk MOS transistors.
Physical Models for SOI
SOI MOSFET simulations are based upon the physical operation of the device, which
exhibits both MOS and bipolar phenomena. As a result a more complex set of physical
models will be required than for either MOS or bipolar technologies. Table 4-1 shows these
models.

Table 4-1 SOI Physical Models

Model Description

Mobility Klaassen’s Model (KLA) is recommended to account for lattice

scattering, impurity scattering, carrier-carrier scattering and
impurity clustering effects at high concentration. The
Shirahata mobility model (SHI) is needed to take into account
surface scattering effects at the silicon/oxide interface, which is a
function of the transverse electric field. High electric field velocity
saturation is modelled through the field dependent mobility model
(FLDMOB). You can tune model parameters using the MOBILTY
statement syntax.
Interface Charge In SOI transistors, there exist two active silicon to oxide interfaces
on the wafer. The top interface, under the top gate, is similar to
MOS technology. The bottom interface is quite different and
typically contains significantly more charge. You can set different
interface charges in SPISCES using the INTERFACE statement with
region specific parameters.
Recombination To take account of recombination effects, we recommend the use
of the Shockley-Read-Hall (SRH) model. This simulates the
leakage currents that exist due to thermal generation. You may also
need to simulate the presence of interface traps at the silicon/oxide
interface. Then, turn on the direct recombination model (AUGER).
The parameters for both models can be tuned in the MOBILITY
statement.

374 ATLAS User’s Manual

Simulating Silicon Devices Using S-Pisces S-Pisces: Silicon Based 2D Simulator

Table 4-1 SOI Physical Models

Model Description

Bandgap Narrowing This model (BGN) is necessary to correctly model the bipolar
current gain when the SOI MOSFET behaves like a bipolar
transistor. Specify the parameters for this model in the MODELS
statement.
Carrier Generation Impact ionization significantly modifies the operation of SOI
MOSFETs. To account for this phenomena, switch on the impact
ionization model (IMPACT) and calibrate for SOI technology. The
calibration parameters are set in the IMPACT statement.
Lattice Heating When you switch a device on, there can be significant current
density within the silicon. This could generate a significant amount
of heat. In bulk MOS devices, the silicon substrates behaves like a
good heat conductor. This generated heat is then quickly removed.
But this isn’t the case with SOI substrates as the buried oxide layer
allows this generated heat to be retained. For SOI MOSFETs, this
can be a significant amount and can drastically affect the operation
of the device. In such cases, take account of this by using the GIGA
module. Note that when you switch on lattice heating by using the
LAT.TEMP parameter in the MODELS statement, you also need to
specify a thermal boundary condition with the THERMCONTACT
statement. See Chapter 7 “Giga: Self-Heating Simulator” for more
details.
Carrier Heating In deep submicron designs, you may need to switch on the
additional energy balance equations. These take into account the
exchange of energy between carriers and between the carriers and
the lattice. See Section 4.2.1 “Simulating MOS Technologies” for
more information.

An example of a set of typical models for a partially depleted SOI

MOSFET could be:

MODEL KLA SHI FLDMOB SRH AUGER BGN

LAT.TEMP
INTERFACE QF=1e10 Y.MAX=0.05
INTERFACE QF=1e11 Y.MIN=0.05
THERMCONTACT NUM=1 ELEC.NUM=4
EXT.TEMP=300
IMPACT SELB

375 ATLAS User’s Manual

Simulating Silicon Devices Using S-Pisces S-Pisces: Silicon Based 2D Simulator

Numerical Methods for SOI

One important issue with SOI device simulation is the choice of numerical methods. In SOI
technology, the potential in the channel (or body) region is commonly referred to as
“floating”. This occurs because there exists no direct contact to it by any electrode. As a
result, when a bias is applied, or increased, on a contact there may be some convergence
problem. This occurs because the guess used in the numerical solution scheme for (, n, p)
may be poor, particularly in the “floating” region. This is particularly true if impact ionization
is used. To overcome the problem of poor initial guess, use the following numerical methods
syntax in isothermal drift-diffusion simulations.
METHOD GUMMEL NEWTON
This method initially performs a GUMMEL iteration to obtain an improved initial guess for the
NEWTON solution scheme. Although this method is more robust, this is slower than using the
NEWTON scheme alone. For more information on the numerical schemes available, see Chapter
20 “Numerical Techniques”.
SOI Physical Phenomena
The physical models and the numerical schemes described above should allow S-PISCES to
study all the important SOI phenomena. These include:
• full or partial depletion effects
• threshold voltage
• subthreshold slopes
• front to back gate coupling effects
• leakage current analysis
• high frequency analysis
• device breakdown
• snapback effects
• the kink effect in the output Ids-Vds characteristics
• negative differential resistance

376 ATLAS User’s Manual

Chapter 5
Blaze: Compound Material 2D
Simulator
Overview Blaze: Compound Material 2D Simulator

5.1 Overview
If you are unfamiliar with the general operation of the ATLAS framework, read Chapter 2
“Getting Started with ATLAS” before reading this chapter.
BLAZE is a general purpose 2D device simulator for III-V, II-VI materials, and devices with
position dependent band structure (e.g., heterojunctions). BLAZE accounts for the effects of
positionally dependent band structure by modifications to the charge transport equations and
Poisson’s equation. BLAZE is applicable to a broad range of devices such as: HBTs, HEMTs,
LEDs, lasers, heterojunction photodetectors (e.g., APDs, solar cells) and heterojunction
diodes.
This chapter is composed of several sections. Section 5.1.1 “Basic Heterojunction
Definitions” gives an overview of the basic heterojunction band parameters. Section 5.1.2
“Alignment” explains heterojunction alignment. Heterojunction charge transport is covered
in Section 5.1.3 “Temperature Dependence of Electron Affinity”. This section includes the
details of how BLAZE modifies the basic transport models to simulate heterodevices. Section
5.3 “The Physical Models” covers the BLAZE models specific to compostionally dependent
materials. Detailed information about the material systems encountered in heterojunction
simulation is covered in Section 5.4 “Material Dependent Physical Models”. This includes
the relationships between the compound elemental concentrations and bandgap, dielectric
constant, low field mobility, and other important material and transport parameters. Finally,
Section 5.5 “Simulating Heterojunction Devices with Blaze” covers how to define materials
and models for heterojunction devices with BLAZE.
See Appendix B “Material Systems” for the defaults parameters and for some of the materials
discussed in this chapter.

378 ATLAS User’s Manual

Overview Blaze: Compound Material 2D Simulator

5.1.1 Basic Heterojunction Definitions

Figure 5-1 shows the band diagrams and band parameters for a basic p-n heterojunction
device under equilibrium conditions. This diagram illustrates two materials with different
bandgaps and electron affinities and a Schottky barrier in contact to the n-type material.

E(x)
Eo

qVo El(x)
p

n m
Ec(x)
gp
Ei(x) b
Ef
Ev(x)
Metal
v
gn

x
0
Figure 5-1: Band Diagram of p-n Heterojunction
Referring to Figure 5-1:
• Ec(x), Ev(x) and Ei(x) are the spatially dependent conduction band, valence band and
intrinsic energy levels respectively.
• Ef and Eo are the Fermi and Vacuum level energies.
• Egp and Egn are the p-type and n-type material bandgaps.
• Ev is the fraction of the difference between Egp and Egn that appears in the valence band
at the heterojunction interface.
• Ec (not labelled) is the fraction of the difference between Egp and Egn that appears in the
conduction band at the heterojunction interface.
• qVo is the built-in potential of the heterojunction.
• p and n are the electron affinities of the p-type and n-type materials.
• m is the work function of the metal.
• b is the barrier height between the metal and the semiconductor.
The basic band parameters for defining heterojunctions in BLAZE are bandgap parameter,
EG300, AFFINITY, and the conduction and valence band density of states, NC300 and NV300.
To define these parameters for each material, use the MATERIAL statement. Other transport
parameters relating these basic definitions to compound elemental concentrations
(X.COMPOSITION and Y.COMPOSITION) can also be defined.

379 ATLAS User’s Manual

Overview Blaze: Compound Material 2D Simulator

See Section 5.4 “Material Dependent Physical Models” for a description of these
relationships for each material system and the Section 5.5 “Simulating Heterojunction
Devices with Blaze” for their usage.
The work function of metals is defined using the CONTACT statement.

Note: In the following, the affinites are defined by the vacuum energy minus the lowest conduction band edge
energy.

5.1.2 Alignment
As shown from Figure 5-1, the difference between the two materials bandgaps creates
conduction and valence band discontinuities. The distribution of bandgap between the
conduction and valence bands has a large impact on the charge transport in these
heterodevices. There are two methods for defining the conduction band alignment for a
heterointerface The first method is the Affinity Rule, which uses the ALIGN parameter on the
MATERIAL statement. The second method is to adjust the material affinities using the
AFFINITY parameter on the MATERIAL statement.
The Affinity Rule
This is the default method in BLAZE for assigning how much of the bandgap difference
appears as the conduction band discontinuity. The affinity rule assigns the conduction band
discontinuity equal to the difference between the two materials electron affinities (AFFINITY
on the MATERIAL statement). The affinity rule method is used by default for all materials
where the ALIGN parameter has not been defined on the MATERIAL statement.
Using the ALIGN parameter in the MATERIAL statement
Experimental measurements of band discontinuities can differ from what is assigned using
the affinity rule with the standard material electron affinities. Therefore, BLAZE allows Ec to
be calculated by specifying the ALIGN parameter on the MATERIAL statement. ALIGN specifies
the fraction of the bandgap difference, which will appear as the conduction band
discontinuity. This bandgap difference is between the material, which is specified by the
ALIGN parameter, and the smallest bandgap material in the overall structure (the reference
material). Use the ALIGN parameter to modify the electron affinity of the material and BLAZE
will create the desired conduction band offset.
In many applications, a Schottky barrier or more than two different semiconductor materials
are present. Keep the reference material bandgap and these assigned affinities in mind when
defining offsets for multiple materials or Schottky barrier heights. Examples for multiple
materials and Schottky barriers are given in the following section. In the following examples,
Eg, X, Ec, and Ev are positive.

380 ATLAS User’s Manual

Overview Blaze: Compound Material 2D Simulator

Manually Adjusting Material Affinity

You can use the AFFINITY parameter on the MATERIAL statement to adjust the conduction
band offset. The following examples describe the procedure for aligning heterojunctions
using this method in BLAZE.
EXAMPLE 1

qVo
1 El(x)

2
Ec(x)
g1
Ei(x)
Ef
Ev(x)
v
g2

Material1 Material2
Figure 5-2: Band diagram of heterojunction with band offset.
Figure 5-2 shows a heterojunction consisting of two semiconductors with different bandgaps
Eg1 and Eg2 and electron affinities 1 and 2. This example is similar to the bandstructure of a
HFET or HEMT. For this example, Eg1 < Eg2 and 2 < 1.
Allocating the conduction band offsets using the affinity rule:
E c =  1 –  2 5-1

E g = E g 2 – E g 1 5-2

and
E v = E g – E c 5-3

Ec is the amount of the conduction band discontinuity at the heterointerface. Ev is the
amount of the valence band discontinuity.

Note: The Affinity Rule is used to calculate the conduction band offset for a material if the ALIGN parameter is not
specified on the MATERIAL statement for that material.

381 ATLAS User’s Manual

Overview Blaze: Compound Material 2D Simulator

Using the ALIGN parameter on the MATERIAL statement

Let’s assign 80% of the bandgap difference between Material1 and Material2 to the
conduction band offset. Then, define the ALIGN parameter on the MATERIAL statement for
Material2 using
MATERIAL NAME=Material2 ALIGN=0.80
Then:
E c =  E g2 – E g1   0.80 5-4

Internally, the affinity of Material2 is adjusted so that Ec equals this value. This value of
electron affinity will override any electron affinity specification for Material2. This has an
impact on any calculation where this material’s electron affinity is used and considered when
specifying Schottky barriers contacted to this materials. See “EXAMPLE 4” on page 388 for
more details on Schottky barrier considerations.
Manually Adjusting Material Affinity
MATERIAL NAME=Material2 AFFINITY=VALUE
where VALUE is adjusted to provide for the desired conduction band offset as calculated by the
affinity rule, that is, relative to all other assigned affinities. This value of electron affinity will
override any prior specification of electron affinity for Material2. This has an impact on any
calculation where this material’s electron affinity is used and must be considered when
specifying Schottky barriers contacted to this materials. See “EXAMPLE 4” on page 388 for
more details on Schottky barrier considerations

Note: If you do not specify the ALIGN parameter on the MATERIAL statement, BLAZE will use the Affinity Rule
and either the default electron affinity or the affinity assigned using the AFFINITY parameter on the
MATERIAL statement to calculate the conduction band offsets.

382 ATLAS User’s Manual

Overview Blaze: Compound Material 2D Simulator

EXAMPLE 2

qVo12 qVo23
El(x) 2

1 3

g2
Ec(x)
Ef

v23
Ei(x) g3
v12
g1

Ev(x)
Material 1 Material 2 Material 3

Figure 5-3: Band diagram of three material system (lowest Eg in center)

Figure 5-3 details a heterostructure device consisting of three semiconductors with different
bandgaps g1, g2, and g3, and electron affinities 1, 2, and 3. This is similar to the band
diagram of a Double Heterojunction Bipolar Transistor. For this example, g1 > Eg2 < g3 and
1 < 2 > 3.
Allocating the conduction band offsets using the affinity rule:
E c12 =  2 –  1 5-5

E g12 = E g 1 – E g 2 5-6

and
E v12 = E g12 – E c12 5-7

for the heterojunction between Material1 and Material2 and:

E c23 =  2 –  3 5-8

E g23 = E g 2 – E g 3 5-9

and
E v23 = E g23 – E c23 5-10

383 ATLAS User’s Manual

Overview Blaze: Compound Material 2D Simulator

for the heterojunction between Material2 and Material3.

Notice the reference material (i.e., material) with the smallest bandgap. In this case,
Material2, is located between the two larger bandgap materials, Material1, and
Material3.
Let’s assign 80% of the bandgap difference between Material1 and Material2 to the
conduction band offset for this heterojunction. Then, define the ALIGN parameter on the
MATERIAL statement for Material 1 using:
MATERIAL NAME=Material1 ALIGN=0.8
then:
E c12 =  E g1 – E g2   0.80 5-11

Internally, the affinity of Material1 is adjusted so that Ec12 equals this value.
Let’s assign 70% of the bandgap difference between Material3 and Material2 to the
conduction band offset for this heterojunction. Defining the ALIGN parameter on the
MATERIAL statement for Material3 using:
MATERIAL NAME=Material3 ALIGN=0.70
then:
E c23 =  E g3 – E g2   0.70 5-12

Internally, the affinity of Material3 is adjusted so that Ec23 equals this value.
These new values of electron affinity for Material1 and Material3 will override any
electron affinity specification for these materials. This has an impact on any calculation
where these materials’s electron affinity is used and must be considered when specifying
Schottky barriers contacted to these materials. See “EXAMPLE 4” on page 388 for more
details on Schottky barrier considerations.
Manually Adjusting Material Affinity
You can assign the conduction band offsets for each heterojunction by setting the electron
affinities for Material1 and Material3 using the AFFINITY parameter on the MATERIAL
statement.
Let’s assume an electron affinity for Material2 of 4eV (~ that of GaAs). Let’s decide that
between Material1 and Material2, the conduction band offset is 0.3eV and that between
Material3 and Material2, the conduction band offset is 0.2eV. Then, for Material1:
MATERIAL NAME=Material1 AFFINITY=3.7
and for Material3:
MATERIAL NAME=Material3 AFFINITY=3.8
This is the easiest method to define the conduction band offsets for multiple materials. This
value of electron affinity will override any electron affinity specification. This has an impact
on any calculation where this material’s electron affinity is used and must be considered when
specifying Schottky barriers contacted to this materials. See “EXAMPLE 4” on page 388 for
more details on Schottky barrier considerations.

384 ATLAS User’s Manual

Overview Blaze: Compound Material 2D Simulator

Note: The band offsets are always defined with reference to the conduction band. Therefore, if a specific valence
band offset is required, the appropriate conduction band offset should be calculated from the desired
valence band offset and the materials bandgap.

EXAMPLE 3

Eo
qVo12
qVo23
1
El(x)
2
Ec(x) 3
g1
Ei(x)
Ef
Ev(x)
v12
g2
g3

v23

Material1 Material2 Material3

Figure 5-4: Band diagram of three material system (lowest Eg not in center)
Figure 5-4 details a heterostructure device consisting of three semiconductors with different
bandgaps g1, g2, and Eg3 and electron affinities 1, 2, and 3. This is similar to the
“EXAMPLE 2” on page 383, except that the narrow bandgap material is not located in
between the other larger bandgap materials. This will add extra complexity to the conduction
and valence band offset calculations. For this example, g1 < g2 < g3 and 3 < 2 < 1.
Allocating the conduction band offsets using the affinity rule:
E c12 =  1 –  2 5-13

E g12 = E g2 – E g1 5-14

and
E v12 = E g12 – E c12 5-15

for the heterojunction between Material1 and Material2 and:

E c23 =  2 –  3 5-16

385 ATLAS User’s Manual

Overview Blaze: Compound Material 2D Simulator

E g23 = E g3 – E g2 5-17

and
E v23 = E g23 – E c23 5-18

for the heterojunction between Material2 and Material3.

Using the ALIGN parameter on the MATERIAL statement
Notice that the reference material (i.e., the material with the smallest bandgap), Material1,
is not shared between the two larger bandgap materials, Material2, and Material3. This
will be important in calculating the conduction band offsets for the heterojunction formed by
Material2 and Material3.
Let’s assign 80% of the bandgap difference between Material1 and Material2 to the
conduction band offset for this heterojunction. Since the reference material is one of the
materials of this heterojunction, we can proceed as before. Define the ALIGN parameter in the
MATERIAL statement for Material2 using:
MATERIAL NAME=Material2 ALIGN=0.8
then
E c12 =  E g2 – E g1   0.80 5-19

Let’s assign 70% of the bandgap difference between Material3 and Material2 to the
conduction band offset for this heterojunction. Since the reference material is not one of the
materials in this heterojunction, another procedure will be used. Since BLAZE always uses the
bandgap of the reference material (the smallest bandgap material in overall structure) when
calculating the conduction band offset using the ALIGN parameter on the MATERIAL
statement, the actual value for the ALIGN parameter needs to be calculated as follows:
ALIGN =  E g32  E g31    E c32  E g32  5-20

Once calculated, you can use this value for the ALIGN parameter on the MATERIAL statement
for Material3. For this example, let’s assume that
E g32 = 0.2 5-21

E g = E g2 – E g1 5-22

and
E g31 = 0.4 5-23

then for a desired conduction band offset fraction of 0.70:

ALIGN =  E g32  E g31    E c32  E g32  =  0.2  0.4   0.70 = 0.35 5-24

386 ATLAS User’s Manual

Overview Blaze: Compound Material 2D Simulator

You can assign the proper value of the ALIGN parameter reflecting the desired conduction
band offset as:
MATERIAL NAME=Material3 ALIGN=0.35
This assigns 70% of the bandgap difference between Material3 and Material2 as the
conduction band offset.

Note: Calculating ALIGN in this manner is only necessary when the reference material is not in contact with the
material where the ALIGN parameter will be specified.

These new values of electron affinity for Material2 (from the first heterojunction band
offset calculation) and Material3 (from the second heterojunction band offset calculation)
will override any electron affinity specification for these materials. This has an impact on any
calculation where these materials’s electron affinity is used and must be considered when
specifying Schottky barriers contacted to these materials. See “EXAMPLE 4” on page 388
for more details on Schottky barrier considerations.
Manually Adjusting Material Affinity
You can assign the conduction band offsets for each heterojunction by setting the electron
affinities for Material2 and Material3 using the AFFINITY parameter on the MATERIAL
statement. The electron affinity for Material2 is adjusted relative to Material1 and
Material3 is adjusted relative to Material2 by the amount of the desired conduction band
offset for each heterojunction. Since Material1 affinity is larger than that for Material2
and Material2 affinity is larger than that for Material3, the affinities for Material2 and
Material3 are reduced to provide the desired conduction band offsets.
Let’s assume an electron affinity for Material1 of 4eV (~ that of GaAs). Let’s decide that
between Material1 and Material2, the conduction band offset is 0.3eV and that between
Material2 and Material 3, the conduction band offset is 0.2eV. Then, for Material2:
MATERIAL NAME=Material2 AFFINITY=3.7
and for Material3:
MATERIAL NAME=Material3 AFFINITY=3.5
This is the easiest method to define the conduction band offsets for multiple materials. This
has an impact on any calculation where this materials electron affinity is used and must be
considered when specifying Schottky barriers contacted to this materials. See “EXAMPLE 4”
on page 388 for more details on Schottky barrier considerations.

387 ATLAS User’s Manual

Overview Blaze: Compound Material 2D Simulator

EXAMPLE 4

E(x)
Eo

qVo El(x)
1

2 m
Ec(x)
g1
Ei(x) b
Ef
Ev(x)
Metal
v
g2

x
0

Figure 5-5: Schematic band diagram for an abrupt heterojunction

Figure 5-5 details a heterostructure device consisting of two semiconductors with different
bandgaps g1 and Eg2 and electron affinities 1 and 2 and a Schottky barrier. For this
example, g1 < g2 and 2 < 1.
This example will first define the heterojunction band offsets and then the Schottky barrier
height. Schottky contact barrier heights are calculated by BLAZE using the metal work
function and the semiconductor electron affinity as:
b = m – s 5-25

where b is the Schottky barrier height, m is the work function of the metal, and s is the
semiconductor electron affinity. m is set using the WORKFUN parameter in the CONTACT
statement. Therefore, the semiconductor electron affinity as modified or defined during the
heterojunction alignment process plays an important role in determining the value of the
metal workfunction needed to provide the desired barrier height. Let’s assume for this
example that a Schottky barrier height of 0.2eV is desired and calculate the appropriate metal
workfunction for each case.
Using the Affinity rule for the heterojunction
E c =  1 –  2 5-26

and
E v = E g – E c 5-27

388 ATLAS User’s Manual

Overview Blaze: Compound Material 2D Simulator

Ec is the amount of the conduction band discontinuity at the heterointerface. Ev is the
amount of the valence band discontinuity.

Note: The Affinity Rule is used for a material if the ALIGN parameter is not specified on the MATERIAL
statement for that material.

Let’s use an electron affinity for Material1 of 4eV and for Material2 of 3.5eV. Since the
affinity of the material on which the Schottky barrier is formed was not modified with this
method of alignment, the metal work function needed to provide for a Schottky barrier height
of 0.2eV is:
 m = 3.5 + 0.2 = 3.7 5-28

You can now assign this value to the WORKFUN parameter on the CONTACT statement as:
CONTACT NUMBER=1 WORKFUN=3.7
This produces a Schottky barrier height of 0.2eV between the metal and Material 2 using the
ALIGN parameter on the MATERIAL statement:
Let’s assign 80% of the bandgap difference between Material1 and Material2 to the
conduction band offset, an electron affinity for Material1 of 4eV, and g of 0.2eV. Then,
define the ALIGN parameter on the MATERIAL statement for Material2 using:
MATERIAL NAME=Material2 ALIGN=0.80
Then:
E c =  E g2 – E g1   0.80 = 0.2  0.80 = 0.16 5-29

Internally, the affinity of Material2 is reduced by 0.16eV so:

 2 =  1 – E c =  4 – 0.16  = 3.84 5-30

You can use this value of electron affinity to assign the proper value of WORKFUN on the
CONTACT statement to provide for a Schottky barrier height of 0.2eV.

 m = 3.84 + 0.2 = 4.04 5-31

You can now assign this value to the WORKFUN parameter on the CONTACT statement as:
CONTACT NUMBER=1 WORKFUN=4.04
producing a Schottky barrier height of 0.2eV between the metal and Material2.

389 ATLAS User’s Manual

Overview Blaze: Compound Material 2D Simulator

Manually Adjusting Material Affinity

Let’s assign a conduction band offset between Material1 and Material2 of 0.15eV, an
electron affinity for Material1 of 4eV, and the desired Schottky barrier height of 0.2eV. The
affinity for Material2 is calculated from the affinity of Material1 and the desired
conduction band offset as:
 2 =  1 – 0.15 = 4 – 0.15 = 3.85 5-32

This is then used to assign the value of AFFINITY using:

MATERIAL NAME=Material2 AFFINITY=3.85
You can now use this value of electron affinity to calculate the metal workfunction necessary
to produce a Schottky barrier height of 0.2eV as:
 m = 3.85 + 0.2 = 4.05 5-33

You can now assign this value to the WORKFUN parameter on the CONTACT statement as:
CONTACT NUMBER=1 WORKFUN=4.05
This produces a Schottky barrier height of 0.2eV between the metal and Material2.

5.1.3 Temperature Dependence of Electron Affinity

The electron affinity of a semiconductor is generally a function of lattice temperature. This
temperature dependence is usually modelled implicitly in ATLAS by making it a fraction of
the temperature dependence of the bandgap. The default value of the fraction is 0.5 but you
can assign it a value in the range [0,1] using the CHI.EG.TDEP.
If you specify the ALIGN parameter, then CHI.EG.TDEP will be ignored. In that case, the
affinity is adjusted to keep the conduction band offset at the specified ratio of the bandgap
differences at the heterojunction at the particular temperature required.
Example 1
For a material where the affinity is given only at 300 K (e.g., GaAs) the electron affinity at
temperature T would be

  T  =   300  – CHI · EG · TDEP   Eg  T L  – Eg  300   5-34

Example 2
For HgCdTe, the affinity depends on bandgap and consequently temperature (see Equation 5-
164). In this case, you must include the following statement.
MATERIAL MATERIAL=HGCDTE CHI.EG.TDEP=0

390 ATLAS User’s Manual

Overview Blaze: Compound Material 2D Simulator

Example 3
If you are using F.BANDCOMP to specify the electron affinity and your function includes its
full temperature dependence, then specify CHI.EG.TDEP = 0 on the MATERIAL statement.
If your function excludes temperature dependence of affinity, then you can use the value of
CHI.EG.TDEP to adjust this.

Note: The ASYMMETRY parameter on the MATERIAL statement will be used to assign a fraction of bandgap
narrowing to the affinity. See Section 3.2.9 “Bandgap Narrowing” for more information.

391 ATLAS User’s Manual

Transport Models Blaze: Compound Material 2D Simulator

5.2 Transport Models

5.2.1 The Drift Diffusion Transport Model
Drift-Diffusion with Position Dependent Band Structure
The current continuity equations for electrons and holes, and the Poisson Equation (see
Equation 3-1) are the same as for the homogeneous case. Although the changing dielectric
constant is taken into account. The current density expressions, however, must be modified to
take into account the nonuniform band structure [169]. This procedure starts with the current
density expressions:

J n = –  n n n 5-35

J p = –  p n p 5-36

where n and p are quasi-Fermi potentials.

1
 n = --- E FN 5-37
q

1
 p = --- E FP 5-38
q

The conduction and valence band edge energies can be written as:
Ec = q  0 –   – 5-39

Ev = q  0 –   – –Eg 5-40
where:
• 0 is some reference potential.
•  is the position-dependent electron affinity.
• Eg is the position-dependent bandgap.
• 0 can be selected in the form:

 kT L N cr  r + E g kT L N vr
 0 = -----r + --------- ln -------
- = ------------------ – --------- ln -------- 5-41
q q n ir q q n ir

where nir is the intrinsic carrier concentration of the arbitrarily selected reference material,
and r is the index that indicates all of the parameters are taken from reference material.
Fermi energies are expressed in the form:
n- – kT ln n
E FN = E c + kT L ln ----- 5-42
Nc L

392 ATLAS User’s Manual

Transport Models Blaze: Compound Material 2D Simulator

n
E FP = E v + kT L ln ------ – kT L ln n 5-43
N v

The final terms in Equations 5-42 and 5-43 are due to the influence of Fermi-Dirac statistics.
These final terms are defined as follows:
F1  2   n  E FN – E c
- = F 1  2  ------
–1 n
 n = --------------------------  n = --------------------- 5-44
n kT L  N c
e

F1  2   p  E v – E FP
- = F 1  2  ------
–1 p
 p = --------------------------  p = --------------------- 5-45
p kT L Nv
e

where Nc and Nv are position-dependent and n=p= 1 for Boltzmann statistics.

By combining Equations 5-35 to 5-45, you can obtain the following expressions for current
densities.

 kT L  kT L N 
J n = kT L  n n – q n n   + --------- ln  n + --- + --------- ln -----c- 5-46
 q q q n ir

 kT L +E kT L N 
J p = kT L  p n – q p p   + --------- ln  p + ---------------g + --------- ln -----v- 5-47
 q q q n ir

5.2.2 The Thermionic Emission and Field Emission Transport Model

You can activate alternative current density expressions for electron and hole current
[313, 319], which take into account thermionic emission dominated current in abrupt
heterojunctions. These equation applies only at the node points along the interface of the
heterojunction and take the form:
– E C
J n = q n  1 +    n – n exp  -------------- 
+ –
5-48
  kT   L

– E V
J p =  – q  p  1 +    p – p exp  -------------- 
+ –
5-49
  kT L  

where Jn and Jp are the electron and hole current densities from the "-" region to the "+"
region. n and p are the electron and hole thermal velocities. Ec is conduction band energy
change going from the “-” region to the “+” region. Ev is the valence band energy change
going from the “-” region to the “+” region. The  parameter represents the contribution due
to thermionic field emission (tunneling).
The thermal velocities n andp are given by:

* 2
An TL
 n = --------------
- 5-50
qN C

393 ATLAS User’s Manual

Transport Models Blaze: Compound Material 2D Simulator

* 2
Ap TL
 p = --------------
- 5-51
qN V

where TL is the lattice temperature, NC is the conduction band density of states, NV is the
valence band density of states, and A *n and A*p are the electron and hole Richardson constants.

The minimum valued Richardson constants from the “-” region or “+” region are used for the
calculation of the thermal velocities [319]. You can specify the Richardson constants with the
ARICHN and ARICHP parameters of the MATERIAL statement. If the Richardson constants
aren’t specified, the following expressions will be used:
2 *
* 4qk m
An = ----------------------n- 5-52
3
h

2 *
* 4qk m
Ap = ----------------------p- 5-53
3
h

where mn* and mp* are the electron and hole effective masses. The electron and hole effective
masses can be specified with the M.VTHN and M.VTHP parameters of the MATERIAL statement.
If the effective masses aren’t specified, then they will be calculated from the conduction and
valence band densities of states using Equations 3-31 and 3-32. To enable the thermionic
transport model, you must specify THERMIONIC on the INTERFACE statement. You must also
set the S.S parameter on the INTERFACE statement to indicate that the model applies to
semiconductor-semiconductor interfaces.
The tunneling factor  in Equation 5-48 is zero when the tunneling mechanism is neglected.
When you account for tunneling by specifying the TUNNEL parameter in the INTERFACE
statement, the tunneling factor, ,is calculated by using the expression:
+
EC XE
+
 E C – E x  0.5

exp  ------------------- exp  ----------  2m n  E C – E x   dx dE x
 
1 – 4 *
 = ------ 5-54
kT  kT   h 
E min
 0


where Ex is the energy component in the X direction and Emin = max[Ec(0-),Ec(W)] as

described in Figure 5-6.

394 ATLAS User’s Manual

Transport Models Blaze: Compound Material 2D Simulator

Figure 5-6: Band Diagram of p-n heterojunction

The band edge is approximated as a linear function of position in the wide-band gap material.
The slope corresponds to the electric field at the interface.

5.2.3 Non-local Heterojunction Tunneling Model

A weakness of the TUNNEL model described above is that it is a local model. Specifically, it
calculates the tunneling current based on the value of the field at the interface and the current
is injected entirely at the interface. Tunneling is possible over a range of energies. And
assuming that the process is elastic, it will give rise to carrier generation or recombination
throughout much of the depletion region. To calculate this tunneling current properly, you
need to take into account the non-linear shape of the potential barrier as shown in Figure 5-6.
ATLAS has a non-local tunneling model that calculates the tunneling path and the
contribution to the tunneling current for each energy at which elastic tunnelling is possible. In
order to use this model, specify one or more QTREGIONs covering the junction. The tunneling
paths of the QTREGION should begin on the narrow bandgap side of the heterojunction and
end at the boundary of the depletion region in the wide bandgap material. The syntax for the
QTREGION statement is described in Section 3.6.6 “Band-to-Band Tunneling”. Tunneling
current is evaluated for all energies at which tunneling is possible, up to the maximum of the
band energy at the interface. The current in a range Ex to Ex+dE is given by Equation 3-509,
where T(E) is evaluated using the WKB approximation. It is then converted into a
recombination or generation rate and inserted into the current continuity equation at each end
of its tunnelling path as shown in Figure 5-6.
To enable this model you specify SHJ.EL or SHJ.HO or both on the MODELS statement.
SHJ.EL is for electron tunneling and SHJ.HO is for hole tunnelling. Convergence is generally
improved by including non-local derivatives in the jacobian. Specify SHJ.NLDERIVS on the
MODELS statement to do this. The recombination/generation rates are automatically output
with any Standard Structure File when one or both of SHJ.EL and SHJ.HO are set.
You can specify the QTREGION parameter on the MODELS statement in order to restrict the
application of the model to the specified QTREGION. The default behavior is to apply the
model to all defined QTREGIONs.

395 ATLAS User’s Manual

Transport Models Blaze: Compound Material 2D Simulator

Table 5-1 MODEL Statement Parameters

Parameter Type Default Units

SHJ.EL Logical False

SHJ.HO Logical False
SHJ.NLDERIVS Logical False

This model is only suitable for calculating tunneling through single heterojunctions. To
calculate tunneling through single or multiple quantum barriers, you need the SIS.EL or
SIS.HO models. This is described below.

5.2.4 Non-local Quantum Barrier Tunneling Model

You can calculate the tunneling current between two semiconducting regions separated by a
quantum barrier using the SIS.EL and SIS.HO models. The barrier layer may be either an
insulating layer or a wide bandgap semiconductor. The model assumes that the charge
tunnels across the whole barrier with the source/sink at the interface with the semiconductor
regions. Tunneling current is evaluated for all energies at which tunneling is possible, up to
the maximum of the band energy at the interface. The current in a range Ex to Ex+dE is given
by Equation 3-509. The transmission probability, T(E), is evaluated at each energy from a
solution of the one-dimensional Schrodinger equation. The default is to use a transmission
matrix method, identical to that used in the QTUNN model of gate oxide tunnelling. To use the
alternative Wentzel-Kramers-Brillouin (WKB) approximation to solve the Schrodinger
equation, specify TUN.WKB on the METHOD statement. The TUN.WKB option will omit some
quantum interference effects, but will result in a much quicker calculation. The tunneling
current is then converted into a recombination or generation rate and inserted into the current
continuity equation at each end of its tunneling path.
In order to use this model, you need to define one or more quantum tunneling regions by
using the QTREGION statement. The QTREGIONs should define tunneling paths which start and
end in the semiconductor regions on opposite sides of the barrier. There is no requirement for
the QTREGIONs to start and end in flat-band regions as is the case for the BBT.NONLOCAL
model. It is possible to simulate coherent tunneling through multiple quantum barriers by
defining a QTREGION which covers all the barriers. You invoke the non-local tunneling by
specifying SIS.EL for electron tunneling, and SIS.HO for hole tunneling, on the MODELS
statement. In case of poor convergence, you set the flag SIS.NLDERIVS on the MODELS
statement. This puts in non-local couplings into the system matrix.
You can also model sequential tunneling by defining a single QTREGION for each barrier. For
structures with a large number of barriers, this can be an onerous task. A more convenient
format for setting up the QTREGION is to specify the QTREGION parameter on each REGION
statement defining a quantum barrier. The QTREGION formed is optimized in terms of system
matrix size because it avoids the use of interpolation. It also injects the tunneling current at
the exact position of the interface and into the lower carrier energy side of the interface.
You set QTREGION=x to set up that REGION as a QTREGION number x. There are some
restrictions with this format. First, it will only work for structures with rectangular meshing.
Second, the interfaces at the boundaries of the QTREGION must be of a special type. You

396 ATLAS User’s Manual

Transport Models Blaze: Compound Material 2D Simulator

enable this by specifying INTERFACE THERMIONIC S.S for semiconducting barriers or

INTERFACE S.I SITHERM for insulating barriers. If the barrier material has been reassigned
as a semiconductor by using the SEMICONDUCTOR on the MATERIAL statement, you specify
INTERFACE S.I SITHERM. You may additionally need to specify localization parameters.
And finally, each REGION statement must use a unique value of the QTREGION number (x) and
the maximum value for x is the total number of QTREGIONs required to be set up using this
method. It is advised to sequentially increment the QTREGION numbers by one.
This format is currently restricted to SIS.EL and SIS.HO models.
You may use a combination of REGION statements and QTREGION statements to set up several
QTREGIONs. The REGION statements must occur before the QTREGION statements, and the
NUMBER parameter of the first QTREGION statement must be given a value of one more than
the number of QTREGIONs set up using the REGION statements.
To restrict the SIS.EL or SIS.HO model to a specific QTREGION specify the region through
the QTREGION parameter on the MODELS statement. The default behavior is to apply the
model to all defined QTREGIONs.

Table 5-2 MODEL Statement Parameters

Parameter Type Default Units

SIS.EL Logical False

SIS.HO Logical False
SIS.NLDERIVS Logical False

Table 5-3 REGION Statement Parameters

Parameter Type Default Units

QTREGION Integer

Example
MODELS SIS.EL QTREGION=3 SIS.NLDERIVS
MODELS SHJ.HO QTREGION=2 SHJ.NLDERIVS
This applies the SIS model for electrons to QTREGION number 3 and the Single
Heterojunction Tunneling model for holes to QTREGION 2. In both cases, non-local couplings
and derivatives are included to improve convergence.

397 ATLAS User’s Manual

Transport Models Blaze: Compound Material 2D Simulator

5.2.5 Quantum Barrier Trap Assisted Tunneling Model

Basic Model
The SIS.EL and SIS.HO models do not consider the possibility of indirect tunneling through
the trap levels. The SIS.TAT model does include this, and has two levels of sophistication.
The simplest model is the extension of the TAT.NONLOCAL model described in Section 3.3.3
“Traps and Defects”. This can be viewed as being a nonlocal SRH recombination process
through the barrier. The trap depth is specified by the TAT.NLDEPTH parameter on the
MODELS statement. It is recommended to use the TAT.RELEI flag on the MODELS statement in
order to make the trap depth relative to the Intrinsic Fermi level. The other two parameters are
the basic recombination lifetimes, SISTAT.TN and SISTAT.TP. You can enable this model for
some or all QTREGIONs by specifying SIS.TAT on the MODELS statement. You can specify the
parameter SIS.NLDERIVS on the MODELS statement to improve the convergence properties by
including non-local Jacobian couplings.

Table 5-4 Parameters for the basic SIS.TAT model.

Statement Parameter Type Default Units

MODELS SIS.TAT Logical False

MODELS SISTAT.TN Real 1.0e-6 s
MODELS SISTAT.TP Real 1.0e-6 s
MODELS TAT.NLDEPTH Real 0.0 eV
MODELS TAT.RELEI Logical False

Advanced Model
The Ielmini trap assisted tunneling model has been described in Section 3.6.7 “Gate Current
Models”, where it is applied to Metal-Insulator-Semiconductor structures, and it has also been
applied to Metal-Insulator-Metal structures in the subsequent section. You can use it for
Quantum barrier structures by specifying ITAT.SC.EL, ITAT.SC.HO, or RTAT.SC on the
MODELS statement along with SIS.TAT. The tunneling current through the barrier is coupled
self-consistently with the current continuity equations on either side. The post-processing
option does not exist, so the parameters ITAT.PP.EL, ITAT.PP.HO, RTAT.PP enable the
equivalent self-consistent modes instead. The barrier material can be either an insulator, a
semiconductor, or an insulator changed to a wide bandgap semiconductor using the
SEMICONDUCTOR parameter on the MODELS statement.
Which model is most appropriate depends on the energy of the trap with respect to the
bandgap of the semiconductors on either side of the barrier. If the energy is above the
conduction band then the ITAT.SC.EL model will model the electron transfer. If the energy
is below the valence band, then the ITAT.SC.HO model will model the hole transfer.
Otherwise, the RTAT.SC option will model transfer of both carriers, including recombination
at the trap centers. The model parameters are exactly the same as in the Ielmini models for
MIS and MIM structures, and the traps can be set up using the TRAPS, DOPING, or DEFECTS
statements. The model can be applied to selected or all QTREGIONs using the QTREGION
parameter on the MODELS statement.

398 ATLAS User’s Manual

Transport Models Blaze: Compound Material 2D Simulator

Table 5-5 Parameters for the Basic SIS.TAT Model

Statement Parameter Type Default

MODELS SIS.TAT Logical False

MODELS ITAT.SC.EL Logical False
MODELS ITAT.SC.HO Logical False
MODELS RTAT.SC Logical False

399 ATLAS User’s Manual

The Physical Models Blaze: Compound Material 2D Simulator

5.3 The Physical Models

Section 3.6 “Physical Models” already describes a comprehensive set of physical models for
silicon. It is known, however, that for III-V, II-VI, and ternary compounds that special
consideration are required such as for mole fraction dependence and material properties. The
following sections describe the material dependent physical models that have been
implemented into BLAZE to account for these effects.
First, a description of the common mobility model equations is given. These equations are
applicable to all materials unless stated in the following material sections. Following this are
sections describing the physical models on a material-per-material basis. For each material,
there are descriptions of the models for bandgap narrowing, electron affinity, density of
states, dielectric permittivity, and low field mobility.

5.3.1 Common Physical Models

Low Field Mobility Models
The default low field mobility models used for most materials in BLAZE are given by the
following:
T L – T MUN
 n0  T L  = MUN  ---------- 5-55
 300

T L – T MUP
 p0  T L  = MUP  ---------- 5-56
 300

where TL is the temperature in degrees Kelvin and the MUN, MUP, TMUN, and TMUP parameters
are user-definable as shown in Table 5-6.

Note: All the mobility models described in Chapter 3 “Physics”, except for the TASCH model, can be used in
BLAZE. But all default coefficients exist only for Silicon and therefore are not suitable for compound
materials.

Table 5-6 User-Specifiable Parameters for Equations 5-55 and 5-56

Statement Parameter Units

MOBILITY MUN cm2/V·s

MOBILITY TMUN

MOBILITY MUP cm2/V·s

MOBILITY TMUP

400 ATLAS User’s Manual

The Physical Models Blaze: Compound Material 2D Simulator

Parallel Electric Field-Dependent Mobility Models

There are two types of electric field-dependent mobility models used in ATLAS/BLAZE.
These models are called Standard Mobility Model and Negative Differential Mobility Model.
Both of these models contain appropriate default values of parameters for different materials.
You need to specify which type of mobility will be used for each material and which material
parameters you want to alter.
The Standard Mobility Model that takes account of velocity saturation is defined according
to:
1  BETAN
1
 n  E  =  n0 ---------------------------------------------------- 5-57
 n0 E BETAN
1 +  ------------------
 VSATN

1  BETAP
1
 p  E  =  p0 ---------------------------------------------------- 5-58
 p0 E BETAP
1 +  ------------------
 VSATP

where VSATN and VSATP are the saturation velocities for electrons and holes, BETAN and
BETAP are constants given in Table 5-7, and n0,p0 are the electron and hole low field
mobilities. This model is activated by specifying the FLDMOB and the EVSATMOD=0 parameter
in the MODEL statement.

Table 5-7 User-Specifiable Parameters for Equations 5-57 and 5-58

Statement Parameter Units

MOBILITY BETAN

MOBILITY BETAP

MOBILITY VSATN cm/s

MOBILITY VSATP cm/s

The Negative Differential Mobility Model of Barnes et. al. [21] has been implemented to
account for certain devices where the carrier drift velocity peaks at some electric field before
reducing as the electric field increases. This model takes account of this through the carrier
mobility with equations of the form:
GAMMAN
 n0 + ------------------  ----------------------
VSATN E
E  ECRITN
 n  E  = ------------------------------------------------------------------------------------- 5-59
GAMMAN
1 +  ----------------------
E
 ECRITN

401 ATLAS User’s Manual

The Physical Models Blaze: Compound Material 2D Simulator

GAMMAP
 p0 + ------------------  ----------------------
VSATP E
E  ECRITP 
 p  E  = ------------------------------------------------------------------------------------- 5-60
GAMMAP
1 +  ----------------------
E
 ECRITP

where VSATN and VSATP are the electron and hole saturation velocities, E0 is a constant, and
n0,p0 are the low-field electron and hole mobilities. To activate this mobility model, specify
EVSATMOD=1 in the MODEL statement.

Table 5-8 User-Specifiable Parameters for Equations 5-59 and 5-60

Statement Parameter Default Units

MOBILITY ECRITN 4.0103 V/cm

MOBILITY ECRITP 4.0103 V/cm

MOBILITY GAMMAN 4.0

MOBILITY GAMMAP 1.0

Note: The Negative Differential Mobility Model introduces an instability in the solution process and is not
recommended for general use. Only activate this model in cases where the device operation directly
depends on negative differential mobility (e.g., a Gunn diode).

For both the standard and negative differential models, an empirical temperature-dependent
model for saturation velocity in GaAs [165] is implemented according to:

VSATN = VSATN · T0 – VSATN · T1 T L 5-61

VSATP = VSATP · T0 – VSATP · T1 T L 5-62

where VSATN and VSATP are expressed in cm/sec and TL is the temperature in degrees Kelvin.
Alternatively, you can set the saturation velocities to constant values using the VSATN and
VSATP parameters of the MATERIAL statement.

402 ATLAS User’s Manual

The Physical Models Blaze: Compound Material 2D Simulator

Table 5-9 User-Specifiable Parameters for Equations 5-61 and 5-62

Statement Parameter Default Units

MOBILITY VSATN.T0 1.13107 cm/s

MOBILITY VSATN.T1 1.2106 cm/s

MOBILITY VSATP.T0 1.13107 cm/s

MOBILITY VSATP.T1 1.2104 cm/sK

Velocity Saturation with Energy Balance Transport Model

When the Energy Balance Transport Model is activated, the mobility can be made a function
of carrier energy. In Section 3.6.1 “Mobility Modeling”, physical models for the dependence
of carrier mobility on carrier energy were introduced. The same models are applicable for use
within BLAZE with one additional model, which applies when the negative differential
mobility model is used.
The carrier temperature dependence is activated when the EVSATMOD=1 parameter is on the
MODEL statement. This model can be derived in a similar fashion as in the case of
EVSATMOD=0 described in “Parallel Electric Field-Dependent Mobility” on page 195. These
expressions, however, require several piecewise approximations, which are too in-depth to
show in this manual. To say that these piecewise expressions provide a continuous velocity
saturation model for mobility versus carrier temperature are completely like the expressions
for drift diffusion given in Equations 3-324 and 3-325.

5.3.2 Recombination and Generation Models

The recombination and generation models for compound semiconductors are the same as the
models previously described in Section 3.6.3 “Carrier Generation-Recombination Models”.
The default parameters for different materials are automatically used, unless new values are
specified. The default parameter values are listed in Appendix B “Material Systems”.

403 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

5.4 Material Dependent Physical Models

5.4.1 Cubic III-V Semiconductors
The basic band parameters (Eg, , , Nc, and Nv) of most III-V cubic can be modeled using
composition dependent interpolation schemes from the parameters of the constituent binary
semiconductors. Table 5-10 lists the compound semiconductors you can model using this
approach. The first column of this table gives the material name used on the REGION, DOPING,
ELECTRODE, PRINT, MODELS, MATERIAL, INTERFACE, IMPACT, MOBILITY, TRAP, DEFECT,
PROBE, LASER, and ODEFECTS statements. The second column of Table 5-10 shows the
association of composition fractions with the constituent elements. The third column
indicates whether this model is used by default (see Section 3.2.5 “Rules for Evaluation of
Energy Bandgap”). If this model is not used, then there is another default model for that
material described in the section listed in the last column of the table.
You can apply the Cubic III-V model to materials, which use an alternative model described
in the Section column of Table 5-10, by setting the CUBIC35 parameter of the MODELS
statement.
In specifying a material name, ATLAS will recognize alternate spellings switching the order
of cations or anions as long as you remember to maintain the associations of elements with
composition fractions as given in Table 5-10 (e.g., Al(x)Ga(1-x)As = Ga(1-x)Al(x)As).

Table 5-10 Cubic III-V Interpolation Model Applicability Chart

ATLAS Name Composition Default Section

GaAs No 5.3.2
AlP Yes
AlAs No 5.3.3
AlSb Yes
GaSb Yes
GaP No 5.3.4
InP No 5.3.4
InSb Yes
InAs No 5.3.4
AlGaAs Al(x)Ga(1-x)As No 5.3.3
InAlP In(1-x)Al(x)P Yes
InGaAs In(1-x)Ga(x)As No 5.3.4
InGaP In(1-x)Ga(x)P No 5.3.4
GaSbP GaSb(1-y)P(y) Yes

404 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Table 5-10 Cubic III-V Interpolation Model Applicability Chart

InAlAs In(1-x)Al(x)As Yes
InAsP InAs(1-y)P(y) No 5.3.4
GaAsP GaAs(1-y)P(y) No 5.3.4
InGaSb In(1-x)Ga(x)Sb Yes
InAlSb In(1-x)Al(x)Sb Yes
AlGaSb Al(x)Ga(1-x)Sb Yes
InAsSb InAs(y)Sb(1-y) Yes
GaAsSb GaAs(y)Sb(1-y) Yes
AlAsSb AlAs(y)Sb(1-y) Yes
InPSb InSb(1-y)P(y) Yes
AlPSb AlP(y)Sb(1-y) Yes
AlPAs AlP(y)As(1-y) Yes
AlGaP Al(x)Ga(1-x)P Yes
InPAsSb InP(x)As(y)Sb(1-x-y) Yes
InGaAsP In(1-x)Ga(x)As(1-y)P(y) No 5.3.4
AlGaAsP Al(x)Ga(1-x)As(1-y)P(y) Yes
AlGaAsSb Al(x)Ga(1-x)As(y)Sb(1-y) Yes
InAlGaAs In(1-x-y)Al(x)Ga(y)As Yes
InAlGaP In(1-x-y)Al(x)Ga(y)P Yes
InAlAsP In(1-x)Al(x)As(1-y)P(y) Yes
InGaAsSb In(1-x)Ga(x)As(y)Sb(1-y) Yes
InAlAsSb In(1-x)Al(x)As(y)Sb(1-y) Yes

Note: Materials in the InGaAsP system by default use the models in Section 5.4.3 “Al(x)Ga(1-x)As System”. Make
sure that the usage of the y composition fraction is different in the two models.

405 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Energy Bandgap
The energy bandgaps of the binary compounds are calculated for transitions in each of the ,
 and L directions using the universal formula for temperature dependent bandgap (see
Equation 3-38). Table 5-7 shows the default parameters for Equation 3-38 for the binary
compounds.

Table 5-11 Default Bandgap Parameters for Cubic III-V Binary Compounds [216]
Binary Eg(0)   Eg (0)   EgL(0) L L
(eV) (meV/K) (K) (eV) (meV/K) (K) (eV) (meV/K) (K)
GaAs 1.519 0.5405 204.0 1.981 0.4600 204.0 1.815 0.6050 204

AlP 3.63 0.5771 372.0 2.52 0.3180 588.0 3.57 0.3180 588

AlAs 3.099 0.885 530.0 2.240 0.700 530.0 2.460 0.605 240

AlSb 2.386 0.42 140.0 1.696 0.39 140.0 2.329 0.58 140

GaSb 0.812 0.417 140.0 1.141 0.475 94.00 0.875 0.597 140

GaP 2.886 0.0 0.0 2.350 0.5771 372.0 2.720 0.5771 372

InP 1.4236 0.363 162.0 2.3840 0.0 0.0 2.0140 0.363 162

InSb 0.235 0.32 170.0 0.630 0.00 0.0 0.930 0.00 0.0

InAs 0.417 0.276 93.0 1.433 0.276 93.0 1.133 0.276 93

For the ternary materials the bandgap in each principal direction is approximated by the
function [2]:
Tabc  x  = xTac +  1 – x Tbc – x  1 – x Cabc 5-63

where Tabc is the ternary bandgap, Tac and Tbc are the binary bandgaps, x is the composition
fraction and Cabc is the bowing factor as given for each of the ternary compounds in each of
the principal directions in Table 5-12.

Table 5-12 Bowing Factors and Alignment for Cubic III-V Ternary Compounds [216]
Ternary C(Eg) C(Eg) C(EgL) dEc/dEg
(eV) (eV) (eV)

AlGaAs 1.31*x-0.127 0.055 0.0 0.65

InAlP -0.48 0.38 0.0 0.5
InGaAs 0.477 1.4 0.33 0.4
InGaP 0.65 0.2 1.3 0.4
GaSbP 2.558 2.7 2.7 0.5
GaSbAs 1.43 1.2 1.2 0.5

406 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Table 5-12 Bowing Factors and Alignment for Cubic III-V Ternary Compounds [216]
InAlAs 0.7 0.0 0.0 0.7
InAsP 0.1 0.27 0.27 0.4
GaAsP 0.19 0.24 0.16 0.4
InGaSb 0.415 0.330 0.4 0.5
InAlSb 0.43 0.0 0.0 0.5
AlGaSb 1.22*x-0.044 0.0 0.0 0.5
InAsSb 0.67 0.6 0.6 0.5
GaAsSb 1.43 1.2 1.2 0.5
AlAsSb 0.84 0.28 0.28 0.5
InPSb 1.9 1.9 1.9 0.5
AlPSb 3.56 2.7 2.7 0.5
AlPAs 0.22 0.22 0.22 0.5
AlGaP 0.0 0.13 0.0 0.5

For the quaternary compounds, the bandgap in each principal direction is approximated by
the function [2]:
x  1 – x  yTabc  x  +  1 – y Tabd  x   + y  1 – y   xTacd  y  +  1 – x Tbcd  y   
Qabcd  x y  = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 5-64
x1 – x + y1 – y

or for compounds with three anions [2]:

xyTabc  u  + y  1 – x – y Tacd  v  +  1 – x – y xTabd  w 
Qabcd  x y  = --------------------------------------------------------------------------------------------------------------------------------------------- 5-65
xy + y  1 – x – y  +  1 – x – y x

where Tabc, Tabd, Tacd, and Tbcd are the ternary bandgaps, x and y are the composition
fractions and u, v and w are given as follows:
u = 1 – x + y  2 5-66

v =  2 – x – 2y   2 5-67
and
w =  2 – 2x – y   2 5-68

For any case (binary, ternary or quaternary), the bandbap used in the drift-diffusion
calculations is taken as the minimum of Eg, Eg or EgL.

407 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Electron Affinity
Table 5-13 shows the electron affinities for the binary compounds in the given direction in k
space.

Table 5-13 Default Electron Affinities and Alignments for Cubic III-V Binary Compounds [216]
Binary Affinity Direction dEc/dEg
(eV)

GaAs 4.07  0.65

AlP 3.98 X 0.7
AlAs 3.85 X 0.65
AlSb 3.60 X 0.5
GaSb 4.06  0.5
GaP 4.0 X 0.4
InP 4.4  0.4
InSb 4.59  0.5
InAs 4.90  0.65

The affinities for the other directions in k space not given in Table 5-13 are calculated by the
following:
 X X 
 =  –  Eg – Eg   dEc  dEg  5-69

   X
 =  –  Eg – Eg   dEc  dEg  5-70

L X X L
 =  –  Eg – Eg   dEc  dEg  5-71
  L   L
where  ,  and  are the affinities in each of the principal directions, Eg , Eg and Eg
are the energy bandgaps in each of the principal directions as calculated above and dEc/dEg is
an alignment parameter describing the proportion of change in bandgap that can be associated
with the conduction band edge. Table 5-13 lists the default values of dEc/dEg. You can
define the value of dEc/dEg by using the DEC.DEG parameter of the MATERIAL statement.
For ternary compounds, the affinities are calculated from the binary values using Equation 5-
63 except Tabc is the ternary affinity, Tac and Tbc are the binary affinities and Cabc is the
bowing factor for affinity which is given by:
Cabc    = – Cabc  Eg dEc  dEg 5-72

where Cabc() is the bowing factor for affinity, Cabc(Eg) is the bowing factor for bandgap as
given in Table 5-12 and dEc/dEg is the alignment parameter as given in Table 5-12.

408 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

For quaternaries, the affinities are approximated by the Equations 5-64 and 5-65 with
affinities replacing energy bandgaps.
Finally for all cases (binary, ternary and quaternary), the affinity used in the drift-diffusion
calculation is chosen as the one in the same direction (, , or L) as chosen for energy
bandgap.
Density of States
For the III-V cubic semiconductor model, the density of states (Nc and Nv) are each
calculated from the effective masses using Equations 3-31 and 3-32. For binary materials, the
density of states effective masses are given in Table 5-14.

Table 5-14 Default Cubic III-V Effective Masses and Permittivities

Binary  X
m ct X   mhh mlh r
mc m cl m ct m cl
(mo) (mo)
(mo) (mo) (mo) (mo) (mo)
GaAs 0.067 0.23 1.3 0.0754 1.9 0.49 0.16 12.91

AlP 0.22 0.155 2.68 0.0 0.0 0.63 0.20 9.8

AlAs 0.15 0.22 0.97 0.15 1.32 0.76 0.15 10.06

AlSb 0.14 0.123 1.357 0.23 1.64 0.94 0.14 12.04

GaSb 0.039 0.22 1.51 0.10 1.3 0.28 0.05 15.69

GaP 0.13 0.253 2.0 0.15 1.2 0.79 0.14 11.10

InP 0.0795 0.88 0.0 0.47 0.0 0.56 0.12 12.61

InSb 0.0135 0.0 0.0 0.25 0.0 0.43 0.015 17.70

InAs 0.026 0.16 1.13 0.05 0.64 0.57 0.025 15.15

For ternary materials, the effective masses are interpolated using the following expression
[2]:
Mabc = xMac +  1 – x Mbc 5-73
where Mabc is the ternary effective mass, Mac and Mbc are the binary effective masses and x
is the composition fraction.
For quaternary compounds the effective masses are interpolated using the following
expression[2]:
Mabcd = xyMac + x  1 – y Mad +  1 – x yMbc +  1 – x   1 – y Mbd 5-74
or for compounds with 3 anions [2]:
Mabcd = xMab + yMab +  1 – x – y Mad 5-75
where Mabcd is the quaternary effective mass, Mac, Mad, Mbc and Mbd are the binary
effective masses and x and y are the composition fractions.
For the calculation of conduction band density of states, the conduction band density of states
mass is calculated for the same direction (, X, or L) as is used for energy bandgap as follows:

mc = mc 5-76

409 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

X 2 X 13
m c =   m ct    m cl   5-77

or
L 2 L 13
m c =   m ct    m cl   5-78

For the calculation of valence band density of states, the valence band density of states mass
is calculated as follows:
32 32 23
m v =  m hh + m lh  5-79

Static Permittivity
Table 5-14 shows the permittivities of the binary compounds. The ternary and quaternary
permittivities are approximated by the interpolation formulas analogous to those used for
effective mass (Equations 5-73, 5-74, and 5-75) replacing masses by permittivities.

410 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

5.4.2 Gallium Arsenide (GaAs) Physical Models

Bandgap Narrowing
Following Klausmeier-Brown [142], the bandgap narrowing effects are important only for p-
type regions. By default, BLAZE uses the bandgap narrowing values shown in Table 5-15.

Table 5-15 Default Bandgap Narrowing Values

Concentration (cm-3) Bandgap Narrowing (meV)

1.01018 31.0

2.01018 36.0

4.01018 44.2

6.01018 48.5

8.01018 51.7

1.01019 54.3

2.01019 61.1

4.01019 64.4

6.01019 61.9

8.01019 56.9

1.01020 53.2

2.01020 18.0

Note: This table is only used for GaAs. No data is available at present for other materials. The C-INTERPRETER
function for bandgap narrowing, however, allows a user-defined model for bandgap narrowing to be applied
for these materials. See Appendix A “C-Interpreter Functions” for more information about on these functions.

411 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Low Field Mobility

You can make the mobility in GaAs concentration dependent by setting the CONMOB parameter
of the MODEL statement. In this model, mobility is interpolated from the values in Table 5-16.

Table 5-16 Default Concentration Dependent Mobility for GaAs

Mobility in GaAs (cm2/v-s)

-3)
Concentration (cm
Electrons Holes

1.01014 8000.0 390.0

2.01014 7718.0 380.0

4.01014 7445.0 375.0

6.01014 7290.0 360.0

8.01014 7182.0 350.0

1.01015 7300.0 340.0

2.01015 6847.0 335.0

4.01015 6422.0 320.0

6.01015 6185.0 315.0

8.01015 6023.0 305.0

1.01016 5900.0 302.0

2.01016 5474.0 300.0

4.01016 5079.0 285.0

6.01016 4861.0 270.0

8.01016 4712.0 245.0

1.01017 4600.0 240.0

2.01017 3874.0 210.0

4.01017 3263.0 205.0

6.01017 2950.0 200.0

8.01017 2747.0 186.9

1.01018 2600.0 170.0

2.01018 2060.0 130.0

412 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Table 5-16 Default Concentration Dependent Mobility for GaAs

Mobility in GaAs (cm2/v-s)

-3)
Concentration (cm
Electrons Holes

4.01018 1632.0 90.0

6.01018 1424.0 74.5

8.01018 1293.0 66.6

1.01020 1200.0 61.0

If MODEL ANALYTIC is specified, the program will use [142]:

8000 – 940
 n p = 940 + --------------------------------------------------- 5-80
0.75
1 +  --------------------------
Nt
 16
2.8  10

where Nt is the total impurity concentration.

5.4.3 Al(x)Ga(1-x)As System

The Al(x)Ga(1-x)As material system is commonly used for the fabrication of heterojunction
devices. These materials are available in BLAZE by specifying the material name GaAs,
AlGaAs, or AlAs. As a ternary material system, different material properties are obtained by
adjusting the molar fraction of Aluminum and Gallium. This mole fraction is represented by
the x as written in Al(x)Ga(1-x)As. GaAs material parameters are identical to the those of
AlGaAs with the mole fraction x set equal to zero. AlAs material parameters are identical to
the those of AlGaAs with mole the fraction x set equal to one. Fundamental in the proper
simulation with the AlGaAs material system is the relationship between this mole fraction x,
and the material parameters for that composition. In the following sections, the relationship
between mole fraction and material parameters for the AlGaAs material system will be
described. You can specify the x-composition fraction of a material in the REGION statement
with the X.COMP parameter.
Bandgap
There are three primary conduction bands in the AlGaAs system, depending on mole fraction,
that determine the bandgap. These are named Gamma, L, and X. The default bandgaps for
each of these conduction band valleys are as follows:
E g = EG300 + x.comp   1.155 + 0.37  X.COMP  5-81

E gL = 1.734 + x.comp   0.574 + 0.055  X.COMP  5-82

E gX = 1.911 + x.comp   0.005 + 0.245  X.COMP  5-83

The bandgap used for any given Al concentration is the minimum as calculated from these
equations. EG300 is the bandgap at 300K and specified on the material statement.

413 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

x.composition is the Aluminum mole fraction and can be user-defined in the REGION
statement.
The temperature dependence of the bandgap is calculated according to:
2
300 2 TL
E g  T L  = E g  300  + EGALPHA  --------------------------------------
- – ----------------------------------
- 5-84
300 + EGBETA T L + EGBETA

The value of Eg(300) is taken as the minimum of Eg, Egx, and EgL. The default temperature
dependent bandgap parameters for AlGaAs are listed in Table 5-17.

Table 5-17 Default Bandgap Parameters for Al(x)Ga(1-x)As

Statement Parameter Default Units

MATERIAL EG300 1.59 eV

MATERIAL EGALPHA 5.40510-4 eV/K

MATERIAL EGBETA 204 K

Electron Affinity
As indicated in the introduction, the semiconductor electron affinity  is a key parameter for
determining the alignment of heterojunctions. For AlGaAs,  is a function of Eg and is given
by:
 AlGaAs = 4.07 – 0.85   E g  X.COMP  – E gGaAs  5-85

Density of States and Effective Mass

The valence and conduction band densities of states, NC and NV, are calculated from the
effective masses according to the following equations:
3
* ---
 2m e kT L 2
N c = 2  ----------------------- 5-86
2
 h 

3
* ---
 2m h kT L 2
N v = 2  ----------------------- 5-87
2
 h 

For the AlGaAs system the conduction band and valence band effective masses, for electrons
and holes, are given by [169]:
0.067 + 0.083x  0  x  0.45 
me = 5-88
0.85 – 0.14x  x  0.45 

m lh = 0.087 + 0.063x 5-89

414 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

m hh = 0.62 + 0.14x 5-90

32 32 23

mh =  m + m  5-91
lh hh

Dielectric Permittivity
The default static dielectric constant for AlGaAs is given by:
 AlGaAs = 13.18 – 2.9  X.COMP 5-92

Low Field Mobility

The default low field electron mobility for AlGaAs is a function of the composition fraction,
x, within the system. The following equations outline this relationship.

Table 5-18 Low Field Mobility Equations

AlGaAs Low Field Mobility Mole Fraction Range

4  0  X.COMP  0.429 
 n = 8000 –  1.818 10  X.COMP 

4  0.429  X.COMP  0.46 

 n = 90 + 1.1435 10   X.COMP – 0.46  2

4  0.46  X.COMP  0.5 

 n = 90 + 3.75 10   X.COMP – 0.46  2

 n = 200 –  2   X.COMP – 0.46    0.5  X.COMP  1.0 

5.4.4 In(1-x)Ga(x)P System

The In(1-x)Ga(x)P system is commonly used in optoelectronics, especially solar cells, that
are lattice matched to GaAs. Particularly, important is are composition fractions around 0.5.
Bandgap
The default energy bandgap for InGaP latticed matched to GaAs used in Blaze is given by
Equation 5-93 [175].

Eg = 1.35 + X · COMP 0.73 + X · COMP X · COMP 0.7 5-93

Electron Affinity
The default electron affinity for InGaP lattice matched to GaAs used in BLAZE is given by
Equation 5-94 [122].

415 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

 = 4.38 – 0.58 X · COMP 5-94

Density of States Effective Mass

The electron density of states effective mass is given by Equations 5-95 and 5-96 [122].

m c = 0.088 X · COMP  0.74 5-95

m c = 0.63 + 0.13 X · COMP X · COMP  0.74 5-96

The heavy hole and light hole density of states effective masses are given by Equations 5-97
and 5-98 [122].

m hh = 0.6 + 0.19 X · COMP 5-97

m lh = 0.09 + 0.05 X · COMP 5-98

Dielectric Permittivity
The default static dielectric permittivity for InGaP lattice matched to GaAs used in BLAZE is
given by Equation 5-99 [122].

 = 12.5 – 1.4 X · COMP 5-99

5.4.5 In(1-x)Ga(x)As(1-y)P(y) System

The In(1-x)Ga(x)As(1-y)P(y) material system is commonly used for the fabrication of
heterojunction devices. These include laser diodes, photodiodes, Gunn diodes, and high speed
heterostructure transistors. As a quaternary material, two different mole fraction parameters,
x and y, are necessary to specify any particular combination. This produces a wide array of
InGaAsP materials and characteristics. Of particular interest in this system are materials that
are lattice matched to InP.
The default material characteristics in BLAZE for the InGaAsP system correspond to
composition fractions x and y that yield InGaAsP material that is lattice matched to InP. The
xcomposition and ycomposition are specified in the REGION statement with the X.COMP and
Y.COMP parameters. The relationship between x and y that satisfy this condition is given by:

0.1896  Y.COMP
x = -------------------------------------------------------------------------- 0  Y.COMP  1 5-100
0.4176 –  0.0125  Y.COMP 

Many of the parameter models for the In(1-x)Ga(x)As(y)P(1-y) system are functions of the
composition fraction Y.COMP only. The composition fraction, X.COMP, can be deduced from
the preceding relationship. Again, the default material characteristics in BLAZE for the
InGaAsP system correspond to composition fractions x and y that yield InGaAsP material
that is lattice matched to InP.

416 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Note: Don’t use this material system to form GaAs by setting x=1 and y=1, specify GaAs as the material instead.

Bandgap
The default energy bandgap for the InP lattice matched In(1-x)Ga(x)As(y)P(1-y) system used
in BLAZE is given by:
E g  InGaAsP  = 1.35 + X.COMP   0.642 +  0.758  X.COMP  +  0.101  Y.COMP – 5-101
1.101   Y.COMP–  0.28  X.COMP – 0.109  Y.COMP + 0.159   X.COMP  Y.COMP

Electron Affinity
The electron affinities for materials in the InP lattice matched InGaAsP system are derived
from conduction band offsets and from the assumption that the affinity of InP is 4.4eV. The
default conduction band edge offset between lattice matched InGaAsP and InP is then:
E c = 0.268  Y.COMP + 0.003   Y.COMP  2 5-102

Density of States and Effective Mass

The density of states is defined, as before, as a function of the effective masses of electrons
and holes according to Equation 3-31. For the InGaAsP system, the default conduction and
valence band effective masses, for electrons and holes, are given by the following.
For the conduction band:
*
m e = 0.08 –  0.116  Y.COMP  +  0.026  X.COMP – 0.059   X.COMP  Y.COMP + 5-103
 0.064 – 0.02  Y.COMP    X.COMP  2 +  0.06 + 0.032  X.COMP    Y.COMP  2

For the valence band the hole effective mass is defined by:
2
---
* 1.5 1.5 3
mh =  m lh + m hh  5-104

where the default light hole effective mass is given by:

m lh = 0.120 –  0.116  Y.COMP  + 0.03   X.COMP  2 5-105

and the default heavy hole effective mass is a constant and is given by:
m hh = 0.46 5-106

Dielectric Permittivity
The default static dielectric constant for lattice matched InGaAsP to InP is given by
 InGaAsP =  14.6   1 – X.COMP   Y.COMP + 12.5   1 – X.COMP   5-107
 1 – Y.COMP + 13.18  X.COMP  Y.COMP+ 11.11  X.COMP   1 – Y.COMP 

417 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Low Field Mobility

The default low field mobility parameters for electrons and holes for lattice matched InGaAs
are given by linear interpolations from the binary compounds GaAs and InP. The following
formulas are used:
 n1 = 33000 +  8500 – 33000   X.COMP 5-108

 p1 = 460 +  400 – 460   X.COMP 5-109

 n2 = 4600 +  300 – 4600   X.COMP 5-110

 p2 = 150 +  100 – 150   X.COMP 5-111

 n0 =  n1 +  1 – Y .COMP    n2 –  n1  5-112

 p0 =  p1 +  1 – Y .COMP    p2 –  p1  5-113

5.4.6 The Si(1-x)Ge(x) System

Advances in the growth of Silicon and Si(1-x)Ge(x) alloys have allowed the potential for
using bandgap engineering to construct heterojunction devices such as HBTs and HEMTs
using these materials. BLAZE supports the SiGe material system by providing composition
dependent material parameters. These parameters are accessed by specifying the material
name SiGe.
The following sections describe the functional relationship between Ge mole fraction x, and
the SiGe material characteristics necessary for device simulation.
Bandgap [154]
Bandgap is one of the most fundamental parameters for any material. For SiGe, the
dependence of the bandgap on the Ge mole fraction, x.composition, is divided into ranges as
follows:
for x 0.245
E g = 1.08 + X.COMP  0.945 – 1.08   0.245 5-114

for 0.245 x 0.35

E g = 0.945 +  X.COMP – 0.245    0.87 – 0.945    0.35 – 0.245  5-115

for 0.35  x 0.5

E g = 0.87 +  X.COMP – 0.35    0.78 – 0.87    0.5 – 0.35  5-116

for 0.5 < x  0.6

418 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

E g = 0.78 +  X.COMP – 0.5    0.72 – 0.78    0.6 – 0.5  5-117

for 0.6x 0.675

E g = 0.72 +  X.COMP – 0.6    0.69 – 0.72    0.675 – 0.6  5-118

for 0.675  x 0.735

E g = 0.69 +  X.COMP – 0.675    0.67 – 0.69    0.735 – 0.675  5-119

for 0.735  x 1

E g = 0.67 5-120

The temperature dependence of the bandgap of SiGe is calculated the same as for Silicon
except that EGALPHA and EGBETA are a function of Ge mole fraction x as follows:
2
2 TL
300
E g  T L  = E g + EGALPHA -------------------------------------- – ----------------------------------
- 5-121
300 + EGBETA T L + EGBETA

–4
EGALPHA =  4.73 + X.COMP   4.77 – 4.73   10 5-122

EGBETA = 636 + X.COMP   235 – 636  5-123

where Eg is dependent upon the mole fraction as above.
Electron Affinity
The electron affinity of SiGe is taken to be constant (4.17) with respect to composition.
Density of States
The density of states for SiGe is defined differently compared to the previous materials by not
being a function of the effective masses. Instead the density of states have been made to
depend upon the Ge mole fraction, x.composition,according to:
19 19 19
N c = 2.8 10 + X.COMP   1.04 10 – 2.8 10  5-124

19 18 19
N v = 1.04 10 + X.COMP   6.0 10 – 1.04 10  5-125

Dielectric Function
The compositional dependence of the static dielectric constant of SiGe is given by
 = 11.8 + 4.2  X.COMP 5-126

419 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Low Field Mobility

No specific SiGe low field mobility models have been implemented into BLAZE.
Velocity Saturation
In SiGe, the temperature dependent velocity saturation, used in the field dependent mobility
model is defined by the following equations.

VSATN = 1.38  10   tan h  ---------- 

7 175
5-127
  TL  

VSATP = 9.05  10   tan h  ---------- 

6 312
5-128
  TL  

Note: All other defaults used for SiGe are taken from Silicon.

5.4.7 Silicon Carbide (SiC)

Silicon carbide materials are of interest for high power, high temperature applications. The
main characteristics of silicon carbide are that they have a very wide bandgap, high thermal
conductivity, high saturation velocity, and high breakdown strength. Silicon carbide is
commercially available in three polytypes called 6H-SiC, 3H-SiC, and 4H-SiC. ATLAS
supports all three these polytypes. The following paragraphs describe the material defaults for
these materials.
Band Parameters for SiC
SiC band parameter equations are identical to those used for Silicon but with the values
adjusted for 3C-SiC, 4H-SiC, and 6H-SiC. The physical band parameter values are shown in
Tables B-15 and B-16.

SiC Mobility Parameters

420 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Isotropic Mobility
By default, mobility is assumed to be entirely isotropic in nature. That is, there is no
directional component. The default low field mobilities of electrons and holes for 3C-SiC,
4H-SiC, and 6H-SiC are shown in Table 5-19.

Table 5-19 Silicon Carbide Low Field Mobility Defaults

Statement Parameter 6H-SiC 3C-SiC 4H-SiC Units

MOBILITY MUN 330 1000 460 cm2/V·s

MOBILITY MUP 60 50 124 cm2/V·s

Anisotropic Mobility
The mobility behavior within SiC is now known to be anisotropic in nature, which
dramatically alters the electrical performance of a device. A anisotropic model has been
implemented into ATLAS to correctly model this behavior. Following the ideas of Lindefelt
[163] and Lades [152], the mobility within the drift diffusion equations has been made a
tensor property. As a result the mobility has become:

1 0 0
 = 0 1 0 5-129
0 0 2

where 1 represents the mobility defined in one plane and 2 the mobility defined in a second
plane. In the case of SiC, 1 represents the mobility of the direction <1100> while 2
represents the mobility of the direction <0001>. These mobilities are defined for both holes
and electrons.
Anisotropy is handled by allowing you to specify mobility parameters in two directions
normal to each other. To enable the model, specify the N.ANGLE or P.ANGLE or both
parameters of the MOBILITY statement. When you specify the parameter(s), it signifies that
the mobility parameter values and model flags in the enclosing MOBILITY statement apply
along the specified angle, with respect to the X axis, for electrons (N.ANGLE) and holes
(P.ANGLE). The parameter values and model flags specified in a second MOBILITY statement
(without the N.ANGLE or P.ANGLE specification) apply to all normal directions (in the x-y
plane) to the angle specified in the other statement.
When calculating the current along an element edge, the direction vector of the edge is used
to calculate the weighting of the two mobilities specified by the two MOBILITY statements.
The weights (w1 and w2) are taken to be the dot product of the current direction vector with
the angle specified on the N.ANGLE or P.ANGLE parameter for the associated parameters and
models, the dot product of the current direction vector, and the normal to the specified angle
for the models and parameters not associated with the N.ANGLE or P.ANGLE parameter
specification.
The net weighted mobility is calculated by a weighted application of Mattheissen's rule as
given by Equation 5-130.

421 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

1
 = -------------------- 5-130
w1 w2
------ + ------
1 2

Defining Anisotropic Mobility in ATLAS

To define a material with anisotropic mobility, specify two MOBILITY statements. In each
statement, the N.ANGLE and P.ANGLE parameters are used to specify the direction where that
particular mobility is to apply. The following example shows how this is done.

# FIRST DEFINE MOBILITY IN PLANE <1100>

#
MOBILITY MATERIAL=4H-SiC VSATN=2E7 VSATP=2E7 BETAN=2 BETAP=2 \
MU1N.CAUG=10 MU2N.CAUG=410 NCRITN.CAUG=13E17 \
DELTAN.CAUG=0.6 GAMMAN.CAUG=0.0 \
ALPHAN.CAUG=-3 BETAN.CAUG=-3 \
MU1P.CAUG=20 MU2P.CAUG=95 NCRITP.CAUG=1E19 \
DELTAP.CAUG=0.5 GAMMAP.CAUG=0.0 \
ALPHAP.CAUG=-3 BETAP.CAUG=-3
#
# NOW DEFINE MOBILITY IN PLANE <0001>
#
MOBILITY MATERIAL=4H-SiC N.ANGLE=90.0 P.ANGLE=90.0 VSATN=2E7 VSATP=2E7 \
BETAN=2 BETAP=2 MU1N.CAUG=5 MU2N.CAUG=80 NCRITN.CAUG=13E17 \
DELTAN.CAUG=0.6 GAMMAN.CAUG=0.0 \
ALPHAN.CAUG=-3 BETAN.CAUG=-3 \
MU1P.CAUG=2.5 MU2P.CAUG=20 NCRITP.CAUG=1E19 \
DELTAP.CAUG=0.5 GAMMAP.CAUG=0.0 \
ALPHAP.CAUG=-3 BETAP.CAUG=-3
In this example, the second MOBILITY statement applies to the direction 90° with respect to
the X axis (or along the Y axis). The first MOBILITY statement without the N.ANGLE and
P.ANGLE parameter assignments applies along the X axis.

Carrier-Carrier Scattering Contribution to Mobility

422 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

The expression for the carrier-carrier scattering contribution to mobility given by Lades [151]
is the same as the Conwell-Weisskopf expression given in Equation 3-215. The default
values of the coefficients D.CONWELL and F.CONWELL for 4H-SiC and 6H-SiC are set to the
values given in Table 5-20.

Table 5-20 Default Values of Conwell-Weisskopf Mobility Model Parameters for SiC

Statement Parameter Default (4H-SiC) Default (6H-SiC) Units

MOBILITY D.CONWELL 6.91020 3.0451020 (cm V s)-1

MOBILITY F.CONWELL 7.4521013 7.4521013 cm-2

Doping Dependent Low Field Mobility for 4H-SiC and 6H-SiC

The analytic low field model (Caughey-Thomas) is given in Equations 3-202 and 3-203.
Specific parameters for 4H-SiC and 6H-SiC are given by Lades [151]. These are listed in
Table 5-21.

Table 5-21 Default Values of Analytic Mobility Model Parameters for SiC.

Statement Parameter Default (4H-SiC) Default (6H-SiC) Units

MOBILITY MU1N.CAUG 0.0 0.0 cm2/(V·s)

MOBILITY MU1P.CAUG 15.9 6.8 cm2/(V·s)
MOBILITY MU2N.CAUG 415.0 947.0 cm2/(V·s)
MOBILITY MU2P.CAUG 124.0 99.0 cm2/(V·s)
MOBILITY ALPHAN.CAUG 0.0 0.0
MOBILITY ALPHAP.CAUG 0.0 0.0
MOBILITY BETAN.CAUG -1.8 -1.8
MOBILITY BETAP.CAUG -1.8 -1.8
MOBILITY GAMMAN.CAUG 0.0 0.0
MOBILITY GAMMAP.CAUG 0.0 0.0
MOBILITY DELTAN.CAUG 0.61 0.59
MOBILITY DELTAP.CAUG 0.34 0.31
MOBILITY NCRITN.CAUG 1.941017 1.111018 cm-3
MOBILITY NCRITP.CAUG 1.761019 2.11019 cm-3

The parameters for the low field mobility for the CVT model have also been changed so they
give the same mobility as the ANALYTIC model.

Temperature Dependent Saturated Velocity for 4H-SiC and 6H-SiC

423 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

The parallel field dependent mobility model for electrons enabled by specifying FLDMOB on
the MODELS statement has a saturated velocity given by Equation 3-308. For 4H-SiC and 6H-
SiC, a different model for saturated velocity is preferable. This is given by
TL DELTAN.FLD
v n = ALPHAN.FLD  ------------------------------ 5-131
 TNOMN.FLD

and applies to electrons in 4H-SiC and 6H-SiC. If the BETA exponents of this model are
temperature dependent, as in the Canali modification of Equation 3-310, then the specific
default values of N.BETA0 and N.BETAEXP for SiC are used. All defaults are given in Table
5-22.

Table 5-22 Default Values of Saturated Electron Velocity Model Parameters for SiC.

Statement Parameter Default (4H-SiC) Default (6H-SiC) Units

MOBILITY ALPHAN.FLD 2.2107 1.9107 cm/s

MOBILITY TNOMN.FLD 300 300 Kelvin
MOBILITY DELTAN.FLD -0.44 -1.0
MOBILITY N.BETA0 1.2 1.7
MOBILITY N.BETAEXP 1.0 1.25

Anisotropic Impact Ionization of 4H-SiC

To enable a model [103] for the anisotropic impact ionization rate of 4H-SiC, specify the
ANISO parameter of the IMPACT statement. In this model, the following equations describe the
ionization rate:

 2 2 Ex Ey 
2
  E x E y  = aEXP  – c 1 – A c  ------------  5-132
  bx by  

A = ln  a y  a x  5-133

2 2 –1  2
E E 
c = --b- =  -----x- + -----y- 5-134
E b 2
by 
2
x

2 2 2 2
c E c E
-----------x- -----------y-
2 2
bx by
a = ax ay 5-135
Here, Ex and Ey are the electric field magnitude in the x and y directions. The values
parameters ax, ay, bx, and by depend upon the crystal orientation. The crystal orientation can
be either 0001 or 11 2 0 as specified by either the SIC4H0001 or the SIC4H1120 flag of the
IMPACT statement. By default, SIC4H0001 is selected.

424 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

The values associated with the y coordinates (ay and by) correspond to the selected
orinetation. The values of the unselected orientation are then associated with the x
coordinates (ax and bx). Table 5-23 shows the default values of ax, bx, ay, and by, where AE*
and BE* are for electrons and AH* and BH* are for holes.

Table 5-23 Anisotropic Impact Ionization Model Defaults

Parameter Statement Default Units

AE0001 IMPACT 1.76108 cm-1

BE0001 IMPACT 3.30107 V/cm
AH0001 IMPACT 3.41108 cm-1
BH0001 IMPACT 2.50107 V/cm
AE1120 IMPACT 2.10107 cm-1
BE1120 IMPACT 1.70107 V/cm
AH1120 IMPACT 2.96107 cm-1
BH1120 IMPACT 1.60107 V/cm

Thermal Parameters
The equations governing these effects are identical to those for Silicon but with adjusted
coefficients. See Appendix B “Material Systems” for a list of all these parameters.
Two level Incomplete Ionization Model
The model for incomplete ionization described in Section 3.1.2 “Carrier Continuity
Equations” assumes a single ionisation energy for donor impurities and a single ionization
level for acceptor impurities. In 4H-SiC and 6H-SiC, the dopant locations may have either a
locally cubic arrangement of surrounding atoms, or a locally hexagonal arrangement [151].
These different arrangements result in different ionization energies for the same dopant
species. In 6H-SiC, there are 2 cubic environment sites for every hexagonal environment site
and in 4H-SiC there is one cubic site for every hexagonal site. The ionized n-type doping is
given by

+ N d  1 –  N d
N d = --------------------------------------------------------------------------------------------
- + ----------------------------------------------------------------------------------------------------------- 5-136
E Fn – E c + ED.CUBIC E Fn – E c + ED.HEXAGONAL
1 + GCB exp  ------------------------------------------------------- 1 + GCB exp  ---------------------------------------------------------------------
 KT L   KT L 

and the ionized p-type doping given by

- N a  1 –  N a
N a = --------------------------------------------------------------------------------------------
- + ----------------------------------------------------------------------------------------------------------
- 5-137
E v – E Fp + EA.CUBIC E v – E Fp + EA.HEXAGONAL

1 + GVB exp -------------------------------------------------------  
1 + GVB exp --------------------------------------------------------------------- 
 KT L   KT L 

425 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

where  = 1
--- for 4H-SiC and  = 2
--- for 6H-SiC. The ionization energies for donor
2 3
impurities, ED.CUBIC and ED.HEXAGONAL, are given default values corresponding to
Nitrogen doping. The ionization levels for acceptor impurities are given default values
corresponding to Boron doping. The default values are from [151] and are given in Table 5-
24.

Table 5-24 Default Values of Parameters for Site Specific Ionization Levels

Statement Parameter Default (4H-SiC) Default (6H-SiC) Units

MATERIAL ED.CUBIC 0.09 0.14 eV

MATERIAL ED.HEXAGONAL 0.05 0.08 eV
MATERIAL EA.CUBIC 0.32 0.32 eV
MATERIAL ED.HEXAGONAL 0.32 0.32 eV

The ionised densities obtained using this model can be quite different from those obtained by
using a single, averaged, ionization energy. To use this model, specify INCOMPLETE on the
MODELS statement and it will be used in any 4H-SiC or 6H-SiC regions.

5.4.8 GaN, InN, AlN, Al(x)Ga(1-x)N, In(x)Ga(1-x)N, Al(x)In(1-x)N, 

and Al(x)In(y)Ga(1-x-y)N
The following sections describe the relationship between mole fraction, X.COMP, and the
material parameters and various physical models specific to the Al/ln/GaN system.
Bandgap
By default, the bandgap for the nitrides is calculated in a two step process. First, the
bandgap(s) of the relevant binary compounds are computed as a function of temperature, T,
using Equations 5-138 through 5-140 [293].
–3 2
0.909  10 T -
Eg  GaN  = 3.507 – --------------------------------------- 5-138
T + 830.0

–3 2
 10 T -
Eg  InN  = 1.994 – 0.245
--------------------------------------- 5-139
T + 624.0

–3 2
Eg  AlN  = 6.23 – 1.799  10 T - 5-140
---------------------------------------
T + 1462.0

Note: The bandgap for InN contradicts recent observations [298,62,314] of a bandgap of less than 1.0 eV. Despite
these observations, we keep the above specified value to maintain consistency with ternary and quaternary
blends. You should carefully consider whether to use the default values, or supply your own values using
one of the standard methods of user specification. The simplest method is to specify EG300 on the
MATERIAL statement.

426 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Then, the dependence on composition fraction, x, is described by Equations 5-141 and 5-143
[216].
Eg  Inx Ga 1 – x N  = Eg  InN x + Eg  GaN   1 – x  – 3.8x  1 – x  5-141

Eg  Al x Ga 1 – x N  = Eg  AlN x + Eg  GaN   1 – x  – 1.3x  1 – x  5-142

Eg  Al x In 1 – x N  = Eg  AlN x + Eg  InN   1 – x  5-143

Electron Affinity
The electron affinity is calculated such that the band edge offset ratio is given by [216].
Ec 0.7
---------- = -------- 5-144
Ev 0.3

You can overide this ratio by specifying the ALIGN parameter of the MATERIAL statement.
Permittivity
The permittivity of the nitrides as a function of composition fraction, x, is given by linear
interpolations of the values for the binary compounds as in Equations 5-145 and 5-146 [8].
  In x Ga 1 – x N  = 15.3x + 8.9  1 – x  5-145

  Al x Ga 1 – x N  = 8.5x + 8.9  1 – x  5-146

Density of States Masses

The nitride density of states masses as a function of composition fraction, x, is given by linear
interpolations of the values for the binary compounds as in Equations 5-147 through 5-150
[293].
m e  In x Ga 1 – x N  = 0.12x + 0.2  1 – x  5-147

m h  In x Ga 1 – x N  = 0.17x + 1.0  1 – x  5-148

m e  Al x Ga 1 – x N  = 0.314x + 0.2  1 – x  5-149

m h  Al x Ga 1 – x N  = 0.417x + 1.0  1 – x  5-150

427 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Low Field Mobility

The Albrecht Model
You can choose to model low field mobility following the work of Albrecht et.al [6] by
specifying ALBRCT on the MODEL statement or ALBRCT.N or ALBRCT.P or both on the
MOBILITY statement for separate control over electrons and holes. This model is described as
follows:
–3  2
1 AN · ALBRCT  N  -------------------------------------
TL

--------------------- = ----------------------------------------- 5-151
  N T L  N0N  ALBRCT  T0N  ALBRCT
TL 2 –2  3
ln 1 + 3  -------------------------------------  -------------------------------------
N
 T0N  ALBRCT  N0N  ALBRCT
TL 32
+ BN · ALBRCT   ------------------------------------- +
 T0N  ALBRCT
CN · ALBRCT
-----------------------------------------------------------------------
-
exp  T1N  ALBRCT  T L  – 1

where µ(N,T) is the mobility as a function of doping and lattice temperature, N is the total
doping concentration, and TL is the lattice temperature. AN.ALBRCT, BN.ALBRCT, CN.ALBRCT,
N0N.ALBRCT, T0N.ALBRCT and T1N.ALBRCT are user-specifiable parameters on the
MOBILITY statement. You can use a similar expression for holes with the user-defined
parameters AP.ALBRCT, BP.ALBRCT, CP.ALBRCT, N0P.ALBRCT, T0P.ALBRCT and
T1P.ALBRCT.
Table 5-25 shows the default values for the parameters of the Albrecht Model.

Table 5-25 Default Parameter Values for the Albrecht Model

Parameter Default Parameter Default Units

AN.ALBRCT 2.6110-4 AP.ALBRCT 2.6110-4 V*s/(cm2)

BN.ALBRCT 2.910-4 BP.ALBRCT 2.910-4 V*s/(cm2)
CN.ALBRCT 170.010-4 CP.ALBRCT 170.010-4 V*s/(cm2)
N0N.ALBRCT 1.010-17 N0P.ALBRCT 1.010-17 cm-3
T0N.ALBRCT 300.0 T0P.ALBRCT 300.0 K
T1N.ALBRCT 1065.0 T1P.ALBRCT 1065.0 K

Farahmand Modified Caughey Thomas

You can use a composition and temperature dependent low field model by specifying the
FMCT.N and FMCT.P in the MOBILITY statement. FMCT stands for Farahmand Modified
Caughey Thomas. This model [74] was the result of fitting a Caughey Thomas like model to
Monte Carlo data. The model is similar to the analytic model described by Equations 3-202
and 3-203. This modified model is described by Equations 5-152 and 5-153 for electrons and
holes.

428 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

BETAN · FMCT
 n  T N  = MU1N · FMCT  ----------
T
+ 5-152
 300

T DELTAN · FMCT
 MU2N · FMCT – MU1N · FMCT   ----------
 300
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
EPSP · FMCT
ALPHAN · FMCT T  300 
N
1 + --------------------------------------------------------------------------------------------------
T GAMMAN · FMCT
NCRITN · FMCT  ----------
 300

T BETAP · FMCT
 p  T N  = MU1P · FMCT  ---------- + 5-153
 300

T DELTAP · FMCT
 MU2P · FMCT – MU1P · FMCT   ----------
 300
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
EPSP · FMCT
ALPHAP · FMCT T  300 
N
1 + --------------------------------------------------------------------------------------------------
T GAMMAP · FMCT
NCRITP · FMCT  ----------
300

In these equations, T is the lattice temperature and N is the total doping. Table 5-26 shows the
user-definable parameters.

Table 5-26 User-specifiable parameters for the Faramand modified Caughey Thomas model
for Nitrides.

Parameter Statement Type Units

MU1N.FMCT MOBILITY Real cm2/(V*s)

MU1P.FMCT MOBILITY Real cm2/(V*s)
MU2N.FMCT MOBILITY Real cm2/(V*s)
MU2P.FMCT MOBILITY Real cm2/(V*s)

429 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Table 5-26 User-specifiable parameters for the Faramand modified Caughey Thomas model
for Nitrides.

Parameter Statement Type Units

ALPHAN.FMCT MOBILITY Real

ALPHAP.FMCT MOBILITY Real
BETAN.FMCT MOBILITY Real
BETAP.FMCT MOBILITY Real
GAMMAN.FMCT MOBILITY Real
GAMMAP.FMCT MOBILITY Real
DELTAN.FMCT MOBILITY Real
DELTAP.FMCT MOBILITY Real
EPSN.FMCT MOBILITY Real
EPSP.FMCT MOBILITY Real
NCRITN.FMCT MOBILITY Real cm-3
NCRITP.FMCT MOBILITY Real cm-3

Tables 5-27 and 5-28 show the default parameters as taken from the Monte Carlo fits for
various nitride compositions.

Table 5-27 Default Nitride Low Field Mobility Model Parameter Values [74]

MU1N.FMCT MU2N.FMCT
MATERIAL ALPHAN.FMCT BETAN.FMCT
(cm2/V.s) (cm2/V.s)

InN 774 3138.4 0.68 -6.39

In0.8Ga0.2N 644.3 1252.7 0.82 -1.81

In0.5Ga0.5N 456.4 758.1 1.04 -1.16

In0.2Ga0.8N 386.4 684.2 1.37 -1.36

GaN 295.0 1460.7 0.66 -1.02

Al0.2Ga0.8N 132.0 306.1 0.29 -1.33

Al0.5Ga0.5N 41.7 208.3 0.12 -0.6

Al0.8Ga0.2N 47.8 199.6 0.17 -0.74

AlN 297.8 683.8 1.16 -1.82

430 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Table 5-28 Default Nitride Low Field Mobility Model Parameter Values [74]

MATERIAL DELTAN.FMCT GAMMAN.FMCT EPSN.FMCT NCRITN.FMCT

InN -1.81 8.05 -0.94 1017

In0.8Ga0.2N -1.30 4.84 -0.41 1017
In0.5Ga0.5N -1.74 2.21 -0.22 1017
In0.2Ga0.8N -1.95 2.12 -0.99 1017
GaN -3.84 3.02 0.81 1017
Al0.2Ga0.8N -1.75 6.02 1.44 1017
Al0.5Ga0.5N -2.08 10.45 2.00 1017
Al0.8Ga0.2N -2.04 20.65 0.01 1017
AlN -3.43 3.78 0.86 1017

For composition fractions not listed in Tables 5-27 and 5-28, the default parameters are
linearly interpolated from the nearest composition fractions on the table. You can override
these defaults by specifying any of the parameters listed in Table 5-26 on the MOBILITY
statement. Currently, this model has only been calibrated for electrons. We only recommend
that you use this model for holes when you define a set of real default parameters.
High Field Mobility
You can select nitride specific field dependent mobility model by specifying GANSAT.N and
GANSAT.P on the MOBILITY statement. This model [74] is based on a fit to Monte Carlo data
for bulk nitride, which is described in Equations 5-154 and 5-155.

N1N · GANSAT – 1
E
 n0  T N  + VSATN --------------------------------------------------------------------------
N1N · GANSAT
ECN · GANSAT
 n = --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
- 5-154
N2N · GANSAT N1N · GANSAT
1 + ANN · GANSAT  --------------------- +  ---------------------
E E
 ECN · GANSAT  ECN · GANSAT

N1P · GANSAT – 1
E
 p 0  T N  + VSATP --------------------------------------------------------------------------
N1P · GANSAT
ECP · GANSAT
 p = --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
- 5-155
N 2P · GANSAT N1 P · GANSAT
1 + ANP · GANSAT  --------------------- +  ---------------------
E E
 ECP · GANSAT  ECP · GANSAT

In these equations, µ0(T,N) is the low field mobility and E is the electric field. Table 5-29 lists
the user-definable parameters.

431 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Table 5-29 User Definable Low Field Nitride Mobility Model Parameters
Parameter Statement Type Units

N1N.GANSAT MOBILITY Real

N1P.GANSAT MOBILITY Real
N2N.GANSAT MOBILITY Real
N2P.GANSAT MOBILITY Real
ANN.GANSAT MOBILITY Real
ANP.GANSAT MOBILITY Real
ECN.GANSAT MOBILITY Real V/cm
ECP.GANSAT MOBILITY Real V/cm

Table 5-30 shows the default parameters as taken from the Monte Carlo fits for various nitride
compositions.

Table 5-30 Default Nitride Field Dependent Mobility Model Parameter Values [74]

VSATN ECN.GANSAT
MATERIAL N1N.GANSAT N2N.GANSAT ANN.GANSAT
(cm/S) (kV/cm)

InN 1.3595107 52.4242 3.8501 0.6078 2.2623

In0.8Ga0.2N 0.8714107 103.4550 4.2379 1.1227 3.0295

In0.5Ga0.5N 0.7973107 148.9098 4.0635 1.0849 3.0052

In0.2Ga0.8N 1.0428107 207.5922 4.7193 1.0239 3.6204

GaN 1.9064107 220.8936 7.2044 0.7857 6.1973

Al0.2Ga0.8N 1.1219107 365.5529 5.3193 1.0396 3.2332

Al0.5Ga0.5N 1.1459107 455.4437 5.0264 1.0016 2.6055

Al0.8Ga0.2N 1.5804107 428.1290 7.8166 1.0196 2.4359

AlN 2.1670107 447.0339 17.3681 0.8554 8.7253

For composition fractions not listed in Table 5-30, the default parameters are linearly
interpolated from the nearest composition fractions on the table.
You can override these defaults by specifying any of the parameters listed in Table 5-29 in the
MOBILITY statement. Currently, this model has only been calibrated for electrons. We only
recommend that you use this model for holes when you define a set of real default parameters.

432 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Also note that these models exhibit negative differential mobility and may exhibit poor
convergence. In such cases, you may do well to use the simpler model given in Equations 3-
306 and 3-307 with a reasonable value of saturation velocity.
Figure 5-7 shows the modeled velocity-field curves for pure GaN for various low-field
mobilities.

Figure 5-7: Velocity-Field Curves for Various Low-Field Mobilities

Curve Fit Velocity Saturation Mobility Models
We have used the Deckbuild Optimizer to optimize parameters for the default GANSAT
mobility model, Equation 5-154 for electrons and the silicon saturation model, Equation 3-
306. We have built these models in accessable by setting logical flags on the MOBILITY
statement as listed in Table 5-31.

Table 5-31 Comparison of fit velocity saturation models

MOBILITY statement Reference Description

GANSAT.N [74] Monte Carlo InAlGaN electrons

CHEN.N [44] Monte Carlo GaN electrons
CHEN.P [44] Monte Carlo GaN holes
OZGUR.N [199] Monte Carlo ZnO electrons

433 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Figure 5-8 shows a comparison of the various velocity saturation mobility models.

Figure 5-8: Comparison of velocity saturation characteristics of various fit models.

The default parameters for the various electron saturation models using Equation 5-154 are
given in Table 5-32.

Table 5-32 Default parameters for fit models using Equation 5-154
MODEL MUN0 VSATN ECN.GANSAT N1N.GANSAT N2N.GANSAT ANN.GANSAT
cm2/(V*s) cm/s V/cm

GANSAT.MU 1000 1.9e7 220e3 7.2044 0.7857 6.1973

CHEN.N 1100 1.9e7 198e3 4.56 0.664 5.131
OZGUR.N 1000 1.9e7 256e3 4.12 -0.578 6.7

The default parameters for the hole saturation model using Equation 3-306 is listed in Table
5-33.

Table 5-33 Default parameters for fit models using Equation 3-306
MODEL MUP0 BETAP VSATP
cm2/(V*s) cm/s

CHEN.P 100 0.8 1.4e7

434 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Impact Ionization Parameters

Table 5-34 shows the extracted default values for the Selberherr impact ionization model
from [200] for GaN.

Table 5-34 Default Impact Ionization Parameters (IMPACT statement)

Parameter Units GaN InN AlN InGaN AlGaN

AN1 cm-1 2.52108 2.52108 2.52108 2.52108 2.52108

AN2 cm-1 2.52108 2.52108 2.52108 2.52108 2.52108
BN1 V/cm 3.41107 3.41107 3.41107 3.41107 3.41107
BN2 V/cm 3.41107 3.41107 3.41107 3.41107 3.41107
AP1 cm-1 5.37106 5.37106 5.37106 5.37106 5.37106
AP2 cm-1 5.37106 5.37106 5.37106 5.37106 5.37106
BP1 V/cm 1.96107 1.96107 1.96107 1.96107 1.96107
BP2 V/cm 1.96107 1.96107 1.96107 1.96107 1.96107
BETAN 1.0 1.0 1.0 1.0 1.0
BETAP 1.0 1.0 1.0 1.0 1.0
EGRAN V/cm 0.0 0.0 0.0 0.0 0.0

Supplied Impact Ionization Rate Tables

Using the tabular Selberherr model, we have provided build-in tabular representations of
results from several recent publications. We have provided convenient access using logicals
on the IMPACT statement. The association between logical parameter name and publication is
contained in Table 5-35.

Table 5-35 Association Between IMPACT Statement Parameter and Reference

IMPACT statement Reference Description

N.CHEN [44] Monte Carlo

P.CHEN [44] Monte Carlo
N.MCCLINTOCK [181] measured
P.MCCLINTOCK [181] measured
N.OGUZMAN [200] Monte Carlo

435 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Table 5-35 Association Between IMPACT Statement Parameter and Reference

P.OGUZMAN [200] Monte Carlo
N.TUT [275] measured Al(0.4)Ga(0.6)N
P.TUT [275] measured Al(0.4)Ga(0.6)N

Figures 5-9 and 5-10 show a comparison of the various tabular ionization rate models against
the default GANSAT.N and GANSAT.P models for the case of extrapolation as indicated by the
TBL.EXTRAP parameter of the IMPACT statement.
Figures 5-11 and 5-12 show a comparison of the various tabular ionization rate models
against the default GANSAT.N and GANSAT.P models for the case of disabled extrapolation as
indicated by ^TBL.TXTRACT on the IMPACT statement.

Figure 5-9: Comparison of electron ionization rates as a funciton of reciprocal field with extrapolation
enabled.

436 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Figure 5-10: Comparison of hole ionization rates as a funciton of reciprocal field with extrapolation enabled.

437 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Figure 5-11: Comparison of electron ionization rates as a funciton of reciprocal field with extrapolation
disabled.

438 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Figure 5-12: Comparison of hole ionization rates as a funciton of reciprocal field with extrapolation disabled.
Recombination Parameters
Table 5-36 shows the default values for radiative rates for the the binary wurtzite nitride
compounds.

Table 5-36 Default Radiative Rates for Nitrides (MATERIAL COPT)

Material COPT Units Reference

GaN 1.110-8 cm3/s [189]

InN 2.010-10 cm3/s [333]

AlN 0.410-10 cm3/s [297]

Note: Currently, no default formula have been implemented for ternary and quaternary compounds for COPT.

439 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Default Models for Heat Capacity and Themal Conductivity

By default, the heat capacity and thermal conductivities used in self-consistent heat flow
simulations (GIGA and GIGA3D) for the GaN/AlGaN/InGaN system are given by the
following equations [216].
2
20 –   D  T 
C L  T  = C L  300K  ----------------------------------------------  L 5-156
2
20 –   D  300K 


 L  T  =  L  300K   --------------
T
5-157
 300K

Here, T is the local temperature and the other parameters are given in Table 5-37 [216].

Table 5-37 Default Heat Capacity and Thermal Conductivity Parameters for GaN
Parameter Unit L CL L D 
(W/Kcm) (Ws/gK) (g/cm3) (K)

GaN 1.30 0.49 6.15 600 -0.28

AlN 2.85 0.6 3.23 1150 -1.64
InN 0.45 0.32 6.81 660 0.0

Values for ternary and quaternary compositions are obtained by linear interpolation as a
function of composition.
Adachi's Refractive Index Model [216]
For the III-V nitride compounds, Adachi's refractive index model is expressed by Equation 5-
158.

h – 2  h 
A  -------  2 – 1 + ------- – 1 – -------  + B
h
nr    = 5-158
 Eg  E
 g Eg 

where Eg is the bandgap,  is the optical frequency and A and B are material composition
dependent parameters. For AlxGa1-xN, the compositional dependence of the A and B
parameters are given by the expressions in Equations 5-159 and 5-160.
2
A  x  = 9.827 – 8.216X – 31.59X 5-159

2
B  x  = 2.736 + 0.842X – 6.293X 5-160
For InxGa1-xN, the compositional dependence of the A and B parameters are given by the
expressions in Equations 5-161 and 5-162.
A  x  = 9.827  1 + X  – 53.57X 5-161

B  x  = 2.736  1 – X  – 9.19X 5-162

440 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

GANFET Convenience Model

We have introduced a convenience flag for enabling a specific set of defaults useful for
simulation of most GaN FET type devices. The GANFET parameter of the MODELS statement
enables POLAR, CALC.STRAIN, the Albrecht mobility model for electrons, and SRH and
defaults POLAR.SCALE to 1.0.

5.4.9 The Hg(1-x)Cd(x)Te System

The following sections describe the relationship between mole fraction, X.COMP, and the
material parameters of the of the Hg(1-x)Cd(x)Te system. This data has been taken from
[303].
Bandgap
Equation 5-163 is used in the calculation of bandgap as a function of composition fraction.
2
Eg = – 0.302 + 1.93 X · COMP – 0.810 X · COMP 5-163

3
+ 0.832 X · COMP + 5.354  10   1 – 2 X · COMP T L
–4
 

Here, TL is lattice temperature.

Electron Affinity
Equation 5-164 is used in the calculation of electron affinity as a function of composition
fraction.
 = 4.23 – 0.813  Eg  T L  – 0.083  5-164

Permittivity
Equation 5-165 is used in the calculation of relative permittivity as a function of composition
fraction.
2
 = 20.5 – 15.5 X · COMP + 5.7 X · COMP 5-165

Density of States Masses

Equations 5-166 and 5-167 are used to calculate the density of states effective masses of
electrons and holes.
–1
-------- = – 0.6 + 6.333  ------------------- + ----------------------------
me 2 1
5-166
m  Eg  TL  Eg  TL  + 1 
0

me- = 0.55
------- 5-167
m0

441 ATLAS User’s Manual

Material Dependent Physical Models Blaze: Compound Material 2D Simulator

Mobility
Equations 5-168 and 5-169 are used to calculate the electron and hole mobilities as functions
of composition fraction.

7.5   0.6
– 2  ---------------------------
0.2
8 0.2   
 e = 9  10  --------------------- T L  X · COMP 5-168
 · COMP
X

 h = 0.01 e 5-169

5.4.10 CIGS (CuIn(1-x)Ga(x)Se2), CdS, and Zn0

The following table gives the default material parameter values for CIGS, CdS, and Zn0 [86,
155, 254].

Table 5-38 Default Parameters ofr CIGS, CdS, and Zn0

Parameter Zn0 CdS CIGS

r 9 10 13.6
n 100 100 100
p 25 25 25
Eg 3.3 2.48 Equation 5-170
 4.5 4.18 4.58
Nc 2.21018 2.411018 2.21018
Nv 1.81019 2.571019 1.81019

Eg =  1.036 + 4.238 10

–5
 T L + – 8.875 10– 5 170.0   tanh  170.0 0.5  T L  – 1.0   1 – X · COMP5-170

–5
+  1.691 + 8.820  10  T L – 16.0  10 – 5  189.0  tanh  189.8 0.5  T L  – 1.0 X · COMP – 0.02
 X · COMP  1 – X · COMP 

Zn0 Velocity Saturation

The electron velocity saturation model described in Figure 5-8 and Table 5-31 [199] can be
used by setting the OZGUR.N parameter of the MOBILITY statement.

442 ATLAS User’s Manual

Simulating Heterojunction Devices with Blaze Blaze: Compound Material 2D Simulator

5.5 Simulating Heterojunction Devices with Blaze

5.5.1 Defining Material Regions with Positionally-Dependent Band Structure
Step Junctions
The easiest way to define a device with positionally dependent band structure is to specify
two adjacent semiconductor regions with dissimilar bandgap. In this case, BLAZE would
simulate an abrupt heterojunction between the two materials.
For example, you want to simulate an abrupt heterojunction parallel to the X axis at a location
of y=0.1 microns. For values of y greater than 0.1 specify, GaAs. For values of y less than 0.1,
specify AlGaAs with a composition fraction of 0.3. The following statements would specify
this situation.
REGION Y.MIN=0.1 MATERIAL=GaAs
REGION Y.MAx=0.1 MATERIAL=AlGaAs x.COMP=0.3
This fragment specifies that the two regions form an abrupt heterojunction at Y=0.1. The first
region is composed of GaAs while the second is composed of AlGaAs.
These two material names are used by BLAZE to choose default material models for the two
regions. For a complete list of the materials available in ATLAS/BLAZE, see Appendix B
“Material Systems”. For the AlGaAs region a composition fraction of 0.3 is specified.
Graded Junctions
A grading can be applied to this heterojunction with a simple modification. For example:
REGION Y.MIN=0.1 MATERIAL=GaAs
REGION Y.MAx=0.1 MATERIAL=AlGaAs x.COMP=0.3 GRAD.34=0.01
specifies that the composition fraction of the AlGaAs region decreases from 0.3 at y=0.1
microns to 0.0 at y=0.11 microns. The GRAD parameter specifies the distance over which the
mole fraction reduces to zero. The GRAD parameter is indexed such that GRAD.12 corresponds
to the Y.MIN side of the region, GRAD.23 corresponds to the x.MAX side of the region,
GRAD.34 corresponds to the Y.MAX side of the region, and GRAD.41 corresponds to the x.MIN
side of the region. In most cases, the GRAD.n parameter acts to increase the size of the region.
By default, the GRAD.n parameters are set to zero and all heterojunctions are abrupt. Note that
the GRAD parameter acts just like the other region geometry parameters in that later defined
regions overlapping the graded part of the region will overlap the grading. If in the previous
example the grading had been applied to the GaAs region, it would be over lapped by the
AlGaAs region. This would have produced an abrupt interface. A solution would be to limit
Y.MAX in the AlGaAs region to 0.09. Make sure you specify regions in the proper order to
avoid such problems.
Along similar lines, you can use the overlapping of regions to an advantage in forming graded
heterojunctions between two materials in the same system with different non-zero
composition fractions. For example:
REGION Y.MIN=0.1 MATERIAL=AlGaAs x.COMP=0.3 GRAD.12=0.02
REGION Y.MAx=0.11 MATERIAL=AlGaAs x.COMP=0.1
specifies a graded heterojunction with a composition of 0.3 at y = 0.1 falling to 0.1 at y =
0.11.

443 ATLAS User’s Manual

Simulating Heterojunction Devices with Blaze Blaze: Compound Material 2D Simulator

5.5.2 Defining Materials and Models

Materials
For example to set the bandgap for the material, InP, use the following syntax.
MATERIAL MATERIAL=InP EG300=1.35
Models
BLAZE has two ways of simulating the physical effects of variations in semiconductor
composition. For relatively gradual variations in composition, the standard modifications to
the drift-diffusion equations can be considered adequate for simulation purposes. For abrupt
heterojunctions, it has been suggested that thermionic emission may be the dominant factor in
the behavior of heterojunction behavior.
Individual material parameters and models can be defined for each material or region. These
models are set in the MATERIAL, MODEL, and IMPACT statements.
This statement uses the MATERIAL parameter to select all regions composed of the material
"InP". The bandgap in these regions will be set to 1.35. There are two ways to set the
parameters of a particular region. The first is to use the region index. For example:
MODEL REGION=1 BGN
In this case, the bandgap narrowing model is enabled in the region indexed number 1.
The second is to use the region name. For example:
IMPACT NAME=substrate SELB
This example turns on the Selberherr Impact Ionization Model in the region named
substrate. You can then set the parameters for all regions and materials by omitting the
MATERIAL, REGION, or NAME parameters, as in the following:
MODEL BGN
This statement sets the bandgap narrowing model for all regions and materials
Parser Functions
To use the C-INTERPRETER functions, you need to know the C programming language. See
Appendix A “C-Interpreter Functions” for a description of the parser functions.
To specify a completely arbitrary spatial variation of varying composition fraction, use a
parser function. To define the parser function for composition fraction, write a C function
describing the composition fraction as a function of position. A template for the function
called COMPOSITION is provided with this release of ATLAS. Once you define the
COMPOSITION function, store it in a file. To use the function for composition, set the
F.COMPOSIT parameter to the file name of the function.

444 ATLAS User’s Manual

Chapter 6
3D Device Simulator
3D Device Simulation Programs 3D Device Simulator

6.1 3D Device Simulation Programs

This chapter aims to highlight the extra information required for 3D simulation as compared
to 2D. You should be familiar with the equivalent 2D models before reading this chapter.
This chapter describes the set of ATLAS products that extends 2D simulation models and
techniques and applies them to general non-planar 3D structures. The structural definition,
models and material parameters settings and solution techniques are similar to 2D. You
should be familiar with the simulation techniques described in Chapter 2 “Getting Started
with ATLAS” and the equivalent 2D product chapters before reading the sections that follow.
The products that form 3D device simulation in ATLAS are:
• DEVICE3D – silicon compound material and heterojunction simulation
• GIGA3D – non-isothermal simulation
• MIXEDMODE3D – mixed device-circuit simulation
• TFT3D – thin film transistor simulation
• QUANTUM3D – quantum effects simulation
• LUMINOUS3D – photodetection simulation
The 3D modules, THERMAL3D, are described in Chapter 17 “Thermal 3D: Thermal
Packaging Simulator”.
In a similar manner to the 2D products, GIGA3D, MIXEDMODE3D, LUMINOUS3D and
QUANTUM3D should be combined with both DEVICE3D or BLAZE3D depending on the
semiconductor materials used.

6.1.1 DEVICE3D
DEVICE3D provides semiconductor device in 3D. Its use is analogous to the 2D simulations in
S-PISCES and BLAZE. See Chapter 4 “S-Pisces: Silicon Based 2D Simulator” for more
information on S-PISCES. See Chapter 5 “Blaze: Compound Material 2D Simulator” for more
information on BLAZE.

6.1.2 GIGA3D
GIGA3D is an extension of DEVICE3D that accounts for lattice heat flow in 3D devices.
GIGA3D has all the functionality of GIGA (see Chapter 7 “Giga: Self-Heating Simulator”)
with a few exceptions. One exception is that additional syntax has been added to account for
the three dimensional nature of thermal contacts. You can specify the Z.MIN and Z.MAX
parameters on the THERMCONTACT statement to describe the extent of the contact in the Z
direction. Another exception is that there is no BLOCK method available in the 3D version.

6.1.3 TFT3D
TFT3D is an extension of DEVICE3D that allows you to simulate amorphous and
polycrystalline semiconductor materials in three dimensions. TFT3D is completely analogous
to the TFT simulator described in Chapter 14 “TFT: Thin-Film Transistor Simulator”. The
complete functionality of the TFT simulator is available in TFT3D for three dimensional
devices.

446 ATLAS User’s Manual

3D Device Simulation Programs 3D Device Simulator

6.1.4 MIXEDMODE3D
MIXEDMODE3D is an extension of DEVICE3D or BLAZE3D that allows you to simulate
physical devices embedded in lumped element circuits (SPICE circuits). MIXEDMODE3D is
completely analogous to the MIXEDMODE simulator, which is described in Chapter 12
“MixedMode: Mixed Circuit and Device Simulator”. The complete functionality of the
MIXEDMODE simulator is available in MIXEDMODE3D for three dimensional devices.

6.1.5 QUANTUM3D
QUANTUM3D is an extension of DEVICE3D or BLAZE3D, which allows you to simulate of the
effects of quantum confinement using the Quantum Transport Model (Quantum Moments
Model). QUANTUM3D is completely analogous to the QUANTUM model but applies to three
dimensional devices. See Chapter 13 “Quantum: Quantum Effect Simulator” for more
information on QUANTUM.

6.1.6 LUMINOUS3D
LUMINOUS3D is an extension of DEVICE3D that allows you to simulate photodetection in
three dimensions. LUMINOUS3D is analogous to the LUMINOUS simulator, which is described
in Chapter 10 “Luminous: Optoelectronic Simulator” with a few significant differences
described in Section 6.3.11 “LUMINOUS3D Models”.

447 ATLAS User’s Manual

3D Structure Generation 3D Device Simulator

6.2 3D Structure Generation

All 3D programs in ATLAS supports structures defined on 3D prismatic meshes. Structures
may have arbitrary geometries in two dimensions and consist of multiple slices in the third
dimension.
There are two methods for creating a 3D structure that can be used with ATLAS. One way is
through the command syntax of ATLAS. Another way is through an interface to DEVEDIT3D.
A direct interface from ATHENA to 3D ATLAS impossible. But DEVEDIT3D provides the
ability to read in 2D structures from ATHENA and extend them non-uniformly to create 3D
structures for ATLAS.
ATLAS Syntax For 3D Structure Generation
Mesh generation
Section 2.6.3 “Using The Command Language To Define A Structure” covers the generation
of 2D and 3D mesh structures using the ATLAS command language. The Z.MESH statement
and the NZ and THREE.D parameters of the MESH statement are required to extend a 2D mesh
into 3D.
Conventionally, slices are made perpendicular to the Z axis. The mesh is triangular in XY but
rectangular in XZ or YZ planes.
Region, Electrode, and Doping definition
Section 2.6.3 “Using The Command Language To Define A Structure” also covers the
definition of 2D regions, electrodes and doping profiles. To extend the regions into 3D, use
the Z.MIN and Z.MAX parameters. For example:
REGION NUM=2 MATERIAL=Silicon X.MIN=0 X.MAX=1 Y.MIN=0 Y.MAX=1
Z.MIN=0 Z.MAX=1
ELECTRODE NAME=gate X.MIN=0 X.MAX=1 Y.MIN=0 Y.MAX=1 Z.MIN=0
Z.MAX=1
DOPING GAUSS N.TYPE CONC=1E20 JUNC=0.2 Z.MIN=0.0 Z.MAX=1.0
For 2D regions or electrodes defined with the command language, geometry is limited to
rectangular shapes. Similarly, in 3D regions and electrodes are composed of rectangular
parallelopipeds.
DevEdit3D Interface
DEVEDIT3D is a graphical tool that allows you to draw 3D device structures and create 3D
meshes. It can also read 2D structures from ATHENA and extend them into 3D. These
structures can be saved from DEVEDIT3D as structure files for ATLAS. Also, save a
command file when using DEVEDIT3D. This file is used to recreate the 3D structure inside
DEVEDIT3D, which is important, since DEVEDIT3D doesn’t read in 3D structure files.
ATLAS can read structures generated by DEVEDIT3D using the command:
MESH INF=<filename>
The program is able to distinguish automatically between 2D and 3D meshes read in using
this command.

448 ATLAS User’s Manual

3D Structure Generation 3D Device Simulator

Defining Devices with Circular Masks

DEVEDIT3D makes a triangular mesh in the XY plane and uses Z plane slices. This means,
that normally the Y direction is vertically down into the substrate. But in the case of using
circular masks, you need to rotate the device.
With defining devices using circular masks in DEVEDIT3D, the XY plane should be the
surface of the device and the Z direction should be into the substrate.

449 ATLAS User’s Manual

Model And Material Parameter Selection in 3D 3D Device Simulator

6.3 Model And Material Parameter Selection in 3D

Models and material parameters are chosen in 3D in common with other 2D modules using
the MODELS, IMPACT, MATERIAL, MOBILITY, INTERFACE, and CONTACT statements. The
following models are available in 3D device simulation programs. All of these models are
documented in Chapter 3 “Physics” or in the 2D product chapters.

6.3.1 Mobility
• Table for 300K (CONMOB)
• Thomas’s model (ANALYTIC)
• Arora’s model (ARORA)
• Klaassen’s model (KLAASSEN)
• Lombardi’s model (CVT)
• Yamaguchi’s model (YAMA)
• Parallel field dependence (FLDMOB)
• Parallel field dependence with negative differential mobility (FLDMOB EVSATMOD=1)

6.3.2 Recombination
• Shockley Read Hall (SRH)
• Concentration dependent lifetime SRH (CONSRH)
• Klaassen’s concentration dependent lifetime SRH (KLASRH)
• Auger (AUGER)
• Klaassen’s concentration dependent Auger recombination model (KLAAUG)
• Optical recombination (OPTR)
• Bulk and interface traps (TRAP, INTTRAP)
• Continuous defect states (DEFECT)

6.3.3 Generation
• Selberherr’s impact ionization (IMPACT SELB)
• Crowell’s impact ionization (IMPACT CROWELL)
• Hot electron injection (HEI)
• Fowler Nordheim tunneling (FNORD)

6.3.4 Carrier Statistics

• Boltzmann (default)
• Fermi (FERMI)
• Band gap narrowing (BGN)
• Incomplete ionization (INCOMPLETE)
• Quantum mechanical effects (QUANTUM)
• Heterojunction thermionic field emission

450 ATLAS User’s Manual

Model And Material Parameter Selection in 3D 3D Device Simulator

6.3.5 Boundary Conditions

• Ohmic and Schottky
• Current boundary conditions
• Lumped element boundary conditions
• Distributed contact resistance

6.3.6 Optical
• Photogeneration with ray tracing (LUMINOUS3D).

6.3.7 Single Event Upset Simulation

• Single event upset simulation.
All these models, with the exception of SINGLEEVENTUPSET, are documented in the Chapter
3 “Physics” or in the 2D product chapters of this manual.

6.3.8 Boundary Conditions in 3D

External Passive Elements
You can attach external lumped resistors, capacitors and inductors to any contact. The syntax
is the same as for the 2D products, which is:
CONTACT NAME=drain RES=1e3 CAP=1e-12 L=1e-6
You can also apply distributed resistances to contacts. The algorithm used for estimating
contact area for 3D distributed contact resistance multiplies the contact perimeter in a given Z
plane by the displacement in the Z direction. This algorithm will only work properly for
planar contacts that do not vary in the Z direction. They may however abruptly terminate or
start in the Z direction.
The units of lumped external passive elements are ohms for resistors, Farads for capacitors
and Henrys for inductors. Distributed contact resistance is defined in ohms.cm3.
Thermal Contacts for GIGA3D
Thermal contacts for non-isothermal simulation in GIGA3D are defined in an analogous
manner to the 2D thermal contacts in GIGA. The Z.MIN and Z.MAX parameters are used to
define the extent of the thermal contact in the Z plane. The units of the thermal resistance
parameter ALPHA are scaled in 3D to W/(cm.K). For more information about GIGA3D, see
Chapter 7 “Giga: Self-Heating Simulator”.

6.3.9 TFT3D Models

Models for simulating thin-film transistors made from amorphous or polycrystalline
semiconductors are supported in TFT3D. The definition of the continuous defect states in the
bandgap is performed using the same parameters as in 2D simulations with TFT. The models
for continuous defect (or trap) densities are documented in Chapter 14 “TFT: Thin-Film
Transistor Simulator”.

451 ATLAS User’s Manual

Model And Material Parameter Selection in 3D 3D Device Simulator

6.3.10 QUANTUM3D Models

Models for simulating quantum effects semiconductors are supported in QUANTUM3D. The
definition of the quantum moments solver is the same as in 2D simulations with QUANTUM.
The models for simulating quantum effects and the parameters to control the model are shown
in Chapter 13 “Quantum: Quantum Effect Simulator”.

6.3.11 LUMINOUS3D Models

Many of the models for simulating photodetection in LUMINOUS3D are similar to those for
simulating photodetection in LUMINOUS. For more information about LUMINOUS, see Chapter
10 “Luminous: Optoelectronic Simulator”. This section, however, shows several important
differences between LUMINOUS and LUMINOUS3D.

452 ATLAS User’s Manual

Numerical Methods for 3D 3D Device Simulator

6.4 Numerical Methods for 3D

6.4.1 DC Solutions
There are several differences between the default numerical methods applied in 2D ATLAS
and those applied in 3D ATLAS. For example, with respect to non-linear iteration strategies,
the current version of the 3D simulator does not support the BLOCK method. The NEWTON and
GUMMEL iteration strategies are supported for 3D simulations, whereas NEWTON, GUMMEL and
BLOCK are all supported for 2D simulations.
In solving the linear subproblem, the default approach in 3D is to use an iterative solver. This
is believed to be the most efficient method for general 3D problems. In 2D, the direct solver is
used by default. You may find it desirable to use direct methods in 3D problems due to
improved convergence of computational efficiency. You can select the direct method by
specifying DIRECT in the METHOD statement. Also, in 3D there are two linear iterative solution
methods available. The defaults are ILUCGS (incomplete lower-upper decomposition
conjugate gradient system) and BICGST (bi-conjugate gradient stabilized). Historically, tests
have shown that the current implementation of ILUCGS is slightly more stable than BICGST
and is the default iterative solver in 3D. You can define the BICGST solver by specifying
BICGST in the METHOD statement.

Note: Iterative solvers are recommended for large problems, typically greater than 5000 node points, due to lower
solution times and memory usage.

6.4.2 Transient Solutions

In transient mode, a semi-implicit scheme is used in addition to the default TR-BDF
algorithm [3]. This algorithm is recommended for complex simulations such as Single Event
Upset. To select this method, use:
METHOD HALFIMPLICIT
It is available for solving the drift-diffusion equations and if LAT.TEMP is specified on the
MODELS statement. If a carrier temperature solution is requested (HCTE.EL or HCTE.HO or
both), then HALFIMPLICIT cannot be used.

6.4.3 Obtaining Solutions In 3D

ATLAS3D programs can perform DC and transient analysis in an equivalent manner to 2D.
The SOLVE statement is used to define the solution procedure. The syntax used is described
in Section 2.9 “Obtaining Solutions”.

453 ATLAS User’s Manual

Numerical Methods for 3D 3D Device Simulator

6.4.4 Interpreting the Results From 3D

The log files produced by 3D ATLAS can be plotted in TONYPLOT exactly as those that result
from S-PISCES or BLAZE. The only difference is the units of the currents produced. Log files
from 3D simulations save current in Amperes, whereas the 2D simulations use Amperes/
micron.
The solution files produced by 3D ATLAS should be plotted using TONYPLOT3D. These files
cannot be read directly into the 2D TONYPLOT program. TONYPLOT3D contains features that
allows you to make slices of the 3D structure, which can be plotted in 2D TONYPLOT. For
more information about TONYPLOT3D, see the TONYPLOT3D USER ’S MANUAL.

6.4.5 More Information

Many examples using 3D ATLAS have been installed on your distribution tape or CD. You
can find more information about 3D ATLAS by reading the text associated with each
example. You can also find find some more information at www.silvaco.com.

454 ATLAS User’s Manual

Chapter 7
Giga: Self-Heating Simulator
Overview Giga: Self-Heating Simulator

7.1 Overview
GIGA extends ATLAS to account for lattice heat flow and general thermal environments.
GIGA implements Wachutka’s thermodynamically rigorous model of lattice heating [294],
which accounts for Joule heating, heating, and cooling due to carrier generation and
recombination, and the Peltier and Thomson effects. GIGA accounts for the dependence of
material and transport parameters on the lattice temperature. GIGA also supports the
specification of general thermal environments using a combination of realistic heat-sink
structures, thermal impedances, and specified ambient temperatures. GIGA works with both
S-PISCES and BLAZE and with both the drift-diffusion and energy balance transport models.
See the “Drift-Diffusion Transport Model” on page 98 and the “Energy Balance Transport
Model” on page 100 for more information on these models.
Before continuing with this chapter, you should be familiar with ATLAS. If not, read Chapter
2 “Getting Started with ATLAS”, along with Chapter 4 “S-Pisces: Silicon Based 2D
Simulator” or Chapter 5 “Blaze: Compound Material 2D Simulator”.

7.1.1 Applications
A major application of GIGA is the simulation of high-power structures including bipolar,
MOS, IGBT, and thyristor devices. Another important application is the simulation of
electrostatic discharge (ESD) protection devices. Thermal effects are also important in SOI
device operation due to the low thermal conductivity of the buried oxide, and in devices
fabricated in III-V material systems due to the relatively low thermal conductivity of these
materials.
Recent studies have demonstrated that accounting self-consistently for lattice heating is
necessary for accurate simulation of bipolar VLSI devices. This is due to the sensitive
temperature dependence of the carrier injection process. Since bipolar devices are key
components of CMOS technologies and many devices can be impacted by parasitic bipolar
effects, the applications of GIGA are very general.
For other more information on GIGA’s application, see Section 7.3 “Applications of GIGA”.

7.1.2 Numerics
GIGA supplies numerical techniques that provide efficient and robust solution of the
complicated systems of equations that result when lattice heating is accounted for. These
numerical techniques include fully-coupled and block iteration methods. When GIGA is used
with the energy balance equations, the result is a “six equation solver”, which defines the
state-of-the-art for general purpose device simulation.
For more information on numerical techniques, see Chapter 20 “Numerical Techniques”.

456 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

7.2 Physical Models

7.2.1 The Lattice Heat Flow Equation
GIGA adds the heat flow equation to the primary equations that are solved by ATLAS. The
heat flow equation has the form:
T L
C --------- =   T L  + H 7-1
t

where:
• C is the heat capacitance per unit volume.
•  is the thermal conductivity.
• H is the heat generation
• TL is the local lattice temperature.

The heat capacitance can be expressed as C = Cp, where Cp is the specific heat and  is the
density of the material.
Specifying the LAT.TEMP parameter in the MODELS statement includes the lattice heat flow
equation in ATLAS simulations.
GIGA supports different combinations of models. For example, if the HCTE and LAT.TEMP
parameters are specified in the MODELS statement and both particle continuity equations are
solved, all six equations are solved. If HCTE.EL is specified instead of HCTE, only five
equations are solved and the hole temperature Tp is set equal to lattice temperature TL.

7.2.2 Specifying Thermal Conductivity

Standard Model
The value of thermal conductivity, k, for each region should be specified in the MATERIAL
statement. Because thermal conductivity is generally temperature dependent, the following
four models are available.

Table 7-1 Standard Thermal Conductivity Models

Equation Units MATERIAL Flag

k  T  = TC.CONST  W  cm · K  TCON.CONST

TC.NPOW  W  cm · K  TCON.POWER
k  T  =  TC.CONST    T L  300 

2
k  T  = 1   TC.A +  TC.B  T L +  TC.C  T L   W  cm · K  TCON.POLYN

k  T  =  TC.E    TL – TC.D   W  cm · K  TCON.RECIP

457 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

The TC.CONST, TC.NPOW, TC.A, TC.B, TC.C, TC.D and TC.E parameters are all user-
specifiable in the MATERIAL statement. As described in Table 7-1, the choice of models is also
user-specified in the MATERIAL statement.
To clarify model selection and parameter value definition, the following syntax examples are
given.
Enabling TCON.POWER:
material mat=Diamond TCON.POWER TC.NPOW=2 TC.CONST=1.4
Enabling TCON.POLYN:
material mat=GaN TCON.POLYN TC.A=1 TC.B=2 TC.C=4
Enabling TCON.RECIP:
material mat=silicon TCON.RECIP TC.B=2 TC.D=4
The thermal model and parameter selection can also be completed through REGION
statements as follows:
MATERIAL REGION=5 TC.A=<n> TC.B=<n> TC.C=<n>
MATERIAL REGION=6 TC.A=<n> TC.B=<n> TC.C=<n>
It is also advised that you include PRINT on the MODELS statement to verify model selection
and parameter values used.
Compositionally Dependent Thermal Conductivity (TCON.COMP) [206]
The key material dependent parameters in Equation 7-1 are the thermal conductivity and the
heat capacity. In general, both thermal conductivity and heat capacity are both composition
and temperature dependent. In later sections, we will show how you have a great deal of
flexibility in specifying these model dependencies.
For most materials, we provide reasonable default values for these material parameters,
which you can examine by specifying PRINT on the MODELS statement. These models allow
user specification but are not material composition dependent. As such, we have provided
default models for some of the more popular materials that are both composition and
temperature dependent. These models described here apply to the following materials: Si, Ge,
GaAs, AlAs, InAs, InP, GaP, SiGe, AlGaAs, InGaAs, InAlAs, InAsP, GaAsP and InGaP.
The default built-in model derives temperature dependency in a manner very similar to the
user specifiable models discussed later in this section.
The composition depencencies of the ternary compounds and the binary, SiGe, are derived
from interpolation from the binary compounds (or the values for Si and Ge in the case of
SiGe).
The basic model for thermal conductivity is given by:
TL 
  T L  = K 300   ---------- 7-2
300

where (TL) is the temperature dependent thermal conductivity, TL is the lattice temperature
and K300 and  are material dependent parameters.

458 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

Table 7-2 shows the default values for K300 and  for Si, Ge and the binary III-V compounds.

Table 7-2 Default Parameter Values for Thermal Conductivity

Material K300(W/Kcm) 

Si 1.48 -1.65
Ge 0.60 -1.25
GaAs 0.46 -1.25
AlAs 0.80 -1.37
InAs 0.273 -1.1
InP 0.68 -1.4
GaP 0.77 -1.4

The parameters K300 and  are interpolated as a function of composition fraction x using
Equations 7-3 and 7-4.
AB 1
K 300 = --------------------------------------------------------------- 7-3
1 – x 1 – x x
x - + -------------------
 ------------ + ----------- 
K A B C 
300 K 300

AB A B
 =  1 – x  + x 7-4

The C parameter in Equation 7-3 is a bowing factor used to account for the non-linear aspects
of the variation of thermal conductivity with composition. Table 7-3 shows the default values
of the bowing factor, C, for the various ternary compounds and SiGe.

Table 7-3 Default Bowing Parameter Values for Thermal Conductivity

Material C(W/K cm)

SiGe 0.028
AlGaAs 0.033
InGaAs 0.014
InAlAs 0.033
InAsP 0.033
GaAsP 0.014
InGaP 0.014

459 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

C-Interpreter Defined Thermal Conductivity

You can use the C-INTERPRETER to define the thermal conductivity, , as a function of the
lattice temperature, position, doping and fraction composition. This is defined using the
syntax:
MATERIAL REGION=<n> F.TCOND=<filename>
where the <filename> parameter is an ASCII file containing the C-INTERPRETER function.
For more information about the C-INTERPRETER, see Appendix A “C-Interpreter Functions”.
Anisotropic Thermal Conductivity
The flux term in Equation 7-1 models the thermal conductivity  as being isotropic by default.
You can specify an anisotropic thermal conductivity aniso. In GIGA2D, the anisotropic
value is applied in the Y direction. In GIGA3D, it is applied in the Z direction by default. In
this case, the thermal conductivity tensor is

1 0 0
 = 0 1 0 7-5
0 0 2

You can also change the anisotropic direction to be the Y direction instead of the Z direction
by specifying the YDIR.ANISO parameter or by specifying ^ZDIR.ANISO.
The thermal conductivity is temperature dependent, and the anisotropic thermal conductivity
must have a model selected for its dependence on temperature. To do this, select
TANI.CONST, TANI.POWER, TANI.POLYNOM or TANI.RECIP. These are analagous to
TCON.CONST, TCON.POWER, TCON.POLYNOM and TCON.RECIP respectively. You can specify
different temperature dependency models for the (T) and aniso(T). The parameters for the
chosen model for aniso(T) are specified by using the relevant parameters from A.TC.CONST,
A.TC.A, A.TC.B, A.TC.C, A.TC.D, A.TC.E and A.TC.NPOW. These are equivalent to the
isotropic parameters but with a prefix of A. to distinguish them. No built-in models are
available for the anisotropic component of thermal conductivity. To give the anisotropic
component the same temperature dependence as for the built-in model (Equation 7-2), specify
the TANI.POWER.
As in the case of anisotropic permittivity, the discretization of the flux term is modified. For
the simple cases of a rectangular mesh and a diagonal thermal conductivity tensor, the
discretization chooses the value of thermal conductivity appropriate to the direction.
If the coordinate system where the thermal conductivity tensor is diagonal and the ATLAS
coordinate system are non-coincident, then the coordinate transformation can be specified as
a set of vectors
X = (X.X, X.Y, X.Z)
Y = (Y.X, Y.Y, Y.Z)
Z = (Z.X, Z.Y, Z.Z)
where the vector components can be specified on the MATERIAL statement. The default is that
X = (1.0, 0.0, 0.0)
Y = (0.0, 1.0, 0.0)

460 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

Z = (0.0, 0.0, 1.0)

and you should specify all necessary components. You do not need to normalize the vectors to
unit length. ATLAS will normalize them and transform the thermal conductivity tensor
according to the usual rules. If you specify any component of X, Y or Z, then a more complete
form of discretization will be used.
This is similar to the complete discretization carried out for the case of a generally anisotropic
dielectric permittivity described in Section 3.12 “Anisotropic Relative Dielectric
Permittivity”.
To enable the more complete form of discretization, use the TC.FULL.ANISO parameter on
the MATERIAL statement.
Table 7-4 shows the parameters used in Anisotropic Thermal Conductivity.

Table 7-4 MATERIAL Statement Parameters

Parameter Type Default Units

A.TC.CONST Real 0.0 (W/cm/K)

A.TC.A Real 0.0 cm K /W
A.TC.B Real 0.0 cm /W
A.TC.C Real 0.0 cm / W /K
A.TC.D Real 0.0 K
A.TC.E Real 0.0 W/cm
A.TC.NPOW Real 0.0 -
TANI.CONST Logical False
TANI.POWER Logical False
TANI.POLYNOM Logical False
TANI.RECIP Logical False
TC.FULL.ANISO Logical False
YDIR.ANISO Logical False
ZDIR.ANISO Logical True
X.X Real 1.0
X.Y Real 0.0
X.Z Real 0.0
Y.X Real 0.0
Y.Y Real 1.0

461 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

Table 7-4 MATERIAL Statement Parameters

Parameter Type Default Units

Y.Z Real 0.0

Z.X Real 0.0
Z.Y Real 0.0
Z.Z Real 1.0

7.2.3 Specifying Heat Capacity

Standard Model
For transient calculations, specify heat capacities for every region in the structure. These are
also functions of the lattice temperature and are modeled as:
HC D
C = HC A + HC  B T L + HC  C T L 2 + -------------- 3
-  J  cm  K  7-6
2
TL

Default values of HC.A, HC.B, HC.C, and HC.D are provided for common materials. You can
specify these values in the MATERIAL statement.
The following statements will be used to specify the temperature dependent heat capacities of
the regions previously defined.
MATERIAL REGION=5 HC.A=<n> HC.B=<n> HC.C=<n> HC.D=<n>
MATERIAL REGION=6 HC.A=<n> HC.B=<n> HC.C=<n> HC.D=<n>

Table 7-5 User Specifiable Parameters for Equation 7-6

Statement Parameter Units

MATERIAL HC.A J/cm3/K

MATERIAL HC.B J/cm3K2
MATERIAL HC.C J/cm3K3
MATERIAL HC.D JK/cm3

To select the standard model for heat capacity, specify HC.STD on the MATERIAL statement.

462 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

Compositionally Dependent Heat Capacity [206]

The temperature dependence of heat capacity can be expressed by:

TL 
 ---------- – 1
 300
C  T L  =  C 300 + C 1 ------------------------------------
- 7-7
TL  C
 ---------- + ------------ 1
 300 C 300

where C(TL) is the temperature dependent heat capacity,  is the mass density, C300, C1 and 
are material dependent parameters. Table 7-6 shows the default values of , C300, C1 and 
for the binary compounds and Si and Ge.

Table 7-6 Default Parameter Values for Heat Capacity

Material  C300  C1  
(g/cm3) (J/K/kg) (J/K/kg)

Si 2.33 711 255 1.85

Ge 5.327 360 130 1.3
GaAs 5.32 322 50 1.6
AlAs 3.76 441 50 1.2
InAs 5.667 394 50 1.95
InP 4.81 410 50 2.05
GaP 4.138 519 50 2.6

For the ternary compounds and SiGe, the parameters , C300, C1 and  are interpolated versus
composition using a simple linear form as in Equation 7-8.
P ab =  1 – x P a + xP b 7-8

When the lattice heat flow equation is solved, the global device temperature parameter is the
maximum temperature at any mesh point in the device.
To select the compositionally dependent heat capacity, specify HC.COMP on the MATERIAL
statement.
C-Interpreter Defined Thermal Capacity
You can use the C-INTERPRETER to define the thermal capacity, TCAP, as a function of the
lattice temperature, position, doping and fraction composition. This is defined using the
syntax:
MATERIAL REGION=<n.region> F.TCAP=<filename>
where the <filename> parameter is an ASCII file containing the C-INTERPRETER function.
For more information about the C-INTERPRETER, see Appendix A “C-Interpreter Functions”.

463 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

You can override the default values for , C300, C1, and  by specifying HC.RHO, HC.C300,
HC.C1, HC.BETA respectively on the MATERIAL statement.

7.2.4 Specifying Heat Sink Layers

You can define regions to only include thermal calculations. These regions will typically
consist of layers associated with heat sinks. They are defined using the REGION statement.
Even though in reality the heat sink materials are typically metal conductors, it is easier to
specify these layers with the material type, INSULATOR. This is because as insulators the
program will only solve heat flow and not attempt to solve current continuity in these layers.
The region number is subsequently used as an identifier when thermal conductivities and heat
capacities are assigned to these regions.
The following statements specify two layers of a heat sink for inclusion in the thermal
calculation.
REGION NUM=5 Y.MIN=0.5 Y.MAX=2.0 INSULATOR
REGION NUM=6 Y.MIN=2.0 Y.MAX=3.0 INSULATOR

7.2.5 Non-Isothermal Models

Effective Density Of States
When lattice heating is specified with the drift-diffusion transport model, the effective density
of states for electrons and holes are modeled as functions of the local lattice temperature as
defined by Equations 3-31 and 3-32.
When lattice heating is specified with the energy balance model, the effective densities of
states are modeled as functions of the local carrier temperatures, Tn and Tp, as defined by
Equations 3-148 and 3-149.
Non-isothermal Current Densities
When GIGA is used, the electron and hole current densities are modified to account for
spatially varying lattice temperatures:

J n = – q  n n   n + P n T L  7-9

J p = – q  p p   p + P p T L  7-10

where:
Pn and Pp are the absolute thermoelectric powers for electrons and holes. Pn and Pp are
modeled as follows:
k 5 N
P n = – -----B  --- + ln -----c- + KSN + z n 7-11
Q 2 n 

kB 5 N
P p = -----  --- + ln -----v- + KSP + z p 7-12
Q 2 p 

464 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

Here, kB is Boltzmann’s constant. KSN and KSP are the exponents in the power law
relationship between relaxation time (or mobility) and carrier temperature. They are also used
in the energy balance model. Their values typically range between -1 and 2 in Silicon and
they are set on the MODELS statement.
You can also model them using the C-Interpreter functions F.KSN and F.KSP if a more
complicated dependence is required.
The quantities n and p are the phonon drag contribution to the thermopower. These are only
significant in low doped material and at low temperatures. ATLAS has a built in model for
this phonon drag contribution, which you enable by specifying the PHONONDRAG parameter on
the MODELS statement. The built-in model is

kB T L PDEXP · N
z n =  ----- PDA · N  ---------- 7-13
 Q  300

for electrons and

kB T L PDEXP · P
z p =  ----- PDA · P  ---------- 7-14
 Q  300

for holes. A theoretically derived value for PDEXP.N and PDEXP.P is -7/2 [109] but
experimentally obtained results generally give a value of around -5/2 [84]. The values of
PDA.N and PDA.P depend on doping level and can also depend on sample size. Therefore, it
is recommended that you choose values to fit to your sample. Alternatively, empirical data or
more complicated formulae can be incorporated by using the C-Interpreter functions F.PDN
and F.PDP on the MATERIAL or MODELS statements [23], [101], [284].
The thermopower can be thought of as having 3 components. The first is the derivative of the
Fermi Potential with respect to temperature. For Maxwell-Boltzmann statistics, this is
k 3
– -----B  --- + ln Nc
------- 7-15
Q 2 n

for electrons and

k
-----B  3 -------
--- + ln Nv 7-16
Q 2 p

for holes.
ATLAS incorporates this effect indirectly through the boundary conditions. It does not
contribute directly to the temperature gradient term in the expressions for current, see
Equations 7-9 and 7-10.

465 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

The second term is due to carrier scattering and is

k
– -----B  1 + KSN  7-17
Q

for electrons and

k
-----B  1 + KSP  7-18
Q

for holes. The third term is the phonon drag contribution -n and p, and the second and third
terms are included directly into the temperature gradient term in the expressions for current.
Seebeck Effect
This is the phenomenon in which an open circuited voltage is generated across a
semiconductor having a temperature gradient. To observe this in practice, you need two
materials each with a different thermopower. In simulation, you can directly observe the
Seebeck effect because you can measure the open circuit voltage between two contacts at
different temperatures. Current boundary conditions are applied to one contact and solved for
zero current.
This will result in a voltage difference between the two contacts.
The Seebeck coefficient is given as the ratio of the Voltage difference between the contacts to
the temperature difference, and the most convenient units for expressing it are mV/K. To
theoretically calculate the Voltage difference generated by the temperature gradient, integrate
the thermopower as a function of temperature between the temperatures on the contact
T1
 n N D P n  T  +  p N A P p  T  dT

V 1 – V 0 = – ----------------------------------------------------------------------
n ND + p NA
- 7-19
T0

for a compensated material, where T0 and T1 are the temperatures on the contacts and V0 and
V1 are the voltages. The effective Seebeck coefficient for the device is obtained by dividing
the voltage difference so obtained by the temperature difference (T1 - T0).
Table 7-7 shows the default values for Equations 7-11 and 7-12.

Table 7-7 User-Specifiable Parameters for Equations 7-11 and 7-12

Statement Parameter Default Units

MODELS F.KSN

MODELS F.KSP

MODELS KSN -1 None

MODELS KSP -1 None
MODELS PHONONDRAG False

466 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

Table 7-7 User-Specifiable Parameters for Equations 7-11 and 7-12

Statement Parameter Default Units

MODELS F.PDN

MODELS F.PDP

MATERIAL F.PDN

MATERIAL F.PDP

MATERIAL PDA.N 0.2

MATERIAL PDA.P 0.2
MATERIAL PDEXP.N -2.5
MATERIAL PDEXP.P -2.5

7.2.6 Heat Generation

When carrier transport is handled in the drift-diffusion approximation the heat generation
term, H, used in Equation 7-1 has the form:
2 2
Jn Jp
H = ----------- - – T L  J n P n  – T L  J p P p 
- + ----------- 7-20
q n n q p p
 n  p
+ q  R – G  T L  ---------  –  n – T L  ---------  +  p
T n p T n p
 n 
– T L  --------- + P n divJ n – T L  --------p- + P p divJ p
 T  n p  T  n p

In the steady-state case, the current divergence can be replaced with the net recombination.
Equation 7-20 then simplifies to:

2 2
Jn Jp
H = ----------- - + q  R – G    p –  n + T L  P p – P n   – T L  J n P n + J p P p 
- + ----------- 7-21
q n n q p p

where:
2 2
Jn Jp
----------- - is the Joule heating term,
- + -----------
q n n q p p

q  R – G    p –  n + T L  P p – P n   is the recombination and generation heating and cooling

term,

– T L  J n P n + J p P p  accounts for the Peltier and Joule-Thomson effects.

467 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

A simple and intuitive form of H that has been widely used in the past is:

H =  Jn + Jp   E 7-22

GIGA can use either Equations 7-21 or 7-22 for steady-state calculations. By default, Equation
7-22 is used. Equation 7-21 is used if the HEAT.FULL parameter is specified in the MODELS
statement. To enable/disable the individual terms of Equation 7-21, use the JOULE.HEAT,
GR.HEAT, and PT.HEAT parameters of the MODEL statement.
If the general expression shown in Equation 7-20 is used for the non-stationary case, the
 
derivatives  --------n- and  --------p- are evaluated for the case of an idealized non-degenerate
T L n p T L n p
semiconductor and complete ionization.
The heat generation term, H, is always set equal to 0 in insulators. For semiconductors, two
numerical approaches to discretizing the Joule heating term have been implemented. The first
method approximates the term as a special type of flux term. The second method
approximates the term as a source term. These may give different results in some cases. The
first method is the default, while the second method may be enabled by specifying
METHOD ^FLUX.JOULE
to explicitly clear the inclusion of Joule heating as a flux term. This option is included
primarily for compatibility with VICTORY DEVICE.
2
For conductors H = ---------------
V 
-.

When electron and hole transport are modeled in the energy balance approximation (by
specifying HCTE on the MODELS statement), the following expression is used for H:
H = W n + W p + E g U 7-23

where U, Wn, and Wp are defined by Equations 3-150 through 3-152.

If the energy balance model is enabled for only electrons or only holes, then a hybrid of
Equations 7-23 and 7-21 or 7-22 is used. For example if the energy balance equation is solved
for electrons, but not for holes, then H is evaluated as follows if HEAT.FULL is specified.
2
Jp
H = W n + E g U + ------------ – T L J p P p 7-24
q p p

A simpler form for the heating will be used if HEAT.FULL is not specified.

H = Wn + Eg U + Jp  E 7-25

If the energy balance equation is solved for holes, but drift-diffusion is solved for holes, then
the heating term is analagous to Equations 7-24 and 7-25 with hole terms swapped for
electron terms and vice versa.

468 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

The first terms of Wn and Wp are output to structure files as Joule heating. The last term in
Equation 7-24 is output as Peltier Thomson Heat power. The remaining terms are output as
recombination heat power.

7.2.7 Thermal Boundary Conditions

At least one thermal boundary condition must be specified when the lattice heat flow equation
is solved. The thermal boundary conditions used have the following general form:
u
  J tot  s  =   T L – T ext  7-26

where  is either 0 or 1, Jutot is the total energy flux and s is the unit external normal of the
boundary.
The projection of the energy flux onto s is:

 J u  s = –  T
--------L- +  T L P n +  n  J n  s +  T L P p +  p   J p  s  7-27
 tot  n

When  = 0, Equation 7-26 specifies a Dirichlet (fixed temperature) boundary condition:

T L = TEMPER 7-28

where you define TEMPER in the THERMCONTACT statement as shown in the next section. You
can specify dirichlet boundary conditions for an external boundary (which may coincide with
an electrode) or for an electrode inside the device.
When  = 1, Equation 7-26 takes the form:

 J u  s = -------
1
-  T – TEMPER  7-29
 tot  R th L

where the thermal resistance, Rth, is given by:

1
R th = ------------------ 7-30
ALPHA

and ALPHA is user-definable on the THERMCONTACT statement.

Specifying Thermal Boundary Conditions

Setting thermal boundary conditions is similar to setting electrical boundary conditions. The
THERMCONTACT statement is used to specify the position of the thermal contact and any
optional properties of the contact. You can place thermal contacts anywhere in the device
(including sidewalls). Equation 7-29 is used if a value is specified for . Otherwise, Equation
7-28 is used.
The following command specifies that thermal contact number 1 is located between x=0 m
and x=2 m at y=0 m, and that the temperature at the contact is 300K.
THERMCONTACT NUM=1 X.MIN=0 X.MAX=2 Y.MIN=0 Y.MAX=0 TEMP=300

469 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

You can use simpler statement if the coordinates of a thermal contact coincide with the
coordinates of an electrical contact. In this case, it is permissible to specify the location of the
thermal contact by referring to the electrode number of the electrical contact. For example,
the statement:
THERMCONTACT NUM=1 ELEC.NUM=3 TEMP=400
specifies that thermal contact number 1 is located in the same position as electrode number 3
and that the contact temperature is 400K.
Specifying the BOUNDARY parameter gives you flexibility in applying thermal boundary
conditions. This parameter is set by default and means that the thermal boundary condition
will only be set on the outside surface of the thermal contact, where it forms the exterior of
the device. If the parameter is cleared by specifying ^BOUNDARY in the THERMCONTACT
statement, the boundary conditions will be applied to the interior of the thermal contact and to
the part of the surface of the thermal contact that forms an interface with the interior of the
device. In a 2D model, you may have a thermal contact that is internal to the device and in
this case you would need to specify ^BOUNDARY. Otherwise, the contact will be ignored.
For example:
THERMCONTACT ELEC.NUM=1 ^BOUNDARY TEMPER=450 ALPHA=2.5
where the ELECTRODE extends into the device and will set the interior points of the
thermalcontact to 450K and will apply the flux boundary condition on all faces of the thermal
contact that interface with the interior of the device.

Table 7-8 User Specifiable Parameters for Equations 7-28 and 7-29

Statement Parameter Units Default

THERMCONTACT ALPHA W/(cm2K) 

THERMCONTACT BOUNDARY - True

THERMCONTACT TEMPER K 300

Representing a thermal environment, in terms of thermal impedances, leads to efficient

solutions. But, thermal impedance representations are typically only approximations.
Detailed thermal modeling (e.g., the effect of heat sink design changes) typically requires
using detailed modeling of thermal regions with specified external ambient temperatures.

Note: You can’t alter the value of a thermal resistor within a sequence of SOLVE statements. Rerun the input file
whenever a thermal resistor is changed.

470 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

7.2.8 Temperature Dependent Material Parameters

GIGA automatically uses the built-in temperature dependence of the physical models that are
specified. When lattice heating is specified, temperature dependent values are evaluated
locally at each point in the device. When lattice heating is not solved, the models only provide
global temperature dependence. In other words, all points in the device are assumed to be at
the specified temperature.
The non-isothermal energy balance model uses the same carrier temperature dependencies of
the mobility and impact ionization models as in the pure energy balance case, but with
coefficients that depend on the local lattice temperature. Impact ionization coefficients
depend on lattice temperature. Almost all other models and coefficients depend on lattice
temperature.
When lattice heating is used, there is no point in specifying models that do not include
temperature dependence. For mobilities, don’t specify CONMOB. Instead, specify ANALYTIC or
ARORA. For impact ionization coefficients, specify the SELBERHERR model.
GIGA can account for the temperature dependence of the minority carrier lifetimes for
electrons or holes or both. The LT.TAUN (electrons) and LT.TAUP (holes) parameters of the
MATERIAL statement are used to select this model. This model is turned on whenever the
value of LT.TAUN or LT.TAUP is greater than 0, which is the default.
The temperature dependence of electron and hole lifetimes in the SRH recombination model
have the forms:

T L LT · TAUN
 n = TAUN0  ---------- 7-31
300

T L LT · TAUP
 p = TAUP0  ---------- 7-32
 300

Table 7-9 User-Specifiable Parameters for Equations 7-31 and 7-32

Statement Parameter Default Units

MATERIAL TAUN0 110-7 s

MATERIAL TAUP0 110-7 s

MATERIAL LT.TAUN 0
MATERIAL LT.TAUP 0

See Equations 3-335 and 3-336 for information regarding concentration dependent lifetimes.
There is an alternative model for the temperature dependence of the recombination lifetimes.
This is

 n  T L  = TAUN0 exp  TCOEFF · N  T L  300 – 1.0   7-33

471 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

 p  T L  = TAUP0 exp  TCOEFF · P  T L  300 – 1.0   7-34

To enable this model, use SRH.EXPTEMP on the MODELS statement along with SRH or CONSRH.
You can set the parameters TCOEFF.N and TCOEFF.P on the MATERIAL statement (see Table
7-10).

Table 7-10 Material Parameters for Exponential Temperature Dependence Model

Statement Parameter Type Default

MATERIAL TCOEFF.N Real 2.55

MATERIAL TCOEFF.P Real 2.55

7.2.9 Phase Change Materials (PCMs)

Phase Change Materials, PCMs, are materials that display a change in resistivity with
temperature that can be associated with a phase change in the material between crystalline
and amorphous states. Typically, the material exhibits a low resistivity in the crystalline state
and a relatively higher resistivity in the amorphous state. Once a state (crystalline/amorphous
or low/high resistivity) is obtained, it can be "frozen" into the material by rapidly lowering the
temperature. This characteristic enables PCM materials to exhibit hysterisis under such
conditions and makes them suitable to store information. Although this model was designed
for PCM materials, it is also applicable to other materials exhibiting similar characteristics.
To enable the PCM model, specify PCM in the MODEL statement. When you enable this model,
the resistivity as a function of lattice temperature will be described (see Figure 7-1).

PCM.ARHO
Resistivity

A
.AE TAU
(cm)

M
PC CM.A
P
AU
. CT A
PCM.CRHO M E
PC M.C
PC

PCM.CTC PCM.ATC
Temperature (K)
Figure 7-1: PCM Temperature Dependent Resistivity Model
Here at low temperatures, less than or equal to PCM.CTC. The resistivity is constant and equal
to the value specified by PCM.CRHO. At high temperatures above PCM.ATC, the resistivity is
constant and equal to the value of PCM.ARHO. Between the temperatures PCM.CTC and
PCM.ATC, the resistivity varies either linearly or logarithmically with temperature depending
on the value of the PCM.LOG parameter of the MODEL statement.

472 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

The resistivity as a function of temperature is as given in Equations 7-35 through 7-37 for the
case where PCM.LOG is not true.

 = PCM CRHO  T < PCM · CTC  7-35

 = PCM ARHO  T > PCM · ATC  7-36

 = -------------------------------------------------------------------------------------------------------------
T – PCM · CTC   PCM · ARHO – PCM · CRHO 
- + PCM · CRHO  PCM · CTC < T < PCM · ATC  7-37
 PCM · ATC – PCM · CTC 

For the case where PCM.LOG is true, the resistivity between the critical temperatures is given
by Equations 7-38 and 7-39.
x
 = 10 7-38

 T – PCM · CTC 
-  log  PCM ARHO  – log  PCM CRHO  
x =  -------------------------------------------------------- 7-39
· ·
 PCM · ATC – PCM · CTC 10 10

+ log  PCM · CRHO   PCM · CTC < T < PCM · ATC 

Hysterisis is built into the model in two ways. In static simulations, the resistivity is updated
according to Figure 7-1 only for increasing temperature. This is suitable since phase changes
in memory devices are generally made during increasing temperature ramps.
For time domain simulations, the hysterisis is described using the Johnson-Mehl-Avrami
model as given in Equations 7-40 and 7-41. Here, t is the time at the last time step, dt is the
size of the time step, T is the lattice temperature, ̂  T  is the temperature dependent static
resistivity given by Equations 7-35 through 7-39, (t,T) is the resistivity at the last time step,
and (t+d+T) is the resistivity at the new time step.
Equation 7-40 applies to increasing temperature. Equation 7-41 applies to decreasing
temperature.

  t + t T  =   t T  +  ̂  T  –   t T    7-40

 PCM · AP – P CM · AEA 
 1 – exp  – dt  PCM · ATAU  exp  ----------------------------
- 
 kT 
 

  t + t T  =   t T  +  ̂  T  –   t T    7-41

 PCM · CP – P CM · CEA- 
 1 – exp  – dt  PCM · CTAU  exp  ----------------------------
 kT  
 

473 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

Table 7-11 shows the parameters of Equations 7-35 through 7-41.

Table 7-11 Material Parameters for PCM Materials

Statement Parameter Type Default Units

MATERIAL PCM.AEA Real 3.2 eV

MATERIAL PCM.AP Real 1.0
MATERIAL PCM.ARHO Real 108 cm

MATERIAL PCM.ATAU Real 210-36 s

MATERIAL PCM.ATC Real 620 K

MATERIAL PCM.CEA Real 3.7 eV
MATERIAL PCM.CP Real 1.0
MATERIAL PCM.CRHO Real 108 cm

MATERIAL PCM.CTAU Real 210-36 s

MATERIAL PCM.CTC Real 600 K

MATERIAL PCM.LATHEAT Real 622 J

Latent heat can be also be introduced into the source term of the heatflow equation (Equation
7-1) by specifying PCM.LATHEAT on the MATERIAL statement. The latent heat is introduced as
described in Equation 7-42 for increasing temperatures and Equation 7-43 for decreasing
temperatures where H(t+dt,T) is the the latent heat at the current time step and H(t, T) is the
ˆ  T  is the temperature dependent static latent heat
latent heat at the previous time step. H
given by Equation 7-44 for increasing temperature and Equation 7-46 for decreasing
temperature.

Note: The time integration performed in Equations 7-40 and 7-41 above and Equations 7-42 and 7-43 below
critically depends on the choice of the parameters PCM.AEA, PCM.CEA, PCM.ATAU, and PCM.CTAU
relative to the operating temperatures and time steps. In some cases, you may find that the simulator
produces no discernable changes in resistivity. In this case, you should carefully reconsider the
specifications of these parameters.

ˆ  T  – H  t T   
H  t + t T  = H  t T  +  H 7-42

 PCM · AP – P CM · AEA- 
 1 – exp  – dt  PCM · ATAU  exp  ----------------------------
 kT  
 

474 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

ˆ  T  – H  t T   
H  t + t T  = H  t T  +  H 7-43

 PCM · CP – P CM · CEA 
 1 – exp  – dt  PCM · CTAU  exp  ----------------------------
- 
 kT 
 

ˆ  T  =  --------------------------------------------------------
H
T – PCM CTC
·

- PCM · LATHEAT  PCM · CTC  T  PCM · ATC  7-44
 PCM · ATC – PCM · CTC

ˆ  T  = –  --------------------------------------------------------
H
PCM ATC – T
·

- PCM · LATHEAT  PCM · CTC  T  PCM · ATC  7-45
 PCM · ATC – PCM · CTC

You can initialize the resistivity of a PCM material to a user-specified value by setting the
RESISTIVITY parameter of the MATERIAL statement. If unspecified, the initial resistivity will
be set to the value of PCM.CRHO.
To simplify the numerics of PCM simulation, set the PCM parameter of the METHOD statement.
This parameter effectively sets the METHOD parameters as described in Table 7-12.

Table 7-12 PCM Method Settings

Parameter Setting

MIN.TEMP 1e-9
MAX.TEMP 1e9
DVMAX 1e9
DIRECT TRUE
CARRIERS 0
NEWTON TRUE
TAUTO FALSE

If you want to simulate a mixture of semiconductors and PCM materials, then set both PCM
and CARRIERS=2 on the METHOD statement.
You can redefine the resistivity relations given by Equations 7-35 through 7-44 with a user-
specified function using the C-interpreter. You can assign the F.PCM parameter of the
MATERIAL statement to the file name of a valid C interpreter function expressing the variation
of resistivity with temperature and history.

475 ATLAS User’s Manual

Physical Models Giga: Self-Heating Simulator

7.2.10 C-Interpreter Defined Scattering Law Exponents

KSN and KSP are exponents in the relationship between relaxation time and carrier energy. In
the case for electrons:
KSN
n   E – Ec  7-46

They are generally calculated from the dependence of mobility on carrier temperature, see
Equations 3-145 and 3-146 in Section 3.4.1 “The Energy Balance Equations”.
The value of KSN and KSP therefore depend on the scattering mechanisms, which are
determining the relaxation time. Generally, KSN and KSP will themselves depend on Carrier
Energy. The C-Interpreter functions F.KSN and F.KSP are available in ATLAS to model this
dependence. The Energy input is 1.5 kB Tn,p if HCTE is specified, or 1.5 kB TL if only
LAT.TEMP is specified. To use these C-Interpreter functions, the syntax is
MODELS F.KSN=<filename> F.KSP=<filename>
where the <filename> parameter is an ASCII file containing the C-INTERPRETER function.
See Appendix A “C-Interpreter Functions” for more information regarding the C-
INTERPRETER and its functions.

476 ATLAS User’s Manual

Applications of GIGA Giga: Self-Heating Simulator

7.3 Applications of GIGA

7.3.1 Power Device Simulation Techniques
This section contains a series of techniques, which you may find useful when simulating
typical power device structures. Not all of the features described below are specific to GIGA
and are common to the ATLAS framework.
Floating Guard Rings
No special syntax is needed for the simulation of uncontacted doping areas used in floating
guard rings. The program is able to simulate guard ring breakdown with the standard impact
ionization models. In some extreme cases, convergence may be slow due to poor initial
guesses. If convergence is slow, both GUMMEL and NEWTON should be used in the METHOD
statement.
Floating Field Plates
You should use the ELECTRODE statement to specify the field plate regions as electrodes. If
these plates do not contact any semiconductor, then you can set these electrodes to float in the
same manner as EEPROM floating gates. The following statement line specifies that the field
plate region PLATE1 is a floating field plate.
CONTACT NAME=PLATE1 FLOATING
If the plates do contact the semiconductor, then do not use this syntax. Instead, current
boundary conditions are used at the electrode with zero current. See the “Floating Contacts”
on page 66 for more information about floating electrodes.
External Inductors
Inductors are commonly used in the external circuits of power devices. You can use the
CONTACT statement to set an inductor on any electrode. The following statement sets an
inductance on the drain electrode of 3 H/m.
CONTACT NAME=DRAIN L=3E-3
The next statement is used to specify a non-ideal inductor with a resistance of 100 /m.
CONTACT NAME=DRAIN L=3E-3 R=100

7.3.2 More Information

Many examples using GIGA have been installed on your distribution tape or CD. These
include power device examples but also SOI and III-V technologies. You can find more
information about the use of GIGA by reading the text associated with each example.

477 ATLAS User’s Manual

Chapter 8
Laser: Edge Emitting Simulator
Overview Laser: Edge Emitting Simulator

8.1 Overview
LASER performs coupled electrical and optical simulation of semiconductor lasers. LASER
works with BLAZE and allows you to
• Solve the Helmholtz equation to calculate the optical modes, their fields and intensity
patterns.
• Calculate optical gain, depending on the photon energy and the quasi-Fermi levels.
• Calculate the carrier recombination due to light emission (i.e., stimulated emission).
• Solve Photon Rate equations to calculate modal photon densities.
• Calculate laser light output power, light-voltage and current voltage characteristics.
To perform a LASER simulation, you need to know ATLAS and BLAZE first. If you don’t, read
Chapters 2 “Getting Started with ATLAS” and 5 “Blaze: Compound Material 2D Simulator”.

479 ATLAS User’s Manual

Physical Models Laser: Edge Emitting Simulator

8.2 Physical Models

To simulate semiconductor lasers, the basic semiconductor equations (see Equations 3-1 to 3-
4) are solved self-consistently with an optical equation that determines the optical field
intensity distribution. LASER uses the following coordinate system:
• The X axis is perpendicular to the laser cavity and goes along the surface (from left to
right).
• The Y axis is perpendicular to the laser cavity and goes down from the surface.
• The Z axis goes along the laser cavity.
The X and Y axes are the same as in other ATLAS products. The electrical and optical
equations are solved in the X and Y plane (i.e., perpendicular to the laser cavity).

8.2.1 2D Vector Helmholtz Equation

In general, components of electromagnetic field are coupled together into a system of
equations, which is represented by vector Helmholtz equation:
2
1 
  ---   H = ------- H 8-1
 c
2

In Cartesian coordinates, assuming that laser waveguide is homogeneous in z-direction,

dielectric permittivity tensor is diagonal and material in non-magnetic, this equation can be
written as follows:

 2 2
  Hx 1 H x 
  ------  yy  H y 1  H y 
2
  ------ 2
 2 +    + 1 – ------
- –  + -  yy H x =  H x
------
yy  y   y
 zz    zz   xy yy 
 y zz   x 2
x c
 8-2
 2 2
1 H y  H y 
    ------  xx  H x 1  H x 
2
  ------ 2

 xx  x   x  + + 1 – ------
- –  + -  H =  Hy
------
 zz  y 2     xy xx  x     y 2 xx y
 zz zz c

This is a non-Hermitian complex eigenvalue problem for the square of propagation constant 
and eigen vector =(Hx, Hy). The system is discretized with finite difference method and
solved with a sparse iterative eigenvalue solver, using frequency  as a parameter. To launch
vector Helmholtz solver for laser simulation, set V.HELM parameter on the LASER or MODELS
statements. The total number of transverse modes to be computed is set by the NMODES
parameter.
After the eigenvalue problem is solved, all other components of optical field can be found:

i H H
H z = ---  ---------x- + ---------y- 8-3
 x y 

H H
E x = -----------------  ---------z – ---------y-
i
8-4
 xx  0 y  z 

480 ATLAS User’s Manual

Physical Models Laser: Edge Emitting Simulator

H H
E y = – -----------------  ---------z – ---------x-
i
8-5
 yy  0  x z 

E x E y
E z = ----------   xx --------- +  yy ---------
i
8-6
 zz  x y 

The intensity pattern for each transverse optical mode is then computed using

 
I m  x y  =  E m  H m + E m  H m   ẑ 8-7
 

and normalized by area:

 Im  x, y  dS = 1 8-8

Effective refractive index and effective absorption for each mode are defined as
c
n eff m = e   m   ---- . 8-9


 eff m = 2m   m  8-10

8.2.2 2D Scalar Helmholtz Equation

It is often the case that for each mode, one of the components is much larger than the other.
This allows you to neglect the coupling between Hx and Hy components in Equation 8-2. We
therefore obtain a system of two independent scalar equations for each component:

 2
  Hx 1 H x 
  ------
2
2
 2 +  
yy  y   y  + -  H =  Hx
------ (TM)
2 yy x
x  zz  c
 8-11
 2
1 H y  H y ------
    ------
2
- 2
  + +  H =  Hy (TE)
xx  x   zz  x   y 2 2 xx y
 c

To use this approach, set S.HELM parameter on the MODELS statement instead of V.HELM. In
addition, this method allows you to solve for only TE or only TM polarization. Thus,
removing unnecessary modes. This is achieved by switching off HELM.TE or HELM.TM
parameters. By default both parameters are true.

481 ATLAS User’s Manual

Physical Models Laser: Edge Emitting Simulator

8.2.3 1D Scalar Helmholtz Equation

More simplification to Equation 8-2 may occur when dielectric permittivity is one-
dimensional: (x,y)=(x) or (x,y)=(y). If we assume that the field components are also one-
dimensional: Hx,y(x,y)=Hx,y(x) or Hx,y(x,y)=Hx,y(y) (i.e., we are dealing with infinite slabs).
In this case, equation for each component will also become one dimensional. A 1D scalar
solver can be launched by changing parameter HELM.GEOM on LASER or WAVEGUIDE
statements, which has a default value HELM.GEOM=2DXY.
In the case of (x,y)=(x), set the HELM.GEOM=1DX parameters and the following equations
will be solved:

 2
  Hx 
2
2
 2 + -  H =  Hx
------ (TM)
2 yy x
 x c
 8-12

1 H y 
    ------
2
2
 + -  H =  Hy
------ (TE)
 xx  x   zz  x  2 xx y
 c

In the case of (x,y)=(y), set the HELM.GEOM=1DY parameter and the following equations will
be solved:


1 H x ------
    ------
2
- 2

 y  y
yy  +  H =  Hx (TM)
2 yy x
  zz  c
 2 8-13
 H 2
 y  2
+ -------  xx H y =  H y (TE)
 2 2
 y c

To solve the 1D non-Hermitian problems, LASER uses LR algorithm, which scales as O(N2),
enabled by the parameter HEML1D=LR on the METHOD statement and is a default. An
alternative is more stable but a more time consuming QR algorithm, which scales as O(N3)
and is set by HEML1D=QR.

482 ATLAS User’s Manual

Physical Models Laser: Edge Emitting Simulator

8.2.4 1.5D Effective Index Solver

Effective index solver attempts to solve a 2D scalar Helmholtz equation by combining 1D
solutions in X and Y directions. It is suitable for quasi-2D structures, such as slab waveguides
and stripe lasers, where the variation of dielectric constant along one of the directions is slow.
Specifically, in the case of (x,y)  (y), using a variable separation ansatz Hx(x,y)=X(x)Y(y;x),
Equation 8-11 for TM modes can be written as:

2
  Y 2  2
 + -------  xx Y =  ---- n eff  x  Y
  y2 c2 c 
 (TM) 8-14
 1 X  2
  xx   ------  +  ---- n eff  x  X =  X
2

  x  zz  x   c 

The first 1D eigen problem is solved at each X-slice in Y direction to find effective index
neff(x), while the second equation is solved once in the middle of the device to find
propagation constant . Similarly, for TE modes Hy(x,y)=X(x)Y(y;x) and the equations read:

2
1 Y 
    ------ 
2
 Y
  y    y + ------
-  Y = ---
- n  x 
yy 2 yy c eff 
 zz c
 2
(TE) 8-15
  X  2
+ ---- n eff  x  X =  X
2
  
 x 2 c

To launch this method, set HELM.GEOM=15DY. In the case of (x,y)  (x), set
HELM.GEOM=15DX. For this solver, in addition to the total number of modes, you may also
choose the number of modes to be taken in Y direction in the first equation, by setting
EFFMODES parameter on LASER or WAVEGUIDE statements. By default, only fundamental
mode is taken in Y direction.

8.2.5 2D Cylindrical Helmholtz Solver

When V.HELM parameter is switched on and CYLINDRICAL parameter is present on MESH
statement, a 2D cylindrical Helmholtz solver will be launched. In cylindrical coordinates,
assuming that LASER is cylindrically symmetric, dielectric permittivity is isotropic, and
material is non-magnetic. The vector Helmholtz equation can be written for Hr and Hz
components as follows. (Equations for the case of m=0 or anisotropic dielectric constant look
somewhat different but are not shown here for the sake of brevity.)

 2 2
 1  H r 1  H r  3   1  H r  m – 1  H H r 2 H z
2
- Hr +   1 ---   z – 2
 
– --
-  – --
-  + – ----
- + --
-  + -------------------- – -----  = ----- H r
2  z 2  r  z     r 2 z    r z  r  z 2
 r r c
 2
8-16
 2
 Hz H z m 2
 1  Hz 1 1  1
 ---- ---    1 H r 2
 – --
-  – ---  – - + + -------- H z +  ---  = ----- H z
  r2  z 2  r  r    r 2 
z  z 2
 r c

483 ATLAS User’s Manual

Physical Models Laser: Edge Emitting Simulator

Here, m is an orbital (azimuthal) number, which can be set by ORBIT.NUM on the LASER or
WAVEGUIDE statement. This is a non-Hermitian complex eigenvalue problem for the square of
eigen frequency  and eigen vector =(Hr, Hz). The intensity of light is directed
perpendicular to the R-Z plane.

8.2.6 Dielectric Permittivity

For anisotropic relative dielectric permittivity, LASER uses the following model [128]
(i=x,y,z)
2 2
 i  x y  = n i  x y  =  n MAT i  x y  +  n GAIN i  x y  + n ABS  x y  + n FC  x y  + n TEMP  x y   8-17
where:
• n MAT i  x y  =  MAT i  x y  is the material high frequency complex refractive index. You
can specify the material refractive index or corresponding material dielectric permittivity
using the parameters on the MATERIAL statement shown in Table 8-1.
• change of complex refractive index due to gain is given by the following expression,
where j = – 1 , gi(x,y) is the local optical gain in i=x,y,z direction and the ALPHAR
parameter on the MATERIAL statement represents a Kramers-Kronig relation between
changes of real and imaginary parts:
c 8-18
n GAIN i  x y  = –  ALPHAR + jg i  x y    -------
2

• change of complex refractive index due to bulk absorption outside of active region is
given by the ALPHAA parameter specified in the MATERIAL statement. Additionally or
instead of this, you can specify the ABS.BULK parameter on the MODELS statement to
compute bulk absorption coefficient at each node using multiband k.p band-structure
• change of complex refractive index due to free-carrier absorption is turned on by
ABS.FCARRIER parameter on the MODELS statement and is given by the following
expressions, where FC.RN, RC.RP, FC.EXPRN, FC.EXPRP, FC.AN, FC.AP, FC.EXPAN,
and FC.EXPAP are parameters on the MATERIAL statement:
FC.EXPRN FC.EXPRP
e  n FC  x y   = –  FC.RN  n + FC.RP  n  8-19

FC.EXPAN FC.EXPAP c
m  n FC  x y   =  FC.AN  n + FC.AP  n   ------- 8-20
2

• change of complex refractive index due to change of lattice temperature is given by the
following expression, where DINDEXDT is the parameter on the MATERIAL statement and
T(x,y) is the local temperature:
n TEMP  x y  = DINDEXDT   T  x y  – T 0  8-21

484 ATLAS User’s Manual

Physical Models Laser: Edge Emitting Simulator

Table 8-1 Rule Hierarchy for Determination of High Frequency Material Permittivity

EPS.XX, EPS.YY, and EPS.ZZ specify the real parts of XX, YY and ZZ components of dielectric
permittivity constant. EPSIM.XX, EPSIM.YY, and EPSIM.ZZ specify the imaginary parts.

EPS.ISO and EPSIM.ISO specify real and imaginary parts of dielectric constant in isotropic case or
for unspecified tensor components in anisotropic case. EPSINF can be used instead of EPS.ISO.
REAL.INDEX and IMAG.INDEX specify real and imaginary parts of refractive index, which may be
more convenient than dielectric permittivity constant.
If F.INDEX is defined in the MATERIAL statement, then the bulk permittivity is calculated from the
square of the index returned from the C-Interpreter function defined in the file specified by F.INDEX.
If INDEX.FILE is defined in the MATERIAL statement, then the bulk permittivity is calculated as the
square of the index interpolated from the table specified in the file pointed to by INDEX.FILE.
If the material is listed in Table B-31, the permittivity is calculated as the square of index interpolated
from the built-in tables for that material.

After complex refractive index is calculated, local absorption can be found from the
following equation:
2
  x y  = ------- m  n  x y   8-22
c

8.2.7 Local Optical Gain and Spontaneous Emission

LASER has four gain/spontaneous emission models set by the GAINMOD parameter. GAINMOD
can take the value of 1 for Standard model (Equations 3-630 and 3-633), 2 for Empirical
model (Equations 3-629 and 3-635), 3 for Takayama model (Equations 3-629 and 3-636) and
5 for multiband k.p model.
Multiband k.p model #5 can be used for bulk regions (Equations 3-646 and 3-647) and for
quantum wells (Equations 13-25 and 13-26). For multiband k.p model parameter ZB.ONE,
ZB.TWO, ZB.THREE, or WZ.THREE should also specified on the MODELS statement. There is
already a ABS.BULK parameter on MODELS statement, which computes absorption, using
exactly the same mechanism. ABS.BULK is disabled in regions, where gain model #5 is used.
A user defined gain can also be used by specifying the name of a C-Interpreter function
F.LGAIN on the MATERIAL statement.

485 ATLAS User’s Manual

Physical Models Laser: Edge Emitting Simulator

8.2.8 Stimulated Emission

Carrier recombination due to stimulated light emission is modeled as follows:

  -------------
c j
R st  x y  = - g  x y  m  x y   S m 8-23
n eff m j
j = x y z m

where
• Rst(x,y) is the recombination rate due to stimulated light emission.
• The first summation is performed over all directions.
• The second summation is performed over all laser modes.
• neff,m is the effective refractive index.
• Sm is the modal photon density.
• gj(x,y) is local gain for electric field polarized in j direction
• jm(x,y) is given by
j 2
j E m  x y 
 m  x y  = 2  e  n j  x y    ---------------------------- 8-24
 Im  x y  dS
Gain confinement factor can be defined as the integral of jm(x,y) over active region [176,
291]. LASER can calculate gain confinement factor either by integrating over all quantum
well regions:


j j
m =  m  x y  dS 8-25
QWELL

or integrating over the area with positive gain:


j j
m =  m  x y  dS 8-26
gain  0

Total confinement factor can be defined as sum of confinement factors over all directions:


j
m = m 8-27
j = x y z

486 ATLAS User’s Manual

Physical Models Laser: Edge Emitting Simulator

8.2.9 Photon Rate Equations

Optical gain provides a link between optical and electrical models. The optical gain depends
on the quasi-Fermi levels and in turn impacts dielectric permittivity (see Equation 8-11), and
by the coupling between the stimulated carrier recombination rate (Rst) and the density of
photons (S) as described by Equation 8-23.
To determine Sm, LASER solves the system of photon rate equations:
dS m
- =  -------- G m – -------------
1 S + R
c
--------- 8-28
dt  n eff  ph m m sp m

where the modal gain Gm is given by:

   gj  x y m    m  x y   dS
j
Gm = 8-29
j = x y z

and the modal spontaneous emission rate R sp m is given by:

   r sp  x y m    m  x y   Em  dS
j j
R sp m =   8-30
j = x y z

where  is the spontaneous emission coupling factor, defined by parameter

EMISSION.FACTOR on MATERIAL statement (default is 1), rjsp(x,y,m) is spontaneous
emission rate per unit energy per unit volume, Em= /ph,m is the line width of the laser
mode and ph,m is the modal photon lifetime, which represents laser losses and is given by
[200, 246].
1 c
------------- = --------   mir +  abs + LOSSES + EXTRACTION  8-31
 ph m n eff

Here, abs is the absorption loss. To include absorption loss into photon rate equation,
specify the ABSORPTION parameter on the LASER statement. It is defined as:

    j  x y 
j
 abs =   m  x y  dS 8-32
j = x y z

where total absorption j(x,y) is given by Equation 8-22.

The parameter LOSSES is defined on LASER statement and corresponds to internal laser losses
without regards to their origin. The parameter EXTRACTION is defined on LASER statement
and is used to phenomenologically account for extraction of light from the cavity if mirror
losses are zero (as in the case of cylindrical cavity).
The mirror loss, mir , can be defined in two ways. One way is to define the percentage
reflectivity of both mirrors. The other way is to define the individual facet mirror
reflectivities. If the former is chosen, then set the MIRROR.LOSS parameter on the LASER
statement to the percentage reflectivity for the facet cavity mirrors. This assumes that both
front and back mirrors are identical and gives a mirror loss calculated by:

487 ATLAS User’s Manual

Physical Models Laser: Edge Emitting Simulator

 mir = ------------------------------------------------------ ln  --------------------------------------------

1 1
8-33
2 CAVITY.LENGTH  MIRROR.LOSS2

where CAVITY.LENGTH is set to the length of the laser cavity on the LASER statement.
If you choose the latter model for mir , set the RR and RF parameters on the LASER statement
instead. These parameters are the rear and front facet reflectivities respectively. The mirror
loss is then calculated by:
1 1
 mir = -------------------------------------------------------- ln ------------------ 8-34
2  CAVITY.LENGTH RF  RR

The user-specified parameters for the loss models in Equation 8-31 are given in Tables 8-2
and 8-3.

Table 8-2 User Specifiable Parameters for Equation 8-28

Statement Parameter Default Units

LASER LOSSES 0 cm-1

LASER EXTRACTION 0 cm-1

Table 8-3 User-Specifiable Parameters for LASER Loss Models

Statement Parameter Default Units

LASER MIRROR.LOSS 90.0 %

LASER CAVITY.LENGTH 100.0 m

8.2.10 Optical Power

The optical modal power emitted from the front mirror is calculated by [200, 272]
hS m c ln  1   R f R r  
P f m = -----------------  ------------------------------------------------------------------------ 8-35
2n eff 1 + R  R  1 – R    1 – R 
f r r f

where Sm is the photon density of mode m,  is the frequency of emitted light, Rf and Rr are
the front and rear mirror reflectivities. When light is extracted through a coupling to a nearby
waveguide, as in the case of cylindrical cavities, you can set mirror reflectivities to 100%, and
use the parameter EXTRACTION instead. In this case, the optical power emitted is given by
hS m c
P f m = -----------------  EXTRACTION 8-36
2n eff

The total optical intensity emitted from the front mirror is given by

I  x y  =  Pfm Im  x y  8-37
m

488 ATLAS User’s Manual

Physical Models Laser: Edge Emitting Simulator

8.2.11 Gain Saturation

To simulate non-linear gain saturation in LASER, use the simple model described by [34]
g  x y 
g  x y  = ----------------------------------------------------------- 8-38
1 + I  x y   GAIN.SAT

where GAIN.SAT is a user-specifiable parameter on the MATERIAL statement, g´(x,y) is the

local gain. To enable this model, specify GAIN.SAT on the LASER statement. Non-linear,
absorption loss is similarly modeled using the following expression.
  x y 
  x y  = ---------------------------------------------------------------------------------- 8-39
1 + I  x y   ABSORPTION.SAT

where ´(x,y) is the local absorption and ABSORPTION.SAT is a user-definable parameter on

the MATERIAL statement. To enable this model, specify ABSORPTION.SAT on the LASER
statement.

489 ATLAS User’s Manual

WAVEGUIDE Statement Laser: Edge Emitting Simulator

8.3 WAVEGUIDE Statement

It is often useful to know and to be able to optimize the optical intensity patterns in a laser
waveguide before doing a full-scale laser simulation. For that reason, vector Helmholtz solver
can also be used as a standalone optical mode solver. To find optical modes at any point
during a device simulation, use the WAVEGUIDE statement and specify the V.HELM, OMEGA, or
PHOTON.ENERGY parameters and the number of transverse modes NMODE. If any of the regions
has a gain model turned on, it will be included in the imaginary part of the dielectric constant.
Thus, you can check how the optical modes are modified by the increase of gain in the system
or by changing temperature profile. The WAVEGUIDE statement and its parameters are
completely independent of the LASER statement. Therefore, it can be used by itself or
simultaneously with the laser simulation.
If you want to solve for optical modes after each bias point as a post-processing, you may use
just one WAVEGUIDE statement in the beginning of voltage sweep and include the TRACE
parameter. A useful output characteristics would be real and imaginary parts of propagation
constants and modal effective indexes, which can be stored into the IV log file by setting
WVGD.PROP and WVGD.REFR on the output statement.
Additionally, the WAVEGUIDE statement also allows you to get a dispersion of the propagation
constant . To run a dispersion calculation, specify DISPERSION, the range of photon
frequencies OMEGA.INIT (or E.INIT) and OMEGA.FINAL (or E.FINAL), the number of
samples NDISP, and the log file name DISP.NAME, where dispersion will be stored.

490 ATLAS User’s Manual

Simulation Parameters Laser: Edge Emitting Simulator

8.4 Simulation Parameters

Simulation of lasers includes three steps:
1. Defining a domain
2. Defining a type of laser simulation
3. Analyzing the output.
Numerical parameters are used to regulate convergence and computation time.

8.4.1 Specifying Simulation Domain

LASER uses rectangular mesh. If ATLAS mesh is rectangular, it will be used by default.
Alternatively, you can set up a separate laser mesh with LX.MESH and LY MESH statements.
By default, the solution domain spans over the whole device. Use the parameters LX.MIN,
LX.MAX, LY.MIN, and LY.MAX to limit the domain. Additionally, you may switch off
HELMHOLTZ parameter on the REGION or ELECTROD statement to exclude the region or
electrode from the domain.
For devices with mirror symmetry at X=0, specify only half of the device and use parameter
REFLECT on LASER or WAVEGUIDE statement. The optical modes will be either symmetric or
anti-symmetric with respect to Hx, Ey, Ez (and anti-symmetric or symmetric with respect to
Hy, Hz, Ex). You can block either type by switching off HELM.SYM or HELM.ASYM parameters.
By default, both symmetries are calculated.
The default boundary conditions on all boundaries are Perfect Electric Conductor (PEC),
which is set by HELM.PEC parameter on the LASER or WAVEGUIDE statements. Alternatively,
you can use HELM.PMC to set Perfect magnetic Conductor (PMC) boundary conditions on all
boundaries. If both HELM.PEC and HELM.PMC are switched off, magnetic field components
will be set to zero at the boundaries.

8.4.2 Physical Parameters

Relative dielectric permittivity is given by Equation 8-11. In many cases, say for index-
guided lasers, imaginary part of dielectric permittivity can be neglected and the system of
equations become pure real. This is regulated by the parameter INDEX.MODEL=0 in either the
LASER or WAVEGUIDE statements and is a default. If imaginary part is important, set
INDEX.MODEL=1. As usual, the trade-off for complex refractive index model is a slightly
longer computation time.
LASER will find only certain number of modes with the highest propagation constant. The
number of transverse modes can be specified by the parameter NMODE in the LASER or
WAVEGUIDE statement.
The simplest way to run LASER is to do a single lasing frequency simulation. In this case,
specify PHOTON.ENERGY or OMEGA in the LASER statement, which will be fixed throughout
the simulation and used as a parameter to Helmholtz equation and finding gain. Note that in
this case, the number of laser modes is equal the number of transverse modes set by NMODE
parameter.
A FLOAT parameter (false by default) on the LASER statement will do a slight modification by
letting the lasing frequency to float to the energy where gain, averaged over the device, is
maximum.

491 ATLAS User’s Manual

Simulation Parameters Laser: Edge Emitting Simulator

If multiple transverse modes need to be included in simulation, specify LMODES parameter and
set the range of photon energies with parameters E.INIT and E.FINAL or OMEGA.INIT and
OMEGA.FINAL. In this case, for each transverse mode, LASER will find a number of
longitudinal modes, which will have different frequencies but the same field and intensity
pattern. The longitudinal modes correspond to eigen modes of Fabry-Perot resonator. To limit
the number of longitudinal modes, use the MAX.LMODES parameter (default is 20). If a large
range of lasing modes is required, but the number of lasing modes has to be kept low to save
computation time, you can use DELTAE (same as ESEP) parameter to specify energy
separation between laser modes. In the case of multiple transverse mode simulation, the total
number of laser modes is less or equal to NMODE*MAX.LMODES.

Note: If multiple longitudinal modes are to be accounted for, use frequency dependent gain models by setting
GAINMOD equal to 1 or 5 for the active lasing region.

The CAVITY.LENGTH parameter specifies is the length of the laser cavity in the Z direction.
This will affect mirror losses and longitudinal mode energies.
Laser simulation can be done self-consistently with lattice temperature model (Equation 7-1).
By default, stimulated emission is not used as a generation-recombination heat source. To use
it as a heat source, specify STIM.HEAT together with GR.HEAT on the MODELS statement.

8.4.3 Numerical Parameters

Once the number of modes is known and mode energies are found, modal characteristics are
calculated and included into photon rate equations for each laser mode. These equations are
solved self-consistently with the rest of ATLAS equations (e.g, Poisson, electron and hole
continuity) using the Newton scheme. ATLAS also has an older version of LASER, which
uses 1.5D effective index solver together with relaxation method of solving photon rate
equations. The older version works only when neither V.HELM or S.HELM is specified.

Note: Newton scheme works only when V.HELM or S.HELM is present on the LASER statement.

You have a choice of including not all, but only certain number of laser modes with strongest
amplitudes in the total Jacobian. This is done using the parameter NMOD.NEWTON, which has
the default value of 50 for longitudinal mode simulation and NMODE for single frequency
simulation.
When the laser simulation starts, only electrical characteristics and local gain everywhere in
the device is calculated. If at some bias point, usually just below laser threshold, local gain
becomes positive, photon rate equations are added to the Jacobian and photon densities will
be computed. If subthreshold behavior is of interest, you can enforce solution of photon rate
equations by setting the START parameter on the LASER statement any time during the
simulation.
To limit the update of the photon density of each laser mode at each Newton iteration, use
parameter MAXCH. It enforces that photon density does not increase or decrease by more than
MAXCH times.
TOLER parameter sets the desired relative tolerance of the photon density calculation. The
default value is 0.01. Setting this parameter to a lower value may slow down the calculation
significantly. Using a larger value will result in a faster but less accurate calculations.

492 ATLAS User’s Manual

Simulation Parameters Laser: Edge Emitting Simulator

Once the Newton iterations have converged, refractive indexes will be updated, the
Helmholtz equation will be re-solved, and new modal quantities will be calculated. Such
iterative scheme of consecutive solutions of Helmholtz and all other equations usually
converges in a couple of steps. You can limit the number of times the Helmholtz equation is
solved for one bias point using parameter ITMAX.

8.4.4 Simulation Output

The output structure file will contain normalized intensity patterns for each transverse mode
and total optical intensity given by Equation 8-37. Additionally, components squared of
normalized optical fields, such as |Ex(x,y)|2 and |Hz(x,y)|2, can be stored in the structure file
by specifying EX.OPT, EY.OPT, EZ.OPT, HX.OPT, HY.OPT, and HZ.OPT on the OUTPUT
statement.
To store the spatial distribution of real and imaginary refractive index defined in Equation 8-
11 at the lasing frequency into the structure file, set REFR.INDEX on the OUTPUT statement.
IV log file will contain current, modal photon density, modal power and modal gain and loss.
You can also store the following quantities:
• Real and imaginary parts of refractive index can be probed at a specific location and
stored into IV log file. For this, specify REFR.REAL and REFR.IMAG on the PROBE
statement and also set a location and the name of the probe.
• In order to output eigenvalues (propagation constants or eigen frequencies) of the
Helmholtz equation on the screen, specify PRT.EVAL on LASER or WAVEGUIDE statement.
• If real and imaginary parts of propagation constants and modal effective indexes are of
interest, they can be stored into the IV log file by setting LAS.PROP (same as WVGD.PROP)
and LAS.REFR (same as WVGD.REFR) on the output statement.
• In order to output gain confinement factor defined in Equation 8-25 and Equation 8-27
into IV log file, use the parameters LAS.WELLCONX, LAS.WELLCONY, LAS.WELLCONZ,
and LAS.WELLCONT. In order to output gain confinement factor defined in Equation 8-26
and Equation 8-27, use the parameters LAS.GAINCONX, LAS.GAINCONY, LAS.GAINCONZ,
and LAS.GAINCONT.
If SPEC.NAME="filename" is specified, the modal spectrum file
"modal_filename_X.log" will be generated after each bias point, where X=1,2,3,... .
Alternatively, modal spectrum can also be saved using the SPECTRUM="filename"
parameter on the SAVE statement, which will generate a "modal_filename.log" file. (In the
latter case, if quantum wells are present, the additional file "filename.log" will be stored,
which will contain gain and spontaneous emission of each quantum well).

493 ATLAS User’s Manual

Simulation Parameters Laser: Edge Emitting Simulator

8.4.5 Generation of Near-Field and Far-Field Patterns

LASER can generate the far-field pattern if you specify the PATTERNS parameter of the SAVE
statement. The PATTERNS parameter specifies the file name prefix of two output files. The
first of these files, with the .nfp suffix, contains the near-field pattern. This is essentially a
copy of the near field pattern contained in the structure file output. The second of these files,
with the .ffp suffix, contains the far-field pattern.
The far-field pattern is essentially a 2D fast Fourier transform (FFT) of the near-field pattern.
The sampling for this transform is set by the minimum spacing in the laser mesh and the
overall size of the laser mesh. The output samplings for the near-field and far-field patterns
are controlled by the NEAR.NX, NEAR.NY, FAR.NX, and FAR.NY parameters of the LASER
statement. These specify the numbers of samples in the X and Y directions in the near and far
field. By default, the values are set to 100.
To specify the range of angles, where far field pattern is requested, use the parameters
FAR.SPANX and FAR.SPANY with the defaults of 180°. The angles will span from
-FAR.SPANX/2 to FAR.SPANX/2.

494 ATLAS User’s Manual

Chapter 9
VCSEL Simulator
Overview VCSEL Simulator

9.1 Overview
VCSEL (Vertical Cavity Surface Emitting Lasers) performs electrical, thermal, and optical
simulation of vertical cavity surface emitting lasers using accurate, robust, and reliable fully
numerical methods and non-uniform meshes. With VCSEL, you can simulate advanced
structures containing multiple quantum wells and oxide apertures including the effects, such
as gain guiding, lattice heating, current crowding, and spatial hole burning. VCSEL works
with BLAZE and GIGA and allows you to do the following:
• Test whether the structure complies to general VCSEL device requirements by simulating
cold cavity reflectivity experiments. Calculate the reflectivity of a VCSEL cavity as a
function of the incident light wavelength in a user-specified range. Then, determine the
resonant frequency of the cold cavity.
• Solve the Helmholtz equation (see Equation 9-1) in cylindrical coordinates to calculate
optical intensities of the multiple transverse modes for index guiding and gain guiding
structures.
• Calculate the optical gain in multiple quantum well systems depending on the photon
energies, quasi-Fermi levels, temperature, and optical intensity distribution. Calculate the
carrier recombination due to the spontaneous and stimulated emission.
• Solve the photon rate equations (see Section 9.2.4 “Photon Rate Equations”) for multiple
transverse modes to calculate the photon density in each mode and the total photon
density.
• Calculate the light output power and the wavelength for each transverse mode.
• Speed up solution of the Helmholtz equation by using perturbational treatment.
• Simulate more general cavities that support multiple longitudinal modes. This obtains
modal gain spectra and determines the dominant mode with maximum gain.
To perform a VCSEL simulation, you need be familiar with ATLAS and BLAZE first. Please
see Chapters 2 “Getting Started with ATLAS” and 5 “Blaze: Compound Material 2D
Simulator” for more information.

Note: To avoid the confusion in terms, we refer to a VCSEL device as VCSEL and to a VCSEL simulator as
VCSEL:ATLAS in this chapter.

496 ATLAS User’s Manual

Physical Models VCSEL Simulator

9.2 Physical Models

When modeling VCSELs, it is essential to take into account interaction of optical, electrical,
and thermal phenomena that occur during the VCSEL operation. Modeling only optical
properties is already an involved task. Different methods developed recently to improve the
accuracy of optical solution produce results that vary from method to method [24]. The
variation is especially strong between vectorial models that tend to produce a spread in results
much larger than simpler scalar models.
On the other hand, the complexity of the vectorial methods makes the self-consistent electro-
thermo-optical simulation impractical from the point of view of the calculation time required.
For VCSEL simulation in VCSEL:ATLAS, we adopted an approach that can account for
mutual dependence of electrical, optical, and thermal phenomena.
Basic semiconductor equations (see Equations 3-1 to 3-4) are solved self-consistently with
the Helmholtz Equation, Lattice Heat Flow Equation (see Equation 7-1), and the Photon Rate
Equation (see Equation 9-12). Since we are using a scalar method, we neglect polarization
effects and reduce Maxwell equations to scalar Helmholtz equation. The method used for the
solution of Helmholtz equation is based on a well developed effective index model, which
showed its validity for a much wider range of problems than originally expected [304]. This
method fine tuned for simulation of various VCSEL structures is often referred to as Effective
Frequency Method (EFM). EFM is a fast and flexible method that allows further development
to include dispersion, diffraction losses, and graded interfaces.
VCSEL:ATLAS uses a cylindrical coordinate system to take advantage of the cylindrical
symmetry of the VCSEL devices. Figure 9-1 shows the relationship between ATLAS’s XY
coordinate system and the r--z cylindrical system used in VCSEL:ATLAS. Note that the X
coordinate axis is equivalent to the radius r, and the Y coordinate axis is equivalent to the Z
axis in the cylindrical system.

9.2.1 Reflectivity Test Simulation

Reflectivity experiment is an important test conducted to evaluate the quality of a VCSEL
device. In this experiment, the light is normally incident on a VCSEL device and its
reflectivity is measured as a function of wavelength. Most manufactured VCSEL resonators
are tested this way to ensure adequate optical performance. Numerical simulation of this
experiment enables you to calibrate the material parameters used in the optical model and
eliminate possible specification errors at the early stage of numerical analysis of a VCSEL
device.
Another purpose of the numerical reflectivity test is to ensure that the analyzed device
complies with general VCSEL requirements. These requirements include the following:
1. Presence of a high reflectivity band in the reflectivity spectrum of a VCSEL cavity.
2. The reflectivity over 99% in the high reflectivity band.
3. Presence of a drop in reflectivity within the high reflectivity band. The drop in reflectivity
(increase in absorptance) occurs at the resonant wavelength of the cavity.
VCSEL:ATLAS calculates reflectivity and absorptance using transfer matrix approach
(Section 10.3 “Matrix Method”). In the simulation, the incident light propagates along the
axis of symmetry of the device.

497 ATLAS User’s Manual

Physical Models VCSEL Simulator

The VCSEL.INCIDENCE parameter specifies the position of the light origin with respect to the
device (top or bottom). Specific output associated with the reflectivity test includes
reflectivity and absorptance spectra plotted in the corresponding files,
reflectivity(absorption)_top.log and reflectivity(absorption)_bottom.log.
To run the reflectivity test simulation, specify the VCSEL.CHECK parameter and other
parameters in Table 9-1 on a VCSEL statement. Section 9.3.3 “Enabling VCSEL Solution”
will discuss parameter specification in detail.

Table 9-1 User-Specifiable Parameters for the Reflectivity Test

Statement Parameter Default Units

VCSEL VCSEL.CHECK False

VCSEL VCSEL.INCIDENCE 1
VCSEL OMEGA s-1
VCSEL EINIT eV
VCSEL EFINAL eV
VCSEL PHOTON.ENERGY eV
VCSEL INDEX.TOP 1.0
VCSEL INDEX.BOTTOM 1.0
VCSEL NSPEC 100

If the structure does not meet requirements 1-3, VCSEL:ATLAS will close with an error.
Otherwise, VCSEL:ATLAS uses Brent's minimization procedure to find the resonant
frequency of the cold cavity. VCSEL:ATLAS uses the obtained value as a reference
frequency in the solution of the Helmholtz equation.

498 ATLAS User’s Manual

Physical Models VCSEL Simulator

Figure 9-1: The relationship between rectangular and cylindrical coordinate systems used by VCSEL

9.2.2 Helmholtz Equation

VCSEL:ATLAS solves the Helmholtz Equation in cylindrical coordinates r, z, and  using
effective frequency method [304].
2
2 
 E  r z   + -------   r z   E  r z   = 0 9-1
2
c

where  is frequency, (r, z, , ) is the complex dielectric permittivity, E(r, z, ) is the

optical electric field, and c is the speed of light in vacuum.
Using the expansion around a reference frequency 0:
2

2 0 0
-------   r z  w   ---------   r z   0  + 2 -------   r z   0    –  0  9-2
2 2 2
c c c

we transform the Helmholtz equation into:

2 2 2
  + k 0   r z   E  r z   = k 0   r z  E  r z   9-3

where k0=0/c and  is a dimensionless frequency parameter given by:

0 –   –  0 2  Im   
 = 2 ----------------- = 2 --------------- – j ----------------------- 9-4
0  0

We assume that the field is separable:

E  r z   = f  r z    r   9-5

499 ATLAS User’s Manual

Physical Models VCSEL Simulator

This assumption of separability is the main approximation of the method. The applicability of
the method can be extended beyond this approximation by adding a non-separable component
of the field on the right hand side of Equation 9-5. For dispersive materials, you can make
another modification of the effective frequency method by taking into account material
dispersion in Equation 9-2. For separable fields, we obtain the longitudinal wave equation for
function f:
2
 2 2
-------- + k 0   r z   f  r z   = v eff  r  k 0   r z  f  r z   9-6
2
z

at each lateral position (r, ). Due to the cylindrical symmetry, we can ignore the azimuthal
variations. In this case, we only need to solve Equation 9-6 for each cylindrically symmetric
region characterized by a particular distribution of  in the vertical direction z. Complex
Newton method combined with the characteristic matrix approach (Section 10.3 “Matrix
Method”) yields eigenvalues eff and eigenfunctions f(z) of the longitudinal wave equation.
The transverse wave equation takes the form:
2
1 d  d l 2 2
--- ----- r ----- – ----- + v eff  r k 0   r  lm  r  = n lm K 0   r  lm  r  9-7
r dr  dr r 2

where
L

   r z f
2
  r =  r  z  dz 9-8
0

VCSEL:ATLAS uses standard eigentechniques to solve an ordinary differential Equation 9-7.

Solutions of this equation are LPlm modes.
For each mode, the imaginary part of the eigenvalue determines the overall mode gain(loss)
factor, which is negative below the lasing threshold. The lasing occurs when:
Im  n lm  = 0 9-9
th

The dielectric permittivity is given by:

 0 g  r z   0  ALPHAA + FCN  n + FCP  p 

  r z  =  0 +  – ALPHAR + j  ------------------------- – j --------------------------------------------------------------------------------------- 9-10
k0 k0

where:
• 0 is the bulk high frequency permittivity. You can specify the bulk high frequency
permittivity using the EPSINF parameter of the MATERIAL statement. If unspecified, the
rule hierarchy in Table 9-2 will be used to define the high frequency permittivity.
• ALPHAR is a line width broadening factor.
• kc
• g (r, z) is the local optical gain.
• ALPHAA is the bulk absorption loss and is specified in the MATERIAL statement (specify
ABSORPTION in the VCSEL statement to include absorption loss).

500 ATLAS User’s Manual

Physical Models VCSEL Simulator

• FCN and FCP are the coefficients of the free-carrier loss and are set using the MATERIAL
statement (specify FCARRIER in the VCSEL statement to include this loss mechanism).

Table 9-2 User-Specifiable Parameters for Equation 9-10

Statement Parameter Default Units

MATERIAL ALPHAR 4.0

MATERIAL ALPHAA 0.0 cm-1

MATERIAL EPSINF

MATERIAL FCN 3.0×10-18 cm2

MATERIAL FCP 7.0×10-18 cm2

VCSEL ABSORPTION FALSE

VCSEL FCARRIER FALSE

9.2.3 Local Optical Gain

For a discussion of gain models, see Sections 3.9.3 “The Standard Gain Model” through 3.9.5
“Takayama's Gain Model”.
Stimulated Emission
Carrier recombination due to stimulated light emission is modeled as follows:

 --------------
c 2
R st  r z  = g  r z  E m  r z   Sm 9-11
NEFF m
m

where Rst is the recombination rate due to stimulated light emission, NEFF is the group
effective refractive index, and Sm is the photon number. The m subscript in this equation and
all subsequent equations refers to a modal quantity. For example, Sm in Equation 9-11 is the
photon number for mode m. The NEFF parameter is user-defined but has a default value of
3.57 (see Table 9-3).

Table 9-3 User-Specifiable Parameters for Equation 9-11

Statement Parameter Default Units

VCSEL NEFF 3.57

501 ATLAS User’s Manual

Physical Models VCSEL Simulator

9.2.4 Photon Rate Equations

Optical gain provides the link between optical and electrical models. The optical gain
depends on the quasi-Fermi levels and in turn impacts dielectric permittivity (see Equation 9-
12), and by the coupling between the stimulated carrier recombination rate (Rst) and the
density of photons (S) as described by Equation 9-11.
To determine Sm, VCSEL solves the system of photon rate equations:
dS m  c 1 c LOSSES
- =  -------------- G m – ----------
--------- – ---------------------------- S m + R sp 9-12
dt  NEFF  ph m NEFF  m

where the modal gain Gm is given by

   gm  r z   Em  r z 
2
Gm = rddrdz 9-13

and the modal spontaneous emission rate Rspm is given by

R sp =
m    rsp  r z m rddrdz . 9-14

Em(r,z) is the normalized optical field.

Table 9-4 User Specifiable Parameters for Equation 9-12

Statement Parameter Default Units

VCSEL LOSSES 0 cm-1

The modal photon lifetime, phm , in Equation 9-12 represents the losses in the laser. The
losses per mode are given by
1 c c
---------- = --------------   a +  fc +  mir  = -------------- G m –  0   lm 9-15
 ph NEFF m m NEFF
m

a is the bulk absorption loss, fc is the free-carrier loss, mir is the mirror loss, and  lm is a
dimensionless frequency parameter. These are defined as:

   ALPHAA  Em  r z 
2
a =  rddrdz 9-16
m

    FCN n + FCP p 
2
 fc =  E m  r z  rddrdz 9-17
m

NEFF
 mir = G m –  a –  fc –  0   lm -------------- 9-18
m m c

502 ATLAS User’s Manual

Physical Models VCSEL Simulator

Spontaneous Recombination Model

For a discussion of spontaneous recombination models, see Sections 3.9.1 “The General
Radiative Recombination Model” and 3.9.2 “The Default Radiative Recombination Model”.
Optical Power
The optical power emitted by the front facet is given by Equation 9-19.
hS m c  mir
P f = ----------------- ------------------------------------------------------------------------ 9-19
NEFF 1 + R  R  1 – R    1 – R 
f r r f

503 ATLAS User’s Manual

Simulating Vertical Cavity Surface Emitting Lasers VCSEL Simulator

9.3 Simulating Vertical Cavity Surface Emitting Lasers

The key to simulating a Vertical Cavity Surface Emitting Laser is the input deck. The input
deck describes the physical device structure, the physical models to be used, and the specific
measurements to be performed. The deck itself can be roughly divided into three sections:
structure specification, model definition, and obtaining solutions. The following sections will
describe these concepts in detail.

9.3.1 Specifying the Device Structure

VCSEL devices are the most complicated devices to be addressed by device simulation. This
is due to the many layers involved in making distributed Bragg reflectors (see “Specifying
Distributed Bragg Reflectors” on page 506) and the possible quantum well (see “Specifying
Quantum Wells” on page 507) active layers. This complexity is mitigated by the fact that
these devices are mostly epitaxial. Also, many of the layers are periodic. Recognizing these
simplifications, we’ve designed a custom syntax for defining VCSEL devices.
This syntax makes specification of periodic structures (e.g., super-lattices) simple, which
completely avoids problems of aligning various layers to each other. This new syntax also
simplifies meshing the structure.
Despite this simplification, specifying a VCSEL device can still be complex. We recommend
that you look at the standard examples. See Section 2.4 “Accessing The Examples” for more
information about these examples.
The order of statements is important when specifying the device structure. The following
order of statements should be strictly followed.
1. Specify the MESH statement. This initializes the structure definition and allows
specification of cylindrical symmetry.
2. Set up the mesh in the X direction by using the X.MESH statements. The mesh in the Y
direction is usually specified along with the device layers in the REGION statements. You
can, however, introduce the Y mesh lines by using the Y.MESH statements after the
X.MESH statements.
3. Specify the device regions using REGION and DBR statements. These statements specify
the layer thickness, composition, doping, order of placement, and meshing information in
the Y direction. The order of REGION and DBR statements is up to you and is usually
defined by the order the layers appear through the structure.
4. You can use quantum well models for the gain and spontaneous recombination (see
Sections 3.9.7 “Unstrained Zincblende Models for Gain and Radiative Recombination”
and 3.9.10 “Strained Wurtzite Three-Band Model for Gain and Radiative
Recombination”) by specifying QWELL on the REGION statement for the appropriate
regions (see Section 13.6 “General Quantum Well Model”).
5. Specify the locations of the electrodes using ELECTRODE statements.
6. Use DOPING statements to specify additional doping or composition fraction information.
This usually isn’t necessary since most, if not all doping and composition information can
be specified during the specification of regions.

504 ATLAS User’s Manual

Simulating Vertical Cavity Surface Emitting Lasers VCSEL Simulator

Setting Up Cylindrical Symmetry

VCSELs use cylindrical symmetry to account for the 3D nature of VCSEL devices. To
specify cylindrical symmetry in the MESH statement, use the CYLINDRICAL parameter. An
example of specifying cylindrical symmetry is down below.
Example
MESH CYLINDRICAL
The axis of symmetry for cylindrical is the Y axis at X=0. When using cylindrical symmetry,
no X coordinates should be specified with negative values since X represents radius.
The usual abbreviation rules for parameter names can be used. See Section 2.5 “The ATLAS
Syntax” for more information.
Specifying the Mesh
To specify the mesh in the X direction, use the X.MESH statement. Each X.MESH statement
specifies the location of one mesh line in the X direction and the spacing between mesh lines
at that location.
The location is specified by the LOCATION parameter in microns. The spacing is specified by
the SPACING parameter, which is also in microns. The number of mesh lines specified is
limited to built-in limits in the overall size of the mesh. There should be, however, at least two
X.MESH statements. There shouldn’t be X.MESH specifications with negative values in the
LOCATION parameter. Such specification would be ill-defined due to the cylindrical
symmetry.
You should also specify mesh lines at any defining edges of regions, electrodes, or other
geometrical aspects in the X direction Also for similar reasons, use the Y.MESH statements to
specify mesh lines in the Y direction.
To have mesh lines inserted automatically at defined region edges, use the REGION and DBR
statements. To enable this feature, specify AUTO in the MESH statements. See Section 21.34
“MESH” for information about the MESH statement.
Specifying Regions
A region is a volume (in cylindrical coordinates a disk or annulus) that has a uniform material
composition. The region is specified by a REGION statement. In the REGION statement, the
material composition is specified by the MATERIAL parameter.
The MATERIAL parameter can take on values of any material name as described in Appendix
B “Material Systems”. For Ternary and Quaternary materials, you can also specify
composition fractions X or Y or both, using the X.COMPOSE or Y.COMPOSE parameters or
both. You can use both the DONORS and ACCEPTORS parameters to specify uniform densities
of ionized donors or acceptors or both.
The thickness of the region in the Y direction in microns is specified by the THICKNESS
parameter. You can also use the X.MIN, X.MAX, Y.MIN, and Y.MAX parameters to specify the
location and extent of the region. Except in special cases, we don’t recommend this because
of the difficulties aligning the numerous layers involved.
Unless it’s specified, the region will encompass the entire range of mesh in the X direction as
specified in the X.MESH statements. Mesh lines in the Y direction will then be added at the
upper and lower edges of the region. The mesh spacing at the edges is specified in microns by
the SY parameter.

505 ATLAS User’s Manual

Simulating Vertical Cavity Surface Emitting Lasers VCSEL Simulator

The TOP/BOTTOM parameters specifies the relative ordering of regions in the Y direction.
Specifying TOP indicates that the region starts at the top of all previously specified regions
(i.e., at the minimum previously specified Y coordinate) and extends to the thickness in the
negative Y direction. Remember for device simulation, the Y axis conventionally extends
down into the wafer for increasing values of Y. If no previous regions are specified, TOP
indicates the region starts at Y=0 and extends to the thickness in the negative Y direction.
BOTTOM indicates that the region starts at the bottom of all previously specified regions (i.e.,
at the maximum previously specified Y coordinate) and extends to the thickness in the
positive Y direction. If no previous regions are specified, BOTTOM indicates the region starts at
Y=0 and extends to the thickness in the positive Y direction.
Conventionally, we have found it convenient to start the active layers at the Y=0 location (as
done in the standard examples). In ordering, REGION statements:
1. Start with the upper DBR by using the TOP parameter.
2. Use the TOP parameter on the top side contact layers.
3. Use the BOTTOM parameter on the active layers.
4. Use the BOTTOM parameter on the lower DBR.
5. Use the BOTTOM parameter on the substrate/lower contact regions.
Specifying Distributed Bragg Reflectors
A Distributed Bragg Reflector (DBR) is a periodic structure composed of regions of two
alternating material compositions. A DBR can be specified by a series of REGION statements
alternating between two different material composition. DBRs, however, typically consist of
many of such layers and specifying them with REGION statements can be tedious, which could
be prone to errors. You can use the DBR statement to simplify the specification instead.
Generally, the DBR statement (alias SUPERLATTICE) can be used to specify any superlattice
composed of layers of two alternating material compositions. For more information about this
statement, see Section 21.9 “DBR”.
The MAT1 and MAT2 parameters specify the material names of the two materials, which is used
the same way the MATERIAL parameter of the REGION statement is used to specify a single
material.
The X1.COMP, Y1.COMP, X2.COMP, and Y2.COMP parameters are used to specify the X and Y
composition fractions of the two materials for ternary and quaternary materials.
The thicknesses of the two layers are specified by the THICK1 and THICK2 parameters of the
DBR statement. The doping for the layers are specified by the NA1 and NA2 parameters for
acceptors. The ND1 and ND2 parameters for donors.
The mesh spacing for the DBR can be specified using the SPA1 and SPA2 parameters, just like
the SY parameter of the REGION statement. You can also use the N1 and N2 parameters to
specify the integer number of the mesh divisions in the introduction of Y direction for each of
the layers.
The total number of layers is specified by the integer HALF.CYCLE parameter. The
HALF.CYCLE parameter specifies the total number of layers. Thus, an odd values specifies
that one more layer of the material 1 is to be used than of material 2.
The TOP and BOTTOM parameters of the DBR statement are used just like the REGION statement,
except the first material to be added to the top/bottom of the device is always of material 1.
DBR and REGION statements can be intermingled indiscriminately.

506 ATLAS User’s Manual

Simulating Vertical Cavity Surface Emitting Lasers VCSEL Simulator

Specifying Oxide Apertures

Once you specify all the epitaxial layers of the VCSEL, you can then specify an oxide
aperture by using a REGION statement with the MATERIAL parameter assigned to an insulator
and specify the geometry by using the X.MIN, X.MAX, Y.MIN, and Y.MAX parameters.
Specifying Quantum Wells
You can implement single and multiple quantum wells as described in Section 13.6 “General
Quantum Well Model”.
Specifying Electrodes
Device electrodes are specified using the ELECTRODE statement. Note that electrodes in
device simulation are treated as boundary conditions. Therefore, there’s no advantage to
resolve the full outline of the electrode. Usually, the easiest way is to place electrodes at the
top and bottom of the structure. To do this, specify the TOP or BOTTOM parameters of the
ELECTRODE statement. The extent along the top or bottom of the device can be described by
the X.MIN or X.MAX or both parameters. If you omit these parameters the electrode will
extend all the way along the corresponding surface of the device.
Use the NUMBER parameter to assign electrodes to a number. Electrodes should be assigned
consecutive numbers starting at 1. They can also be given a name by using the NAME
parameter. They can also be addressed later by their name or number when modifying the
electrode characteristics, using the CONTACT statement or assigning biasing information in the
SOLVE statements.

9.3.2 Specifying VCSEL Physical Models and Material Parameters

Much of the simulation results depend on how the physical models are described. For more
information about physical models, see Section 3.6 “Physical Models”.
Selecting Device Models
For most laser devices including VCSELs, we advise that you at least enable predominant
recombination mechanisms, since these usually define the threshold characteristics and
compete with stimulated emission for carrier recombination. Thus, directly impact the
efficiency of the device.
Typically, three bulk recombination mechanisms should be addressed: Schockley-Read-Hall
(SRH), Auger, and radiative recombination. You can enable these mechanisms by specifying
SRH, AUGER, and OPTR parameters on a MODEL statement.
The parameters of these models are specified on MATERIAL statements. These include TAUN
and TAUP, which specify the SRH lifetimes. AUGN and AUGP, which specify the Auger
coefficients. And COPT, which specifies the radiative recombination rate coefficient.
These models and associated parameters are discussed in more detail in Section 3.6.3 “Carrier
Generation-Recombination Models”.

507 ATLAS User’s Manual

Simulating Vertical Cavity Surface Emitting Lasers VCSEL Simulator

Specifying Material Parameters

For LASER or VCSEL simulation there are several other material parameters that should be
considered. First and foremost, specify the high frequency dielectric relative permittivity of
each material region. This can be specified as given by Table 9-2.
The high frequency dielectric enters in the Helmholtz equation and directly affects which
laser mode may be present. In particular, try to consider the design of the DBRs carefully.
Typically, the objective of the design of the DBRs is to choose the layer thicknesses so that
they are one quarter wavelength thick relative to the local dielectric. During this design
process, you need to also carefully consider the band-gap of the active region, since this
affects the energy and wavelength of the maximal gain. A good design will try to operate near
the peak gain, while satisfying the conditions for oscillation between the DBR mirrors.
Another important consideration for choosing material parameters is the selection of band
edge parameters such as band gap and electron affinity. Although defaults do exist for many
material systems it is almost always better for you to specify the best estimate. Also, consider
the mobilities since they directly affect series resistance.
Section 3.6.1 “Mobility Modeling” has more information about these parameters. For more
about MATERIAL statements, see Section 21.32 “MATERIAL”.

9.3.3 Enabling VCSEL Solution

VCSEL Solution Mesh
The solution to Helmholtz equation is performed on a spatially discrete domain or mesh. Use
of matrix method for the solution of the longitudinal wave equation allows the reduction in
the number of mesh points in y (z in VCSEL) to the number of material boundaries. This
significantly improves calculation time while maintaining the accuracy. The mesh points in
the transverse direction x (r in VCSEL) coincide with the device mesh.
Specifying VCSEL Parameters
Specify the VCSEL statement to enable the VCSEL simulator. Once you enable it, the
semiconductor device equations are solved self-consistently with the photon rate equations
and the Helmholtz equation.
To enable reflectivity test simulation, do the following:
1. Specify the VCSEL.CHECK parameter in a VCSEL statement.
2. Specify the VCSEL.INCIDENCE parameter to monitor cold cavity reflectivity for light
incident on a structure from either top or bottom of the structure.
• VCSEL.INCIDENCE = 1 - The light incident from the top.
• VCSEL.INCIDENCE = 0 - The light incident from the bottom.
• VCSEL.INCIDENCE = 2 or >2 - Both directions of light incidence are considered. The
program compares cavity resonant frequencies obtained for each direction of
incidence. If results do not agree within a certain tolerance, VCSEL:ATLAS will
close with an error.
By default, light is incident from the top of the structure.
3. Specify INDEX.TOP for the refractive index of the medium above the structure.
4. Specify INDEX.BOTTOM for the refractive index of the medium below the structure.
Default medium above and below the structure is air.

508 ATLAS User’s Manual

Simulating Vertical Cavity Surface Emitting Lasers VCSEL Simulator

5. Specify the PHOTON.ENERGY or OMEGA parameters for (initial) photon energy or

frequency.
6. Specify the EINIT and EFINAL paprameters. These parameters set the photon energy
range in reflectivity test. If not specified, they take the following default values:
EINIT=0.8 PHOTON.ENERGY and EFINAL=1.2 PHOTON.ENERGY.
7. Specify NSPEC for the number of sampling points between EINIT and EFINAL. The
default value is NSPEC=100.
You can specify the following optional parameters in the VCSEL simulation:
• NMODE specifies the number of transverse modes of interest. Default value is NMODE=1.
• PROJ enables faster solution of the photon rate equations below threshold. You should
disable this solution scheme when the bias reaches lasing threshold. Specify ^PROJ in a
VCSEL statement to do that.
• PERTURB enables faster solution of the Helmholtz equation. When specified, the program
solves the longitudinal wave equation only once. Perturbational approach is used to
account for the refractive index changes [304].
• LOSSES specifies any additional laser losses.

9.3.4 Numerical Parameters

The following numerical parameters control the performance of the VCSEL simulation.
• TOLER sets the desired relative tolerance of the photon density calculation. The default
value is 0.01. Setting this parameter to a lower value may slow down the calculation
significantly. Using a larger value will result in a faster but less accurate calculations.
• ITMAX sets maximum number of external VCSEL iterations during photon density
calculation. The default value is 30.
• TAUSS is an iteration parameter used in the calculation of photon densities. Using a larger
value of this parameter can speed up the calculation but may cause convergence
problems.
• MAXCH is the maximum allowed relative change of the photon density between VCSEL
iterations. Using a larger value of this parameter can speed up the calculation but may
cause convergence problems.

9.3.5 Alternative VCSEL Simulator

If you need to model a vertical cavity laser that supports multiple longitudinal modes, use the
LASER statement to enable VCSEL solution.
Specify the LASER statement to enable the laser simulator. Once you enable it, the
semiconductor device equations are solved self-consistently with the laser rate equations. The
major difference between this simulator and the one discussed above is in the approach to the
solution of the Helmholtz equation.
The solver used in this case allows multiple longitudinal modes, but it is not as efficient in
dealing with structural variations in the transverse direction.
There are several kinds of parameters that appear in the LASER statement. Most of the
parameters available are discussed above with respect to VCSEL statement.

509 ATLAS User’s Manual

Simulating Vertical Cavity Surface Emitting Lasers VCSEL Simulator

Specify laser physical parameters and models, and specify the DBR1.START, DBR1.FINAL,
DBR2.START, and DBR2.FINAL parameters in the LASER statement.
The mirror loss is defined as:

 mir = ------- ln  -----------

1 1
9-20
2L R f R r

where L is the cavity length given as:

L=DBR2.START-DBR1.FINAL
Rf and Rr are the front and rear mirror reflectivities and are calculated from the solution of
Helmholtz equation in the range specified by DBR1.START, DBR2.START, and DBR2.FINAL.
LASER simulates single or multiple longitudinal modes as well as single or multiple
transverse modes. Also with LASER, you can simulate index guiding only or gain guiding.
Simulation of gain guiding takes more computation time but should be considered when
spatially varying gain effects are anticipated.
If you to calculate the laser spectrum, specify the multiple longitudinal modes model and
additional parameters. Do the following:
1. Specify the LMODES parameter in the statement. This enables the multiple longitudinal
mode model.
2. Specify the EINIT and EFINAL parameters. These parameters set the photon energy range
within, which LASER will take into account multiple longitudinal modes. Make sure
initial photon energy is within this range and is specified for the active lasing region.
3. Specify the photon energy separation (ESEP parameter). If this isn’t specified, LASER will
automatically calculate the number of longitudinal modes based on the cavity length and
the energy range. We recommend that you allow LASER to choose the photon energy
separation.
4. Specify the spectrum file name (SPEC.SAVE). LASER will produce a structure file
containing spectrum data after calculation of each bias point. LASER will automatically
append _dcN.log to the specified file name (where N is the number of the bias point) for
steady-state solutions or _trN.log for a transient simulation. The first bias point where
LASER is active will have N=1. This is rarely the first bias point during simulation. Use
TONYPLOT to examine these files. If you specify the SPECSAVE parameter, it will only
save the spectrum files on every las.SPECSAVE solution.
The index in the spectrum file name will still increase by one each time. The spectrum data
can be stored in a single file for the transient simulation, only if the MULTISAVE-LASER
statement is set to FALSE. To enable multiple transverse mode solutions, specify the number
of transverse modes by using the NMODE parameter from the LASER statement.
By default, VCSELs simulate using the index guiding model. To simulate using the gain
guiding, specify INDEX.MODEL=1.

510 ATLAS User’s Manual

Simulating Vertical Cavity Surface Emitting Lasers VCSEL Simulator

9.3.6 Far Field Patterns

Sometimes it is convenient to describe the beam as a diffraction free beam [72, 278]. We first
model the near field pattern as a Gaussian by fitting to the near-field pattern calculated by
VCSEL as described above. The equation for the Gaussian pattern is given in Equation 9-21.
2
E  r  = exp –  ----
r
9-21
 w

Here, r is the radial coordinate and w is the fitting parameter. You can model the far-field
pattern as a Gaussian by specifying GAUSS on the VCSEL statement. The Bessell beam far-
field pattern described by Equation 9-22 can be selected by specifying BESSEL on the VCSEL
statement.

r 2
E  r  = J  A · BESSEL  r  exp –  ---- 9-22
 w

Here J is the zero order Bessel function and A.BESSEL is a user specifiable parameter with
units of 1 on the VCSEL statement.
You can choose a combination Gaussian/Bessel funciton by specifying GBESSEL on the
VCSEL statement. This function is described by Equation 9-23.

E  r  = J  A · BESSEL  r  9-23

You can save near and far field patterns by specifying a value for the PATTERNS parameter in
the SAVE statement. The value of the PATTERNS parameter is a character string representing
the root name of a file for saving the near and far field patterns. The near field pattern is saved
to a file with the string .nfp appended to the root name. The far field pattern is saved to a file
with the string .ffp appended to the root name. Use TONYPLOT to examine these files.

511 ATLAS User’s Manual

Semiconductor Laser Simulation Techniques VCSEL Simulator

9.4 Semiconductor Laser Simulation Techniques

The most common technique for simulating laser diodes is to simulate forward characteristics
with a gradually increasing bias. The forward voltage is usually specified. You can use,
however, current boundary conditions and include external elements.
To save computational time, we recommend that you use the VCSEL statement. The effective
frequency method employed for the solution of the Helmholtz equation in this case allows
you to consider structures with oxide apertures or other structural variations in transverse
direction. We also recommend that you test the structure using reflectivity experiment
simulation prior to applying the bias.
If you use the multiple longitudinal mode model, computational time will be longer and
strongly dependent on the number of longitudinal modes involved in the calculation.
Simulating multiple transverse modes takes more computation time and should only be
preformed when the higher order modes are of interest.

512 ATLAS User’s Manual

Chapter 10
Luminous: Optoelectronic
Simulator
Overview Luminous: Optoelectronic Simulator

10.1 Overview
LUMINOUS is a general purpose light propagation and absorption program integrated into the
ATLAS framework. LUMINOUS and LUMINOUS3D calculate optical intensity profiles within
the semiconductor device. The intensity profiles are then converted into photogeneration
rates, which are directly integrated into the generation terms in the carrier continuity
equations. This unique coupling of tools allows you to simulate electronic responses to
optical signals for a broad range of optical detectors. These devices include but are not limited
to pn and pin photodiodes, avalanche photodiodes, Schottky photodetectors, MSMs,
photoconductors, optical FETs, optical transistors, solar cells, and CCDs.
The following sections address various types of optoelectronic devices. Go to the sections
that are most relevant to your application. We strongly recommend, however, that you read
the other sections too.

Note: You should be familiar with ATLAS and either S-PISCES or BLAZE before you can use LUMINOUS. If
not, read Chapter 2 “Getting Started with ATLAS” and either Chapter 4 “S-Pisces: Silicon Based 2D
Simulator” or Chapter 5 “Blaze: Compound Material 2D Simulator”.

The propagation of light can be described by any one of 4 physical models or several user-
defined approaches. The physical models for light propagation include the following.
• Ray tracing (RT), which is described in Section 10.2 “Ray Tracing”. Ray tracing is a
general method of resolving 2D and 3D non-planar geometries but ignores coherence and
diffraction effects.
• The transfer matrix method (TMM), which is described in Section 10.3 “Matrix Method”.
The matrix method is a 1D method that includes interference effects. This method is
recommended for large area devices such as thin film solar cells.
• The beam propagation method (BPM), which is described in Section 10.4 “Beam
Propagation Method in 2D”. BPM is a general 2D method that includes diffraction effects
at an increased computational expense.
• Finite difference time domain (FDTD) methods, which are described in Section 10.5
“Finite Difference Time Domain Analysis”. FDTD is the most general 2D and 3D method
and accounts for both diffraction and coherence by direct solution of Maxwell’s
equations. This is the most computationally expensive approach.
The user-definable methods are described in Section 10.6 “User-Defined Photogeneration”.
The remaining sections in this chapter discuss the various other common topics in
photodetector simulation.

514 ATLAS User’s Manual

Overview Luminous: Optoelectronic Simulator

10.1.1 Hybrid Source Specifications

In some cases, it may be useful to specify different propagation models for different portions
of the optical spectrum. You can do this by specifying the same beam index (specified by
NUMBER on the BEAM statement) for more than one BEAM statement. In doing so, you should be
careful to ensure that the portions of the spectrum specified in subsequent BEAM statements by
WAVEL.START and WAVEL.END or E.START and E.END do not overlap.

Note: The overlap conditions of the spectral sampling is not currently checked in the LUMINOUS software as we
anticipate other applications for the general capability of hybrid beams.

When the solution for the intensity of a specified beam on the SOLVE statement is requested,
the intensity applies to all parts of the hybrid total. This allows for application of for example
ray tracing at longer wavelengths and FDTD at short wavelengths for a multispectral source.
This can produce significant advantages in flexbly optimizing input decks for total run time.

515 ATLAS User’s Manual

Ray Tracing Luminous: Optoelectronic Simulator

10.2 Ray Tracing

Optoelectronic device simulation is split into two distinct models that are calculated
simultaneously at each DC bias point or transient timestep.
1. Optical ray trace using real component of refractive index to calculate the optical
intensity at each grid point
2. Absorption or photogeneration model using the imaginary component of refractive index
to calculate a new carrier concentration at each grid point.
This is followed by an electrical simulation using S-PISCES or BLAZE to calculate terminal
currents.

Note: LUMINOUS assumes that the refractive indices, both real and imaginary, are constant within a particular
region.

10.2.1 Ray Tracing in 2D

Defining The Incident Beam
An optical beam is modeled as a collimated source using the BEAM statement. The origin of
the beam is defined by parameters X.ORIGIN and Y.ORIGIN (see Figure 10-1). The ANGLE
parameter specifies the direction of propagation of the beam relative to the X axis. ANGLE=90
is vertical illumination from the top. MIN.WINDOW/MAX.WINDOW parameters specify the
illumination window. As shown in Figure 10-1, the Illumination Window is “clipped” against
the device domain so that none of the beam bypasses the device.
The beam is automatically split into a series of rays so that the sum of the rays covers the
entire width of the illumination window. When the beam is split, ATLAS automatically
resolves discontinuities along the region boundaries of the device.

Figure 10-1: Optical Beam Geometry

516 ATLAS User’s Manual

Ray Tracing Luminous: Optoelectronic Simulator

Although the automatic algorithm is usually sufficient, you can also split the beam up into a
number of rays using the RAYS parameter. Each ray will then have the same width at the beam
origin and the sum of the rays will cover the illumination window. Even when you specify the
RAYS parameter, ATLAS will automatically split the rays to resolve the device geometry.
Ray Splitting At Interfaces
Rays are also split at interfaces between regions into a transmitted ray and a reflected ray.
Figure 10-2 illustrates the difference between rays that are split to resolve the geometry and
transmitted/reflected rays split at a region interfaces.

Figure 10-2: Reflected and Transmitted Rays

In Figure 10-2, the incident rays come in from the top left. They intersect a interface between
two material regions with differing refractive indices. Within this interface lies a geometric
point where the normal to the interface changes. This implies that the angles of reflection and
transmission will be different for light incident to the left of the point from light incident on
the right. Thus, the incident rays are split to resolve the interface point. The second level of
splitting occurs at the interface itself. Here, the incident rays are split into reflected and
transmitted rays.

517 ATLAS User’s Manual

Ray Tracing Luminous: Optoelectronic Simulator

10.2.2 Ray Tracing in 3D

In 3D, a simpler algorithm is used for ray tracing because the 2D algorithms become
intractable in 3D. Therefore, the algorithm doesn’t automatically split rays to resolve
topological features of the device. You should specify enough rays to resolve such features to
the desired accuracy. But there’s a trade off between the computation time and accuracy in
specifying the number of rays.
The 3D source geometry is very similar to the geometry shown in Figure 10-1. In 3D, the
source origin is specified by three coordinates: X.ORIGIN, Y.ORIGIN, and Z.ORIGIN. Unlike
Figure 10-1, there are two angles that describe the direction of propagation.

Figure 10-3: Propagation in 3D

Here, the angle PHI (aliased to ANGLE) describes the angle or propagation with respect to the
XZ plane. The angle THETA describes the angle of rotation specified about the Y axis. These
angles are on the BEAM statement.
In 3D, the clipping of the window of propagation is done in two directions. In the X direction,
this is described by XMIN and XMAX. XMIN and XMAX are aliases of MIN.WINDOW and
MAX.WINDOW. In the Z direction, the window is described by ZMIN and ZMAX.
The numbers of rays in the X and Z directions are described by NX and NZ. Ray samples are
taken at regular intervals over the source window.
The discrete sampling of the source beam into rays in LUMINOUS3D is unlike that done in
LUMINOUS. In LUMINOUS, the source beam is automatically broken up into a set of rays that
resolve the device topology and variations in the interior of the device. In LUMINOUS3D, this
process is more complex and is a computationally intractable. As such in LUMINOUS3D,
specify a discrete sampling of the source beam (see Figure 10-4).

518 ATLAS User’s Manual

Ray Tracing Luminous: Optoelectronic Simulator

Figure 10-4: Source beam sampling

In Figure 10-4, the extent of the source sampling is specified by the XMIN, XMAX, ZMIN and
ZMAX parameters of the BEAM statement. Even samples are taken along each of the beam front
principal axes. The number of samples in the X and Z directions are given by the NX and NZ
parameters of the BEAM statement.

519 ATLAS User’s Manual

Ray Tracing Luminous: Optoelectronic Simulator

10.2.3 Reflection and Transmission

Figure 10-5 shows the relationship between the angles of incidence (i), reflection (r), and
transmission (t) at the interface between two media. These coefficients are calculated as a
function of the refractive indices in the two media.

Figure 10-5: Angles of incidence, reflection and transmission

The reflection and transmission coefficients of the light for parallel and perpendicular
polarization are calculated as shown in Equations 10-1 to 10-6.
n 2 cos  i – n 1 cos  t
E r = ---------------------------------------------- E i (parallel p or TE polarization) 10-1
n 1 cos  t + n 2 cos  i

2n 1 cos  i
E t = ---------------------------------------------- E i (parallel p or TE polarization) 10-2
n 1 cos  t + n 2 cos  i

n 1 cos  i – n 2 cos  t
E r = ---------------------------------------------- E i (perpendicular s or TM polarization) 10-3
n 1 cos  i + n 2 cos  t

2n 1 cos  i
E t = ---------------------------------------------- E i (perpendicular s or TM polarization) 10-4
n 1 cos  i + n 2 cos  t

Er 2
R =  ----- (perpendicular s or TM polarization) 10-5
 Ei

E t 2 n 2 cos  t
T =  ----- ----- ------------- (perpendicular s or TM polarization) 10-6
Ei n 1 cos  i

520 ATLAS User’s Manual

Ray Tracing Luminous: Optoelectronic Simulator

where Ei is electric field of the incident wave, Er is the field of the reflected wave, Et is the
field of the transmitted wave. R is the reflection coefficient, T is the transmission coefficient,
n1 is the refractive index on the incident side and n2 is the refractive index on the transmission
side.

Note: Here, we have adopted Yee’s notation for TM (Transverse Magnetic) and TE (Transverse Electric) fields
[321].

The angles of reflection and transmission are given in Equations 10-7 and 10-8.
r = i 10-7

n 1 sin  i = n 2 sin  t 10-8

where i is the angle of incidence, t is the angle of transmission, and r is the angle of
reflection.
Specifying Reflections
By default, no reflected rays are traced. The transmission coefficients are properly calculated
for the transmitted rays. The REFLECTS=<i> parameter is used to set an integer number of
reflections to consider. Note that setting a very large number of reflections can lead to
extremely long simulation times for the ray trace.
One convenient way to overcome the long CPU times is to use the MIN.POWER parameter.
This terminates each ray when the optical power falls to the fraction of the original power
defined by this parameter.
Front Reflection
By default, reflection and refraction at the first interface (the initial interface with the device)
are ignored. The first reflection coefficient is zero and the transmission coefficient is one. The
polarization and angle of the transmitted ray at the first interface is identical to the
polarization and angle of the incident beam.
If the FRONT.REFL parameter of the BEAM statement is specified, the transmission coefficient
is calculated using Equations 10-1 to 10-6. When the transmission coefficient is calculated,
it’s assumed that the material outside the device domain is a vacuum. The transmitted rays are
attenuated by the transmission coefficient but the reflected ray is not traced.
Back Reflection
By default, the reflection at the back of the device are ignored. No reflected ray is traced once
the back of the device is reached. If the BACK.REFL parameter is specified, the backside
reflection coefficient is calculated (again assuming a vacuum outside the device) and the
back-side reflected ray is traced.
Sidewall Reflection
By default, the reflection from the sides of the device are ignored. No reflected ray is traced
back into the structure. As above, BACK.REFL is used to enable the sidewall reflections
assuming a vacuum outside the device.

521 ATLAS User’s Manual

Ray Tracing Luminous: Optoelectronic Simulator

User Specified Reflection

You can specify the value of the reflection coefficient at an interfaces by specifying the
REFLECT parameter of the INTERFACE statement. The value of REFLECT must be between 0.0
and 1.0 inclusive, and the OPTICAL parameter must also be specified on the statement. You
should also specify the values of P1.X, P1.Y, P2.X, and P2.Y.

10.2.4 Periodic Boundaries

In certain cases, you may find it convenient to take advantage of periodicity to reduce
computation. You can do so in ray tracing by specifying PERIODIC on the BEAM statement.
When you specify PERIODIC, all rays intersecting the extreme edges of the device in the x
and z directions perpendicular to these axes will continue at the opposite edge of the device
without reflection or refraction.
Discontinuous Regions
You can simulate devices electrically where a single region is defined as two or more
separated areas. The ray tracing algorithm, however, doesn’t support such structures. If a
structure has two separate areas with the same region number, we recommend you using
DEVEDIT to renumber the regions, perhaps even creating a new region number for each area.

Note: This limitation is only for two separated areas with the same region number and not for two regions with
different region numbers of the same material. This latter case can be simulated.

10.2.5 Diffusive Reflection [327]

For optical ray-tracing semiconductor interfaces may be modelled as having diffusive
reflections. To enable this model, specify the DIFFUSIVE parameter on the INTERFACE
statement. When using this model, you must also specify the region of incidence. To do this,
use the REGION parameter of the INTERFACE statement.
You can also specify the fraction of the incident power that experiences specular reflection
using the SPECULAR parameter of the INTERFACE statement. The fraction of the incident
power that is lost to absorption can be specified using the ABSORPTION parameter of the
INTERFACE statement.
By default, the diffusive reflection has a Lambertian distribution (having constant luminance
with respect to angle). You can also choose a Gaussian or Lorenz distribution by specifying
either GAUSS or LORENZ respectively on the INTERFACE statement. The Gaussian function is
described by Equation 10-9.
 
2 
–  ----------------------------------------------------2-
1  2 DISPERSION 
F    = ------------------------------------- e 10-9
DISPERSION

Here, F() is the scattering function.  is the dispersion angle and DISPERSION is a user
definable parameter on the INTERFACE statement specifying the relative spread of the
dispersion.

522 ATLAS User’s Manual

Ray Tracing Luminous: Optoelectronic Simulator

The Lorenz function is described by Equation 10-10.

DISPERSION
F    = ---------------------------------------------------- 10-10
2 2
 + DISPERSION

When the DISPERSION parameter is set to zero no disperesion takes place.

The SCATTERED parameter on the INTERFACE statement specifies the number of rays per
scattering event.

10.2.6 Light Absorption and Photogeneration

The cumulative effects of the reflection coefficients, transmission coefficients, and the
integrated loss due to absorption over the ray path are saved for each ray. The generation
associated with each grid point can be calculated by integration of the Generation Rate
Formula (Equation 10-11) over the area of intersection between the ray and the polygon
associated with the grid point.
P – y
G =  0 ------- e 10-11
hc

where:
• P is the ray intensity factor, which contains the cumulative effects of reflections,
transmissions, and loss due to absorption over the ray path.
• o is the internal quantum efficiency, which represents the number of carrier pairs
generated per photon observed.
• y is a relative distance for the ray in question.
• h is Planck’s constant
• is the wavelength.
• c is the speed of light.
• is the absorption coefficient given by Equation 10-12.
4
 = ------ k 10-12


where k is the imaginary part of the optical index of refraction.

Photogeneration on a Non-uniform Mesh
The photogeneration algorithm used integrates the optical intensity around each node point.
This is done to ensure that the total photogeneration rate isn’t grid sensitive. A uniform
photogeneration rate is defined as a constant value of (photogeneration rate at any
node)*(element area around the node). In TONYPLOT, a uniform photogeneration rate may
appear to vary across a non-uniform mesh density.

523 ATLAS User’s Manual

Ray Tracing Luminous: Optoelectronic Simulator

Photogeneration at Contacts
The photogeneration associated with nodes that are also defined as electrodes is a special
case. The electrical boundary conditions require that the carrier concentration at electrode
nodes equals the doping level. This means that photogeneration at nodes that are electrodes
must be zero. But just setting these nodes to zero photogeneration will typically cause an
apparent drop in quantum efficiency.
The photogeneration rate at the contact nodes is calculated as usual. This photogeneration
rate, however, is applied to the neighboring node inside the semiconductor. This means for a
uniform mesh and photogeneration rate, if the photogeneration rate is 1.01017 pairs/cm-3s,
then the nodes at the contacts will have zero photogeneration and the next node into the
semiconductor will have 2.01017 pairs/cm-3s.

10.2.7 Outputting the Ray Trace

For example, in LUMINOUS the rays generated during ray tracing are stored in all subsequent
structure files. In LUMINOUS3D, this isn’t the case. To save the rays in LUMINOUS3D, set the
RAYTRACE parameter of the BEAM statement to the name of a file where the results of the ray
trace are to be stored.

10.2.8 Monte Carlo Ray Trace

The Monte Carlo ray trace algorithm is a means of simulating random textured interfaces
found in photodetectors, such as solar cells. The idea behind the algorithm is to characterize
interactions between light and structural interfaces using probabilistic estimation through
random number generation.
The algorithm is described as follows:
First, you should specify MONTE.CARLO on the BEAM statement. You can then specify the
number of rays using the RAYS parameter. Remember that the more rays specified the larger
the statistical sampling and presumably the better the characterization. You should try
increasing the number of rays until you see convergence in the value of available
photocurrent to a value invariant with the number of rays.
In 3D, you may also use the NX and NZ parameters to specify the number of rays, but this
approach simply forms the product NX*NZ to obtain the total number of rays.
Individual rays are started at uniformly random locations over the source domain usually
specified by the MIN.WINDOW and MAX.WINDOW parameters or defaulted by clipping to the
device (see Figure 10-1). In 3D, the source domain is defined by the XMIN, XMAX, ZMIN, and
ZMAX parameters.
Each ray is traced in the direction of propogation until it is incident with a structural interface
between regions or a region and the outside of the device. When an interface is encountered,
the reflection and transmission probabilities are calculated using Equations 10-5 and 10-6.
Rather than trace both the transmitted and reflected ray as is done in standard ray tracing, a
uniform random number is generated and compared against the transmission (or reflection)
probability to determine which ray to trace. Only one ray is traced per interaction at an
interface.

524 ATLAS User’s Manual

Ray Tracing Luminous: Optoelectronic Simulator

Interfaces can also be given textural characteristics using the angular distribution functions,
ADFs, as described by the INTERFACE statement (see Equations 10-40 through 10-45). When
a ray is incident on an interface with an assigned ADF, the surface normal is randomly
modified to obey the probability distribution of the ADF. The modified angle is used to
calculate the transmission and reflection probabilities as well as the transmission and
reflection angle. The transmission and reflection probabilities are used in the standard way
described above to determine which ray is traced.
The key to improved collection efficiency of textured surfaces is that light reflecting from a
angled facet may be reflected back into the textured surface and improve the net transmission
of light. These multi-reflected rays are accounted for in the MC ray trace algorithm by
examining the direction of the transmitted or reflected ray relative to the surface normal. If a
reflected ray is reflected in the direction of the interface or a transmitted ray is transmitted in
the direction of the source, the algorithm is recursively reapplied until the ray is reflected
from or transmitted through the interface.

525 ATLAS User’s Manual

Matrix Method Luminous: Optoelectronic Simulator

10.3 Matrix Method

Matrix method presents a fast and accurate way to simulate electromagnetic wave
propagation through a layered medium. Among a variety of approaches to matrix theory, the
most commonly used are the characteristic matrix and the transfer matrix descriptions.
Using the matrix method is not limited to LUMINOUS only. It also finds its application in
VCSEL and LED modules. ATLAS uses the characteristic matrix approach that relates total
tangential components of the electric and magnetic fields at the multilayer boundaries. More
often used transfer matrix, which relates tangential components of the electric field of left-
travelling and right-travelling waves, has several disadvantages that limit its applicability.
The structure of a multilayer completely determines the characteristic matrix of this
multilayer. The transfer matrix also contains information about the media on both sides of the
multilayer. The characteristic matrix approach is more general and can be expanded to deal
with graded interfaces, in which case, the transfer matrix method is inappropriate to use.

10.3.1 Characteristic Matrix

First, we describe the characteristic matrix of a single layer. As mentioned earlier, the matrix
relates tangential components of the electric E(z) and magnetic H(z) fields at the layer
boundaries z=0 and z=d.
E0  E d
= M 10-13
H0 Hd

The matrix itself is

cos  j sin   Y
M = 10-14
jY sin  cos 

where
2
 = ------ nd cos  10-15


is the phase shift for the wave propagating through the layer, n is the complex refractive
index, and  is the angle of wave propagation in the layer (Figure 10-6). Y is the optical
admittance of the layer, which is for parallel (p or TE) and perpendicular (s or TM)
polarizations, is given by:

s 
Y = -----0- n cos  10-16
0

p 
Y = -----0- n  cos  10-17
0

where  and µ0 are permittivity and permeability of free space.

526 ATLAS User’s Manual

Matrix Method Luminous: Optoelectronic Simulator

The characteristic matrix of a multilayer is a product of corresponding single layer matrices.

If m is the number of layers, then the field at the first (z=z0) and the last (z=zm) boundaries are
related as follows:
E  Z0  E  Zm  cos  i j sin  i  Y i
= M 1 M 2 M m  Mi = 10-18
H  Z0  H  Zm  jY i sin  i cos  i

10.3.2 Reflectivity, Transmissivity, and Absorptance

The detailed derivation of amplitude reflection (r) and transmission (t) coefficients is given in
[227]. The resulting expressions are shown in Equations 10-19 and 10-20.
Y 0 M 11 + Y 0 Y s M 12 + M 21 – Y s M 22
r = -------------------------------------------------------------------------------------
- 10-19
Y 0 M 11 + Y 0 Y s M 12 + M 21 + Y s M 22

2Y 0
t = -------------------------------------------------------------------------------------
- 10-20
Y 0 M 11 + Y 0 Y s M 12 + M 21 + Y s M 22

where Y0 and Ys are the characteristic admittances of the media on both sides of the
multilayer. Mij are the elements of the characteristic matrix of the multilayer. Light is incident
from the medium with admittance Y0. The energy coefficients (reflectivity, transmissivity, and
absorptance) are given by:
2
R = |r| 10-21

Re  Y s  2
T = ------------------ t 10-22
Re  Y 0 

Re  Y s 
A =  1 – R  1 – ------------------------------------------------------------------------------------------
- 10-23
Re   M 11 + Y s M 12   M 21 + Y s M 22  

These expressions take into account imaginary part of refractive index. Therefore, they are
more general than Fresnel formulae (Equations 10-1 through 10-6). LUMINOUS automatically
uses these expressions when appropriate.

527 ATLAS User’s Manual

Matrix Method Luminous: Optoelectronic Simulator

n(z)
ns
n0 n

 z
 z=0 z=d

Figure 10-6: Refractive index profile of a single-layer coating

10.3.3 Transfer Matrix and Standing Wave Pattern

The characteristic matrix of a multilayer structure fully describes the optical properties of the
multilayer. It is not completely, however, independent of external parameters. If the angle of
incidence of light on a multilayer structure changes, the characteristic matrix changes too.
Equations 10-15, 10-16, and 10-17 show the dependence of the single layer matrix elements
on the angle of wave propagation in the layer. This implicit dependence on the angle of light
incidence makes the characteristic matrix approach similar to the transfer matrix method for
most practical purposes. In fact, when the multilayer structure does not contain any graded
layers, the two methods are equivalent.
The transfer matrix directly relates the electric field amplitudes of transmitted and reflected
waves to the amplitude of the incident wave. This property of the transfer matrix makes it the
method of choice in many optics applications.
The transfer matrix elements are related to the characteristic matrix elements by the following
expressions:
Q 11 =  M 11 + M 21  Y i + Y o M 12 + Y o  Y i M 22   2 10-24

Q 12 =  M 11 + M 21  Y i – Y o M 12 – Y o  Y i M 22   2 10-25

Q 21 =  M 11 – M 21  Y i + Y o M 12 – Y o  Y i M 22   2 10-26

Q 22 =  M 11 – M 21  Y i – Y o M 12 + Y o  Y i M 22   2 10-27

where Yi and Yo are the admittances of the input and output media defined according to
Equations 10-16 and 10-17 for (s) and (p) polarization components respectively.

528 ATLAS User’s Manual

Matrix Method Luminous: Optoelectronic Simulator

The amplitude reflection and transmission coefficients are:

r = Q 21  Q 11 10-28

t = 1  Q 11 10-29

In some devices that use multilayer structures, it is sufficient to know the transmittance and
reflectance of the structure in order to fully characterize the device response to incident light.
Other devices, however, such as thin film detectors, require the knowledge of light intensity
distribution within the multilayer structure. Therefore in addition to the light transfer
properties of such a structure, you need to be able to calculate the standing wave pattern
formed by interference of traveling wave components in each layer. Standard intensity based
ray-tracing calculations do not account for coherent effects and are unable to simulate
standing wave patterns. Beam Propagation Method and full-wave Maxwell solvers are
computationally expensive and are not useful for structures containing large number of
layers. In this case, the numerical solution of the Helmholtz equation using the transfer matrix
method is an attractive alternative.
Although this approach is inherently one-dimensional, it is known to produce highly accurate
results for practical devices in two dimensions. Important requirements are:
• Lateral feature sizes in the device are much larger than typical layer thicknesses in a
multilayer.
• Incident light is normal to the surface or is at small angles to the surface normal. When
these requirements are met, the device can be divided into several lateral regions in which
the structural variations occur in vertical direction z only.
For each lateral region, the one-dimensional Helmholtz equation is solved using the transfer
matrix method.
First, we find reflected component of electrical field at z=0:
E r = rE i

where Ei and Er are the amplitudes of incident and reflected waves respectively. The electric
field and magnetic field at z=0 are:
E  0  = Ei + Er 10-30

H  0  =  E r – E i Y i 10-31

Then, the fields at layer interfaces can be found step by step using one layer characteristic
matrices. The electric field at any point within a certain layer is given by:
E  z  = E  z j  cos    – iH  z j  sin     Y  j  10-32

H  z  = H  z i  cos    – iE  z i  sin     Y  i  10-33

529 ATLAS User’s Manual

Matrix Method Luminous: Optoelectronic Simulator

where E(zj) and H(zj) are electric and magnetic fields at layer interface between layers (j) and
(j-1), Y(j) is the optical admittance of the layer (j) and phase is given by:
 = 2n cos     z – z j    10-34

This way, intensity profile is generated throughout the device structure. The photogeneration
rate is given by:

 Ez 2
G  z  = ------  -------------- 10-35
hc 2

To enable transfer matrix method for calculation of intensity distribution and photogeneration
profiles in thin film detectors, specify TR.MATRIX parameter on a BEAM statement.
For convenience, the parameters are shown in Table 10-1. The SUB.INDEX and SUB.EXTINCT
parameters allow specification of the complex index of refraction at the exit side of the layer
stack. Similarly, you can use the AMBIENT.INDEX parameter of the BEAM statement to specify
the real index on the input side of the layer stack.

Table 10-1 External Index Parameters

Statement Parameter Default

BEAM AMBIENT.INDEX 1.0

BEAM SUB.INDEX 1.0
BEAM SUB.EXTINCT 0.0

530 ATLAS User’s Manual

Matrix Method Luminous: Optoelectronic Simulator

10.3.4 Transfer Matrix with Diffusive Interfaces

Frequently, thin-film detectors are constructed with rough or irregularly textured interfaces
between materials. These interfaces exhibit partly specular and partly diffused reflection and
transmission characteristics.
An approach has been proposed [143, 144, 332] that models both coherent (specular) and
incoherent (diffused) light propagation and absorption. With this approach, the specular light
is modeled using the transfer matrix method while the diffuse light is modeled using ray
tracing.
The connection between the specular and diffused portions of the light is described by haze
functions defined at each irregular interface as shown in Figure 10-15.

Figure 10-7: Relationship between the haze functions and the specular and diffused light at a textured
interface
Equations 10-36 and 10-39 give the transmissive haze function HT and the reflective haze
function HR.
4SIGMA  CT  n 1 cos  1 – n 2 cos  2 NT
H T = 1 – exp –  --------------------------------------------------------------------------------------------- 10-36
  

4SIGMA  CR  cos   n 1 NR
H R = 1 – exp –  ----------------------------------------------------------------- 10-37
  

531 ATLAS User’s Manual

Matrix Method Luminous: Optoelectronic Simulator

Here:
•  is the optical wavelength.
•  is the angle of incidence.
• n1 is the index of refraction on the incident side of the interface.
• n2 is the index of refraction on the transmitted side of the interface.
• CR, CT, NR, NT, and SIGMA are user-defined parameters described in Table 10-2.
Here, CR, CT, NR, and NT can be considered tuning parameters. SIGMA characterizes the mean
variation of rough features on the interface.

Table 10-2 Default Values for User Specifiable Parameters of Equations 10-47 and 10-48

Parameter Statement Default Units

CR INTERFACE 1.0
CT INTERFACE 0.5
NR INTERFACE 2.0
NT INTERFACE 3.0
SIGMA INTERFACE 20.0 nm

You can describe the CT and CR parameters as a wavelength function in tabular form using the
TCT.TABLE and TCR.TABLE parameters of the INTERFACE statement. These parameters can
be names of the files containing ASCII descriptions of the dependencies. The first row of
each file is a single number describing the number of samples. Subsequent rows contain pairs
of numbers containing specifying the wavelength in microns and the value of CT or CR at that
wavelength.
The hazefunctions, HT and HR, give the fraction of transmitted or reflected light that is
diffused as given in Equations 10-38 and 10-39.
I dT
H T = --------------------
- 10-38
I dT + I sT

I dR
H R = --------------------
- 10-39
I dR + I sR

Here:
• IdT is the transmitted diffused intensity.
• IsT is the transmitted specular intensity.
• IdR is the reflected diffused intensity.
• IsR is the reflected specular intensity.

532 ATLAS User’s Manual

Matrix Method Luminous: Optoelectronic Simulator

The transmissive haze function, HT, gives the ratio of the transmitted diffused light intensity
to the specular light intensity. The reflective haze function, HR, gives the ratio of the reflected
diffused light intensity to the specular light intensity. Figure 10-8 shows some example
evaluations of the haze functions.
The transmissive and reflective hazefunctions can be written to log files by assigning the
OUT.HT and OUT.HR parameters of the INTERFACE statement to the desired output file
names. When outputting the haze functions, you should also specify the range of wavelengths
and number of samples using the LAM1, LAM2, and NLAM parameters on the INTERFACE
statement.

Figure 10-8: Haze Functions as a Function of Wavelength Example

The transmitted and reflected incoherent (diffused) light is also described by an Angular
Distribution Function (ADF). The ADF is specified on the INTERFACE statement as either
CONSTANT, TRIANGLE, GAUSS, LORENZ, LAMBERT, or ELLIPSE. Equations 10-40 through 10-
45 show the functional forms for the various ADFs.
CONSTANT: ADF    = 1 10-40

TRIANGLE: ADF    = 1 – 2   10-41

2
–
GAUSS: ADF    = exp  ----------------------------------------- 10-42
 2DISPERSION

533 ATLAS User’s Manual

Matrix Method Luminous: Optoelectronic Simulator

LORENZ: 1
ADF    = exp ---------------------------------------------------- 10-43
2 2
 + DISPERSION

LAMBERT: ADF    = COS    10-44

COS   
LORENZ: ADF    = ---------------------------------------------------------------------------------------- 10-45
2
COS    +  --------------------------------- SIN 
2 0.5 2
 SEMIMINOR

Here, the angles are in radians and the DISPERSION and SEMIMINOR parameters are specified
on the INTERFACE statements.
You can also specify the ADFs in tabular form. You can assign the ADF.TABLE parameter to
the file name of the file containing an ASCII description of the angular dependence of ADF.
The first row of this table contains the number of samples. Subsequent rows contains two
numbers. The first number is the angle in degrees from 0° to 90°, and the second number is
the optional unnormalized value of the ADF at that angle.
To enable the diffusive transfer matrix model, specify DIFFUSE on the BEAM statement.
The ADFs are normalized in each case for evaluation of diffusive generation by
photoabsorption as given in Equation 10-46.
– y
2 -----------------
COS 
Gy = 2
o P o ADF   e d 10-46

Here:
is the absorption coefficient.
Po is the intensity at the interface (either IdT or IdR).
y is the distance from the interface.
To enable the model specify DIFFUSIVE on the BEAM statement and specify the diffusive
characteristics of each interface as appropriate using INTERFACE statements.
The ADFs can be output by assigning the OUTADF parameter of the INTERFACE statement to
the name of the output file. When you output the ADF, you should also specify the range of
wavelengths and number of samples using the LAM1, LAM2, and NLAM parameters of the
INTERFACE statement.

534 ATLAS User’s Manual

Beam Propagation Method in 2D Luminous: Optoelectronic Simulator

10.4 Beam Propagation Method in 2D

You can simulate most optoelectronic devices using ray tracing based on geometrical optic
principles. But when diffraction or coherent effects are important, ray tracing methods are no
longer sufficient. In this case, you need a method that takes into account the wave nature of
light.
A method that does this is called the Beam Propagation Method (BPM). BPMs such as [287]
rely on paraxial appoximation. The BPM in LUMINOUS, however, has been extended to solve
a more general Helmholtz Wave Equation (Equation 10-47).
2
2 n 2  E  r t 
 E  r t  – ----- ---------------------- = 0 10-47
2 2
c t

Here, E is the electric field of an optical wave, n is the complex refractive index of the
material, and c is the speed of light in vacuum.
The Helmholtz Wave Equation is consistent with the Rayleigh-Sommerfeld formulation of
scalar diffraction theory [102]. It can also be derived from Maxwell’s equations assuming
homogeneous and isotropic media and neglecting non-linear effects.
These assumptions put certain restrictions on the applicability of BPM. For example, material
regions with gradual change in refractive index or photodetector saturation modeling aren’t
supported.

10.4.1 Using BPM

The BPM in LUMINOUS isn’t designed to completely replace the fast and robust ray tracing
algorithm. In fact, we recommend you start the analysis of your particular application using
the ray tracing method (see Sections 10.2.1 “Ray Tracing in 2D” to 10.2.3 “Reflection and
Transmission”).
Only use BPM when you think light diffraction or coherent effects are important. Here are the
following guidelines when to use it:
• The source is monochromatic (no coherent effects are observed for multispectral
sources).
• The application is 2D (currently, a 3D BPM isn’t implemented due to calculation time
considerations).
• Spatial distribution of irradiance or photogenerated carrier density is important.

Note: Carrier diffusion lengths should be considered.

• The structure isn’t periodic in the X direction.

BPM should be useful for:
• Multilayer structures with layer thicknesses comparable to wavelength.
• Narrow incident beams (when transverse beam size is comparable to wavelength).

535 ATLAS User’s Manual

Beam Propagation Method in 2D Luminous: Optoelectronic Simulator

Currently, BPM doesn’t support either user-specified reflection coefficient models

(F.REFLECT) or photogeneration rate vs. position or time. The SCAN.SPOT parameter of
SOLVE statement (see Section 21.57 “SOLVE”) and anti-reflection coatings (see “Anti-
Reflective (AR) Coatings for Ray Tracing and Matrix Method” on page 566) are also not
included.
If you need to use BPM, include the BPM parameter in the BEAM statement (see Section 21.3
“BEAM”). BEAM parameters related specifically to ray tracing (such as RAYTRACE, RAYS,
THINEST, MAX.WINDOW, and MIN.WINDOW) are ignored when BPM is specified.
Unlike ray tracing, the incident beam is specified on the top of the device. The origin of the
beam is then assumed to be at the X.ORIGIN in the BEAM statement, where Y.ORIGIN
parameter is ignored. This is done to avoid unnecessary beam propagation in the material
surrounding the device.
You can specify additional parameters (LONGIT.STEP and TRANSV.STEP) in the BEAM
statement to set respective step sizes for the internal BPM grid. If you don’t define these step
sizes, the default values are used. The default step size in both directions is equal to /16.0,
where  is the wavelength of light in vacuum. Don’t use step sizes larger than /2.0,
especially if coherent effects are of interest. Also, we don’t recommend extremely small step
sizes because it can result in unnecessarily long computation times. The default values are
adequate for most cases.

10.4.2 Light Propagation In A Multiple Region Device

Since Equation 10-47 doesn’t describe light propagation through the region’s boundary, each
region has to be considered separately. Each region’s boundary is composed of straight line
segments (edges) that form a polygon. Light propagating from one of the segments could be
reflected back into the region from other segments of the boundary, while transmitted light
enters other regions. Reflected and transmitted optical fields are calculated using Fresnel
formulae described in Equations 10-1 to 10-6 in Section 10.2.3 “Reflection and
Transmission”.
Once the field reflected from a certain boundary is known, it then propagates back into the
region. The light then reaches all edges of the polygon. At this point, the field incident upon
each boundary is saved (i.e., added to the total field incident on each boundary). Once all
polygon edges are addressed like this, the algorithm then proceeds to the next region.
When the pass through the device is complete, one reflection of light on each boundary has
been accounted. The REFLECTS parameter specifies the number of passes to be done.
Like in ray tracing, if the REFLECTS parameter is large (e.g., 10), the optical part of the
simulation might take considerable amount of time. To limit the number of propagation
routines, use the MIN.POWER parameter. But unlike in ray tracing, where the default value of
this parameter is zero, the default is MIN.POWER=1e-6. This limit helps avoid unnecessary
computations.
Since only propagation through convex polygons can be properly addressed, concave
polygons are automatically split to form two convex regions. If necessary, air polygons are
added at the top of the device structure to fill the empty regions, where the propagation of
light also needs to be considered.

536 ATLAS User’s Manual

Beam Propagation Method in 2D Luminous: Optoelectronic Simulator

10.4.3 Fast Fourier Transform (FFT) Based Beam Propagation Algorithm

Field propagation from a boundary segment through a region is considered in the coordinate
system of that segment. The X axis is along the boundary and Z axis is along the normal (i.e.,
the Z axis corresponds to the usual ATLAS Y axis, see Figure 10-9).

Figure 10-9: Internal Coordinate System

We assume the optical field and the device structure are uniform in the Y direction
(perpendicular to the plane of the Figure 10-9). The rectangular grid covering the whole
region is chosen according to the step sizes defined by LONGIT.STEP (Z) and TRANSV.STEP
(X) parameters. The Helmholtz Equation in a 2D cartesian coordinate system (see Equation
10-48) describes field propagation through the region.
2 2
----------------------  E  x z - + k 2 E  x z  = 0
E  x z - + ---------------------- 10-48
2 2
z x

Provided that the field on the input plane Z=0 is known, Equation 10-48 allows you to find
the field on any subsequent plane Z=Z. The numerical solution of Equation 10-48 is based
on the transformation of input complex field into superposition of plane waves. This
superposition represents the direction cosine spectrum of plane waves, which is obtained by
the FFT of the original field (see Equation 10-49).


F  k x z = 0  =
 E  x z = 0  exp  –ikx X  dx 10-49
–

Each plane wave propagates at a certain angle to the Z axis. The phase accumulated by each
plane wave component before reaching the plane Z=Z is given by:
2 2
F  k x z = Z  = F  k x z = 0  exp  i k – k x Z  10-50

where K=2 n/, n is the complex refractive index. We then can include diffraction and
absorption upon propagation simultaneously. The direction cosine spectrum given by
Equation 10-50 at Z=Z is transformed back into the spatial domain by applying an inverse
FFT.

537 ATLAS User’s Manual

Beam Propagation Method in 2D Luminous: Optoelectronic Simulator

 F  kx Z = Z  exp  ikx X  dkx

1
E  x z = Z  = ------ 10-51
2
–

The field distribution at Z=Z is then found.

These steps are repeated until the field in the whole region is found. The field incident on
each boundary segment and the field at each node within the region are obtained using bi-
linear interpolation.
When the propagation through the same region is considered, the resulting field values at
each node are added to the values saved upon the previous propagation. Since the phase
information is retained in our propagation scheme, adding the field ensures the coherent
effects are taken into account.
The numerical scheme to solve Equation 10-48 requires the rectangular grid in the transverse
direction (X) to cover an area significantly larger than the region of interest.
Zero-padding of the input field distribution is done automatically to specify the input field on
the grid. Super-gaussian filters are used in the spatial domain to suppress possible reflections
from numerical boundary.
Evanescent waves are included in the propagation scheme to avoid discontinuities in the
spectral domain [102].

538 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

10.5 Finite Difference Time Domain Analysis

The finite-difference time-domain (FDTD) method [321] can be used to model optical
interactions between a directed optical source and a semiconductor device in much the same
way as is done with geometric ray tracing. The difference is that the FDTD solves for the
electric and magnetic field propagation directly from the fundamental wave equations. The
FDTD approach correctly models the wave propagation with reflection, diffraction, and
interference effects.

Note: It is important to remember that in these applications, the carrier diffusion lengths are usually much greater
than the wavelength of the light stimulus. Ray tracing is usually an adequate description. But in certain
applications, the physical scale of details of the semiconductor device are too small for ray tracing to
describe the optical interaction. One such case would be a nonplanar antireflective structure. In such cases,
you should consider the FDTD approach.

10.5.1 Physical Model

Maxwell's equations of light propagation are given by:
H –1 1
------- = ------   E – ---  M source +  H  10-52
t  

E –1 1
------ = ------   H – ---  J source + E  10-53
t  

Here, E is the electric field, H is the magnetic field,  is the permittivity,  is the permeability,
Msource is the equivalent magnetic current density, Jsource is the electric current density,  is the
electric conductivity, and * is the equivalent magnetic loss.
In two dimensions Equations 10-52 and 10-53 can be rewritten for transverse electric (TE)
polarization as:
Hx = Hy = Ez = 0 10-54

Hz 1 Ex Ey

---------- = --- ---------- – ---------- –  M s +  Hz  10-55
t  y x

Ex 1 Hz
---------- = --- ---------- –  J s + Ex  10-56
t  y

Ey- = 1 Hz- –  J + Ey 

--- --------- 10-57
---------
t  x s

For transverse magnetic (TM), Equations 10-52 and 10-53 can be written as:
Ex = Ey = Hz = 0 10-58

Hx 1 Ez
---------- = --- --------- –  M s +  Hx  10-59
t  y

539 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

Hy 1 Ez
---------- = --- --------- –  M s +  Hy  10-60
t  x

Ez 1 Hy Hx

--------- = --- ---------- – ---------- –  J s + Ez  10-61
t  x y

For three dimensions, Equations 10-52 and 10-53 can be written as:
E 1 H H
--------x- = --- ---------z – ---------y- – E x 10-62
t  y z

E 1 H H
--------y- = --- ---------x- – ---------z – E y 10-63
t  z x

E 1 H H
---------z = --- ---------y- – ---------x- – E z 10-64
t  x y

H 1 E E
---------x- = --- --------y- – ---------z –  H x 10-65
t  z y

H 1 E E
---------y- = --- ---------z – --------x- –  H y 10-66
t  x z

H 1 E E
---------z = --- --------x- – --------y- –  H z 10-67
t  y x

You can specify the parameters of these equations as described in Table 10-3.

Table 10-3 User Specifiable FDTD Material Parameters

Statement Parameter Default Units

MATERIAL MAG.LOSS 0.0 mho/cm

MATERIAL E.CONDUC 0.0 /cm

MATERIAL PERMITTIVITY 1.0

MATERIAL PERMEABILITY 1.0
MATERIAL J.ELECT 0.0 A/cm2
MATERIAL J.MAGNET 0.0 V/cm2

Rather than using the parameters listed in Table 10-3, it is usually more convenient to use the
complex index of refraction. The implementation of FDTD allows you to use the complex
index of refraction exactly as it is used in ray tracing and the appropriate conversions to the
parameters of Table 10-3 are done automatically.

540 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

In FDTD analysis, Equations 10-54 through 10-61 are solved using a finite-difference
approximation in space and finite-difference in time with suitable boundary conditions to
obtain static predictions of optical intensity and photogeneration rate as a function of the
optical source characteristics.

10.5.2 Beam and Mesh

To initiate FDTD analysis in LUMINOUS2D and LUMINOUS3D, you should specify FDTD on
the BEAM statement. The X.ORIGIN, Y.ORIGIN, Z.ORIGIN, ANGLE, X.MIN, Y.MIN, Z.MIN,
and Z.MAX of the BEAM statement are defined for FDTD exactly as for ray tracing (see
Figures 10-1 and 10-3). In FDTD, an auxiliary mesh is set up to solve the electro-magnetic
equations (see Equations 10-54 through 10-61). The FDTD mesh, as we will refer to it, is
defined in the primed coordinate system (x, y, z) as shown in Figures 10-1 and 10-3. The
FDTD mesh is a tensor product (rectangular) mesh that is not necessarily uniform in the x, y,
and z directions.

Note: For the remainder of this section, we will refer to the FDTD coordinate system using the unprimed x, y, and
z.

The easiest way to handle mesh generation is to use the supplied automatic mesh generator.
To enable automatic mesh generation, specify FD.AUTO on the BEAM statement. The automatic
meshing algorithm creates a minimal mesh that satisfies two conditions. First, the local mesh
spacing must be less than the wavelength divided by the value of the parameter TD.SRATE. If
you also specify the BIG.INDEX parameter, the local spacing will be given by the wavelength
divided and by the parameter value of TD.SRATE divided by the local index or refraction.
In some cases, the prescribed mesh spacing may not be sufficient to resolve small features
like quantum wells. In such cases, you may specify the FD.AUTO.MIN parameter. This
specifies the minimum number of mesh lines introduced into any region in any direction (x, y,
or z). The default value of the FD.AUTO.MIN parameter is 2, so you are guaranteed at least 2
mesh lines in any direction to resolve any device region.
You can specify values of TD.SRATE and BIG.INDEX without specifying FD.AUTO. In such a
case, a single value of mesh spacing is calculated for the whole mesh depending on the largest
index of refraction in the device.
You can also specify the FDTD mesh using the mesh line statements: FDX.MESH and
FDY.MESH. These statements are similar to the X.MESH, Y.MESH, and Z.MESH statements and
allow you to specify location and spacing using the LOCATION and SPACING parameters.
When using FDX.MESH, FDY.MESH, and FDZ.MESH, the locations and spacings are relative to
the FDTD mesh and are specified in microns. This method offers the advantage of non-
uniform meshing and allows for local mesh refinement. In this approach, all FDX.MESH,
FDY.MESH, and FDZ.MESH lines must be stated before the associated BEAM statement.
You may specify a uniform mesh using the SX, SY, and SZ, or SPACING parameters of the
BEAM statement. The limits of the FDTD mesh are chosen by the minimum and maximum
device coordinates relative to the FDTD coordinate system or X.MIN, X.MAX, Z.MIN, and
Z.MAX parameters (if specified) depending on coordinate clipping. The actual spacings in x
and y will be adjusted to fit an integer number of uniform mesh lines within the FDTD mesh
domain.

541 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

You can also define a rectangular uniform mesh in x, y, and z by specifying the desired
number of mesh lines in the x, y, and z directions using the NX, NY, and NZ parameters on the
BEAM statement. TM polarization is the default.
You can specify the polarization of the beam by using the POLARIZATION parameter on the
BEAM statement (POLARIZATION=0.0 implies TE polarization, POLARIZATION not equal to
0.0 implies TM polarization). Conversely, you can specify TM or TE directly on the BEAM
statement.
To initiate FDTD analysis, specify the FDTD parameter on the BEAM statement.

10.5.3 Boundary Conditions [269]

As mentioned previously, Equations 10-54 through 10-61 are solved in a 2D rectangular
domain. You must specify boundary conditions to complete the problem description. In
LUMINOUS2D and LUMINOUS3D, FDTD has 3 types of boundary conditions: Perfect
Electrical Conductor (PEC) boundaries, Perfectly Matched Layer (PML) boundaries, and
source boundaries (plane and point)
Perfect Electrical Conductor (PEC) Boundary
Currently, all boundaries are by default perfect electrical conductor (PEC) boundary
conditions. This implies that field values vanish at the boundaries and incident waves are
completely reflected. This is convenient when simulating periodic problems, however, it may
present difficulties when simulating non-periodic domains. In such cases, the perfectly
matched layer boundary is preferred.
Periodic Boundaries
In many cases, the physical reality can be characterized as periodic in space. The obvious
case is that of the photonic crystal. Here, simulation of the most primitive cell while still
accounting for periodicity is one of the finer points of using finite difference time domain.
Also, certain special cases are conveniently described assuming some degree of periodicity.
For example, you can consider angled incident light. Here, shadowing at the edges of the
device can best be eliminated using periodic boundaries.
In our implementation of FDTD, we allow periodic boundaries in the X and Z directions only.
This is in keeping with the general ATLAS standard of Y being directed into the device.
The periodic boundaries are modeled by algorithmically connecting the solutions from one
edge of the mesh to the coresponding solutions on the other edge. This algorithm produces
truly periodic solutions.
When using this algorithm, you should consider using the device periodic boundaries enabled
by the PERIODIC parameter of the MESH statement. You should be very careful to ensure that
your structure is really periodic. This means that each point along one edge must have exactly
the same compositional characteristics as the corresponding point on the other side.
Everywhere in between arbitrary variation is allowed.
To enable periodic handling of boundaries in the X direction, you can specify X.PERIODIC on
the BEAM statement. Similarly, the Z.PERIODIC parameter will set periodicity in the Z
directions. The parameter PERIODIC on the BEAM statement sets both.
Periodicity is particularly important in evaluation of the far field patterns since these require
two dimensional evaluation of Fourier transforms which also assume periodicity.

542 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

Perfectly Matched Layer (PML) Boundary [27]

The perfectly matched layer boundary is used to absorb outgoing light. This is useful when
we want to simulate an unbounded domain. In other words, we try to absorb (rather than
reflect) all outward bound waves. There are several main concepts for PMLs you should
know.
• PMLs are not really boundary conditions in the conventional sense. These boundary
conditions have a thickness associated with them and are included in the mesh. We hope
to select the absorption coefficient of the layer to allow absorption of the outgoing light to
a specified minimum.
• A PMLs is terminated by a PEC. Thus, all light passing through the PML, after
accounting for absorption, is reflected back into the simulation domain after making two
trips through the PML.
• PMLs are designed to be impedance matched to the simulation domain so that there is no
reflection at the interface between the PML and the simulation domain.
• By default, PMLs absorb only in one direction x or y depending on their location.
Therefore, light propagating parallel to a PML is not affected by the PML. A transverse
value of imaginary index can be specified by the IMAG.INDEX of the BEAM statement.
The absorption coefficient normal to the PML is specified by Equation 10-68.

x – x 0 DEGREE
  x  = ALPHA -----------------
- 10-68
WIDTH

Here, x0 is the coordinate of the interface between the PML and the simulation domain, and x-
x0 represents the depth into the PML. The ALPHA parameter is user-definable on the PML
statement and specifies the maximum absorption coefficient in the layer. The WIDTH
parameter specifies the width of the layer

Note: The PML is added to the simulation domain. WIDTH should be chosen appropriate to the relative
dimensions of the simulation domain.

The DEGREE parameter allows absorption to be increased with the depth into the PML.
Theoretically, DEGREE=0 should produce no reflection at the interface between the PML and
the simulation domain. In practice, however, it is found that such abrupt transitions produce
reflections due to discretization.
Table 10-4 shows the default values for the PML statement.

Table 10-4 PML Definition Parameters

Statement Parameter Default Units

PML ALL False

PML ALPHA 0.0 1/cm
PML BACK False
PML BEAM (all)

543 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

Table 10-4 PML Definition Parameters

PML BOTTOM True
PML DEGREE 2
PML FRONT False
PML LEFT False
PML NSAMP 10
PML PERMEABILITY 1.0
PML PERMITTIVITY 1.0
PML R90 0.0
PML SIGMAX 0.0 mho/cm
PML TOP False
PML RIGHT False
PML WIDTH 0.0 microns
PML XDIR False
PML YDIR False

As mentioned, the PML is terminated by a PEC. Therefore, the reflection coefficient can be
calculated by Equation 10-69.
2 – ALPHA  WIDTH cos 
R    = exp  ---------------------------------------------------------------- 10-69
 DEGREE + 1 

Here, R() is the reflection coefficient of the PML as a function of , the angle of incidence.
For convenience, you can specify the reflection coefficient at normal incidence using the R90
parameter of the PML statement. Given a selected value of DEGREE, LUMINOUS2D calculates
the value of one of R90, WIDTH, and ALPHA if the remaining two parameters are specified.
The placement of PMLs using the PML statement is described in Figure 10-10.

544 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

TOP
X

BACK
LEFT RIGHT

Z BOTTOM

FRONT

Figure 10-10: PML Placement

The ALL parameter specifies that the PML will apply to TOP, BOTTOM, RIGHT, LEFT, FRONT,
and BACK.
You must also specify the number of grid points used to resolve the PML with the NSAMP
parameter of the PML statement. As usual, the number of grid points used to represent the
PML directly impacts computation time.
You can use the BEAM parameter of the PML statement to associate the PML with a previously
defined optical source by beam number. This means that the BEAM statement must precede
any associated PML statements. If the BEAM parameter is not specified, the PML will apply to
all previously defined BEAMs.
Plane Source Boundary
The planar sources in LUMINOUS FDTD use the transmitted field scattered field (TFSF)
formulation. In this formulation, fields scattered from the target in the direction of the source
are allowed to pass through the source. This enables absorbing boundaries, such as PMLs,
behind the source to absorb the scattered fields. Thus, preventing them from further
interaction with the target. This is important to avoid secondary interactions with re-scattered
fields reflecting from the source that might occur if only a PEC boundary (default) is used.
Therefore, it is strongly recommended that you always define a PML on the same edge as a
plane source to avoid difficult to interpret results, such as "available photocurrents" exceeding
the "source photocurrent".
By default, the source boundary is located along the X axis (see Figure 10-1). The source is
simulated as a plane sinusoidal wave. Since the source originates at one of the edges of the
simulation domain, the source is a TE wave if TE is specified on the BEAM statement and TM
otherwise.

545 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

You can define if the stimulus as a cosine or a sine wave by specifying COSINE or SINE on the
BEAM statement. In addition, you can specify a offset phase angle on the PHASE parameter of
the BEAM statement.
Outside of the context of the beam propagation from the FDTD coordinate system origin in
the positive Y direction, the plane wave may originate from any edge of the FDTD domain.
The origin of the plane source is located using the S.TOP, S.BOTTOM, S.LEFT, S.RIGHT,
S.FRONT, and S.BACK parameters of the BEAM statement.
By default, the S.TOP is true and the beam propagates in the same direction as for ray tracing.
The alternate origins are not suggested for general usage but may be useful for optimizing
PML boundaries.
Point Source Boundary
Another convenience that is usually not suggested for general usage is a point source. To
enable a point source, specify POINT on the BEAM statement.
The location of the point source is defined by the X.POINT, Y.POINT, and Z.POINT
parameters of the BEAM statement relative to the X, Y coordinate system in units of microns.

10.5.4 Static Solutions and Error Estimation

The FDTD method is frequently used to analyze problems that are inherently transient. In the
context of LUMINOUS2D and LUMINOUS3D the problems are usually static.
To approximate static solutions, we allow the optical source to propagate from a static source
until all transients are settled out. To help automate the detection of this "steady-state", the
error estimator given by Equation 10-70 is defined.

Ex  – Ex  + Ey  – Ey  + Ez  –Ez 
  ----------------------------------------------------------------------------
nx ny nz


in
Error = ----------------------------------------------------------------------------------------------
- 10-70
Ex  + Ex  + Ez 
------------------------------------
nx ny nz

Here, Ex', Ey', and Ez' represent the field envelope over the last full cycle of the source. Ex'',
Ey'', and Ez'' represent the field envelope over the previous full cycle. If you specify the
parameter TD.ERRMAX on the BEAM statement, you can use this parameter to terminate the
time domain simulation when "steady-state" is achieved.

546 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

10.5.5 Inputs and Outputs

During FDTD analysis various specific information can be extracted. These are described in
the following paragraphs.
Run-Time Output
The run-time output first prints out the coordinates of the FDTD mesh and a mesh summary.
Next, the run-time output prints out the specified time step, the time step limit for Courant-
Friedrich-Levy stability, the time corresponding to the reciprocal of the optical frequency, and
the transition time for the minimum space step. Then, the run-time output contains a summary
of the optical constants for each region. Region 0 is outside the device mesh domain. This
table includes regions corresponding to electrodes.
Finally, the run-time output contains iteration values. The first column is the iteration number.
The second is the simulation time in seconds. The third is the number of wavelengths of the
source. The fourth is the percent of the current wavelength. The fifth is the current error
estimate, and the last is the estimated time to complete (if TD.LIMIT or PROP.LENS is
specified).
Saving FDTD Simulation State
During development of the FDTD capability, we found that saving and loading the current
state of the FDTD domain (i.e., fields) to either file or memory then restoring that state later
for either
• using the same mesh thus saving some computation time in interpolating from the device
simulation domain
• or offering convenience in data extraction for post display (i.e., extraction of cutplanes or
facets).
We are not sure what other possibile uses there may be for saving and restoring state. It is not
of practical importance for providing an initial guess for a subsequent solution. Any
difference between the initial guess and the solution must propogate everywhere through the
FDTD domain. Thus, an incorrect albeit educated initial guess is no better than no initial
guess at all.
To save the "state" to disk, you should assign the parameter OUTSTATE of the BEAM statement
to the root of file names of the files where the domain state is to be saved on disk. Seven files
are actually written with seven unique file names based on the root name. Seven meshes are
required to represent the solution data since solutions are obtained on each face and edge of
the primitive element (rectangle in 2D or 6 sided right-parallelapiped in 3D). The seventh
mesh is for the primitive mesh that describes all the material compositions for the other 6
meshes. Once the OUTSTATE parameter is specified the state is written to disk each time a
FDTD solution processed.
To restore the state, specify the root name of the files that the state with the root name you
specified to OUTSTATE above. You specify the input state on the INSTATE parameter of the
BEAM statement. When you also specify the logical parameter MESHONLY on the BEAM
statement, the simulator will load the mesh from the saved state and not load the fields. If this
parameter is omitted, the fields will be loaded as well. The loading of the saved state takes
place at the very beginning of the FDTD solution process when the FDTD domain is created.

547 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

The current simulation state can also be saved to internal memory. This operation is very fast
since it only requires a associating a user token with a pointer in memory. The disadvantage
to buffering FDTD domains in memory is that they are very big so you may need to consider
your available memory and swap space. Use the Linux "top" function to see how much
memory you are using. Another disadvantage to buffering is that the state is lost between
ATLAS runs.
To save the domain in memory, you should assign the BEAM statement parameter MEMORIZE to
a word of your choosing. After each FDTD solution, the current state is saved off in memory
and associated with the word you assigned to the MEMORIZE parameter.
To restore the state from memory, you assign the RESTORE parameter to the word you had
previously assigned to the MEMORIZE parameter. Like in the case of the INSTATE parameter
the FDTD domain is restored from memory at the beginning of the FDTD solution process.
Disk Management of FDTD File Sizes
The file sizes in FDTD analysis are critical since you can very easily generate enormous
volumes of data quite easily with FDTD. And since most files written by FDTD are only for
visual postprocessing, we provide you the capability to tailor the individual file sizes by
controlling the number of digits of precision maintained in the output files. The STR.PRECIS
parameter controls the precision output to structure files including 2D near and far field
patterns and cut planes. STR.PRESIS specifies the number of digits to maintain after the
decimal place in the structure file. The default value of STR.PRESIS=0 has special
significance since this implies the values will be written to full machine precision.
Time Domain Output
Output structure file "snapshots" can be saved by assigning a root file name to the TD.FILE
parameter of the BEAM statement. This specifies that outfiles will be written periodically with
incremented names to a file with the specified root name. These files can be examined by
TONYPLOT to view the field distributions as a function of time. TD.MANY specifies how many
file names will be used in total before files are written over. TD.EVERY specifies how many
time domain solutions are calculated between each output file.
Typically, the FDTD mesh is much larger/denser than the typical ATLAS device simulation
mesh. This implies that saving the "snapshots" and loading them into TONYPLOT can be
lengthy. To improve the efficiency of visualization, you can subsample the domain on output.
To do this, specify the parameters SUBNX, SUBNY, and SUBNZ or SUBDX, SUBDY, and SUBDZ on
the BEAM statement. The parameters SUBNX, SUBNY, and SUBNZ specify the numbers of evenly
spaced samples output in the X, Y, and Z directions. You can use the parameters SUBDX,
SUBDY, and SUBDZ to specify the sample spacing in X, Y, and Z.
Also to expedite file transfers for post-processing in TONYPLOT, you can crop the simulation
domain using the parameters X.CROP.MIN, X.CROP.MAX, Y.CROP.MIN, Y.CROP.MAX,
Z.CROP.MIN, and Z.CROP.MAX. These parameters are specified in units of microns and by
default encompass the entire device domain including lenselets and PMLs.

548 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

Photogeneration
Once "steady-state" is achieved, the FDTD simulation is stopped and the resultant intensities
are exported (interpolated) back onto the device simulation mesh. These intensities are used
to calculate photogeneration rates used directly in the drift-diffusion equations exactly as it is
done in ray tracing. The photogeneration rate is given by:
2
 E
G = ------  --------- 10-71
hc 2

Near and Far Field Patterns

The FDTD analysis is supported by a very general purpose far field pattern extraction
capability. This capability allows you to extract near and far field patterns from any plane
normal to a coordinate axis in the FDTD domain. Typically, this can be used to extract output
coupling from internal sources since it might be useful for LED analysis, or to extract
reflected sources since it might be useful in radar cross-section analysis.
You can initiate the extraction of near and far field patterns by assigning the FACET parameter
to the root file name for the patterns. The output files will be the root name with the
extensions ".nfp" for near field and ".ffp".
The near and far field patterns are extracted along a plane contained in the FDTD domain
normal to one of the principal axes. The near field pattern is simply the collection of Ex, Ey,
Ez, Hx, Hy, and Hz interpolated at the intersections between the facet and the FDTD mesh.
The plane is defined by one of three parameters X.FACET, Y.FACET, or Z.FACET. X.FACET
defines the location along the X axis of the intersection of the extraction plane with the X
axis. Y.FACET and Z.FACET are similarly defined. Remember, you can only specify one of
X.FACET, Y.FACET, and Z.FACET at a time. The values of X.FACET, Y.FACET, and Z.FACET
are expressed in microns relative to the FDTD coordinate system.
Certain logical placement parameters are provided to simplify pattern extraction along the
extreme faces of the FDTD domain. These parameters are TOP.FACET, BOTTOM.FACET,
LEFT.FACET, RIGHT.FACET, FRONT.FACET, and BACK.FACET. Since each of these
parameters imply a location and direction, they may also only be used one at at time and may
not be specified in conflict with any of the parameters X.FACET, Y.FACET, or Z.FACET. To
clarify TOP.FACET means the same as the assignment Y.FACET=<minimum y coordinate
in microns>.

549 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

Energy Dissipation and Interference

To close the discussion of FDTD, we will mention a couple of things about energy dissipation
and interference effects.
Suppose we illuminate a layer 1 micron thick with a TE sine wave of wavelength 1 micron.
Further assume the wave propagates from the left of the layer against a PEC at the right of the
layer. Figure 10-11 shows the resultant envelope of the light intensity after 1, 2, 3 and 4
wavelengths. The effects of the coherent reflection should be obvious. It is also notable,
however, that the peak of the envelope increases as the square of the number of wavelengths.
This is because there is no power dissipation in the system. The wave amplitude increases as
the number of wavelengths and the power increases as the square of the amplitude. This
illustrates an important consideration in FDTD analysis. That is that if projected waves are
not completely absorbed by the test structure and the PML boundary conditions, the envelope
of intensity increases with time. This has an important impact on the error estimation. The
error estimate that is proportional to the change in intensity envelope divided by the value of
the intensity envelope will be bounded by the square of the simulation time.

Figure 10-11: Finite-Difference Time Domain Non-Dissipative System

Another interesting aspect of Figure 10-11 is that the minimum of the envelope is not at 0 as
expected for such interference but 1. This is because the initial wave does not experience
interference and thus contributes a unit offset to the minimum.

550 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

10.5.6 Accuracy, Stability, and Simulation Time

There are clear trade-offs between speed and accuracy. For spatial sampling, you should
specify FD.AUTO since this best insures that all layers will be resolved, and it will insure that
spatial sampling will obey the rule that the step size is always less than the specified
wavelength divided by the local real index of refraction divided by the parameter TD.SRATE.
The relation between TD.SRATE and error can best be understood by considering the expected
difference between a sine wave and a discrete representation given by TD.SRATE evenly
spaced samples as shown in Figure 10-12.

Figure 10-12: Discretization Errors TD.SRATE=6

Numerically, you can go all the way down to Nyquist rate (TD.SRATE=2). A value of
TD.SRATE=10 produces about a 10% error. Taflove recommends a value of TD.SRATE=20
[269]. The simulation time increases roughly as the square of TD.SRATE for 2D simulation
and the cube of TD.SRATE for 3D. The best way to overcome this is to reduce the size of the
problem by appropriate use of periodic, mirror (PEC), and absorbing (PML) boundary
conditions.
If you are using PML boundaries, we suggest that you make them a similar dimension to the
overall problem. We feel that this gives PMLs equal weight numerically, which is important
since the absorption in PMLs is used in all subsequent postprocessing steps (e.g., available
photocurrent, R, T, A, and optical coupling for LEDs). The automatic meshing (FD.AUTO)
works properly in the PML so you only need specify the width (WIDTH), the reflectivity (R90)
and the degree (DEGREE). We will discuss R90 and DEGREE shortly. Keep in mind each PML
adds mesh to the structure increasing simulation time proportional to the width (WIDTH).
In some cases, PMLs can be removed entirely and replaced with a mirror boundary to reduce
simulation time. This is common when simulating an aluminum mirror. In reality, reflectivity
commonly exceeds 98% and if for example the difference between reflectance of the real
mirror and a perfect mirror is less than the expected accuracy of simulation, you can justify
removal of the PML in favor of a mirror boundary condition.

551 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

Another trick is to reduce the local index of refraction in regions that are not important to the
optical problem (e.g., contacts and metals). This should only be done in cases where the
index in that region is actually dominant for meshing.
The problem size in terms of time discretization is proportional to the total simulation time
divided by the constant time step size.
The step size can be directly set by the parameter DT. It is advisable that the the time step not
exceed the CFL stability limit. This is considered so important that a fail-safe mechanism has
been introduced to place a upper limit on the time step size. This limit can be disabled by
specifying ^DT.SAFE. Specifying time steps above the CFL stability limit is risky in the
sense the simulation may "blow up". The symptoms of this catastrophic condition is that the
residual errors will increase rapidly before stabilizing at some number much greater than one.
This is usually accompanied by huge increases in the fields and absorption dominant in a
single PML or the main simulation domain. Once this state is reached, there is no recovery.
If you experience stability problems, you can simply decrease the time step thus increasing
simulation time. You may also try to change the problem a bit to perhaps save a marginal
situation. The best place to look is in the PMLs. Adjust DEGREE, R90, TD.STATE, or WIDTH
until you reach your desired settings.
You should know that selection of time steps below the CFL stability limit does not insure
stability. The CFL does not consider the effects of the imaginary part of the complex index of
refraction, k. Experience has found that stability is adversely affected by increased k. This
would indicate good stability will favor lower numbers of both the DEGREE and the R90
parameter. For example, you may want to select DEGREE=1 and R90=0.01 in the case that
TD.SRATE=10 since the anticipated discretization errors will exceed the reflectance of the
PML.
As for picking very large time steps, you are eventually limited by Nyquist rate. Sampling
below Nyquist will cause your simulated frequency to alias into a lower frequency. The static
frequency model approach used by FDTD will keep the complex index of the actual
frequency but the propogation frequency will be the remainder of the actual frequency
divided by the Nyquist rate. The results may still be interesting but will certainly be non-
physical.
The last part of the problem size is the total simulation time. This is controlled by a distance
parameter (PROP.LENG) since such a parameter is more user-friendly in the sense that it
relates to intuitively obvious features of the simulation.
The propagation length (PROP.LENG) is the spatial distance in the simulation domain you
expect the light to meaningfully go. This information can be taken to be the maximum
dimension from the source to the furthest meaningful PML measured in microns. If a PML is
important, it should not lie outside the horizon of the simulation time. If you choose a
propogation length that is too short, it should be obvious in the log files and or structure files
produced by the FDTD simulator (TD.LOG and TD.FILE). The log files will show
unconverged behavior. Structure files will show abrupt edges to the simulation horizon.
When you specify PROP.LENG, you should specify BIG.INDEX. The logical parameter
BIG.INDEX specifies that during the conversion from distance to time. The velocity of light
should be scaled by the largest index in the simulation domain. This helps insure that the
PMLs are reached.

552 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

You should also try to consider if reflections may be critical in proper simulation. Reflections
increase optical path length in predictable ways.
The size of the simulation problem is defined by the above considerations. The computational
burden is defined by CPU time and memory space.
A wall clock time estimate is conveniently sent to the run-time output periodically, so you can
tell how long any simulation is likely to take. These estimates are fairly accurate but are
subject to many things that are out of our control. You will, however, be able to decide
whether to abort a simulation based on such projections.
The only way to reduce the estimated completion time of a problem of a given size is to apply
multiple processors or reduce the precision of the simulation. If you have multiple processors
available, you can set the number of processors using the -P option. If you are running in a
non-standard precision, you will see speed-up with standard precision.
The symptoms of excessive memory use are generally sluggish performance of the platform
in question. Process monitors will show that memory is fully consumed and sometimes the
platform will require rebooting. In such cases, you can specify NO.ENV on the BEAM
statement to reduce the memory. This causes the envelope information to no-longer be kept.
The envelope function is used to calculate the residual error. It is also output to structure
files.
When you specify NO.ENV error residual is estimated by the change in intensity, which is
saved instead of the envelopes. This saves roughly 50% on memory. The followng table
gives a summary of memory usage estimates for various cases.

Table 10-5 Approximate Memory Usage for Various Options

3D 2D

^NO.ENV(default) NXNYNZP24.5 NXNYP*8.5

NO.ENV NXNYNZP14.5 NXNYP*6.5

Here, P is precision. Standard precision is 64 bit or 8 bytes.

For the default mode in standard precision in 3D, you have about 200 bytes per mesh point.
A 3D 100010001000 mesh occupies about 200 GB. This is probably about the limit of
what is reasonable from current commonly available resources.
Assuming TD.SRATE=10 and an average index of 4, 1000 mesh points translates to about 25
wavelengths. For a wavelength of 0.5 microns, this is about 12.5 microns.
The good news is that problems larger than several wavelengths can usually be reduced since
coherence effects do not usually (or should not) occur over distances of several wavelengths.
This also implies that large problems should sometimes be reduced by dimension or left to
higher order methods (e.g., ray tracing). Also keep in mind, you may use hybrid optical
sources described in Section 10.1.1 “Hybrid Source Specifications”.
As always, you are free to separate the full problem into smaller more manageable parts. For
example in many cases, you can simulate the effects of a photonic crystal using FDTD over a
small area and apply those effects into larger area devices without applying FDTD to the
entire domain.

553 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

10.5.7 Anisotropic Index Materials and Liquid Crystals

Materials with anisotropic real index of refraction such as liquid crystals can be simulated
using the finite difference time domain method in 3D. By nature, anisotropy is 3D and is only
supported in LUMINOUS3D.
To permit anisotropic simulation, you will first need to specify anisotropy on the BEAM
statement by enabling the ANISO flag. When running in anisotropic mode, you may expect
slower run times and larger memory requirement.
To specify a region as anisotropic, you need to specify N.ANISO on the MATERIAL statement
associated with that region and specify the indices along the principal axes and the rotation
matrix. The parameters N.XX, N.YY, and N.ZZ are used to specify the indices along the
principal crystalline axes and the Euler angles N.ALPHA, N.BETA, and N.GAMMA are specified
also on the MATERIAL statement to give the angular difference between the crystalline and
optical axis.
To simulate twisted-nematic liquid-crystal (TNLC) materials, we must allow for continuous
spatial variation of the index tensor. Full flexibility in the definition of the index tensor as a
funciton of position is very expensive from a memory allocation point of view. As such we
limit the variation in index to occur only along one axis. The selection of which axis is
specified by the N.VS.X, N.VS.Y, or N.VS.Z paramters of the BEAM statement. Keep in mind
that the axes refered to by these parameters are relative to the beam referenced coordinate
system. In this system, the light propogates in the positive Y direction.
Simulation of TNLCs can now be done by assigning a tensor index with a twist angle
imparted as a function of distance through a region. This model is enabled by the TNLC
parameter of the MODEL statement. The Twist angle is given by the relation:

 = A0.TNLC + d TR · TNLC 10-72

where d is the distance along the optical axis in microns specified by the N.VS.X, N.VS.Y, or
N.VS.Z parameter of the BEAM statement. This distance is measured from the nearest extent
of the region in the direction of the optical source. The parameters A0.TNLC and TR.TNLC are
the initial angle in degrees and the twist rate in degrees per micron. The parameters A0.TNLC
and TR.TNLC are specified on the MODELS statement.
Anisotropic PMLs can also be specified by first enabling anisotropic analysis in the BEAM.
statement by setting the ANISO flag. Once that is done, we can simply specify the principal
indices using the NXX, NYY, and NZZ parameters of the PML statement.
You should be careful to make sure that the PML anisotropic indices match the adjacent
material indices along the entire interface.
As an alternative to PMLs, we have provided a capability to specify a matching layer that
smoothly varies the magnitude of the index tensor from for example an isotropic layer or
PML to the anisotropic liquid crystal. To specify such a layer, you must first enable it by
specifying GIAR on the MODELS statement. GIAR stands for graded-index anti-reflective [196].
The variation of real index of refraction is governed by the expression:

n = n1 +  n2 – n1   10 – 15 t + 6 t t  t t t 10-73

554 ATLAS User’s Manual

Finite Difference Time Domain Analysis Luminous: Optoelectronic Simulator

where t is the distance through the region as discussed above divided by the wavelength, and
n1 is the index at points nearest the optical origin, and n2 is the index at the points farthest
from the optical origin.
The values of the extreme indices can be specified for an isotropic layer using the parameters
N1.GIAR and N2.GIAR. For TNLC layer, you need to specify TNLC on the MODELS statement
as well as the values for the extreme values for the ordinary and extraordinary indices using
the parameters NO1.GIAR, NO2.GIAR, NE1.GIAR, and NE2.GIAR.
Currently, absorption is not supported in anisotropic simulations so the electrical simulation
afforded by, for example drift diffusion, is only coincidental to the analysis of the optical
properties using FDTD.

555 ATLAS User’s Manual

User-Defined Photogeneration Luminous: Optoelectronic Simulator

10.6 User-Defined Photogeneration

10.6.1 User-Defined Beams
Standard beam input syntax in LUMINOUS and LUMINOUS3D allows specification of plane
waves with Gaussian or flat-top (top-hat) irradiance profiles. Although many practical
applications require only plane wave illumination, some optoelectronic devices rely on
certain focusing or collimating optical systems that collect incident light. Some of these
optical systems are implemented as lens arrays deposited on top of the device structure, which
can be directly included in LUMINOUS3D. Still, a number of applications that require complex
collimating or focusing optics cannot be approached in this way. To accommodate these
applications, we enabled input of incident beams with complex wavefronts. Currently, the
ray-tracing is not restricted to particular beam properties and allows specification of an
arbitrary collection of rays as an input beam.
Theoretically, a number of ways to specify a beam of arbitrary shape exists. The one chosen
in ATLAS aims to allow you an easy interface with optical system design software. Thus,
collimators or focusing optics designed with conventional optical system tools can be
represented in ATLAS by a ray bundle obtained on the output of the optics tool. ATLAS treats
the ray bundle specified in a file as one beam. You need to set the input file name on the BEAM
statement as a value of INPUTRAYS parameter.
The file INPUTRAYS must have the following format.
NUMBER OF RAYS
NUMBER OF CHANGING PARAMETERS
PARAMETER1 PARAMETER2 PARAMETER3 ...
RAY NUMBER PARAMETER1 VALUE PARAMETER2 VALUE PARAMETER3
VALUE ...
RAY NUMBER PARAMETER1 VALUE PARAMETER2 VALUE PARAMETER3
VALUE ...
... ... ... ...
RAY NUMBER PARAMETER1 VALUE PARAMETER2 VALUE PARAMETER3
VALUE ...
NUMBER OF RAYS stands for the total number of rays specified in the file. It should be the
same as the last RAY NUMBER in the table.
NUMBER OF CHANGING PARAMETERS gives the total number of ray parameters given in the
table. Most often, some parameters are the same for all the rays. For example, wavelength or
ray power can be the same. In this case, there is no need to keep the table column for this
constant parameter. Instead, you can set this parameter directly in the BEAM statement.
NUMBER OF CHANGING PARAMETERS is equal to the number of columns in the table (not
counting the first column of ray numbers).
PARAMETER1, PARAMETER2, PARAMETER3, and so on are numerical codes of parameters. Their
order specifies the order of columns in the table. Numerical codes correspond to the following
parameters.
• 1 - X coordinate of ray origin (in device coordinate system).
• 2 - Y coordinate of ray origin (in device coordinate system).

556 ATLAS User’s Manual

User-Defined Photogeneration Luminous: Optoelectronic Simulator

• 3 - Z coordinate of ray origin (in device coordinate system).

• 4 - ray angle phi in XY plane (0 - along X and 90 - along Y).
• 5 - ray angle theta in XZ plane (in 3D only, 0 - along X and 90 - along Z).
• 6 - ray polarization angle (relative to the plane of incidence, 0 - TE, and 90 - TM).
• 7 - ray relative power.
• 8 - ray wavelength.
• 9 - ray area (µm2 in 3D) or thickness (µm in 2D).
The first column of the table where ray parameters are specified lists ray numbers (integer
values) in ascending order. The first ray number should be 1. The last ray number should be
equal to the NUMBER OF RAYS. Often, it is the case that ray parameters change continuously
from ray to ray (at least in parts of the table). The INPUTRAYS file in ATLAS allows a
shorthand input for these parts. You can omit ray numbers where possible to allow for
automatic fitting of parameter values for these rays.
Linear approximation is used for the omitted rays based on the values of parameters given in
the table. This is better illustrated in the following example.

21
3
1 4 7
1 0.0 60.0 0.2
11 300.0 90.0 1.0
21 600.0 120.0 0.2
Here, NUMBER OF RAYS=21, NUMBER OF PARAMETERS=3, and the order of parameter
columns: X coordinate (1), angle phi (4), and relative ray power (7).
The rays from 2 to 10 are specified using linear interpolation with end values given by rays 1
and 11. The rays from 12 to 20 are also defined similarly. This shorthand input is equivalent
to the following.
21
3
1 4 7
1 0.0 60.0 0.2
2 30.0 63.0 0.28
3 60.0 66.0 0.36
4 90.0 69.0 0.44
5 120.0 72.0 0.52
6 150.0 75.0 0.6
7 180.0 78.0 0.68
8 210.0 81.0 0.76
9 240.0 84.0 0.84

557 ATLAS User’s Manual

User-Defined Photogeneration Luminous: Optoelectronic Simulator

10 270.0 87.0 0.92

11 300.0 90.0 1.0
12 330.0 93.0 0.92
13 360.0 96.0 0.84
14 390.0 99.0 0.76
15 420.0 102.0 0.68
16 450.0 105.0 0.6
17 480.0 108.0 0.52
18 510.0 111.0 0.44
19 540.0 114.0 0.36
20 570.0 117.0 0.28
21 600.0 120.0 0.2
When you use the INPUTRAYS file to define the beam, some of the parameters on the BEAM
statements are disabled because they become incompatible with the user-defined beam. The
parameters that are ignored in this case are CIRCULAR, ELLIPTICAL, GAUSSIAN, all LENS.*
parameters, MAX.WINDOW, MIN.WINDOW, MEAN, RAYS, SIGMA, X.CENTER, X.GAUSSIAN,
X.MEAN, X.RADIUS, X.SIGMA, X.SEMIAXIS, XMAX, XMIN, Y.SEMIAXIS, Z.CENTER,
Z.GAUSSIAN, Z.MEAN, Z.RADIUS, Z.SIGMA, Z.SEMIAXIS, ZMAX, and ZMIN.
The X.ORIGIN, Y.ORIGIN, Z.ORIGIN, PHI, THETA, POLARIZE, REL.POWER, WAVELENGTH,
RAYAREA parameters in the BEAM statement specify values for all rays if the corresponding
parameters are not defined explicitly in the INPUTRAYS file. The parameters Z.ORIGIN and
THETA and the corresponding parameters in INPUTRAYS are for LUMINOUS3D applications
only. They are ignored in LUMINOUS2D.
RAYAREA is a parameter that specifies area in µm2 (thickness in 2D in µm) of each ray in the
BEAM statement. This parameter is used when the specified INPUTRAYS file does not contain
ray area information.
One difficulty that arises when dealing with arbitrary beams has to do with specification of
beam intensity (W/cm2) in the SOLVE statement. For a general ray bundle, illuminating the
device the intensity distribution is not uniform. Thus, intensity cannot be specified by a single
number. There are two ways to avoid this problem. Instead of setting intensity for a ray
bundle obtained on the output of an optical system (collimator or focusing system), you can
go back to the original beam that produced this ray bundle. According to the intensity law of
geometrical optics [33], you can define ray area (thickness) and overall beam intensity for
that input beam. Alternatively, if power (W) carried by each ray is known, you can treat beam
intensity parameter B<n> in the SOLVE statement as a scaling factor. Then, set all ray power
values using the ray relative power parameter in the INPUTRAYS file (or REL.POWER on the
BEAM statement if it is the same for all rays). In this case, you can treat RAYAREA as another
scaling parameter for unit conversions. In any case, make sure you define the ray area
because it’s important for obtaining correct values of photogeneration rates. If you don't
specify ray area, ATLAS uses an estimate based of coordinate values of ray origins. You
should not rely, however, on the accuracy of such an estimate.

558 ATLAS User’s Manual

User-Defined Photogeneration Luminous: Optoelectronic Simulator

You can use INPUTRAYS with POWER.FILE to define multi-spectral complex beams. In this
case, ATLAS ignores ray wavelength specified in INPUTRAYS and uses the wavelength values
from POWER.FILE for each ray. Ray power is then a product of relative ray power from
INPUTRAYS and spectral power density defined in POWER.FILE.

10.6.2 User-Defined Arbitrary Photogeneration

An option exists for you to define the photogeneration rate. A C-INTERPRETER function
written into a text file can be supplied to the program using the F.RADIATE parameter of the
BEAM statement. For example, if a file, myoptics.c, was developed using the template C-
INTERPRETER functions supplied, it can be referenced by using:
BEAM NUM=1 F.RADIATE=myoptics.c
.
.
SOLVE B1=1.0
SOLVE B1=2.0
The file, myoptics.c, returns a time and position dependent photogeneration rate to the
program. This returned value is multiplied at every node point by the value of B1. With this
option, you override all other parameters of the BEAM statement and all the material refractive
indices.

559 ATLAS User’s Manual

User-Defined Photogeneration Luminous: Optoelectronic Simulator

10.6.3 Exponential Photogeneration

Another option is available if distribution of photo-injected carriers is uniform, linear, or has
an exponential dependence on depth. Use the PHOTOGENERATE statement to specify the
desired dependence on depth along a line segment.
The X.ORIGIN, Y.ORIGIN and X.END, Y.END parameters set the coordinates of the beginning
and the end of the line segment. The default values correspond to the top left and bottom left
corners of the device. Other parameters CONSTANT, EXPONENT, FACTOR, LINEAR, and RADIAL
specify the photogeneration rate according to the following formula:

  r
2 
G  l r  =  CONSTANT + LINEAR  l + FACTOR exp  – EXPONENT  l   exp  –  -------------------------  10-74
2
  RADIAL  

where l is the distance along the line segment from the origin point and r is the radial distance
from the line segment (Figure 10-13). PHOTOGENERATE statement ensures TMA compatibility
with PHOTOGEN statement. See Section 21.44 “PHOTOGENERATE” for the list of TMA
compatible parameter names.

Figure 10-13: Specification of the Photogeneration Rate

560 ATLAS User’s Manual

User-Defined Photogeneration Luminous: Optoelectronic Simulator

10.6.4 Tabular Photogeneration (Luminous2D only)

You may also specify the photogeneration rate as a function of location through a simple
tabular interface. This is often useful to importing data from some other generic optical
propagation tool. The values of photogeneration rate are contained in a file. You specify the
file name on the FILE.PHOTOGEN parameter of the BEAM statement.
The file contains ASCII values of the photogeneration rate in units of per second per cubic
cm. These samples are made on a regular mesh in X and Y rastered in increasing values of X
and then Y (i.e., row - column order). The values are space separated in X (along each row)
and separated by carriage returns in Y (between rows along each column).
The spacings in X and Y in microns are specified in the DX.PHOTOGEN and DY.PHOTOGEN
parameters of the BEAM statement. You can specify the origin of the photogeneration in
microns using the X0.PHOTOGEN and Y0.PHOTOGEN parameters of the BEAM statement. If you
do not specify X0.PHOTOGEN and Y0.PHOTOGEN, then the origin will be assumed to coincide
with the minimum X and Y coordinates in the device structure.
For example, the BEAM statement might read:
BEAM X0.PHOTOGEN=0 Y0.PHOTOGEN=0 DX.PHOTOGEN=1 DY.PHOTOGEN=1 \
FILE.PHOTOGEN=generation.table
The contents of the file named "generation.table" might be simply:
1e22 1e22
1e20 1e20
In this case, the photogeneration rate is uniform in the X direction range of X=0 to X=1
micron but varries linearly between 1e22 to 1e20 in the Y direction.

561 ATLAS User’s Manual

Photocurrent and Quantum Efficiency Luminous: Optoelectronic Simulator

10.7 Photocurrent and Quantum Efficiency

One of the important figures of merit of a photodetector is quantum efficiency. Here, quantum
efficiency is defined as the ratio of the number of carriers detected at a given photodetector
electrode divided by the number of incident photons on the detector. Ideally, this ratio should
be 1.0 in detectors without any positive feedback mechanisms, such as avalanche gain.
LUMINOUS doesn’t directly calculate quantum efficiency. It does, however, calculate two
useful quantities printed to the run-time output and saved to the log file. These quantities are
source photocurrent and available photocurrent, which can be viewed in TONYPLOT from log
files produced by LUMINOUS.
Definition of Source Photocurrent
The source photocurrent for a monochromatic source is given in Equation 10-75. Here, Bn is
the intensity of the beam number n, which is user-definable in the SOLVE statement.  is the
source wavelength specified by the WAVELENGTH parameter in the BEAM statement.  is
Planck's constant. c is the speed of light. Wt is the width of the beam including the effects of
clipping (see Section 10.2.1 “Ray Tracing in 2D”). This can be considered as a measure of the
rate of photons incident on the device expressed as a current density.
Bn 
I s = q --------- W t 10-75
hc

Definition of Available Photocurrent

The available photocurrent for a monochromatic source is given by Equation 10-76. Here, all
the terms in front of the summation have the same definitions as for the source photocurrent.
The sum is taken over the number of rays traced, NR. WR is the width associated with the ray.
The integral is taken over the length, Yi, associated with the ray. Pi accounts for the
attenuation before the start of the ray due to non-unity transmission coefficients and
absorption prior to the ray start. i is the absorption coefficient in the material that the ray is
traversing.
The available photocurrent can be thought of as a measure of the photo absorption rate in the
device expressed as a current density. This should be similar but somewhat less than the
source photocurrent. The losses are due to reflection and transmission of light out of the
device structure.
NR
Bn  Yi

 o Pi i e
– i y
I A = q --------- WR dy 10-76
hc
i=1

Depending how you define it, you can calculate quantum efficiency by dividing the current
from one of the device electrodes by either the source photocurrent or the available
photocurrent.

562 ATLAS User’s Manual

Defining Optical Properties of Materials Luminous: Optoelectronic Simulator

10.8 Defining Optical Properties of Materials

For ray tracing, the complex index of refraction of the various material regions in the
structure must be specified. For many of the more common semiconductors and insulators,
there are built-in tables of index versus wavelength. You can specify the index for the
materials lacking reasonable default complex index of refraction.

Note: You can add the INDEX.CHECK parameter to any SOLVE statement to print out the refractive indices
being used for that bias step. The indices are only printed when you perform the ray trace at the first
reference to a given beam on the SOLVE statement.

10.8.1 Setting Single Values For The Refractive Index

The REAL.INDEX and IMAG.INDEX parameters of the MATERIAL statement can be used to set
the real and imaginary indices of a specified material, region or regions. For example, the
statement:
MATERIAL MATERIAL=Air REAL.INDEX=1.0 IMAG.INDEX=0.0
would set the index for all material regions composed of “Air”
The statement:
MATERIAL REGION=1 REAL.INDEX=1.0 IMAG.INDEX=0.0
would set the index for region number 1.
The statement:
MATERIAL REAL.INDEX=1.0 IMAG.INDEX=0.0
would set the index for all regions.

10.8.2 Setting A Wavelength Dependent Refractive Index

The preceding examples set the complex index of refraction for a material regardless of
wavelength. This is probably adequate for monochromatic sources. For multispectral
simulations, the index of refraction should be modeled as having a dependence on
wavelength. There are two ways to do this.
ASCII File Input
The first way is to specify index versus wavelength in a file. This is a text file that contains
ordered triplets of wavelength, real index, and imaginary index. The first entry in the table is
the number of samples. If you set the INDEX.FILE parameter of the MATERIAL statement to
the name of the index file, the linear interpolation from this table will complete to obtain the
index of refraction as a function of wavelength.
A valid index file is shown below.
2
0.5 2.0 0.0
0.6 3.0 0.02
In this example, a real index of 2.5 and an imaginary index of 0.015 would be used for a
wavelength of 0.55 microns.

563 ATLAS User’s Manual

Defining Optical Properties of Materials Luminous: Optoelectronic Simulator

You may also use ASCII file inputs to input real index and complex index independently. You
can assign INDX.REAL to the name of an index file containing ordered pairs representing the
wavelength and real index of refraction. Similarly, you can assign INDX.IMAG to the name of
an index file containing ordered pairs representing the wavelength and the imaginary index.
In all cases the table is preceded by the integer number of samples.
In each case, you may also choose to imbed comment lines delimited by a "#" character at the
beginning of the line that will not be processed when read in.
By default, the units of wavelength are microns and the index is represented by real and
imaginary parts. You can change this interpretation by including two character codes as the
first non-comment line in the file. The following table explains these interpretations.

Table 10-6 ASCII Index File Interpretation Codes.

Code Interpretation

none Wavelength in microns, real index, and imaginary index

"ae" Energy in eV, real index, and absorption coefficient in cm-1
"an" Wavelength in nm, real index, and absorption coefficient in cm-1
"am" Wavelength in um, real index, and absorption coefficient in cm-1
"nm" Wavelength in nm, real index, and imaginary index
"ev" Energy in eV, real index, and imaginary index

C-Interpreter Function
The second way to specify index versus wavelength is by using the C-INTERPRETER function.
The syntax is:
MATERIAL NAME=Silicon F.INDEX=myindex.c
The file myindex.c is an external file conforming to the template supplied with the program.
It returns wavelength dependent real and imaginary indices. Instructions for using the C-
INTERPRETER and finding the template functions are described in Appendix A:Appendix A
“C-Interpreter Functions”.
SOPRA Database
You can choose to use refractive index data from the SOPRA database. To do this, you must
specify the appropriate index file from Table B-40 on the SOPRA parameter of the MATERIAL
statement (see Section B.13.1 “SOPRA Database”).
Outputing Wavelength Dependent Complex Index for TonyPlot
The complex indices of refraction can be written to files suitable for display in TONYPLOT by
specifying the file name root without extension on the OUT.INDEX parameter of the
MATERIAL statement. The files created will use this root appended by "n.log" for real
indices and "k.log" for imaginary indices. For indices input by the SOPRA database, the
root will be appended by "N-n.log" and "N-k.log" where N is a sequence number (e.g., 1,
2, 3 ...) to accommodate the capability of the SOPRA format to allow dependence on
composition fraction and temperature.

564 ATLAS User’s Manual

Defining Optical Properties of Materials Luminous: Optoelectronic Simulator

The Absorption Edge

For various reasons, you may see absorption (non-zero imaginary index) at energies less than
the bandgap. You can ensure that simulation use zero absorption below the band gap by
specifying ABS.EDGE on the BEAM statement.

10.8.3 Quantum Well Absorption

The complex index of absorption can self-consistently include the effects of quantum
confinement. To enable this model, first you should specify the QWELL parameter on the
REGION statements of regions where you want to consider quantum confinement. This will
simply force the calculation of bound state energies in the region.
Second, you need to specify ABS on the same REGION statements to inform the simulator that
you want to use the confined state energies in the calculation of the complex index or
refraction.
Next, you can either use the confined “bandgap” represented by the difference between the
lowestconduction band bound state energy and the highest valence band bound state energy in
the calculation of complex index of refraction using the C interpreter function F.NKEG. This
is specified by the F.NKEG parameter on the MATERIAL statement.
If you do not specify the F.NKEG parameter, the model will use the negative of the gain
function for the quantum well. In this case, you should also specify the gain proper gain
function parameters on the MODEL statement (see Section 3.9 “Optoelectronic Models”).

565 ATLAS User’s Manual

Anti-Reflective (AR) Coatings for Ray Tracing and Matrix Method Luminous: Optoelectronic Simulator

10.9 Anti-Reflective (AR) Coatings for Ray Tracing and Matrix Method
It’s a popular strategy to place Anti-Reflective (AR) coatings on light detecting devices to
improve device quantum efficiency. Such coatings rely on destructive interference of
reflected waves to reduce overall reflection coefficient of light incident on the detecting
device.
Simple AR coatings are typically composed of one layer of a transparent insulating material
that is one quarter optical wavelength thick. Such coatings significantly reduce reflectivity of
light at the design wavelength. Far from the wavelength in question, however, their
performance is poor. Currently, more sophisticated multilayer AR coatings are used for
improved detection of broad-band light.
LUMINOUS enables you to model and design complicated AR coatings optimized for your
specific application. AR coatings are simulated in LUMINOUS using an efficient transfer
matrix method. This method enables you to deal with multilayer AR coatings with virtually
no overhead in computation time. Section 10.3 “Matrix Method” discusses matrix method
and its application to wave propagation in a stratified medium. AR coatings can be modeled
using this approach for Matrix Method analysis (see Section 10.3 “Matrix Method”) or ray
tracing (see Section 10.2 “Ray Tracing”).
Typically, AR coatings affect only optical properties of the device. Therefore, specifying the
coating as a part of the device structure will unnecessarily increase the complexity of the
electrical simulation. In ATLAS, coatings are associated with the interfaces of the device
instead. This allows you to specify certain optical properties of any material boundary while
not affecting electrical properties of the structure.
To define optical properties of a coating associated with that interface, use the INTERFACE
statement and include the OPTICAL parameter. Using this statement, you can define an AR
coating, a dielectric mirror, or an ideally reflecting surface. Note that the coating is absent in
the structure defined for ATLAS.
The AR.THICK parameter defines the thickness of the coating layer. The AR.INDEX parameter
defines the refractive index of the layer. For a single-layer coating or for the first layer of a
multilayer coating, specify the coordinates of the points defining the coated surface P1.X,
P1.Y, P2.X, and P2.Y.

Note: These coordinates should define a line, not a rectangular box of thickness (AR.THICK).

Figure 10-14 shows an example of a single-layer coating. The following shows the syntax
used for a single-layer coating.
INTERFACE OPTICAL AR.INDEX=2.05 AR.THICK=0.0634 P1.X=0.0 P1.Y=0.0
P2.X=10.0 P2.Y=0.0
This defines a 63.4µm layer of real refractive index 2.05 at Y=0.0 µm in a structure.
For a single-layer coating or for the first layer of a multilayer coating, you can also specify
coordinates of the bounding box XMIN, XMAX, YMIN, YMAX. All interfaces within the bounding
box obtain optical properties of the coating. If you don't set the bounding box coordinates, the
default bounding box containing the whole device will be used.

566 ATLAS User’s Manual

Anti-Reflective (AR) Coatings for Ray Tracing and Matrix Method Luminous: Optoelectronic Simulator

The REGION parameter gives you additional flexibility in the description of an AR coating.
REGION specifies the number of a region adjacent to the coating. REGION=0 is the default
value, corresponding to the ambient region outside the device boundaries.
When you define a bounding box for an interface, it could contain other interfaces, which you
do not intend to be AR coated. When you use REGION, you can select only those interfaces
adjacent to that region to have AR coating properties.
You can specify any number of coatings for each device. The COATING parameter in the
INTERFACE statement refers to the number of the coating. If COATING parameter is not set, the
first coating will be used. Coatings should be specified in order (i.e., COATING=3 cannot be
set before COATING=2 is defined).
Different coatings should not overlap. If this occurs, the later coating will be used in the
overlapping part.
Typically, the purpose of a single-layer AR coating is to minimize reflectivity of normally
incident monochromatic light at the design wavelength .
Equations 10-77 and 10-78 define the parameters of choice for a single-layer AR coating
between materials with refractive indexes n1 and n2:

AR · INDEX = n1 n2 10-77


AR · THICK = ------------------------------------- 10-78
4  AR · INDEX

According to Equations 10-77 and 10-78, the coating in the example considered above has
optimal properties for light at 520µm normally incident from air on a silicon detector.
You need several INTERFACE statements to specify a coating composed of multiple layers.
Each statement defines only one layer of a coating. The syntax of the first INTERFACE
statement for a multilayer coating is identical to the specification of a single-layer coating.
Each subsequent layer must have the number of the coating and the number of the layer
specified by COATING and LAYER parameters. Specify the same COATING parameter for all
INTERFACE statements that refer to the layers of the same AR coating. For example, the
following statements specify a two-layer coating:
INTERFACE OPTICAL AR.INDEX=1.5 AR.THICK=0.06 P1.X=0.0 P1.Y=0.0
P2.X=10.0 \
P2.Y=0.0 REGION =2
INTERFACE OPTICAL AR.INDEX=2.05 AR.THICK=0.06 COATING=1 LAYER=2
Here, COATING=1 and LAYER=2 show the second INTERFACE statement describing a second
layer of a two-layer coating number one. Use the LAYER parameter to describe the order of
layers in a coating. If LAYER is not specified, the first (top) layer will be used. Again, you
cannot specify LAYER=3 before LAYER=2. You only need to specify the coordinates of the
interface for the first layer of each coating.
In this example, the first coating layer with refractive index 1.5 is adjacent to the region
number 2 on all boundaries within the specified bounding box. The second layer is outside of
the first layer relative to region number 2, along the same boundaries. You only need to
specify the bounding box and REGION only for the first layer of each coating.

567 ATLAS User’s Manual

Anti-Reflective (AR) Coatings for Ray Tracing and Matrix Method Luminous: Optoelectronic Simulator

Note: You also should also the COATING parameter to a unique index for each device in a MIXEDMODE
simulation.

If a coating is made out of an absorbing material, you can use AR.ABSORB parameter to take
into account absorption. You can also use totally reflective coatings. The coating will behave
as an ideal reflector if you set the AR.INDEX parameter to a value > 1000.
Another way to specify the refractive index of the coating is to use the MATERIAL parameter
on the INTERFACE statement. This enables you to apply any refractive index model supported
by MATERIAL statement in ATLAS. This includes default wavelength dependent refractive
index for supported materials, user-specified index in INDEX.FILE, C-interpreter function in
F.INDEX, or the REAL.INDEX and IMAG.INDEX parameters on the corresponding MATERIAL
statement. When you use the MATERIAL parameter on the INTERFACE statement, the
AR.INDEX and AR.ABSORB parameters will be disabled.
An alternative way to set the properties of the top interface is to use C-INTERPRETER function
with F.REFLECT specified in the BEAM statement. Using this function, you can specify the
reflection coefficient, angle of transmission, and transmitted polarization as a function of
position, wavelength, angle of incidence, and incident polarization.

Figure 10-14: Single Layer AR Coating Under Normal Incidence

568 ATLAS User’s Manual

Anti-Reflective (AR) Coatings for Ray Tracing and Matrix Method Luminous: Optoelectronic Simulator

10.9.1 Anti-Reflective Coatings in LUMINOUS3D

LUMINOUS3D enables you to model and design multilayer AR coatings optimized for your
specific 3D application.
As in LUMINOUS, you can define properties of an AR coating using the INTERFACE statement
with OPTICAL parameter in LUMINOUS3D. The AR.THICK, AR.INDEX, AR.ABSORB, LAYER,
REGION, and COATING parameters retain their meaning and functionality in 3D. The only
difference in description of an AR coating in 3D pertains to specification of the coating
location. The coating location cannot be set by specifying point coordinates. For a single-
layer 3D coating or for the first layer of a multilayer 3D coating, you need to specify
coordinates of the bounding box XMIN, XMAX, YMIN, YMAX, ZMIN, and ZMAX.
The following example
INTERFACE OPTICAL AR.INDEX=2.05 AR.THICK=0.0634 REGION=0
specifies that a 63.4nm layer of real refractive index 2.05 covers the entire outside boundary
of a 3D device. All possible internal boundaries (heterojunctions) are not coated.
In a multilayer coating, the first layer is assumed to be adjacent to the region specified by
REGION.
INTERFACE OPTICAL AR.INDEX=1.5 AR.THICK=0.06 XMIN=0.0 XMAX=2.0
YMIN=0.0 \
YMAX=1.0 ZMIN=0.0 ZMAX=6.0 REGION=2
INTERFACE OPTICAL AR.INDEX=2.05 AR.THICK=0.06 COATING=1 LAYER=2
Here, COATING=1 and LAYER=2 show the second INTERFACE statement describing a second
layer of a two-layer coating number one. Use the LAYER parameter to describe the order of
layers in a coating. If LAYER is not specified, the first (top) layer will be used. Again, you
cannot specify LAYER=3 before LAYER=2. You only need to specify the coordinates of the
interface for the first layer of each coating.
In this example, the first coating layer with refractive index 1.5 is adjacent to the region
number 2 on all boundaries within the specified bounding box. The second layer is outside of
the first layer relative to region number 2, along the same boundaries. You only need to
specify the bounding box and REGION only for the first layer of each coating.
You can also specify the material composition of the AR layer by assigning the AR.MATERIAL
parameter of the INTERFACE statement to any ATLAS material. You can then use the
MATERIAL statement to define the layer refraction characteristics. See also Section 10.9
“Anti-Reflective (AR) Coatings for Ray Tracing and Matrix Method”.

569 ATLAS User’s Manual

Specifying Lenslets, Texturing, and Photonic Crystals Luminous: Optoelectronic Simulator

10.10 Specifying Lenslets, Texturing, and Photonic Crystals

You can specify one of four lenslets to focus the light into the device. The lenslet has no direct
effect on the device simulation mesh. It is only used in the ray trace. Lenslets are described by
user defined parameters of the LENS statement. All lenslet specifications apply to the last
previously specified beam.
Lenslets can be used in either ray tracing or FDTD. If you want to visualize lenslets used in
ray tracing, you will need to assign the OUTFILE parameter of the BEAM statement. The size of
the output file is controlled by the SUBNX, SUBNY, and SUBNZ parameters.
Spherical Lenslets
The spherical lenslet is described by the intersection of a plane (perpendicular to the Y axis
and a sphere). See Figure 10-15. The Y coordinate of the plane is given by PLANE. The origin
of the sphere is described by X.LOC, Y.LOC, and Z.LOC. The sphere radius is described by
RADIUS.

Figure 10-15: Luminous3D Lenslet Specification

For ray tracing purposes, the space between the plane and the device surface is uniformly of a
single index of refraction specified by INDEX. The spherical section projecting above the
plane in the negative Y direction is also considered to consist of the same index of refraction.
The plane and the spherical section form the lenslet. To enable a spherical lens, you should
specify SPHER on the LENS statement.

570 ATLAS User’s Manual

Specifying Lenslets, Texturing, and Photonic Crystals Luminous: Optoelectronic Simulator

Ellipsoidal Lenslets
To specify an ellipsoidal lenslet, define the axial half lengths using the X.SEMI, Y.SEMI and
Z.SEMI parameters of the LENS statement. The ellipsoidal lenslet center is specified by the
X.LOC, Y.LOC, and Z.LOC parameters. The PLANE parameter defines the location of the lens
plane. The INDEX parameter gives the lens real index of refraction. Equation 10-79 gives the
lenslet surface above the lens plane.
2 2 2
---------------------------------
 y – Y · LOC 
 x – X · LOC  - + ---------------------------------  z – Z · LOC 
- = 1
- + --------------------------------- 10-79
2 2 2
X · SEMI Y · SEMI Z · SEMI

To enable the ellipsoidal lens, you need to specify ELLIP on the LENS statement.
Composite Lenslets
You can also specify a composite lenslet. The composite lenslet is composed of a central
planar section, four cylinder sections and four sphere sections, which in shown in Figure 10-
16. The composite lenslet is defined by the XMIN, XMAX, ZMIN, ZMAX, WIDTH, HEIGHT, PLANE
and INDEX parameters. Figure 10-16 shows the meanings of these parameters.

Figure 10-16: Composite Lenslet

To enable the composite lens, you need to specify COMPO on the LENS statement.

571 ATLAS User’s Manual

Specifying Lenslets, Texturing, and Photonic Crystals Luminous: Optoelectronic Simulator

Aspheric Lenslets
Aspheric lenslets are characterized in the normal directions x and z by the following two
equations:
2
X 4 6 8 10
Y x = --------------------------------------------- + C 4 X + C 6 X + C 8 X + C 10 X 10-80
2 2 12 x x x x
Rx +  Rx – X 

2
Z 4 6 8 10
Y z = -------------------------------------------- + C 4 Z + C 6 Z + C 8 Z + C 10 Z 10-81
2 2 12 Z Z Z Z
Rz +  RZ – Z 

To enable the aspheric lens, you need to specify ASPHER on the LENS statement.
To arrive at these equations, you must provide tabulated data in the form of Sag-h pairs
describing the surface of the lenslet in the x and the Z direction. You should specify the name
of the file containing the data for the X direction in the X.SAGS parameter in the BEAM
statement. You should also specify the name of the file containing the data for the Z direction
in the Z.SAGS parameter in the LENS statement.
These files should first contain one line with the integer number of samples. The file should
then contain a number (equal to the number of samples) of additional lines each containing a
pair of real numbers representing the ordered pair of Sag-h data. These pairs must be arranged
by increasing h. A sample at h=0 is implicit and corresponds to Sag=1.
From this data, a Levenberg-Marquardt non-linear least squares algorithm is applied to obtain
the constants r, c4, c6, c8, and c10. This algorithm is iterative. You can specify the number of
iterations used in the SAGITS parameter on the LENS statement with a default value of 10.
To have the results of the least squares fit written to files to be displayed in TONYPLOT,
specify the file names using the X.SOUT and Z.SOUT parameters of the LENS statement.
Other parameters on the LENS statement describing the aspheric lenslet that should be
specified include: X.LOC, Z.LOC, INDEX, and PLANE.

572 ATLAS User’s Manual

Specifying Lenslets, Texturing, and Photonic Crystals Luminous: Optoelectronic Simulator

Pyramid Lenslets
To enable a pyramid shaped lenslet, specify PYRAMID on the LENS statement. The remainder
of the specification is described in Figure 10-17. The level of the lenslet is specified by the
PLANE parameter. The perimeter is specified by the X.MIN, X.MAX, Z.MIN, and Z.MAX
parameters. The peak of the pyramid is located by the X.LOC and Z.LOC parameters. The
height of the pyramid is given by the APEX parameter.

Figure 10-17: Pyramid Lenslet

573 ATLAS User’s Manual

Specifying Lenslets, Texturing, and Photonic Crystals Luminous: Optoelectronic Simulator

Random Textured Lenslets

Random textured lenslets provide the capability to simulate random textured surfaces. The
texturing can be best described by its algorithm.
First, a set of random steps are generated across the device domain in the Z direction. These
random steps start at the minimum Z coordinate and progress until the maximum Z
coordinate is reached. Each random step is made by first generating a Gaussian distributed
random number with mean given by the value of the Z.MEAN parameter and a standard
deviation given by the Z.SIGMA parameter of the LENS statement. This random step size is
added to the previous Z location to yield the next Z location. Step sizes are restricted to
positive numbers. This process is repeated until the Z location reaches the maximum Z
coordinate.
Next, for each Z location the same process is applied in the X direction, where the mean step
size is given by the X.MEAN parameter and the standard deviviation X.SIGMA. This being
done, we now have a complete set of X, Z locations. At each X, Z location, we next generate
a Gaussian distributed random number. This number has a mean given by the value of the
PLANE parameter and a standard devication given by the Y.SIGMA parameter of the LENS
statement. This random number gives the locations of the lenslet surface in the Y direction.
For convenience, the parameter MEAN can be used to specify a single value of both X.MEAN
and Z.MEAN. Similarly, the parameter SIGMA can be used to specify a single value for
X.SIGMA, Y.SIGMA, and Z.SIGMA.
Figure 10-18 shows an example surface generated using a value of MEAN=0.1 and
SIGMA=0.1.

Figure 10-18: An Example Random Textured Lenslet (MEAN=0.1, SIGMA=0.1)

574 ATLAS User’s Manual

Specifying Lenslets, Texturing, and Photonic Crystals Luminous: Optoelectronic Simulator

You can enable the user of a random textured lenslet by specifying the RANDOM parameter on
the LENS statement. You can vary the seed value of the random number generation using the
SEED parameter of the LENS statement. For a fixed value of the SEED parameter, the same
random surface will be generated.
The random lenslet can also be used for 2D ray tracing.
User-Definable Lenslets
User-definable lenslets allow you to define the lenslet surface in a C-Interpreter function (see
Appendix A “C-Interpreter Functions”). The function has a simple interface where given the
X and Z coordinates, you define the y coordinate of the lens surface. To enable the user-
definable lenslet, you need to specify the USER parameter of the LENS statement. You also
need to assign the F.LENS parameter of the LENS statement to the name of the file containing
the C interpreter function. Finally, you will need to specify the index of refraction of the
lenslet using the N and K parameters or the MATERIAL parameter of the LENS statement.
Lenslet Anti-reflective Coatings
It is sometimes useful to specify anti-reflective coatings on lenslets. For FDTD, this is simple
to accomplish by nesting two lenslet specifications a small distance apart. For ray tracing, this
is ineffectual since ray tracing does not directly account for the coherence effects necessary
for proper modeling of reflective properties for thin coatings. In these cases, you should use
the lenslet built-in AR coating capability. This works much like the AR coating capability
described in Section 10.9 “Anti-Reflective (AR) Coatings for Ray Tracing and Matrix
Method”.
To specify an AR coating on a lenslet for ray tracing, you need to specify the AR.INDEX and
AR.THICKNESS parameters of the LENS statement. The AR.INDEX parameter is set to the real
index of refraction of the coating. The AR.THICKNESS parameter is set to the coating
thickness in microns.
Photonic Crystals
You can define photonic crystals by defining periodic repititions of the primitive lens types
described above by specifying the periodicity in the X and Z directions. The parameters
PC.DX and PC.DZ specify the spacing between duplications is the X and Z directions. These
parameters are in microns.
When simulating photonic crystals, make sure that you size the device domain to an integer
number of PC.DX and PC.DZ in the X and Z directions respectively. Also, you should specify
PERIODIC on the BEAM statement to simulate a semi-infinite array.
In the next section, we will describe a practical example of setting up a hexagonal photonic
crystal.

575 ATLAS User’s Manual

Specifying Lenslets, Texturing, and Photonic Crystals Luminous: Optoelectronic Simulator

Hexagonal Photonic Crystals (A Practical Example)

We thought it would be instructive to describe how to specify a hexagonally symmetric
photonic crystal as are often described in literature.
In this discussion, consider a simple equal sided hexagonal array of equal sized cylindrical
elements. First, we look at the symmetries involved to determine the lens description which
only allows rectangular arrays.
The simplest solution is that we must use six LENS statements one for each of the vertices of
the primitive hexagonal cell. Further inspections show that a minumum of four lenses are
needed to describe the hexagonal mesh as described in Figure 10-19.

Figure 10-19: Hexagonal points of translational symmetry

The hexagonal mesh could just as well be represented using six lenses (one for each vertex on
the primitive). The software will not misinterpret the duplicated points, but the lesson here is
that you should always first consider the periodicities involved in whatever mesh is being
analyzed. It is usually best to sketch the situation as in Figure 10-19.
Next, if you desire to use periodic boundaries, you need to consider the planes of symmetry in
the mesh. Ideally, your (symmetric) device boundaries need to coincide with planes of
symmetry in the mesh. For this example, the mesh planes of symmetry are labeled in Figure
10-20.

576 ATLAS User’s Manual

Specifying Lenslets, Texturing, and Photonic Crystals Luminous: Optoelectronic Simulator

Figure 10-20: Hexagonal mesh planes of symmetry

With these considerations in hand, we are now prepared to look at the case of equal sided
hexagons with a side length of 1 micron with cylindrical primitives at each vertex. Figure 10-
21 is useful to construct the primitive lens center points PC.X0 and PC.Z0 and the
translational periods PC.DX and PC.DZ.

Figure 10-21: Hexagonal unit cell constructions

In this case, the values a and b are 0.5 and 0.866.

577 ATLAS User’s Manual

Specifying Lenslets, Texturing, and Photonic Crystals Luminous: Optoelectronic Simulator

The following lines describe such a photonic crystal:

LENS CYLINDER PLANE=0.0 RADIUS=0.2 INDEX=1.5 HEIGHT=0.5 \
PC.X0=-1.0 PC.DX=3.0 PC.Z0=0.0 PC.DZ=1.732
LENS CYLINDER PLANE=0.0 RADIUS=0.2 INDEX=1.5 HEIGHT=0.5 \
PC.X0=1.0 PC.DX=3.0 PC.Z0=0.0 PC.DZ=1.732
LENS CYLINDER PLANE=0.0 RADIUS=0.2 INDEX=1.5 HEIGHT=0.5 \
PC.X0=-0.5 PC.DX=3.0 PC.Z0=-0.866 PC.DZ=1.732
LENS CYLINDER PLANE=0.0 RADIUS=0.2 INDEX=1.5 HEIGHT=0.5 \
PC.X0=0.5 PC.DX=3.0 PC.Z0=-0.866 PC.DZ=1.732
Here, we centered the primitives at X=0 and Z=0. You should be able to understand the
coordinate selections by lookinging at Figure 10-21.
We defined the X and Z mesh limits as follows considering the planes of symmetry as
described in Figure 10-20.
X.MESH LOC=-4.5 SPAC=0.5
X.MESH LOC=4.5 SPAC=0.5

Z.MESH LOC=-4.33 SPAC=0.5

Z.MESH LOC=4.33 SPAC=0.5
The resulting structure is shown in Figure 10-22.

Figure 10-22: Hexagonal photonic crystal structure

578 ATLAS User’s Manual

Frequency Conversion Materials (2D Only) Luminous: Optoelectronic Simulator

10.11 Frequency Conversion Materials (2D Only)

Frequency conversion materials have the property that light absorbed in the material can be
re-emitted at a single wavelength. This property can be used to either up or down convert
light to improve spectral response of photodetectors, for example in solar cells.
In LUMINOUS, these materials are modeled electrically as insulators. The light conversion
capability is enabled by specifying FREQ.CONV on the MODELS statement.
The re-emission process is modeled at discrete locations in the frequency conversion mate-
rial. These locations are described by even sampling in the X and Y directions. You can spec-
ify the numbers of samples in the respective directions using the EMISS.NX and EMISS.NY
parameters of the MATERIAL statement. The sampling of emission centers is described in Fig-
ure 10-23.

Figure 10-23: Frequency Conversion Material Emission Center Sampling

Physically, the remission process occurs in random directions. This process is modeled by
performing ray traces from the emission centers in all directions 0 to 360° over an even sam-
pling. You can specify the number of samples in angle space by the parameter EMISS.NANG
on the MATERIAL statement. The angular sampling is described in Figure 10-24.

579 ATLAS User’s Manual

Frequency Conversion Materials (2D Only) Luminous: Optoelectronic Simulator

Figure 10-24: Frequency Conversion Material Angular Sampling

The rays are traced using the 2D trapezoidal ray tracing described in Section 10.2 “Ray Trac-
ing”. The ray tracing automatically resolves internal and surface topologies. You can specify
that the device domain will be treated as periodic with respect to the re-emitted rays by speci-
fying PFREQ.CONV on the MODELS statement. By default, the value of PFREQ.CONV is TRUE,
so the periodicity can be turned off by specifying ^PFREQ.CONV on the MODELS statement.
You can specify the width of each trapezoid as described in Figure 10-24 using the
EMISS.WIDE parameter of the MATERIAL statement. The width of the trapezoids is limited,
however, by the minimum of the spacing between emission centers (the minimum of dx and
dy in Figure 10-23).
The rays are emitted in monochromatic light at a wavelength specified by the EMISS.LAMB
parameter of the MATERIAL statement.
The intensity of each ray is calculated by integrating the photogeneration rate due to absorp-
tion of primary and re-emitted light in the frequency conversion material. The emission rate
is given by Equation 10-82.

580 ATLAS User’s Manual

Frequency Conversion Materials (2D Only) Luminous: Optoelectronic Simulator


hc EMISS · EFFI
P ray = G  x y  dA ------------------------------------- ------------------------------------------------------------------------------ 10-82
EMISS · LAMB EMISS · LAMB  EMISS · NANG
Emission
Center
Area
where:
• Pray is the initial intensity of each ray.
• G(x,y) is the photogeneration due to absorption in the area of the emission center.
• EMISS.LAMB is the user-specified emission wavelength.
• EMISS.WIDE is the user-specified emission ray width.
• EMISS.NANG is the user specified number of angles for emission.
• EMISS.EFFI is the user specified emission efficiency (number of photons emitted per
electron hole pair generated by absorption).
The emission rate per ray is recalculated iteratively since the generation rate includes the gen-
eration due to re-emitted rays. Convergence of this iterative sequence is controlled by the
FC.ACCURACY parameter of the MODELS statement, which specifies the maximum change in
emission intensity as a divided by the most recently calculated emission intensity.
The user-controllable parameters of the frequency conversion model are given in Table 10-7.

Table 10-7 Frequency Conversion Material Parameter Summary

Parameter Statement Default Units

FREQ.CONV MODELS False

PFREQ.CONV MODELS True
FC.ACCURACY MODELS 1.0e-6
EMISS.LAMB MATERIAL 0.5 microns
EMISS.EFFI MATERIAL 1.0
EMISS.NX MATERIAL 1
EMISS.NY MATERIAL 1
EMISS.NANG MATERIAL 1
EMISS.WIDE MATERIAL 1.0 microns

581 ATLAS User’s Manual

Simulating Photodetectors Luminous: Optoelectronic Simulator

10.12 Simulating Photodetectors

This section describes techniques simulate photodetectors. This section applies to the
simulation of any of the following devices: p-n and p-i-n photodiodes, avalanche
photodiodes, Schottky photodetectors, CCDs, MSMs, photoconductors, optical FETs and
optically triggered power devices.

10.12.1 Defining Optical Sources

Identifying an Optical Beam
You can define up to ten optical sources. Optical sources are described by using the BEAM
statement. All BEAM statements must appear somewhere after the MESH, REGION, DOPING,
and ELECTRODE statements and before any SOLVE statement. The parameters in the BEAM
statement describe a single optical source.
The NUM parameter is used to uniquely identify one optical source. Values between 1 and 10
are valid for the NUM parameter. Optical sources are subsequently referred to by the value of
their NUM parameter. The power of the optical beam is set by using the B<n> parameter of the
SOLVE statement, where n is the beam number defined by NUM.
Origin Plane of the Beam for 2D Optical Sources
To specify the origin of the optical source, use the X.ORIGIN and Y.ORIGIN parameters.
These parameters describe the origin of the optical beam relative to the device coordinate
system. Currently, it is required that the origin lie outside any device region. The ANGLE
parameter specifies the angle of the direction of propagation of the beam with respect to the
device coordinate system. ANGLE=90 specifies vertical (normal) illumination from above.
The width of the optical beam is specified using the MIN.WINDOW and MAX.WINDOW
parameters. These parameters specify the limits of the source beam relative to the beam
origin. If you omit either of the limits, that limit will be clipped to the edge of the device
domain.

Note: It’s extremely important that no section of the origin plane of the beam intersects or is inside the simulation
grid. Otherwise, you’ll get incorrect results. This is important to check in cases where the ANGLE isn’t 90 º or
270 º.

Origin Plane of the Beam for 3D Optical Sources

To specify the origin of a 3D source, use the X.ORIGIN, Y.ORIGIN, and Z.ORIGIN
parameters. These parameters describe the location of the origin relative to the device
coordinate system. The THETA and PHI parameters describe the direction of propagation
relative to the beam origin. The THETA parameter describes rotation of the X axis about the Z
axis. The PHI parameter describes rotation of the Z axis about the Y axis. As with the 2D for
normal illumination, specify a negative Y.ORIGIN and THETA=90.
Specify the window of illumination using the XMIN, XMAX, ZMIN, and ZMAX parameters. Then,
specify the numbers of samples in the X and Z directions using the NX and NZ parameters.
These values should be set to numbers large enough to resolve any salient features of the
topology.

582 ATLAS User’s Manual

Simulating Photodetectors Luminous: Optoelectronic Simulator

To save the rays into a file for visualization, specify the RAYTRACE=<filename> parameter.
The components (ux, uy and uz) of the direction vector representing the direction of the beam
propagation are given by the following:
u y = cos  PHI – 90  10-83

r = – sin  PHI – 90  10-84

u x = r cos  THETA  10-85

u z = r sin  THETA  10-86

Reflections
You can also specify whether to ignore the first reflection using the FRONT.REFL or the
backside and sidewall reflection using the BACK.REFL. You should activate the backside
reflections for devices, which use a back side reflector to improve collection efficiency. To set
the number of reflections solved, use the REFLECTS parameter.
Typically, BACK.REFL should be used if the structure simulated is equivalent to the complete
photodetector geometry as in a discrete device. If the simulation structure is a section of a
larger substrate as in CCD simulation then don’t use BACK.REFL.
Since the ray trace for arbitrary angles of incidence uses reflection and transmission
coefficients, specify the polarization by using the POLARIZATION parameter.
In complex structures, you should limit the ray tracing to trace only those rays with
significant optical power. The MIN.POWER parameter is used to terminate ray traces that drop
below MIN.POWER* (optical source power).

10.12.2 Specifying Periodicity in the Ray Trace

By default, the unspecified boundary conditions for device simulation are mirror or periodic
boundary conditions. For ray tracing, by default, the edges of the device are considered
interfaces with a vacuum. Thus, at the edges of the device, by default, reflections are
calculated. In some cases, it is convenient to consider the edges of the device to be periodic
with respect to ray tracing. This is particularly true when the beam is not normal to the plane.
You should specify PERIODIC in the BEAM statement to obtain such boundaries for ray
tracing.

10.12.3 Defining Luminous Beam Intensity in 2D

When a beam is defined in LUMINOUS, it is, by default, assumed to have a uniform light
intensity across the width of the beam. The intensity is defined in the SOLVE statement with
the B<n> parameter and is in the units of W/cm2.
The beam intensity distribution can also be defined with a Gaussian profile.

1   x – MEAN  2
f  x  = ------------------------------ exp  – ------------------------------- 10-87
SIGMA 2  2  SIGMA  2 

583 ATLAS User’s Manual

Simulating Photodetectors Luminous: Optoelectronic Simulator

The location of the peak of the Gaussian distribution with respect to the beam coordinates is
specified by the MEAN parameter of the BEAM statement. The standard deviation of the
distribution is specified with the SIGMA parameter.
x is the cumulative distance over the beam width where the distribution will be formed.
Define the RAYS parameter to get smooth distribution of intensity profile. This determines the
increment of x. Therefore, the RAYS parameter should be set to a large number as the spacing
between rays is always a constant (width of the beam /RAYS).
The resultant distribution has a unity peak value, which is then scaled by the B<n> value
specified on the SOLVE statement. For example, the following BEAM statement will define a
beam window 2m wide centered at X.ORIGIN and Y.ORIGIN, which has a Gaussian peak of
0.01 W/cm2 and a standard deviation of 0.05 m.
beam num=1 x.origin=5.0 y.origin=-1.0 angle=90.0 wavelength=0.6 \
xmin=-1 xmax=1 GAUSSIAN MEAN=0 SIGMA=0.05 RAYS=200 \
SOLVE B1=1E-2

10.12.4 Defining Luminous3D Beam Intensity

A beam defined in LUMINOUS3D is assumed by default to have a uniform light intensity
across the width of the beam. The intensity is defined in the SOLVE statement with the B,n.
parameter and is in the units of W/cm2.
In LUMINOUS3D, you can also specify a circular or elliptical source. To specify an elliptical
source, specify the appropriate origin location using X.ORIGIN, Y.ORIGIN, and Z.ORIGIN.
This will specify the center of the elliptical source. The major and minor axes of the ellipse
are specified by the XRADIUS and ZRADIUS parameters. Remember, you must also specify the
NX and NZ parameters. These will sample uniformly across the diameters of the ellipse.
You may also specify Gaussian profiles in LUMINOUS3D. To do this, specify the location of
the beam origin. You can specify the Gaussian(s) in the X and Z direction. These can be done
independently or together. These Gaussians are specified by the XMEAN, XSIGMA, ZMEAN, and
ZSIGMA parameters of the BEAM statement.

10.12.5 Monochromatic or Multispectral Sources

The optical source can be either monochromatic or multispectral. For monochromatic
sources, you can use the WAVELENGTH parameter to assign the optical wavelength.
WAVELENGTH uses the units microns to be more consistent with the rest of ATLAS. Note this
if you’re accustomed to the optoelectronic engineering preference for nanometers.
For multispectral sources, spectral intensity is described in an external ASCII file indicated
by the POWER.FILE parameter. This is a text file that contains a list of pairs defining
wavelength and spectral intensity. The first line of the file gives the integer number of
wavelength-intensity pairs in the file. An example of such a file is shown below.
4
0.4 0.5
0.5 1.0
0.6 1.2
0.7 1.1

584 ATLAS User’s Manual

Simulating Photodetectors Luminous: Optoelectronic Simulator

This example specifies that there are four samples and that at a wavelength of 0.4m, the
intensity is 0.5 Watts per square cm per m of wavelength, and so on. With multispectral
sources, specify a discretization of the interpolated information. Values must be specified for
the WAVEL.START, WAVEL.END, and WAVEL.NUM parameters. These values specify the starting
and ending wavelengths and the number of wavelengths to sample. LUMINOUS uses
wavelengths at equal intervals over a specified range of wavelengths.
If you don't specify the values of WAVEL.START, WAVEL.END, and WAVEL.NUM, these
parameters take on the corresponding values from the specified POWER.FILE. For the
example file shown above, LUMINOUS uses the following default values of these parameters:
WAVEL.START=0.4, WAVEL.END=0.7, and WAVEL.NUM=4.
LUMINOUS performs an independent ray trace at each of the sample wavelengths. For
example:
WAVEL.START=0.4 WAVEL.END=0.6 WAVEL.NUM=2
causes ray traces at wavelengths of 0.45 and 0.55. LUMINOUS obtains the intensity associated
with each sample by integrating the values of the spectral intensity file using a piece wise
linear approximation. Each integral is performed over the range between successive
midpoints. In the preceding example, the integration for the sample at 0.45 would be
performed over the range of 0.4 to 0.5.
For a multispectral source, the generation rate (like Equation 10-11) is given by:
WAVEL · END

P   L – y
G = 0  ------------------- e d
hc
10-88
WAVEL · START

where:
•  is the internal quantum efficiency.
• P() is the power spectral density of the source.
• L is a factor representing the cumulative loss due to reflections, transmissions, and
absorption
over the ray path.
• is the wavelength.
•  is Planck's constant.
• c is the speed of light.
•  is the absorption coefficient given by Equation 10-12.
• y is the depth of the device, where x,y forms the two-dimensional mesh.
WAVEL.START and WAVEL.END are the spectral limits specified on the BEAM statement.

Note: The integral in Equation 10-88 may be inexact due to the discrete sampling in conjunction with wavelength
dependence of the absorption coefficient. For constant absorption coefficient, the integral is exact of the
number of discrete wavelength samples specified by WAVEL.NUM on the BEAM statement.

585 ATLAS User’s Manual

Simulating Photodetectors Luminous: Optoelectronic Simulator

The source photocurrent (like Equation 10-75) is given by:

WAVEL · END
B
I s = q -----n- W t
hc  P    d 10-89
WAVEL · START

where P() is the power spectral density of the source. The other parameters have the same
definition as in Equation 10-75.
The available photo current (like Equation 10-76) is given by:
WAVEL · END
NR Xi
B –i y
I A = q -----n-
hc  WR   P    i e P i  dy d 10-90
i=1 0 WAVEL · START

where P() is the power spectral density of the source. The other parameters have the same
definition as in Equation 10-76.
LUMINOUS allows you to chose the units of spectral intensity in an input spectrum file. By
default, the units of spectral intensity in the POWER.FILE are W/(cm2 µm). In this case,
LUMINOUS carries out the integration according to Equations 10-88, 10-89, and 10-90. If the
units of spectral intensity in the POWER.FILE are W/cm2, then you need to use the flag
^INTEGRATE on the BEAM statement. In this case, each value P() in the second column of the
file is a total intensity for a corresponding spectral interval. Integrals over spectrum in
Equations 10-88, 10-89, and 10-90 are reduced to summations of P() values. In both cases,
LUMINOUS treats the subsequently specified value of B<n> on the SOLVE statement as a
unitless multiplier or scale factor for intensity.

This is different from monochromatic case where this parameter has units of W/cm2. To
preserve consistency with monochromatic case, you can also do the same in the multispectral
applications. If you specify the total beam intensity for multispectral sources using B<n>
parameter on the SOLVE statement, you need to set NORMALIZE flag on the BEAM statements.
This ensures that the intensity spectrum is normalized and B<n> has units of W/cm2.
For either the monochromatic or multispectral sources, you can uniformly scale the
wavelength(s) and intensities by using the WAVEL.SCALE parameter and the POWER.SCALE
parameter respectively. These parameters are useful if the intensities or wavelengths are
specified in units other than the default units. To output the samples from the input spectrum
file to a file suitable for display in TONYPLOT, specify the output file name on the OUT.POWER
parameter of the BEAM statement.
Remember that these are built-in tabular spectra and include many samples. You are therefore
advised to use subsample using the WAVEL.START, WAVEL.END, and WAVEL.NUM parameters
as described above.

586 ATLAS User’s Manual

Simulating Photodetectors Luminous: Optoelectronic Simulator

The run-time output includes a couple of headings to represent the integrated intensity as
input and as subsampled. "Iinput" is the integrated input spectrum. This is the integral over
the samples contained in the file specified by the POWER.FILE parameter (or contained in the
built-in AM0 or AM1.5 spectra). "Isampled" is the integral of the sub-sampled spectrum.
You sub-sample the input spectrum when you specify WAVEL.START, WAVEL.END, and
WAVEL.NUM. If you do not specify this sub-sampling, the raw input samples are used and
"Isampled" is identical to "Iinput".
If you sub-sample, "Isample" is always less than or equal to "Iinput". If you chose the
start and end samples to match those contained in the input spectrum, you will get "Iinput"
equal to "Isampled" due to the way we integrate power when we sub-sample.
You can look at the input spectrum in TONYPLOT by specifying a log file name in the value of
the OUT.POWER parameter of the BEAM statement and loading that file into TONYPLOT.
Sub-sampling has advantages to save on time computing optical propagation, such as ray
tracing or FDTD solutions for each specified wavelength. Usually, sub-sampling introduces
only moderate quantization error if the index of refraction does not vary too quickly.

10.12.6 Black Body Sources

You may specify a black body radiation source as described by
2
2hc 1
P  T  = ------------ ---------------------- 10-91
5 hc
 e ---------- – 1
kT

where  is wavelength and T is temperature. To enable the model, you should specify the
black body temperature using the TBLACK parameter of the BEAM statement. The units of
equation P  T  are W/cm2/steradian. For calculation, we normalize to 1 steradian since we
cannot predict the light gatherign characteristics of your optical system. The black body
radiation spectrum will be most useful in distinguished temperature of equivalent targets.

10.12.7 Solar Sources

LUMINOUS and LUMINOUS3D provide built-in definitions of solar spectra for AM0 [230] and
AM1.5 [231] conditions. To specify these spectra, define AM0 or AM1.5 on the appropriate
BEAM statement.

10.12.8 Extracting Dark Characteristics

One of the first tasks in analyzing a new detector design is to examine dark current, device
capacitance, and possibly other unilluminated characteristics. This can normally be done
without using LUMINOUS. Extraction of the characteristics is adequately covered in the
chapters on S-PISCES or BLAZE.
The extraction of reverse bias leakage currents for diodes presents some difficult numerical
problems for device simulators. These problems are associated with limitations on numerical
precision. ATLAS, as well as most other available device simulators, use double precision
arithmetic to evaluate terminal currents. Double precision arithmetic provides roughly 16
decimal digits of precision. Internal scaling allows the measurement of currents down to a
level of between about 10-12 A/micron to 10-16 A/micron. Unfortunately, photodiode leakage
currents are often around or below this level. This means that the currents printed are subject

587 ATLAS User’s Manual

Simulating Photodetectors Luminous: Optoelectronic Simulator

to significant numerical noise and don’t provide an accurate estimate of the device leakage
currents. There are two ways to estimate reverse leakage current.
Integrated Recombination
From a theoretical standpoint, the reverse behavior of diodes can be dominated by one of two
effects: diffusion currents in the neutral regions or recombination currents inside the depletion
region [268]. ATLAS can provide insight into both of these contributing mechanisms. To
estimate recombination current, use the MEASURE statement to calculate the integrated
recombination rate. The following statement can be used:
MEASURE U.TOTAL
When this statement is executed, it prints out the total integrated recombination rate. Multiply
this value by the electron charge (1.602310-19 coulombs) to obtain an estimate of the
recombination current contribution to the reverse diode leakage current.
Extrapolation from High Temperatures
The diffusion current contribution can be estimated by taking advantage of the non-linear
relationship between the diffusion current and temperature. Referring to the expression for
the “Ideal Diode” current given by Equation 10-92, the dominant temperature dependency
arises from the variation of the intrinsic concentration.
qD p p n0 qD n n p0 
- + ------------------- exp   ---------  – 1 
J =  ------------------
qV
10-92
 L L    kT L  
p n

where np0 and pn0 are thermal-equilibrium minority carrier densities on either side of the
junction. This gives an exponential variation of the diffusion current with temperature given
in Equation 10-93.

–Eg  
J  exp  ---------   exp  --------- – 1
qV
10-93
 kT L  kT L 

This relation can be used to estimate the diffusion current contribution at the operating
temperature. The basic idea is to calculate the current at a high temperature where the
problem of numerical precision does not arise. Then, scale the current to the operating
temperature using Equation 10-93. For example, if the device is to operate at 300K, you can
set the temperature to 450K using the TEMPERATURE parameter of the MODEL statement.
Disable any temperature dependence of the energy gap by specifying the band gap using the
EG300 parameter and setting EGALPHA and EGBETA parameters to zero, which are all on the
MATERIAL statement. The following statement illustrates this approach as it might apply to a
silicon diode:
MODEL TEMPERATURE=450
MATERIAL EG300=1.12 EGALPHA=0.0 EGBETA=0.0
ATLAS can then be used to obtain the reverse bias current at the elevated temperature.
Equation 10-94 can be applied to obtain the depletion current contribution at the operating
temperature:

588 ATLAS User’s Manual

Simulating Photodetectors Luminous: Optoelectronic Simulator

 exp  --------qV 
- – 1
E E   kT L 
 g
J = J e  exp -------- – ---------   ---------------------------------
g
10-94
 kT kT  
exp  -------- – 1 
e L qV
  kT e 
where Je is the current measured at the elevated temperature, Eg is the bandgap, Te is the
elevated temperature, TL is the operating temperature, V is the operating bias voltage, and J is
the current estimate at the operating temperature. Once you’ve obtained estimates of the
recombination and diffusion contributions, you can obtain the total leakage current by
summing the two contributions.
Numerical Solution Parameters
ATLAS uses a cut-off value of carrier concentration below, which solutions are not required
to converge. This limit is set by the CLIM.DD. parameter. See Chapter 20 “Numerical
Techniques” for more details on CLIM.DD. For photodetectors, you often need to reduce
CLIM.DD to 105 in order to resolve carrier concentrations in depleted regions before
illumination.

10.12.9 Extracting Detection Efficiency

One of the simpler tasks in characterizing a photodetector design is to measure DC detection
efficiency. This will be done typically as a function of bias voltage, optical intensity, or
wavelength. Each of these analyses can be performed using the SOLVE statement. The Bn
parameter can be used to set the optical intensity of the optical sources described in the
previous section. The following example illustrates obtaining a solution with a specified
optical intensity:
SOLVE B1=1.0
This specifies that a solution is to be obtained for an optical intensity in the beam numbered
“1” of 1.0 Watt/cm2. If this were the first SOLVE statement specified, the ray trace in
LUMINOUS would be initiated. At the start of the solution, the optical intensities of each
optical source with a positive intensity is printed. In addition, the available photocurrent and
source photocurrent are printed. See the prior section on Photocurrent for a definition of these
two quantities.
Internal and External Quantum Efficiency
The available photocurrent divided by the source photocurrent is a measure of the external
quantum efficiency of the detector. The calculated terminal current can be divided by the
source or available photocurrents is used to evaluate the internal quantum efficiency of the
device.

10.12.10 Obtaining Quantum Efficiency versus Bias

The intensities specified in the SOLVE statement apply until another SOLVE statement changes
the intensity of the beam. Sequences of SOLVE statements can be used to vary the optical
intensity at arbitrary intervals. The simple linear ramps of optical intensity can be abbreviated
using the LIT.STEP and NSTEP parameters of the SOLVE statement. The LIT.STEP parameter
specifies the size of the DC step and NSTEP specifies how many steps are desired.
Another option for analyzing DC quantum efficiency is to fix the optical intensity and vary
bias voltages. The bias voltages can be varied in arbitrary discrete steps using several SOLVE

589 ATLAS User’s Manual

Simulating Photodetectors Luminous: Optoelectronic Simulator

statements, or in a linear ramp using individual SOLVE statements. This is useful for
determining the optimum operating bias of devices such as avalanche detectors and photo
transistors.

10.12.11 Obtaining Transient Response to Optical Sources

It is sometimes desirable to examine the time domain response of a detector to time-
dependent (e.g., ramped or pulsed) optical sources. LUMINOUS provides this capability with
the RAMP.LIT parameter. When the RAMP.LIT parameter is specified in a SOLVE statement,
the optical intensity is changed linearly from the most recently set intensity to the intensity set
in the B parameter. If a particular source intensity is not set using the corresponding B
parameter, its intensity is not varied during the transient.
The period of the linear ramp is specified by the RAMPTIME parameter. The transient
simulation stops after the time specified by the TSTOP parameter. If the time given by TSTOP
is greater than that given by RAMPTIME, the source intensities remain constant until the time
given by TSTOP. For transient ramps, the TSTEP parameter should also be set. TSTEP is
typically set to allow several samples within the RAMPTIME. After the first time step,
subsequent time step sizes will be chosen automatically based on estimates of truncation
error. The following is an example of the specification of an optical impulse transient:
SOLVE B1=1.0 RAMPTIME=1E-9 TSTOP=1E-9 TSTEP=1E-11
SOLVE B1=0.0 RAMPTIME=1E-9 TSTOP=20E-9 TSTEP=1E-11
In this example, a triangular impulse in the intensity of the optical source is simulated. The
peak intensity is 1.0 and the impulse width is 2 ns. The response of the device is simulated for
an additional 18 ns. An initial time step of 10 ps is chosen for both parts of the impulse.

10.12.12 Obtaining Frequency Response to Optical Sources

You can also simulate small signal response to optical sources. To obtain a solution for small
signal response, set the SS.PHOT parameter in the SOLVE statement. The BEAM parameter must
also be assigned to the specific optical source index for small signal response of that source.
A single source small signal frequency can be specified using the FREQUENCY parameter. You
can vary the frequency within a single SOLVE statement by using the NFSTEP and FSTEP
parameters. The NFSTEP indicates how many frequency steps are to be simulated, while the
FSTEP indicates the step size. If the MULT.F parameter is specified, the start frequency is
multiplied by the step size for the specified number of steps. If not, the step size will be added
to the start frequency for the specified number of steps. For example:
SOLVE SS.PHOT BEAM=1 FREQUENCY=1e6 NFSTEP=5 FSTEP=10 MULT.F
will invoke solutions as optical frequencies at every decade from 1MHz to 100 GHz.
If the small signal parameters are specified in the same SOLVE statement as a DC bias ramp,
the small signal response is extracted for each bias voltage for each small signal frequency.
This is a useful strategy for analyzing the frequency response of the device as a function of
bias voltage.

Note: When simulating spectral response you should try to sample enough to resolve variations in complex index
of refraction, variations in input spectrum (e.g., AM1.5), and variations due to coherence effects in layers

590 ATLAS User’s Manual

Simulating Photodetectors Luminous: Optoelectronic Simulator

with thicknesses on the order of the wavelength. Such coherence effects are of no consequence in ray
tracing but may be important in the transfer matrix or finite difference time domain methods.

591 ATLAS User’s Manual

Simulating Photodetectors Luminous: Optoelectronic Simulator

10.12.13 Obtaining Spatial Response

To obtain spatial response, move an optical spot along a line segment perpendicular to the
direction of propagation of the source. Each incremental step is equal to the width of the spot.
The total distance over which the source is scanned is defined by the MIN.WINDOW and
MAX.WINDOW parameters of the BEAM statement. The number of steps is defined by the RAYS
parameter of the BEAM statement. The spot width is defined by the ratio:
(MAX.WINDOW - MIN.WINDOW)/RAYS
The spot scan is started by the SCAN.SPOT parameter of the SOLVE statement. This parameter
is set to the beam index of the optical source to be scanned. In other words, the beam defined
by the BEAM statement whose NUMBER parameter is set to the beam index. During the spot
scan, ATLAS obtains solutions and outputs, such as terminal currents, as well as the relative
beam location at each incremental spot location. This information can be used by TONYPLOT
to produce plots of photoresponse as a function of position.
In LUMINOUS3D, the SCAN.SPOT parameter will cause the simulator to power each ray
defined by the NX and NZ parameters sequentially.

10.12.14 Obtaining Spectral Response

The spectral response, defined as device current as a function of the wavelength of the optical
source wavelength, can be obtained. The LAMBDA parameter of the SOLVE statement sets the
source wavelength of the beam in microns. Since the wavelength can be set in each SOLVE
statement, successive solutions can be obtained as a function of wavelength. Each time the
LAMBDA parameter is specified, a new ray trace is run for that new wavelength and the
electrical solution recalculated.
The following statements could be used to extract terminal currents at a series of discrete
wavelengths:
SOLVE B1=1 LAMBDA=0.2
SOLVE B1=1 LAMBDA=0.3
SOLVE B1=1 LAMBDA=0.4
SOLVE B1=1 LAMBDA=0.5
SOLVE B1=1 LAMBDA=0.6
SOLVE B1=1 LAMBDA=0.7
SOLVE B1=1 LAMBDA=0.8
In this example, the spectral response is obtained for wavelengths from 0.2 microns to 0.8
microns. When using LAMBDA, the WAVELENGTH parameter of the BEAM statement will
overridden.
Equivalently, you may find it more convenient to specify a spectral ramp. To specify a ramp
of wavelength, you should assign the BEAM parameter to the beam index of the optical source
to be ramped. Also, you should specify the initial wavelength with the LAMBDA parameter, the
final wavelength with the WFINAL parameter, and the wavelength step size with the WSTEP
parameter as illustrated in the following.
SOLVE B1=1 BEAM=1 LAMBDA=0.2 WSTEP=0.1 WFINAL=0.8
This is completely analogous to the seven statements described above.

592 ATLAS User’s Manual

Simulating Photodetectors Luminous: Optoelectronic Simulator

Note: If you specify POWER.FILE on the BEAM statement, the spectral intensities will be scaled by the value
interpolated from the power file.

10.12.15 Obtaining Angular Response

The response of photodetectors to angle of incidence can be obtained in a manner analogous
to the method used to obtain spectral response.
In this case, the ANGLE parameter of the SOLVE statement specifies an angle added to the value
of the BEAM parameter PHI (ANGLE). This allows collection of response versus angle. The
following examples statements demonstrate the syntax.
SOLVE B1=1 ANGLE=0
SOLVE B1=1 ANGLE=10
SOLVE B1=1 ANGLE=20
SOLVE B1=1 ANGLE=30
SOLVE B1=1 ANGLE=40
SOLVE B1=1 ANGLE=50
SOLVE B1=1 ANGLE=60
Equivalently, you may find it more convenient to specify a angle ramp. To specify a ramp of
angle, you should assign the BEAM parameter to the beam index of the optical source to be
ramped. Also, you should specify the initial angle with the ANGLE parameter, the final angle
with the AFINAL parameter, and the angle step size with the ASTEP parameter as illustrated in
the following.
SOLVE B1=1 BEAM=1 ANGLE=0.0 ASTEP=10.0 AFINAL=60.0
This is completely analogous to the seven statements described above.

593 ATLAS User’s Manual

Simulating Solar Cells Luminous: Optoelectronic Simulator

10.13 Simulating Solar Cells

The LUMINOUS simulators are well suited to the simulation of solar cells. Particularly, the
variety of robust optical propagation models in combination of the capabilities of the device
simulators in the ATLAS framework makes accurate solar cell simulation a possibility. The
following paragraphs focus on problems and solutions specific to the simulation of solar cells.

10.13.1 Solar Spectra

As discussed in Section 10.12.5 “Monochromatic or Multispectral Sources”, LUMINOUS2D
and LUMINOUS3D provide flexible means for defining multispectral sources. In addition, you
can directly access built-in public domain [230, 231] solar spectra for am0 and am1.5
conditions by specifying AM0 or AM1.5 on the BEAM statement. These spectra consist of
numerous samples and we therefore suggest you carefully consider subsampling as discussed
in Section 10.12.5 “Monochromatic or Multispectral Sources”.
The following is an example of how you might define the BEAM statement for analysis of
am1.5 conditions with subsampling.
BEAM NUM=1 AM1.5 WAVEL.START=0.3 WAVEL.END=1.2 WAVEL.NUM=50
Here, we are choosing to simulate am1.5 conditions sampled between 0.3 microns and 1.2
microns at 50 samples. We do not have to worry about sampling the many peaks and valleys
in the solar spectrum as it is automatically integrated into each subsample. The subsampling
effects only the number of wavelengths for which the propagation model is calculated not the
integrated intensity.

10.13.2 Solar Optical Propagation Models

The choice of optical propagation model for your purposes depends largely on the nature of
your solar cell devices. In some cases, you may want to apply several of the different models
to compare the results against measurement for your particular analysis.
In all cases, accurate knowledge of the complex index of refraction of the materials
comprising the solar cell versus wavelength is key in accurate light propagation modeling.
LUMINOUS2D and LUMINOUS3D provide built-in defaults for most materials of importance to
solar cell technology (see Section B.13 “Optical Properties” for more details). Additionally,
LUMINOUS2D and LUMINOUS3D provide flexible means for you to introduce user-supplied
data in the form of ASCII tables or C-Interpreter functions as discussed in Section 10.8
“Defining Optical Properties of Materials”.
If you have little knowledge of the optical properties of the materials in your solar cell, or if
you are less interested in the propagation problem, you may choose to simply define the
photogeneration rate using the C-Interpreter function F.RADIATE. This function allows you
to directly specify the photogeneration rate as a function of position in the device. The
following illustrates the use of the F.RADIATE C-Interpreter function for constant generation
rate.
First, you need to identify the C function defining the generation rate to the simulator. This is
done in the input deck on the BEAM statement using the F.RADIATE parameter as follows:
BEAM NUM=1 F.RADIATE=solar.lib

594 ATLAS User’s Manual

Simulating Solar Cells Luminous: Optoelectronic Simulator

Here, the C-Interpreter function is contained in a separate file named "solar.lib". For
example, the contents of the file "solar.lib" may be as follows:

#include <stdio.h>
#include <stdlib.h>
#include <math.h>
#include <ctype.h>
#include <malloc.h>
#include <string.h>
#include <template.h>

/*
* Generation rate as a function of position
* Statement: BEAM
* Parameter: F.RADIATE
* Arguments:
* x location x (microns)
* y location y (microns)
* z location z (microns)
* t time (seconds )
* *rat generation rate per cc per sec.
*/
int radiate(double x, double y, double z, double t, double *rat)
{
*rat = 1e21;
return(0); /* 0 - ok */
}
Here, we simply define a constant photogeneration rate everywhere in the semiconductor of
1e21 carrier pairs per second per cubic centimeter.
You can easily define this function to specify any arbitrary generation rate as a function of
position and time. You should refer to Appendix A “C-Interpreter Functions” for more
information on the C-Interpreter.
Ray tracing is the principal propagation model for cases where we are not concerned about
interference or diffraction effects in the bulk of the device. This is most useful for relatively
thick bulk devices, such as crystalline silicon devices. It is useful for planar or textured
devices. For textured devices, the texturing must be described explicitly in the device
structure. Details of the ray tracing models are given in Section 10.2 “Ray Tracing”.

595 ATLAS User’s Manual

Simulating Solar Cells Luminous: Optoelectronic Simulator

The following gives an example of the specification of ray tracing for the optical propagation
model for LUMINOUS2D.
BEAM NUM=1 AM1.5 WAVEL.START=0.3 WAVEL.END=1.2 WAVEL.NUM=50
In this case, we will perform 50 ray traces. The omission of other model keywords, such as
TR.MATRIX, FDTD, or BPM, signifies that ray tracing will be performed.
For LUMINOUS2D, the algorithm automatically traces exactly the right number of rays to
completely resolve all discontinuities at the surface and deep into the device. In 3D, you must
specify enough rays to resolve the geometry in the NX and NZ parameters.
With ray tracing, you can also specify anti-reflection coatings using the INTERFACE statement
as discussed in Section 10.9 “Anti-Reflective (AR) Coatings for Ray Tracing and Matrix
Method”.
We recommend the transfer matrix method, as discussed in Section 10.3 “Matrix Method”,
for simulation planar of (1D) devices where bulk interference effects may be of consequence.
In these cases, the transfer matrix method provides fast accurate solutions for 1D analysis
problems only.
To enable transfer matrix analysis, you need to add the TR.MATRIX parameter to the BEAM
statement as in:
BEAM NUM=1 AM1.5 WAVEL.START=0.3 WAVEL.END=1.2 WAVEL.NUM=50
TR.MATRIX
We stress that the transfer matrix method is a one dimensional method and can only address
variations in the Y direction. This should be adequate for much of the simulation problems
addressed for planar thin film devices.
In the same way, you define anti-reflection coatings in ray tracing, you can also use the
INTERFACE statement to define anti-reflection coatings (see Section 10.9 “Anti-Reflective
(AR) Coatings for Ray Tracing and Matrix Method”) for the transfer matrix method.
For textured thin film devices, we have implemented a hybrid transfer matrix– 1D ray tracing
method. In this case, interfaces can be characterized by haze functions defining the proportion
of light transmitted and reflected as diffuse versus specular light and angular distribution
functions describing the angular intensity of diffuse light. These are bulk characteristics and
you should define the device as a planar structure as if you were applying the 1D transfer
matrix method. These haze and angular distribution functions are described in the INTERFACE
statement as illustrated in the following example:
INTERFACE OPTI P1.X=0 P2.X=1 P1.Y=0 P2.Y=0 DIFF ELLIPSE SEM=.31
SIG=20
Here, the optical interface, OPTI, defined along two points –P1.X, P1.Y and P2.X, P2.Y– is
treated as a diffusive (textured) interface as signified by the DIFF parameter. The angular
distribution function is elliptical as indicated by the ELLLIPSE parameter with a semi-major
axis of 0.31.
The interface haze function is characterized by a mean roughness length of 20 nm as specified
by the SIG parameter. This parameter can be correlated to measured surface variations from
SEM photomicrographs.

596 ATLAS User’s Manual

Simulating Solar Cells Luminous: Optoelectronic Simulator

To enable the diffusive transfer matrix model, you should specify DIFFUSE on the BEAM
statement as in the following:
BEAM NUM=1 AM1.5 WAVEL.START=0.3 WAVEL.END=1.2 WAVEL.NUM=50 TR.MAT
DIFFUSE
The diffusive transfer matrix method is discussed in detail in Section 10.3.4 “Transfer Matrix
with Diffusive Interfaces”.
For cases requiring full 2D or 3D analysis of coherence or diffraction effects, the finite
difference time domain (FDTD) analysis should be used. This method provides the most
accurate solutions for optical propagation problems but at the greatest computational expense.
The FDTD method performs direct solutions to the electromagnetic equations of light
propagation. To enable the FDTD model, you need to specify FDTD on the BEAM statement as
in the following example:
BEAM NUM=1 AM1.5 WAVEL.START=0.3 WAVEL.END=1.2 WAVEL.NUM=50 FDTD
There are many additional parameters which should be set for FDTD analysis. These
parameters are described in detail in Section 10.5 “Finite Difference Time Domain Analysis”.

10.13.3 Solar Cell Extraction

Here, we will describe how to extract the most common figures of merit for solar cells. These
extraction methods are solution strategies placed in your input deck after defining the device
structure, physical models, and the optical source as described in the previous section.
Typically, the first step in solar analysis is to illuminate the device. You can turn on the
illumination by assigning the beam intensity on a SOLVE statement as follows:
SOLVE B1=1.0
Here, we assign a value of one "sun" to the intensity of the optical source indexed as number
"1" (B1).
As we had assigned AM1.5 on the BEAM statement, this corresponds to standard test
conditions. In some cases, we may encounter convergence difficulties and may need to ramp
the light intensity as shown in the following:
SOLVE B1=0.01
SOLVE B1=0.1
SOLVE B1=1.0
You may also choose to assign the beam intensity to a value greater than one to simulate solar
concentrator effects. Since we have not yet assigned any bias voltages, the current in the run
time output corresponds to the short circuit current (Isc).
The next step is to extract the illuminated, forward current-voltage characteristic. First, we
should define an output file for capturing the data. For example:
LOG OUTFILE="IV.LOG"

597 ATLAS User’s Manual

Simulating Solar Cells Luminous: Optoelectronic Simulator

This specifies that all current voltage characteristics extracted from any following SOLVE
statement will be captured in the file IV.LOG in the current working directory. To capture the
I-V characteristic, you need to ramp the voltage past the open circuit voltage (Voc) as in the
following:
SOLVE VANODE=0 NAME=ANODE VSTEP=0.1 VFINAL=0.5
SOLVE NAME=ANODE VSTEP=0.01 VFINAL=0.7
Here we assume that the anode is P type and the characteristic is extracted on positive anode
voltages. First, we ramp from 0 to 0.5 volts in 0.1 volt steps. Then, we ramp from 0.51 to 0.7
volts in 0.01 volt steps. We assume that the open circuit voltage is near to 0.7 volts. We chose
the shorter voltage step in the second SOLVE statement to improve accuracy in obtaining Voc.
Essentially, we will resolve Voc to an accuracy comparable to the final voltage step size.
The current-voltage characteristic is observable in TONYPLOT using the following:
TONYPLOT IV.LOG
We can now use the DECKBUILD Extract capability to capture the various solar cell figures of
merit. We first initialize the Extract capability with the captured input file as in the following:
EXTRACT INIT INFILE="IV.LOG"
The following lines demonstrate the extraction of short circuit current (Isc), open circuit
voltage (Voc), maximum power (Pm), voltage at maximum power (Vm), current at maximum
power (Im), and fill factor (FF).
EXTRACT NAME="Isc" Y.VAL FROM CURVE(V."anode", I."cathode") WHERE
X.VAL=0.0
EXTRACT NAME="Voc" X.VAL FROM CURVE(V."anode", I."cathode") WHERE
Y.VAL=0.0
EXTRACT NAME="Pm" MAX(CURVE(V."anode", (V."anode" * I."cathode")))
EXTRACT NAME="Vm" X.VAL FROM CURVE(V."anode",
(V."anode"*I."cathode") ) \
WHERE Y.VAL=$"Pm"
EXTRACT NAME="Im" $"Pm"/$"Vm"
EXTRACT NAME="FF" $"Pm"/($"Jsc"*$"Voc")
Given that we know the area of the device in square cm, we can assign it to a variable and
extract power efficiency as follows:
SET area=1e4
EXTRACT NAME="OPT_INT" MAX(BEAM."1")
EXTRACT NAME="POWER_EFF" (ABS($"Pm")/($OPT_INT*$area))
Here, OPT_INT is the optical intensity in W/sqcm.
It is also common to extract spectral response of solar cells. In this case, the source is
monochromatic so we can omit the inclusion of a source spectrum, such as AM1.5, and
instead define a single wavelength as in the following:
BEAM NUM=1 WAVELENGTH=0.3

598 ATLAS User’s Manual

Simulating Solar Cells Luminous: Optoelectronic Simulator

Here, the wavelength of choice is not important since we will define the extraction
wavelength on the SOLVE statement. Again, we should open an output file to capture the data
as in the following:
LOG OUTFILE="SPECTRUM.LOG"
Then, we can simply capture the spectral response using a wavelength ramp defined on the
SOLVE statement as follows:
SOLVE B1=0.001 BEAM=1 LAMBDA=0.3 WSTEP=0.02 WFINAL=1.2
Here, the B1 parameter sets the intensity to 1 mW. The BEAM parameter identifies the index of
the optical source that is going to be involved in the wavelength ramp. LAMBDA specifies the
initial wavelength in microns, WFINAL specifies the final wavelength in microns, and WSTEP
specifies the wavelength step in microns.
We can then observe the spectral response file using TONYPLOT as follows:
TONYPLOT SPECTRUM.LOG
If you had specified a spectrum file (or AM0 or AM1.5) on the BEAM statement, then the value
of the B1 parameter is multiplied by the intensity interpolated from the spectrum file at the
specified wavelengths.

599 ATLAS User’s Manual

Optical Characterization Luminous: Optoelectronic Simulator

10.14 Optical Characterization

In many cases it is desirable to reduce the complexity of a problem by modeling the problem
over a representative space in a simpler tabular or analytic form. Here, we will call this
reduction of complexity "characterization".
Optical characterization is initiated by the CHARACTERIZE statement (see Section 21.4
“CHARACTERIZE”). The CHARACTERIZE statement is much like the LOG and PROBE
statements in that it descrbes a set of actions to take place at each subsequent solution. In this
case, we produce a characterization file at each solution.
The root file name for characterization of absorption is specified by the ABSORPTION
parameter. File names are generated from the root name appended by a sequence number and
".log". These files can be subsequently examined in TONYPLOT. These files contain
absorption rate versus depth (Y). All transverse dependencies are captured in this simple
form. This simple form is to map into large area devices.
The file names can be sequenced. The maximum number of files in the sequence can be
specified by the SEQ.ABS parameter .
To characterize, you must also specify a source beam index on the BEAM parameter. This is
the same number assigned to the same number as was assigned to the NUMBER parameter of
the BEAM statement.
The range of samples for the characterization is specified by the DEP1 and DEP2 parameters.
DEP1 and DEP2 specify minimum and maximum Y coordinates for subsampling in microns.
Subsampling will be taken uniformly over NDEP samples unless you specify S.DEP1 and
S.DEP2, which specify the spacings in microns at the start and end of sampling.
In this case you do not need to specify NDEP.

600 ATLAS User’s Manual

Chapter 11
LED: Light Emitting Diode
Simulator
Overview LED: Light Emitting Diode Simulator

11.1 Overview
LED is a simulator within the ATLAS framework that provides general capabilities for
simulation of light emitting diode (LED) devices. When used with the BLAZE simulator,
LED supports the simulation of DC and transient electrical and light emitting characteristics
of LEDs. When used with the GIGA simulator, LED permits the self-consistent analysis of
thermal effects on both electrical and optical emission characteristics. You can also use the
LED simulator with the MIXEDMODE simulator to analyze the circuit level performance of
LED devices.
The LED simulator supports simulation of zincblende (e.g., AlGaAs/GaAs, InGaAsP/InP),
wurtzite (e.g., GaN/AlGaN/InGaN), and organic/polymer material systems. It can also
account polarization effects and the effects of strain on both emission spectra and
piezoelectric polarization.
Extraction of electrical characteristics, such as DC, transient and small signal response, is
augmented by the capability to extract light emitting characteristics, such as emission power
versus current, emission spectra, emission wavelength, efficiency and output coupling (a
function of emission angle).
Rigorous 2D reverse ray tracing modeling is used to characterize output coupling
characteristics in including far field patterns.
The finite difference time domain (FDTD) method can alternatively be applied to determine
output coupling in 2D or 3D particularly for devices with photonic features, such as photonic
crystals or gratings.
The following sections describe how to specify LED devices, how to select and specify
physical models including most importantly the optical models and how to extract the most
pertinent device characteristics for device performance optimization.

602 ATLAS User’s Manual

Defining Light Emitting Devices (LEDs) LED: Light Emitting Diode Simulator

11.2 Defining Light Emitting Devices (LEDs)

There are several ways to define an LED device structures in ATLAS/LED. First, you can
define a structure using the ATHENA process simulator. ATHENA simulates the "construction"
of devices by simulating the process flow and the physical process steps. Second, you can
define the device structure using the DEVEDIT device builder tool. The DEVEDIT tool allows
you to specify the device structure in a simple visual point and click environment. Finally,
you can define device structures using ATLAS framework commands (statements). For LED
devices, the ATLAS command language approach is probably the most convenient since the
device structures are somewhat simple due to their epitaxial nature.
In the following sections, we will concentrate on definition of the device structure using the
ATLAS command syntax. We recommend that you read Chapter 2 “Getting Started with
ATLAS” where we describe general features of the ATLAS simulator, commands and
structure definition before you proceed. Chapter 3 “Physics” also provides insight into the
details of the electrical and optical physical models used by ATLAS/LED. Chapter 5 “Blaze:
Compound Material 2D Simulator” describes some general physics of simulation of
compound semiconductors and heterojunctions.
Within the ATLAS command syntax, there are two approaches for defining device structures.
The general purpose structure definition approach (see Section 2.6.3 “Using The Command
Language To Define A Structure”) requires more user-specification and requires that you
describe the entire mesh. The auto-mesh approach (see Section 2.6.4 “Automatic Meshing
(Auto-meshing) Using The Command Language”) is simpler to use, requires less user-
specification and is particularly well-suited to specification of epitaxial (especially multiple-
quantum well or super-lattice) devices. It is for these reasons that we recommend the auto-
mesh approach when defining LED devices .
In the following sections, we will focus mainly on this approach. We will point out important
differences with the other methods: ATHENA, DEVEDIT or manual-mesh as these differences
may arise.
The first command in any ATLAS input deck is the MESH command. When loading a structure
from ATHENA, DEVEDIT or a previous ATLAS simulation, you should assign the INFILE
parameter of the MESH statement to the name of the file containing the device structure. For
example:
MESH INFILE=led.in
would cause the simulator to load a device structure from the file named led.in in the
current working directory. When defining the device structure in the ATLAS auto-meshing
syntax, the first statement should contain the AUTO parameter as shown in the following
statement:
MESH AUTO
This will invoke the auto meshing capability.

Note: The INFILE parameter is not specified since this would indicate that the device was both being loaded as
well as constructed in the ATLAS synta.x Therefore, would produce an ambiguity or conflict.

603 ATLAS User’s Manual

Defining Light Emitting Devices (LEDs) LED: Light Emitting Diode Simulator

When using auto-meshing, you must only specify the mesh in the X direction using X.MESH
statements. The mesh spacing in the Y direction is specified in the REGION statements as
described later. The following example shows how you would specify an X mesh 1 m long
with a uniform mesh spacing of 0.1 m.
X.MESH LOCATION=0 SPACING=100
X.MESH LOCATION=1000 SPACING=100
Using auto-meshing, you should next specify the device regions using REGION statements.
The REGION statement can be used to specify the material types of individual "regions", as
characterized by a spatial extent in the X and Y directions, and their composition, doping,
strain and certain other characteristics describing how the region is to be modeled.
In the following example, we will describe a simple GaN/InGaN diode structure suitable for
LED device simulation.
REGION MATERIAL=GaN THICKNESS=0.25 ACCEPTOR=1e19 BOTTOM NY=20
REGION MATERIAL=InGaN THICKNESS=0.003 DONOR=2e14 X.COMP=0.2 BOTTOM
NY=20 LED
REGION MATERIAL=GaN THICKNESS=0.25 DONOR=1e18 BOTTOM NY=20
In this example, we describe a three layer LED. The top layer is composed of GaN
(MATERIAL=GaN), it has a thickness of 0.25 microns (THICKNESS=0.25) and has an acceptor
concentration of 1e19 (ACCEPTOR=1e19). The middle layer is composed of InGaN
(MATERIAL=InGaN), it has a thickness of 0.003 microns (THICKNESS=0.003) cm-3, has an In
composition fraction of 0.2 (X.COMP=0.2), and has a donor concentration of 2e14
(DONOR=2e14) cm-3. The bottom layer is also composed of GaN (MATERIAL=GaN), it has a
thickness of 0.25 microns (THICKNESS=0.25), and has a donor concentration of 1e18
(DONOR=1e18) cm-3.
The BOTTOM parameter in each of the REGION statements specifies that each region resides
directly below the preceeding region. NY specifies the minimum number of Y grid lines used
to resolve the region. The LED parameter in the second REGION statement specifies that region
will be treated as a light emitting region during the post-processing stages of the simulation,
and certain light emitting characteristics of that region will be extracted.
The final step in defining the LED using auto-meshing is to specify the electrodes. To specify
electrodes, use the ELECTRODE statement. The following example specifies two electrodes
extending all the way across the top and the bottom of the device.
ELECTRODE NUMBER=1 NAME=anode TOP
ELECTRODE NUMBER=2 NAME=cathode BOTTOM
Here, the NUMBER parameter is used to give the electrodes numerical tags they can be referred
to in subsequent operations. Similarly, the NAME parameter is used for future references (see
Section 21.57 “SOLVE”).

604 ATLAS User’s Manual

Specifying Light Emitting Diode Models LED: Light Emitting Diode Simulator

11.3 Specifying Light Emitting Diode Models

In this section, we discuss methods and models particular to LED analysis.

11.3.1 Specifying Polarization and Piezoelectric Effects

In some material systems (such as the GaN/InGaN system), the built-in fields due to
polarization and strain (piezoelectric effect) can play significant roles in the LED emission
characteristics. ATLAS introduces these polarization fields as sheet charges at the top and
bottom edges of regions. To include the effects of polarization on a region, you should specify
the POLARIZATION parameter on the corresponding REGION statement.
This will only include spontaneous polarization. If you want to also include piezoelectric
polarization, you can also specify the CALC.STRAIN parameter. If you specify CALC.STRAIN,
the simulator will automatically calculate strain from the lattice mismatch and will calculate
the piezoelectric polarization and apply it to the region. You can also specify the value of the
STRAIN parameter, which specifies the axial strain in the region.
With STRAIN and POLARIZATION set, the simulator will apply a piezoelectric polarization
calculated using the strain value assigned by the STRAIN parameter.

11.3.2 Choosing Radiative Models

The next step in simulating LED devices is the selection of appropriate radiative models. For
LED devices, the most relevant radiative model is the model for spontaneous (radiative)
recombination. For radiative recombination, ATLAS/LED offers several options. Section 3.9
“Optoelectronic Models” discusses these options in more detail.
First, is the general radiative model (see Equation 3-368). To enable this equation, specify
OPTR in a MODELS statement associated with the active region(s) of the LED device (i.e., those
regions with the LED parameter specified in the REGION statement). For example:
MODELS MATERIAL=InGaN OPTR
will enable the standard radiative recombination model. The disadvantage of using the
standard model is that it provides no information about the spectral content of the LED light
emission. In fact, using this model to properly specify emission intensity, you must specify
the emission wavelength in the SOLVE statement for extraction of LED luminous intensity. To
specify the emission wavelength for the general radiative recombination model given by
Equation 3-366, assign the value of the wavelength to the L.WAVE parameter as shown in the
following SOLVE statement, where L.WAVE is in units of microns. For example:
SOLVE L.WAVE=0.51
The COPT parameter of the MATERIAL statement specifies the rate constant for the general
radiative recombination model. For example:
MATERIAL COPT=1.25e-10

COPT assigned a value of 1.2510-10 cm3/s to the radiative rate constant. You can calibrate
the value of this constant using the more physical radiative models discussed next. This
calibration will be discussed at the end of this section.
Beyond the basic model just described, ATLAS/LED provides three more physical models for
radiative rates. For zincblende materials, ATLAS provides a one band model [318] and a two
band [162] model. For wurtzite materials, ATLAS/LED provides a three band model [50].

605 ATLAS User’s Manual

Specifying Light Emitting Diode Models LED: Light Emitting Diode Simulator

The one band zincblende model by Yan et.al. [318] accounts for optical transitions between a
single valence band and the conduction band. To activate this model, specify ZB.ONE in the
MODELS statement. See Section 3.9.7 “Unstrained Zincblende Models for Gain and Radiative
Recombination” for a discussion of the Yan model.
The two band zincblende model [162] accounts for optical transitions between light and
heavy hole valence bands and the conduction band. To activate this model, specify ZB.TWO in
the MODELS statement. See Section 3.9.7 “Unstrained Zincblende Models for Gain and
Radiative Recombination” for a discussion of the ZB.TWO model.
The three band wurtzite model [50] is based on k.p modeling and accounts for optical
transitions between the conduction band and heavy and light hole and the crystal field split-
off valence bands. See Section 3.9.10 “Strained Wurtzite Three-Band Model for Gain and
Radiative Recombination” for a discussion of the WZ.KP model. To activate this model,
specify WZ.KP in the MODELS statement. For the on-going example, we chose the following
WZ.KP model (since InGaN is a wurtzite material).
MODEL MATERIAL=InGaN WZ.KP
You should note that any of the ZB.ONE, ZB.TWO, or WZ.KP model can be used to analyze
spectral characteristics, such as emission wavelength and emission spectra.
You can also choose to analyze LED efficiency degradation due to competing recombination
mechanisms by enabling these other non-radiative mechanisms. Particularly, you can enable
Schockley-Read-Hall or Auger mechanisms. To activate these models, specify SRH and
AUGER in the MODELS statement as follows:
MODELS SRH AUGER
Make sure you do not enable two radiative mechanisms for the same region, such as OPTR and
one or more of the ZB.ONE, ZB.TWO, or WZ.KP models. If you do, you may over account for
the same radiative recombination. For more information about recombination models, see
Section 3.6.3 “Carrier Generation-Recombination Models”.
As mentioned before, you can calibrate the radiative rate constant (using COPT of the
MATERIAL statement) of the basic radiative recombination model (enabled by the OPTR
parameter of the MODELS statement) to the more physical models (ZB.ONE, ZB.TWO, or
WZ.KP). To do this, first prepare a simulation with the activated physical model. In this
simulation, save a structure file around the device operating conditions. You can then extract
the proper value of COPT to use with the OPTR model by dividing the radiative recombination
rate at any point by the product of the electron and hole concentrations.

11.3.3 Using k.p Band Parameter Models in Drift Diffusion

For the WZ.KP model, the band calculations performed in Equations 3-671, 3-672, 3-673, and
3-675 can be used in the drift diffusion part of the simulation. To use this capability, specify
the K.P parameter on the MODELS statement as follows:
MODELS WZ.KP K.P

606 ATLAS User’s Manual

Data Extraction LED: Light Emitting Diode Simulator

11.4 Data Extraction

11.4.1 Extracting Luminous Intensity
When the LED parameter is specified on any REGION statement, the luminous intensity
emitted by the LED device is automatically calculated and written to the log file. The file that
this information is output to is specified by the OUTFILE parameter of the LOG statement. The
following example outputs the log file to the file led.log.
LOG OUTFILE=led.log
You can then load led.log into TONYPLOT for post-processing visualization, where you can
plot characteristics, such as luminous intensity versus current (see Figure 11-1).

Figure 11-1: Example Luminous Intensity Plot

Note: Luminous intensity is calculated by integrating the luminous spectrum obtained by models containing
spectral content (e.g., MODELS WZ.KP SPONT). This does not include other non-spectral models (e.g.,
MODELS OPTR).

607 ATLAS User’s Manual

Data Extraction LED: Light Emitting Diode Simulator

11.4.2 Extracting Emission Spectra

The LED simulator can save multiple spectrum files from the SOLVE statement. To save a
spectrum file, specify the SPECTRUM parameter after each solution. SPECTRUM is assigned to
the file name prefix. Each subsequent file will be written to a file named by SPECTRUM and
appended with a version number starting at 0.
The LED simulator can output spectral intensity as a function of energy and wavelength
integrated over all wells specified as LEDs. To output spectral intensity, specify the
SPECTRUM parameter of the SAVE statement. SPECTRUM is assigned to the name of a file that
the LED spectrum is written.
You must also specify either LMIN and LMAX or EMIN and EMAX. LMIN and LMAX specify the
minimum and maximum values of wavelength in microns to be captured. EMIN and EMAX
specify the minimum and maximum values of energy in eV to be captured. Specify either
energy range or wavelength range but not both. To capture the number of descrete samples,
specify the NSAMP parameter of the SAVE statement.
Figure 11-2 shows an example spectrum file plotted in TONYPLOT.

Figure 11-2: Plot of Spectrum File for LED

Note: Emission spectra can only be obtained using a spectral model (e.g., MODELS WZ.KP SPONT).

608 ATLAS User’s Manual

Data Extraction LED: Light Emitting Diode Simulator

11.4.3 Extracting Emission Wavelength

In the LED simulator, you can use the probe to measure wavelength for LED devices. This
will extract the wavelength from the peak of the spectral response over a specified range. To
use this capability, specify WAVELENGTH in the PROBE statement. Use either EMIN and EMAX or
LMIN and LMAX in the PROBE statement to specify the search range. EMIN and EMAX specify the
range in terms of energy in eV, while LMIN and LMAX specify the range in terms of wavelength
in microns.

609 ATLAS User’s Manual

Reverse Ray-Tracing LED: Light Emitting Diode Simulator

11.5 Reverse Ray-Tracing

Reverse ray-tracing is a technique that allows you to obtain optical output characteristics of
an active optoelectronic device based on material properties and device geometry. Direct ray-
tracing is commonly used for modeling of passive optoelectronic components, such as photo-
detectors.
In direct ray-tracing, rays originating at the external light source are traced into the device and
are absorbed to form electron/hole pairs, which are subsequently detected (using the ray-
tracing method in LUMINOUS is described in Sections 10.2.1 “Ray Tracing in 2D” and 10.2.2
“Ray Tracing in 3D”). Unlike direct ray-tracing, the rays in the reverse method originate
inside the active region and are traced until they exit the device. One or multiple origin points
(user specified) for rays are considered, and interference effects for rays originating at the
common source point can be taken into account. When you enable interference, you can
analyze the spectral selectivity of the device structure by performing ray-tracing at multiple
wavelengths.
You can assess light extraction from the device by using the SAVE statement (after biasing).
Five different sets of parameters for reverse ray-tracing are used in the examples throughout
the section to demonstrate the features available for modeling LEDs. The parameter sets are
given in the order of increasing complexity of simulation. Computation times also vary. It
takes approximately 200 times longer to run the last SAVE line compared to the first one.
Set the ANGPOWER parameter in the SAVE statement to start the reverse ray-tracing algorithm.
The name of the output file containing the angular power density vs. output angle dependence
is specified as a value for ANGPOWER. The information in the angpower file includes the
angular power density for TE- and TM- polarized light and total angular power density and
total flux angular density vs. output angle.

Note: In TonyPlot, polar charts show the Y axis directed upward (in the opposite direction to the internal coordinate
system used in ATLAS). Therefore, the plots will appear to be flipped around X axis (top of the structure is at
the bottom of the chart).

The RAYPLOT parameter specifies the name of the output file containing the information on
each ray exiting the device. This file is only created when single origin for all rays is
assumed. The information includes ray output angle, relative ray power (TE-, TM-
polarization, and total), and initial internal angle at the origin (only if INTERFERE parameter
is not specified). 0° angle corresponds to the rays in the X axis direction. 90° angle
corresponds to the rays in the Y axis direction.
To start ray-tracing from one point of origin, specify the origin’s coordinates for rays X and Y
and the wavelength L.WAVE. It is important to choose the origin in the active region of the
device. Rays are not traced if the radiative recombination is zero at the ray origin. All
remaining parameters outlined below are optional and their default values are assumed if the
parameters are not specified.
REFLECTS specifies a number of reflections to be traced for each ray originating at the point
(X,Y) (similar to REFLECTS parameter in the BEAM statement). The default value
(REFLECTS=0) provides for a quick estimate of the coupling efficiency. To obtain a more
accurate result, use REFLECTS>0, especially if you specify MIR.TOP or MIR.BOTTOM. The
choice of this parameter is based on a compromise between calculation time and accuracy.
The maximum allowed value is REFLECTS=10. Number of reflections set to 3 or 4 is often a

610 ATLAS User’s Manual

Reverse Ray-Tracing LED: Light Emitting Diode Simulator

good choice. Example 1 produces a simple ray-tracing analysis of an LED. Figure 11-3
shows the resulting angular distribution of the emitted light intensity.
Example 1
SAVE ANGPOWER=OPTOEX18ANG_1.LOG RAYPLOT=OPTOEX18RAY_1.LOG X=2.0
Y=1.01 L.WAVE=0.9 REFLECTS=4

Figure 11-3: Emitted Light Intensity vs Angle for Example 1

Specify the DIPOLE parameter to turn on a particular angular distribution of the internal
radiating field that corresponds to a preferred in-plane orientation of dipoles. This orientation
is often found in polymer based OLEDs and results in higher optical output coupling. The
default setting (DIPOLE=false) corresponds to isotropic distribution of emitting dipoles and
spherically symmetric radiation pattern.
MIR.TOP specifies that the top surface of the device be treated as an ideal mirror.
MIR.BOTTOM specifies that the bottom surface of the device be treated as an ideal mirror.
SIDE specifies that the rays reaching the sides of the device terminate there and do not
contribute to the total light output. This is often a good assumption for realistic LEDs as these
rays tend to be either absorbed internally or blocked by the casing of the device.
TEMPER is the temperature (needed for using appropriate refractive indexes of the materials).
The default setting of 300 K will be used if TEMPER is not specified.
POLAR specifies polarization of the emitted photons in degrees (linearly polarized light is
assumed). Parallel (TM-mode, POLAR=0.0) and perpendicular (TE-mode, POLAR=90.0)
polarizations result in significantly different output coupling values.
You should use the default value (POLAR=45.0) if there is no preferred direction of
polarization of emitted photons (unpolarized light emission).

611 ATLAS User’s Manual

Reverse Ray-Tracing LED: Light Emitting Diode Simulator

MIN.POWER specifies the minimum relative power of a ray (similar to MIN.POWER parameter
in the BEAM statement). The ray is not traced after its power falls below MIN.POWER value.
This is useful for limiting the number of rays traced. The default value is MIN.POWER=1e-4.
NUMRAYS specifies the number of rays starting from the origin. The default is 180. The
acceptable range 36-3600.
Example 2 shows how some of the parameters described above are used for modeling of a
realistic LED. The angular distribution of light power in Figure 11-4 exhibits almost a
Lambertian pattern for this simple planar LED geometry. Note that optical coupling
coefficient produced in this calculation reflects a 2D nature of the example (the light origin is
not a point but rather an infinite line in Z direction).
Simultaneously, you can calculate the optical coupling efficiency for an axially symmetric 3D
device. Normally, this is the value to be compared with experimental results. To do this,
specify the COUPLING3D parameter (while SIDE is set). For this calculation, the light source is
assumed to be a point located on the axis of symmetry.
Example 2
SAVE ANGPOWER=OPTOEX18ANG_2.LOG RAYPLOT=OPTOEX18RAY_2.LOG X=2.0
Y=1.01 L.WAVE=0.9 SIDE MIR.BOTTOM NUMRAYS=360 REFLECTS=4

Figure 11-4: Emitted Light Intensity vs Angle for Example 2

612 ATLAS User’s Manual

Reverse Ray-Tracing LED: Light Emitting Diode Simulator

The rays are assumed to be incoherent by default. This is a good approximation if the
thickness of the active layer of the device is on the order of a wavelength/index. For layer
thicknesses, smaller than the wavelength coherent effects might be important. When
INTERFERE is set, the rays originating at the common source point are taken to be 100%
coherent. In this case, the phase information upon propagation is preserved. Phase change
upon reflection is also considered. Thus, interference of rays exiting the device at the same
angle is taken into account. In this case, the internal angle information is not written to the
output rayfile.
Example 3 takes into account interference. Figure 11-5 shows the result. You can take
absorption of rays into account by setting the ABSORB parameter. The absorption is assumed
to be specified for each material by the imaginary part of the refractive index. Carrier density
dependent absorption and photon-recycling are not considered at this point.
Example 3
SAVE ANGPOWER=OPTOEX18ANG_3.LOG RAYPLOT=OPTOEX18RAY_3.LOG X=2.0
Y=1.01 L.WAVE=0.9 INTERFERE SIDE MIR.BOTTOM REFLECTS=4

Figure 11-5: Emitted Light Intensity vs Angle for Example 3

Although ray-tracing from one point of origin can give a reasonable estimate of optical output
coupling and angular distribution of light power, it is often desirable to consider multiple
points within the active layer of the device to obtain more accurate results. To set multiple
origin points, use the following parameters.
• XMIN, XMAX, YMIN, and YMAX define a rectangular area containing all origin points.
• XNUM and YNUM specify the number of points along x and y axes within the rectangular
area. If XNUM=1, the X coordinate of all origin points will be set to XMIN so that the points
are chosen along the line X=XMIN. Similarly, you can choose points along the specific y-
line.

613 ATLAS User’s Manual

Reverse Ray-Tracing LED: Light Emitting Diode Simulator

Ray-tracing from multiple origins is realized by repeating single origin algorithm for each
point and by adding up the normalized angular power density values thus obtained. The
luminous power assigned to each source (origin) is proportional to the radiative
recombination at that point. The luminous power of all sources adds up to the value obtained
by integration of radiative recombination over the entire device. Rays originating at different
source points are completely incoherent (even when INTERFERE is set), which is consistent
with the spontaneous character of the radiation produced by an LED. The rayplot file is not
written for multiple origins (even if RAYPLOT parameter is set). Example 4 shows how
multiple origin simulation can be done. Figure 11-6 shows angular distribution of light
obtained in this case.
Example 4
SAVE ANGPOWER=OPTOEX18ANG_4.LOG XMIN=1.0 XMAX=3.0 XNUM=3 YMIN=1.01
YMAX=1.09 YNUM=9 INTERFERE SIDE L.WAVE=0.9 MIR.BOTTOM REFLECTS=4

Figure 11-6: Emitted Light Intensity vs Angle for Example 4

You can also take spectral selectivity of optical output coupling for LEDs into account.
Reverse ray-tracing at multiple wavelengths is considered if the SPECTRUM parameter
specifies a filename for spectral selectivity output while the ANGPOWER parameter is set.
EMIN and EMAX, or LMIN and LMAX specify the energy or wavelength range respectively.
NSAMP specifies the number of spectral components to be considered. We suggest that you
specify the SPECTRUM file only when INTERFERE is set. Otherwise, a warning will be issued
during ray-tracing. When ANGPOWER and SPECTRUM are set in the SAVE statement, the
resulting optical output coupling is averaged over the entire energy (wavelength) range from
EMIN to EMAX (from LMIN to LMAX). The same applies to the quantities in the output angular
distribution file. The spectrum file only shows how output coupling changes with
wavelength. Currently, the shape of the gain curve is not taken into account (flat gain).

614 ATLAS User’s Manual

Reverse Ray-Tracing LED: Light Emitting Diode Simulator

Example 5 shows how to use multiple spectral components. Figure 11-7 shows that the results
obtained after averaging over spectral components and multiple origin points while taking
interference into account are similar to the single point source model, where interference and
multiple spectral content are ignored (Example 2). This demonstrates the applicability of a
simpler model in this case.
Example 5
SAVE ANGPOWER=OPTOEX18ANG_5.LOG SPECTRUM=OPTOEX18SP_5.LOG XMIN=1.0
XMAX=3.0 XNUM=3 YMIN=1.01 YMAX=1.09 YNUM=5 INTERFERE SIDE LMIN=0.86
LMAX=0.94 NSAMP=15 MIR.BOTTOM REFLECTS=4

Figure 11-7: Emitted Light Intensity vs Angle for Parameter Set 5

615 ATLAS User’s Manual

FDTD Analysis of Output Coupling LED: Light Emitting Diode Simulator

11.6 FDTD Analysis of Output Coupling

Finite difference time domain analysis is well suited to calculation of output coupling from
periodic structures, such as LED with photonic-crystal coupling. Performing such analysis is
only a small deviation from the methods discussed in the previous section.
First, you must remember that you are going to simulate a periodic device by applying
periodic boundary conditions. These boundary conditions appear in two places: the device
mesh and the FDTD mesh. Therefore, you need to make sure that the two meshes both have
the same periodicity. If you cannot define such a device, you cannot take advantage of
periodicity.
To model devices as periodic in the device simulator, simply add the PERIODIC parameter to
the MESH statement at the beginning of the simulation deck like this:
MESH PERIODIC
or
MESH PERIODIC THREE.D
Next, you can either load a device structure or syntactically create a structure. But in both
cases, you need to make sure that the device and any optical elements defined above the
semiconductor device are periodic in the X and Z directions if desired.
You also need to include a BEAM statement. The BEAM statement is used to specify the optical
problem. The application of the FDTD method along with the BEAM statement is discussed
fully in Section 10.5 “Finite Difference Time Domain Analysis”.
Your BEAM statement can be easily composed. First, you define the BEAM origin and direction
of propagation. Currently, this is always done for the LUMINOUS simulator.
The BEAM origin should have the Y coordinate above the device with enough space to fit a
plane to evaluate near and far field patterns. Using a coordinate of 1/4 of the smallest
wavelength is adequate for that space. Your BEAM statement should start as follows:
BEAM NUM=1 FDTD X.ORIGIN=0 Z.ORIGIN=0 Y.ORIGIN=-0.2/4 ANGLE=90 \
Here, the X and Z origins do not matter and the Y origin is set, assuming that your device
starts at Y=0. The angle ensures the domain is properly oriented with Y pointing down.
Next, you need to include some parameters for defining the discretization. These parameters
should be defined like this:
BEAM NUM=1 FDTD X.ORIGIN=0 Z.ORIGIN=0 Y.ORIGIN=-0.2/4 ANGLE=90 \
FD.AUTO TD.SRATE=20 PROP.LENG=4 BIG.INDEX TD.ERR=0.01 \
Here, we are saying that the spatial mesh must be sample at least 20 times per wavelength
(TD.SRATE=20). This is the number suggested in literature and we have found it works well.
Please be aware that computation time goes up as the square of this number in 2D and the
cube in 3D.

616 ATLAS User’s Manual

FDTD Analysis of Output Coupling LED: Light Emitting Diode Simulator

Next, we are going to shine the light for an amount of time needed to propagate 4 microns.
This is done in a number of discrete time steps. This distance is very much related to the
maximum dimension of your device. In this case, we are saying that our device is of the order
of 4 microns in size. Actually, here we also need to consider our PMLs that should be of the
same order in size as our device and account for multiple reflections. So a good number
might be 4 or 5 times the maximum dimension of your device.
The next parameter accounts for the fact that the speed of light in media is slower than in
vacuum. This means the calculations based on given by TD.SRATE and PROP.LENG are
automatically multiplied by the largest index of refraction in the device at the wavelength of
operation.
The parameters TD.SRATE, PROP.LENG, and BIG.INDEX control the absolute size of the
problem. The TD.ERR parameter specifies a convergence error that may be met before
PROP.LENG is reached and end the simulation early. You should start out with this set at 0.01
and look at the error log discussed next to determine if you should change that value.
The residual error is calculated as the integrated variation of the envelope of all fields
between two successive wavelengths (cycles) normalized by the integrated envelope.
The next line shows how you output a error log.
BEAM NUM=1 FDTD X.ORIGIN=0 Z.ORIGIN=0 Y.ORIGIN=-0.2/4 ANGLE=90 \
TD.SRATE=20 PROP.LENG=4 BIG.INDEX TD.ERR=0.01 \
TD.LOG=example \
This will cause the error log to be output to the file "example.log". The following figure
shows an error log in TONYPLOT.

Figure 11-8: Error Log

617 ATLAS User’s Manual

FDTD Analysis of Output Coupling LED: Light Emitting Diode Simulator

The level steps are due to the evaluation only occurring every 1 wavelength (cycle). The first
is zero since there is no previous envelope to compare with. The error log also can be used to
display the absorption in the various PMLs and bulk vs. iteration. Plotting these data can be
very useful in optimizing your experiment for speed and accuracy.
The next line shows you how you can extract structure files containing the complete field
solutions produced from FDTD.
BEAM NUM=1 FDTD X.ORIGIN=0 Z.ORIGIN=0 Y.ORIGIN=-0.2/4 ANGLE=90 \
TD.SRATE=20 PROP.LENG=4 BIG.INDEX TD.ERR=0.01 \
TD.LOG=example \
TD.FILE=example SUBNX=100 SUBNY=100 SUBNZ=100 \
Here, we make the simulator output a structure file to the name "example.str" subsampled
to 200200200 samples. We highly recommend you subsample your output files until you
see what your system can handle. This structure file contains all the field information,
structural information including the optical elements such as lenslets and PMLs, and the
optical intensity and photogeneration rate.
You can extract cutplanes through the structure file interface by specifying the TD.ZCUT or
TD.XCUT parameters.
You can assign the FACET parameter to a file root name to save the near and far field patterns
normal to an axis as defined by the X.FACET, Y.FACET, or Z.FACET parameters. Such an
assignment may look like this:
BEAM NUM=1 FDTD X.ORIGIN=0 Z.ORIGIN=0 Y.ORIGIN=-0.2/4 ANGLE=90 \
TD.SRATE=20 PROP.LENG=4 BIG.INDEX TD.ERR=0.01 \
TD.LOG=example \
TD.FILE=example SUBNX=100 SUBNY=100 SUBNZ=100 \
FACET=example Y.FACET=-0.2/4/2 \
Here, we intend to write two files "example.nfp" and "example.ffp" containing the near
field pattern and the far field pattern. The far field pattern is constructed from the 2D fft of
the near field pattern for 3D and the 1D fft for 2D.
You will notice we placed the output facet in between the top of the device and the BEAM
origin as defined earlier.
Finally, we need to specify where the output is to be evaluated. Usually, we consider light
flowing out the top to be successfully coupled to the output. We specify this conditon as
follows:
BEAM NUM=1 FDTD X.ORIGIN=0 Z.ORIGIN=0 Y.ORIGIN=-0.2/4 ANGLE=90 \
TD.SRATE=20 PROP.LENG=4 BIG.INDEX TD.ERR=0.01 \
TD.LOG=example \
TD.FILE=example SUBNX=100 SUBNY=100 SUBNZ=100 \
FACET=example Y.FACET=-0.2/4/2 \
OUT.TOP

618 ATLAS User’s Manual

FDTD Analysis of Output Coupling LED: Light Emitting Diode Simulator

This basically means that all the energy absorbed in the TOP PML is compared against the
total energy absorbed to calculate the output coupling. You may name any PML this way and
can include more than one PML. That is all you need for the BEAM statement.
Next, we need to define the PMLs. PMLs should not intimidate you and they are used for
three reasons:
• They absorb everything for a given wavelength. There is no reflection if the indexes are
properly matched.
• The absorption in PMLs is integrated to evaluate how much energy leaves the device
through that face. This is compared with the absorption integrated in the bulk to give
output coupling.
• If your PMLs are not properly designed (i.e., index matched at the boundaries), you will
see increasing errors. This usually means you are introducing more energy from the
source than is allowed to escape through the PMLs or to be absorbed in the bulk.
For now, keep it simple and define your PMLS to have a width of about size of the device in
that direction and assign the following two parameters:
DEGREE=1 R90=0.0001
You must define either the index or the material that the PML abuts (is coincident along the
edge of) in the device domain. This index must match or the interface will produce
reflections. The index is specified using the REAL.INDEX parameter or material is specified
using the MATERIAL parameter. The index match must be exact over all wavelengths.
Next, we can assign various lenslets and or photonic crystal elements on top of the structure
to augment output coupling. Such descriptions are described in Section 10.10 “Specifying
Lenslets, Texturing, and Photonic Crystals”.
Finally, we specify the extraction of output coupling on the SAVE statement just like we do for
reverse ray tracing (see Section 11.5 “Reverse Ray-Tracing”) except we replace the
assignment of ANG.POWER with an assignment of REF.BEAM to the BEAM index specified by
the NUMBER parameter in the previous BEAM definition
This triggers evaluation of the output coupling that can be examined in the file specified by
the SPECTRUM parameter of the SAVE statement.

619 ATLAS User’s Manual

The LED Statement LED: Light Emitting Diode Simulator

11.7 The LED Statement

The LED statement can be used as a convenient method to perform all of the above mentioned
SAVE operations automatically after each solution step.
The parameter names are the same as the individual SAVE statements. By default, the files
named by the SPECTRUM, ANGPOWER and SPECT.ANGLE parameters of the LED statement will
given unique incremented names at each save. This may generate many files. If you want to
save only the last version of each file, specify the ONEFILE parameter of the LED statement.

11.7.1 User-Definable Emission Spectra

For some materials (particularly organics), the radiative models for emission spectra may be
inadequate. In these cases, you can specify up to two user-definable emission spectra. These
emission spectra are defined using a simple tabular description and is stored in a file separate
from the input deck. The file contains first a line containing one number representing the
number of wavelength-intensity pairs to follow. The wavelength-intensity pairs are lines with
two numbers in them, the wavelength in microns and the relative intensity in Watts. The
absolute intensity is not important since the intensity values are normalized.
The following file fragment demonstrates the format:
12
0.4 .00142
0.42 0.00222
0.44 0.013
.
.
.
Here, the first line contains the number of samples, in this case 12. Then next line specifies
that at 0.4 microns wavelength the relative intensity is 0.00142 Watts etc.
To use the user-defined spectra, assign the values of USER.SPECT or DOPE.SPECT or both to
the appropriate file name. To output the user and dopant spectra to TONYPLOT compatible log
files, specify OUT.USPEC and OUT.DSPEC.

620 ATLAS User’s Manual

Chapter 12
MixedMode: Mixed Circuit and
Device Simulator
Overview MixedMode: Mixed Circuit and Device Simulator

12.1 Overview
MIXEDMODE is a circuit simulator that can include elements simulated using device
simulation and compact circuit models. It combines different levels of abstraction to simulate
relatively small circuits where compact models for single devices are unavailable or
sufficiently accurate. MIXEDMODE also allows you to also do multi-device simulations.
MIXEDMODE uses advanced numerical algorithms that are efficient and robust for DC,
transient, small signal AC and small signal network analysis.
MIXEDMODE is typically used to simulate circuits that contain semiconductor devices for
accurate compact models that don’t exist or circuits where devices play a critical role must be
modeled accurately. Applications of MIXEDMODE include: power circuits that may include
diodes, power transistors, IGBTs, and GTOs, optoelectronic circuits, circuits subject to single
event upset, thin film transistor circuits, high-frequency circuits, precision analog circuits,
and high performance digital circuits.
MIXEDMODE circuits can include up to 200 nodes, 300 elements, and up to ten numerical
simulated ATLAS devices. These limits are reasonable for most applications. But, they can be
increased in custom versions on request to Silvaco. The circuit elements that are supported
include dependent and independent voltage and current sources as well as resistors,
capacitors, inductors, coupled inductors, MOSFETs, BJTs, diodes, and switches. Commonly
used SPICE compact models are available. The SPICE input language is used for circuit
specification.
This chapter describes circuit simulation capabilities rather than device simulation
capabilities. The first part of the chapter contains introductory and background information.
Then, describes presents and explains MIXEDMODE syntax. This is followed by some sample
input decks. The final sections contain a statement reference and a detailed description of the
provided electrical compact models for diodes, BJTs, and MOSFETs.

12.1.1 Background
Circuit simulators such as SPICE [11] solve systems of equations that describe the behavior
of electrical circuits. The devices that are of interest to circuit designers are normally well
characterized. Compact or circuit models are analytic formulae that approximate measured
terminal characteristics. Advanced compact models provide high accuracy with minimum
computational complexity. Device modeling, device characterization and parameter
extraction are concerned with the development and use of accurate and efficient compact
models.
Physically based device simulation solves systems of equations that describe the physics of
device operation. This approach provides predictive capabilities and information about the
conditions inside a device. It can, however, require significant amounts of CPU time.
Information is usually transferred from device simulation to circuit simulation as follows:
electrical characteristics are calculated using a physically-based device simulator. These
calculated electrical characteristics are then used as input by a device modeling and parameter
extraction package such as UTMOST [279]. The extracted parameters are used to
characterize a compact model used by the circuit simulator.

622 ATLAS User’s Manual

Overview MixedMode: Mixed Circuit and Device Simulator

This approach is adequate for many purposes but has limitations. It requires that satisfactory
compact models already exist. The use of compact models always introduces some error.
Models that are adequate for digital circuit simulation may be inadequate for other
applications. Applications and devices for which compact modeling is not always satisfactory
include: precision low power, high power, high frequency circuit simulation, SOI, IGBT,
GTO, TFT, and optoelectronic devices.

12.1.2 Advantages of MixedMode Simulation

The limitations of compact models can be overcome by using physically-based device
simulation to predict the behavior of some of the devices contained in a circuit. The rest of the
circuit is modeled using conventional circuit simulation techniques. This approach is referred
to as mixed-mode simulation, since some circuit elements are described by compact models,
and some by physically-based numerical models.
MIXEDMODE simulation provides several worthwhile advantages. No compact model need be
specified for a numerical physically-based device. The approximation errors introduced by
compact models can be avoided particularly for large signal transient performance. You can
also examine the internal device conditions within a numerical physically-based device at any
point during the circuit simulation. But the cost is increased CPU time over SPICE as CPU
time is comparable to a device simulation excluding the external circuit nodes. MIXEDMODE
simulation normally uses numerical simulated devices typically only for critical devices.
Non-critical devices are modeled using compact models.

623 ATLAS User’s Manual

Using MixedMode MixedMode: Mixed Circuit and Device Simulator

12.2 Using MixedMode

Input file specification for MIXEDMODE is different in many respects to the rest of ATLAS.
But if you’re familiar with SPICE and ATLAS syntax, you should have little difficulty
understanding how these two syntax styles are joined in MIXEDMODE.
Each input file is split in two parts. The first part is SPICE-like and describes the circuit
netlist and analysis. The second part is ATLAS-like and describes the device simulation
model parameters. These two sections of an input file are separated as described in the next
section.
The circuit description includes the circuit topology (called the netlist) and the electrical
models and parameters of the circuit components. The simulation conditions specify the types
of analysis to be performed. These items are described using syntax based on SPICE.
The ATLAS device descriptions provide information about device geometry, doping
distribution, and meshes. Device descriptions can be prepared using the built-in ATLAS
syntax, the ATHENA process simulator, or the DEVEDIT structure specification and meshing
tool. You can also read-in previously calculated device solutions. The device data is read-in
from standard structure format files.
When a simulation has finished, the following information is available:
• I-V data (voltages in all circuit nodes and currents in all circuit branches).
• Internal distributions of solution variables (such as electron, hole, and potential
distributions) within the numerical devices.
The results of previous runs of MIXEDMODE can be used as initial guesses for future
simulations. This is particularly helpful when multiple simulations must be performed from
the same starting point.
The accessing and running of examples for ATLAS are documented in the DECKBUILD
USER ’S MANUAL. We recommend that you run at least one MIXEDMODE example provided
on the distribution tape before trying their own simulations.

12.2.1 General Syntax Rules

The SPICE-like part of any MIXEDMODE input file starts with the .BEGIN statement. The
SPICE-like part of the input file ends with .END. All parameters related to the device
simulation models appear after the .END statement. To start MIXEDMODE3D, specify the 3D
parameter in the BEGIN statement.
The first non-comment statement after initializing ATLAS (go atlas) has to be .BEGIN. The
order of the following netlist and control statements is arbitrary. The last SPICE-like
statement has to be .END.
Unlike the rest of ATLAS, for SPICE-like statements the exact command has to be used,
unique abbreviations are not accepted. Statements are not case sensitive.
There has to be at least one numerical ATLAS device (“A” device) within the netlist.
Comment characters are # and $, but not *.
All ATLAS statements specifying the parameters for the numerical device simulation have to
be specified after .END

624 ATLAS User’s Manual

Using MixedMode MixedMode: Mixed Circuit and Device Simulator

After all ATLAS statements, the simulation has to be explicitly terminated (quit, go
<simulator>).
These rules do not apply to the SET statement for parameterization of the input file, since it is
interpreted by DECKBUILD only.
EXTRACT statements are also an exception similar to SET. Since MIXEDMODE input files are
parsed completely before execution (see Section 12.2.4 “Recommendations” for more
information), extractions can only be done after completion of the simulation. To extract
results from a MIXEDMODE simulation, EXTRACT should be specified after re-initialization of
ATLAS (go atlas).

12.2.2 Circuit and Analysis Specification

The SPICE-like MIXEDMODE statements can be divided into three categories:
• Element Statements: These statements define the circuit netlist.
• Simulation Control Statements: These statements specify the analysis to be performed.
• Special Statements: These statements typically related to numerics and output (first
character being a dot “.”).
The specification of the circuit and analysis part has to be bracketed by a.BEGIN and an.END
statements. In other words, all MIXEDMODE statements before.BEGIN or after .END will be
ignored or regarded as an error. The order is between.BEGIN and.END is arbitrary.
Netlist Statements
Each device in the circuit is described by an element statement. The element statement
contains the element name, the circuit nodes to which the element is connected and the values
of the element parameters. The first letter of an element name specifies the type of element to
be simulated. For example, a resistor name must begin with the letter, R, and can contain one
or more characters. This means that R1, RSE, ROUT, and R3AC2ZY are all valid resistor names.
Some elements, such as diodes and transistors, must always refer to a model. A set of
elements can refer to the same model. For some elements, such as resistors and capacitors,
model referencing is optional. Each element type has its own set of parameters. For example,
a resistor statement can specify a resistance value after the optional model name. The bipolar
transistor statement (Q) can specify an area parameter. All parameters have corresponding
default values. Independent voltage and current sources have different specifications for
transient, DC, and AC phases of simulation. Transient specifications use the keywords: EXP,
PULSE, GAUSS, SFFM, SIN, and TABLE. AC parameters start with the keyword, AC.
Elements to be simulated numerically are defined as “A” devices (ATLAS devices). At least
one ATLAS device in a circuit is mandatory.
MIXEDMODE supports the use of the following circuit elements:
• Numerically simulated ATLAS devices (“A” devices)
• User-defined two-terminal elements (“B” devices)
• Capacitors (“C” devices)
• Diodes (“D” devices)
• Voltage controlled voltage source (“E” devices)
• Current controlled current source (“F” devices)
• Voltage controlled current source (“G” devices)

625 ATLAS User’s Manual

Using MixedMode MixedMode: Mixed Circuit and Device Simulator

• Current controlled voltage source (“H” devices)

• Independent current sources (“I” devices, may be time dependent)
• JFETs (“J” devices)
• Coupled (mutual) inductors (“K” devices)
• Inductors (“L” devices)
• MOSFETs (“M” devices)
• Optical sources (“O” devices)
• Bipolar junction transistors (“Q” devices)
• Resistors (“R” devices, may be time dependent)
• Lossless transmission lines (“T” devices)
• Independent voltage sources (“V” devices, may be time dependent)
• MESFETs (“Z” devices)
The physical models for linear elements (resistors, capacitors, sources, and so on) are
described Section 12.4.1 “Circuit Element Statements”. The models for diodes: BJTs, JFETs,
MESFETs and MOSFETs are described in Section 12.4.2 “Control and Analysis Statements”.
You can find more extensive documentation, however, in the SMARTSPICE/UTMOST
MODELING MANUALS Volumes 1, 2, and 3.
A node is a point in the circuit where two or more elements are connected. A node can be
described either in terms of a node name or as a node number. The node names and numbers
are arbitrary with the exception of the ground node. The ground node is set either by
specifying “0” as the node number or using the name GND. All voltages at the nodes are
calculated with respect to the voltage at the ground node.
Example
The netlist for the circuit shown in Figure 12-1 is represented by the following MIXEDMODE
input deck fragment.
# independent voltage source, 0.1V, connected to node 0 (GND) and
1:
V0 1 0 0.1
# 1kOhm resistor, connected to node 1 and 2
R1 1 2 1K
# ATLAS device, connected to node 2 (anode) and 0 (cathode),
# current scaled by 5e7, mesh from file dio.str
ADIO 2=anode 0=cathode WIDTH=5e7 INFILE=dio.str

626 ATLAS User’s Manual

Using MixedMode MixedMode: Mixed Circuit and Device Simulator

Figure 12-1: Schematic of Primitive Example Circuit

Control Statements
Control statements are used to specify the analysis to be performed in MIXEDMODE. These
take the place of the SOLVE statements in a regular ATLAS input file. At least one of these
statements must appear in each MIXEDMODE input file. These statements are:
• .DC: steady state analysis including loops
• .TRAN: transient analysis
• .AC: small signal AC analysis
• .NET: small signal parameter extraction (e.g., s-parameters)

If you wish to perform a DC analysis, you should have an associated DC source to relate the
DC analysis to. Likewise, for an AC analysis you should have an associated AC source that
you relate the AC analysis to. For a transient analysis, you should have an associated source
with transient properties that you relate the transient analysis to. Failing to correctly run an
analysis on the same sort of source could yield simulation problems, such an running an AC
analysis on an independent voltage source where no AC magnitude for the source has been
specified.
For more information about these statements, see Section 12.4.2 “Control and Analysis
Statements”.

627 ATLAS User’s Manual

Using MixedMode MixedMode: Mixed Circuit and Device Simulator

Special statements
Other statements beginning with a dot “.” specify special parameters for the circuit
simulation. These include numerical options, file input and output, and device parameter
output. These statements are listed below.
• compact device models (.MODEL)
• the output files (.LOG,.SAVE)
• initial conditions settings (.NODESET,.IC)
• initial conditions from a file (.LOAD)
• numerics (.NUMERIC,.OPTIONS)
• device parameter output (.PRINT)
• miscellaneous (.OPTIONS)
Full descriptions of each statement and associated parameters are found in Section 12.4.2
“Control and Analysis Statements”.

12.2.3 Device Simulation Syntax

The second part of a MIXEDMODE command file (after.END) is used to define physical
models, material parameters, and numerical methods for ATLAS devices referenced in the
“A”-element statements. The following statements may appear in this part of the command
file: BEAM, CONTACT, DEFECT, IMPACT, INTERFACE, INTTRAP, MATERIAL, MOBILITY,
METHOD, MODELS, OUTPUT, PROBE, TRAP, and THERMCONTACT.
You always need to include an indicator to the circuit element name in each device simulation
statement even if there is only one A-device in the circuit. All statements specifying the
device properties and models are just supplemented by the DEVICE=name parameter, where
name is the circuit element in the netlist, and name will always begin with the letter “A”. This
makes it possible to define different material properties and model settings for different
devices within the circuit.
We recommended that you specify the REGION parameter referring to only one region in
IMPACT, MATERIAL, and MODELS statements. If the device consists of more than one region,
several statements with the same device parameters and different region parameters are
recommended.
For example to specify the bipolar set of models to a device the syntax used might be:
MODEL DEVICE=AGTO REGION=2 BIPOLAR PRINT

628 ATLAS User’s Manual

Using MixedMode MixedMode: Mixed Circuit and Device Simulator

12.2.4 Recommendations
Input Parsing
In regular ATLAS (non-MIXEDMODE) simulations, the input is interpreted line by line and
each statement is executed immediately. This is very useful and nicely supported by
DECKBUILD for the interactive development of the input. Circuit simulations, however,
require the complete input before any simulation can be performed. Consequently, the
following occur:
• The complete input is read and parsed before any simulation is initiated.
• An explicit termination of a simulation is required (quit).
• All post processing (extraction and plotting) has to be done after re-initializing ATLAS
again.
No simulation is started until either a QUIT statement or a GO statement is seen in the input
file. Post-processing can be done by restarting ATLAS.
Scale and Suffixes
In the MIXEDMODE part of the input, numerical values of parameters are represented in
standard floating-point notation. The scale suffix may be followed by a unit suffix (e.g., A for
Ampere, V for Volt, and so on). Using a unit suffix can increase the clarity of a command file.
The unit suffix is ignored by the program. The scale suffixes are shown in Table 12-1.

Table 12-1 Scale Suffixes

Factor Name Suffix
10-15 femto- F
10-12 pico- P
10-9 nano- N
10-6 micro- U
10-3 milli- M
103 kilo- K
106 mega- MG
109 giga- G
1012 tera- T

629 ATLAS User’s Manual

Using MixedMode MixedMode: Mixed Circuit and Device Simulator

Numerics
MIXEDMODE solves circuit and device equations simultaneously using fully coupled
algorithms. This provides better convergence and requires less CPU time than alternative
approaches. The number of circuit variables is often small in comparison with the number of
device variables. In this case, the CPU time required for simulation performed using
MIXEDMODE does not increase drastically compared to the sum of the simulation times
required for the individual numerical physically based devices. MIXEDMODE uses the Newton
algorithm for each bias point during steady-state analysis and for each time step during
transient analysis. Different variants of the Newton algorithm are used depending on the
circumstances [179, 180]. The full Newton method [.OPTIONS FULLN] and a modified two-
level Newton method [.OPTIONS M2LN] are available for steady-state simulation. The full
Newton method provides rapid convergence when a good initial guess is available. The
modified two-level Newton algorithm is less sensitive to the initial guess. For transient
simulation, a good initial guess always exists. The full Newton method therefore works very
well. Therefore, it is always used for transient simulation.
When using MIXEDMODE3D, it is recommended that you specify the DIRECT or GMRES solver
in the ATLAS part of the MIXEDMODE input deck on the METHOD statement. Also, use the
NOPROJ parameter in the.OPTIONS statement in the MIXEDMODE of the input deck.
Multi-Device Structure Representation
If more than one ATLAS device is defined in a MIXEDMODE simulation, the structures are
merged together internally. The output solution file is a single file which contains both
structures. The first structure referenced will be on top, all other structures will be attached
below.
Example
A diode and a bipolar transistor are specified as numerical devices with the following element
statements:
ABJT 1=BASE 2=EMITTER 4=COLLECTOR WIDTH=1E4 INFILE=bjt.str
ADIO 3=ANODE 4=CATHODE WIDTH=1.5E5 INFILE=dio.str
After outputting the solution with:
.SAVE MASTER=mas
The solution file for the first DC-point, mas_dc_1, contains both structures with the second
ATLAS device (diode) shifted downwards (see Figure 12-2).
This coordinate shift has to be accounted for eventually when extracting position dependent
solution quantities or when defining spatially dependent properties with the C-INTERPRETER
(See Appendix A “C-Interpreter Functions” for more information).

630 ATLAS User’s Manual

Using MixedMode MixedMode: Mixed Circuit and Device Simulator

Figure 12-2: Display of a MixedMode solution with two Numerical Devices

631 ATLAS User’s Manual

Using MixedMode MixedMode: Mixed Circuit and Device Simulator

Extraction of Results
By default, EXTRACT reads its data from the currently opened log-file when executed along
with ATLAS. Since the extraction of MIXEDMODE log-files require a re-initialization of
ATLAS (see “Input Parsing” on page 629), EXTRACT has to be initialized explicitly with the
correct name of the MIXEDMODE log file. To extract voltages at specific nodes, use the syntax
vcct.node.”circuit node”. To extract circuit elements, use icct.node.”circuit
element”.

Example
Specify the log file:
.LOG OUTFILE=hallo
Subsequent extraction from the transient log-file is done with:
go atlas
extract init inf=”hallo_tr.log”
extract name=”t0” x.val from curve(time,icct.node.”Adio_anode”) \
where y.val=0
It extracts the time, t0, when the transient of the current from the anode electrode of the
adio device in the circuit crosses zero. For more details for the EXTRACT syntax, see the
DECKBUILD USER ’S MANUAL.
Using MixedMode inside the VWF Automation Tools
Like all other Silvaco products, MIXEDMODE is fully integrated into the VWF framework and
can be used for automated experiments. There are, however, some factors to take into
account.
One factor is that the auto-interface feature doesn’t work with MIXEDMODE. All structures
have to be explicitly saved in unique files previous to the MIXEDMODE runs and referred to in
the A- element statements. Another factor is that splits within MIXEDMODE runs are
impossible. To overcome this problem, use the SET statement to define a variable in a process
simulator or in the “dummy” internal run. This variable is used to parameterize the input file.
Example
Capacitance as an independent split variable in a VWF experiment.
go internal
# define the independent split variable in a re-entrant simulator:
set cap=5e-9

go atlas

.BEGIN
# use the variable as parameter in MixedMode:
C1 2 3 $cap

632 ATLAS User’s Manual

Using MixedMode MixedMode: Mixed Circuit and Device Simulator

Another factor is that the automation tools only store files opened by the normal ATLAS LOG
statement in the VWF database but ignore those defined by.LOG. To overcome this, re-
initialize ATLAS and open the relevant log file of the previous MIXEDMODE run with.log as
the append option (so that the file is not reset).
Example
MIXEDMODE log-file definition.
.LOG OUTFILE=hallo
Re-opening the second DC-log file and the transient log file to get them stored in the VWF
database.
go atlas

log outfile=hallo_dc_2.log append

log outfile=hallo_tr.log append
Initial Settings
Initial convergence is critically dependent on the initial settings of the node voltages
(i.e,.IC and.NODESET). There should not be any problem starting from the zero bias case.
Just like starting from a preceding MIXEDMODE solution is simple, since the complete
solution of the circuit and the ATLAS devices is directly available (i.e,.LOAD and.SAVE). But
when loading solutions for the numerical devices from ATLAS, using .OPTIONS
LOADSOLUTIONS, sometimes precise matching of the initial circuit condition is required. In
this case, it is practical to extract the relevant properties in the preceding ATLAS run and use
them to parameterize the MIXEDMODE input.
In the following example, the voltages and current of an ATLAS solution is extracted, and the
results are used for the initial definition of the circuit.
The end of the first part is the stand-alone ATLAS simulation:
# extract the final voltage drop on the anode:
extract name="Von" max(vint."anode")
# extract the gate current:
extract name="I_gate" y.val from curve(vint."anode",i."gate") \
where x.val = $"Von"
extract name="V_gate" y.val from curve(vint."anode",vint."gate") \
where x.val = $"Von"

# now the MIXEDMODE part

go atlas
.BEGIN
# define the gate current source, use extracted value as parameter
I1 0 7 $"I_gate"
#

633 ATLAS User’s Manual

Using MixedMode MixedMode: Mixed Circuit and Device Simulator

# use extracted gate bias and other expressions to calculate

# the node settings:
set Rgl = 10.5
set v7= $V_gate + $I_gate * $Rgl
.NODESET V(1)=2000 V(2)=$"Von" V(3)=$"V_gate" V(4)=$"V_gate" V(5)=-
25 \
V(6)=-15 V(7)=$"v7"

634 ATLAS User’s Manual

A Sample Command File MixedMode: Mixed Circuit and Device Simulator

12.3 A Sample Command File

A sample MIXEDMODE command file is shown below. This file is used to simulate the reverse
recovery of a power diode. Several MIXEDMODE examples are provided with the product
which can be accessed using DECKBUILD.
1. go atlas
2. .BEGIN
3. V1 1 0 1000.
4. R1 1 2 1m
5. L1 2 3 2nH
6. R2 4 0 1MG EXP 1MG 1E-3 0. 20NS 10 200
7. IL 0 4 300
8. ADIODE 3=cathode 4=anode WIDTH=5.E7 INFILE=pd.str
9. .NUMERIC LTE=0.3 TOLTR=1.E-5 VCHANGE=10.
10..OPTIONS PRINT RELPOT WRITE=10
11.$
12..LOAD INFILE=pdsave
13..LOG OUTFILE=pd
14..SAVE MASTER=pd
15.$
16..TRAN 0.1NS 2US
17.$
18..END
19.$
20.MODELS DEVICE=ADIODE REG=1 CONMOB FLDMOB CONSRH AUGER BGN
21.MATERIAL DEVICE=ADIODE REG=1 TAUN0=5E-6 TAUP=2E-6
22.IMPACT DEVICE=ADIODE REG=1 SELB
23.$
24.METHOD CLIM.DD=1.E8 DVMAX=1.E6
25.$
26.go atlas
27.tonyplot pd_tr.log
Description
Line 1: All ATLAS input files should begin with go atlas
Line 2: The.BEGIN and.END statements indicate the beginning and end of the circuit
simulation syntax. These commands are similar to those used in SPICE.
Lines 3-7: Circuit components, topology, and analysis are defined within. Generally, the
circuit component definition consists of three parts: the type of component, the lead or
terminal mode assignments, and the component value or model name. For example, if the first
component definition in this simulation is a DC voltage source, then V1 defines the
component as voltage source number one, 1 and 0 are the two circuit modes for this
component, and 1000 indicates that the voltage source value is 1000 volts. The remaining
circuit components are resistors (R1, R2) inductor (L1) and independent current source (IL).
The reverse recovery of the diode is simulated by dropping the value of output resistor R2
over a small increment of time. The R2 statement contains additional syntax to perform this

635 ATLAS User’s Manual

A Sample Command File MixedMode: Mixed Circuit and Device Simulator

task. Here, the resistor is treated as a source whose resistance decreases exponentially from 1
mOhm to 1 mOhm over the specified time step. This action essentially shorts out the parallel
current source IL, which is also connected to the base of the diode.
Line 8: The ADIODE statement specifies a device to be analyzed by ATLAS. The A part of the
ADIODE command specifies that this is a device statement. The DIODE portion simply defines
the device name. The option INFILE= indicates which device structure file is to be used.
Lines 9-10: These set numerical options for the circuit simulation. WRITE=10 specifies that
every tenth timestep will be saved into the solution file specified on the.SAVE statement.
Line 12: Specifies a file generated by a previous MIXEDMODE simulation to be used as an
initial guess to the voltage.
Line 13-14: Specifies the output log and solution filenames. These names are root names and
extensions will be added automatically by the program.
Line 16: Indicates the type of analysis required. In this case, it is a transient simulation lasting
2 microseconds with an initial timestep of 0.1 nanoseconds.
Line 18: Indicates the end of the circuit description. All following statements will be related
to the ATLAS device.
Lines 20-22: To completely specify the simulation, the physical models used by ATLAS must
be identified. Note that DEVICE=ADIODE must be specified for each line. The MODEL
statement is used to turn on the appropriate transport models. This set includes:
• conmob: the concentration dependent mobility mode,
• fldmob: the lateral electric field-dependent mobility model,
• consrh: Shockley-Read-Hall recombination using concentration dependent lifetimes,
• auger: recombination accounting for high level injection effects,
• bgn: band gap narrowing.
The MATERIAL statement is used to override default material parameters. In this case, the
carrier recombination fixed lifetimes are set. Finally, the Selberherr impact ionization model
is enabled using the IMPACT statement with the SELB option.
Line 24: The METHOD statement specifies numerical options for the device simulation. The
METHOD statement must come after all other device simulation statements.
Line 26: The command GO ATLAS or a QUIT statement is needed to initiate simulation. Since
a plot of the final log file is desired, the GO ATLAS option is used to restart ATLAS after the
end of the MIXEDMODE simulation.
Line 27: The TONYPLOT command is used to plot the resulting log file.

636 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

12.4 MixedMode Syntax

This section is split into two parts. The first part describes circuit element statements, which
describe the netlist. The second part describe the control and analysis statements.

12.4.1 Circuit Element Statements

A – ATLAS device to be simulated using device simulation
Syntax
Axxx n1=name1 n2=name2 [n3=name3 ...] infile=filename [width=val]

Description
This statement defines a device to be represented by a numerical ATLAS model. The device
description with all necessary information (geometry, mesh, doping, models, electrode
names, and so on) must be available in a standard structure file prior to starting a
MIXEDMODE simulation.

Axxx Name of the element. It must begin with A.

n1 Circuit node to which the ATLAS device electrode with the name, name1, is
connected. The ATLAS device must have at least two electrodes. The maximum
number of electrodes allowed in ATLAS is 50. This means that up to 25 ATLAS
devices can be specified. For example, 25 devices with 2 electrodes each, or 10 devices
with 5 electrodes each can be specified. The ATLAS device models should be used
sparingly, because it can be very time consuming. Circuit models should be used for
less important circuit components to conserve CPU time. If an ATLAS device electrode
has been defined as FLOATING on the CONTACT statement, then this electrode should be
connected to a circuit node with a negative node number. MIXEDMODE treats all circuit
nodes with negative node numbers as dummy connection nodes and will not apply a
voltage to these nodes.
infile Name of standard structure format file with device geometry, mesh, doping,
electrodes names, and so on. The number of electrodes and their names should match
those mentioned in this statement. Optionally, this file can contain a solution, which
MIXEDMODE will use as an initial guess (see the .OPTIONS statement for more details).
width Device width. This is an optional parameter (default=1). All currents through ATLAS
device terminals calculated using the 2D ATLAS model will be multiplied by this
parameter to account for the third dimension of the device. width can still be used as a
multiplier to the ATLAS current if a 3D ATLAS structure is used in MIXEDMODE3D.

Example
ABJT1 3=EMITTER 4=BASE 6=COLLECTOR INFILE=BJT1.STR WIDTH=10

Note: Optional parameters for a statement are shown with square brackets (e.g., [n3=name3]).

637 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

B – User-Defined Two-terminal Elements

Syntax
Bxxx n+ n- INFILE=file_name FUNCTION=function_name

Description

Bxxx User-defined two terminal element name. It must begin with B.

n+, n- Positive and negative terminal nodes.
INFILE Name of the text file (file_name) that contains C source code for a
user-defined function that describes element behavior. This file can contain more
than one function description.
function_name Name of the function (“function_name”) from the file.

Example
B1 2 3 infile=ud.c function=rc
MIXEDMODE users who are familiar with C-INTERPRETER can define their own two-terminal
elements using the B statement and a function written in C that defines the behavior of the
element.
User-Defined Model
A user-defined model is specified by defining the following:
• The dependencies of the device terminal current on the terminal voltages
• The derivatives of the terminal current with respect to the terminal voltages
• The device current (I) is described by Equation 10-1.

I = F1  U t  + F2  U t    -------
dU
10-1
 dt 

where:
U is the device voltage, t is time, and F1 and F2 are functions that determine the behavior of
the device. The first term on the right hand side describes the “DC” current and the second
term describes “capacitive” current.
The following four functions are user-specified:
• F1(U,t): the DC current.
• F2(U,t): the capacitance.
• dF1(U,t)/dU: the DC differential conductance.
• Q: the charge associated with F2(U,t).

638 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

To define the element, prepare a text file that contains an appropriate function written in C. A
template for this user-defined function is shown below:
int udef(double v, double temp, double ktq, double time, double
*curr,
double *didv, double *cap, double *charge)
{
/* user-supplied code here */
return(0);
}

Input Parameters
Four input parameters are supplied to the function and can be used in the user-defined code.
The input parameters are:
• v: the voltage across the element (V)
• temp: the temperature (K)
• ktq: the thermal voltage kT/q (V)
• time: transient time (sec); a value of 0 is supplied during DC calculations.

Output Parameters
The four output parameters that must be returned by the function are:
• curr: the value of F1 (Amps)
• didv: the value of dF1(v, time)/dU (A/V)
• cap: the value of F2(v, time)
• charge: the value of the charge (Q)

Example
Consider an element that consists of a resistor R and a capacitor C connected in parallel. The
equation for the total current through this combination is:

I  U t  = ---- + C   -------
U dU
10-2
R  dt 

The quantities that must be user-defined are:

F1  U t  = U
---- 10-3
R

F2  U t  = C 10-4

dF  U t - = --1-
-------------------- 10-5
dU R

Q = CU 10-6

639 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

When R=2k and C=100pF, a user-defined function could have the following form:
intrc(double v, double temp, double ktq, double time, double *curr,
double *didv, double *cap, double *charge)
{
*curr = v/2000.0;
*didv = 1.0/2000.0
*cap = 1.0e-10;
*charge=1.0e-10*v;
return(0); /* 0 - ok */
}
C – Capacitor
Syntax
Cxxx n+ n- value

Description

vcontrolname Name of the voltage source through which the controlling current flows. The
direction of positive controlling current flow is from the positive node,
through the source, to the negative node of vcontrolname.
gain Current gain.

The linear current-controlled current source is characterized by Equation 10-8.

v  n + ,n –  = gain i  vcontro ln ame  10-8
Example
F12 4 5 VIN 0.1

642 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

G – Linear voltage controlled current source

Syntax
Gxxx n+ n- nc+ nc- transconductance

Description

Gxxx Name of the linear voltage controlled current source. It must begin with G.
n+, n- Positive and negative terminal nodes. A positive current flows from the
node, n+, through the source to the node, n-.
nc+, nc- Positive and negative controlling node numbers.
transconductance Transconductance (in 1/Ohms).

The linear voltage controlled current source is characterized by Equation 10-9.

i  n + ,n –  = transconduc tan ce v  nc + ,nc –  10-9
Example
G2 4 5 6 7 5.5
H – Linear current controlled voltage source
Syntax
Hxxx n+ n- vcontrolname transresistance

Description

Hxxx Name of the linear current controlled voltage source. Must begin from H.
n+, n- Positive and negative terminal nodes. A positive current flows from the node,
n+, through the source to the node, n-.

vcontrolname Name of voltage source through which the controlling current flows. The
direction of positive controlling current flow is from the positive node,
through the source, to the negative node of vcontrolname.
transresistance Transresistance (in Ohms).

The linear current controlled voltage source is characterized by Equation 10-10.

v  n + ,n –  = transresis tan ce i  vcontro ln ame  10-10
Example
H12 4 5 V1 0.1K

643 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

I – Independent current source

Syntax
Ixxx n+ n- value [AC acmag] [transient_parameters]

Description

Ixxx Name of the independent current source. It must begin with I.

n+, n- Positive and negative terminal nodes.
value DC value of the source (ampers).
AC Keyword for the AC source value.
acmag AC magnitude.
transient_parameters The transient parameters are described in Section 12.4.3 “Transient
Parameters”.

Example
I1 2 8 0. PULSE 0 200 0 20ns 20ns 100ns 10 100
I2 1 5 1u AC 2u
J – Junction Field-Effect Transistor (JFET)
Syntax
Jxxx nd ng ns [nb] mname [area] [M=val] [L=val] [W=val] [OFF]
[IC=vds,vgs]
[TEMP=val] [DTEMP=val]

Description

Jxxx Name of the JFET element. It must begin with J.

nd, ng, ns, nb Drain, gate, source and bulk terminal nodes. The bulk node doesn’t need to be
specified. If the bulk node is not specified, then the bulk is connected to the
source node
mname Model name. It must refer to a JFET model.
area Area factor. The default is 1.0.
M Multiplier used to describe multiple parallel JFETs.
L Length of the gate in meters.
W Width of the gate in meters.
OFF Sets ON/OFF startup condition for DC analysis. Default is ON.

644 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

IC Initial condition specification for vds and vgs.

TEMP Device operating temperature (oC).
DTEMP Difference (in oC) between the device operating temperature and the circuit
temperature. Default value is 0.

Example
J44 1 4 6 jmodel

Note: See SMARTSPICE/UTMOST MODELING MANUAL, VOLUME 2 for a complete description of the
JFET models.

K – Coupling between two inductors

Syntax
Kxxx Lyyy Lzzz kval

Description
This is not a real circuit element. This statement defines only the coupling between two
inductors.

Kxxx Name. This parameter is not important and is used only to distinguish the statement.
It must begin with K.
Lyyy First inductor element name. It must begin with an L and match one of the inductor
names from the circuit.
Lzzz Second inductor element name. It must begin with an L and match one of the
inductor names from the circuit.
kval Coefficient of mutual coupling, which must be in the range 0 < kval < 1.

The mutual inductance M will be determined from Equation 10-11.

M = kval L1 L2 10-11
Example
K1 L22 LLOAD 0.99

645 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

L – Inductor
Syntax
Lxxx n+ n- value

Description

Lxxx Name of the inductor. It must begin with L.

n+, n- Positive and negative terminal nodes.
value Inductance in henries.

Example
L2 2 3 2.5nH
M – MOSFET
Syntax
Mxxx nd ng ns [nb] mname [L=val] [W=val] [AD=val] [AS=val]
[PD=val] [PS=val] [NRD=val] [NRS=val] [OFF] [IC=vds,vgs,vbs]
[M=val]
[TEMP=val] [DTEMP=val] [GEO=val] [DELVTO=val]

Description

Mxxx MOSFET element name. It must begin with M.

nd, ng, ns, nb Drain, gate, source, and bulk terminal nodes. The bulk terminal node name is
optional. If it is unspecified, ground is used.
mname Model name. It must refer to a MOSFET model.
L=val Channel length in meters.
W=val Channel width in meters.
AD Drain diffusion junction area (meters2). This default is 0.
AS Source diffusion junction area (meters2). This default is 0.
PD Drain diffusion junction perimeter (meters). This default is 0.
PD Source diffusion junction perimeter (meters). This default is 0.
NRD The Number of squares of drain diffusion for resistance calculations. The
default is 0.
NRS The Number of squares of source diffusion for resistance calculations. The
default is 0.
OFF Sets ON/OFF startup condition for DC analysis. Default is ON.

646 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

IC Initial voltage condition specification for vds, vgs, and vbs.

M Multiplier used to describe multiple parallel MOSFETs. The default is 1.
TEMP Device operating temperature (oC).
DTEMP Difference (in oC) between the device operating temperature and the circuit
temperature. Default value is 0.
DELVTO Threshold-voltage shift. When specified on the device line, the value overrides
the value of the model parameter, DELVTO. If not specified, the value of the
model parameter is used.

Example
M1 2 4 8 9 mod1
Mout2 19 20 21 0 nmos L=5u W=2u TEMP=50
M22 3 5 7 8 mosmod1 L=10u W=5u AD=150p AS=150p PD=50u
PS=50u NRD=10 NRS=20

Note: See SMARTSPICE/UTMOST MODELING MANUAL, VOLUME 1 for a complete description of the
MOSFET models.

O – Optical source
Syntax
Oxxx beam value [transient_parameters]

Description

Oxxx Name of an independent optical source. It must begin with O.

beam Beam number. The beam with this number should be described in
the ATLAS section of the command file. See Chapter 10
“Luminous: Optoelectronic Simulator” for a complete description of
optoelectronic simulation.
value DC optical intensity value (W/cm2).
transient_parameters The transient parameters are described in Section 12.4.3 “Transient
Parameters”.

Note: The treatment of optical sources is fully similar to the treatment of independent voltage/current sources. In
other words you can use,.DC statements to simulated DC light responses of the circuit and transient
parameters in order to describe the transient behavior of the optical sources.

Example
O1 1 0.001 pulse 0.001 0.002 0 2ns 2ns 100ns 10 100

647 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

Q – Bipolar junction transistor

Syntax
Qxxx nc nb ne [ns] mname [area] [OFF] [IC=vbe,vce] [M=val]
[TEMP=val] [DTEMP=val]
or
Qxxx nc nb ne [ns] mname [area=val] [areab=val] [areac=val] [OFF]
[IC=vbe,vce] [M=val] [TEMP=val] [DTEMP=val]

Description

Qxxx Name of a bipolar junction transistor. It must begin with Q.

nc, nb, ne, ns Collector, base, emitter, and substrate nodes. The substrate terminal node name
is optional. If it is unspecified, ground is used.
mname Model name. It must refer to a BJT model.
area Emitter area factor. The default value is 1.0.
areab Base area factor. The default is area.
areac Collector area factor. The default is area.
OFF Sets ON/OFF startup condition for DC analysis. Default is ON.
IC Initial voltage condition specification for vbe, vce.
M Multiplier used to describe multiple parallel BJTs. The default is 1.
TEMP Device operating temperature (oC).
DTEMP Difference (in oC) between the device operating temperature and the circuit
temperature. Default value is 0.

Example
Q1 2 3 9 npnmod 1.5 IC=0.6,5.0
Q9 10 11 12 20 mod22 OFF TEMP=50

Note: See SMARTSPICE MODELING MANUAL, VOLUME 2 for a complete description of the BJT models.

648 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

R – Resistor
Syntax
Rxxx n+ n- value [transient_parameters]

Description

Rxxx Name of the resistor element. It must begin with “R”.

n+, n- Positive and negative terminal nodes.
value Resistance in ohms.
transient_parameters The transient parameters are described in Section 12.4.3 “Transient
Parameters”.

Note: Unlike the traditional SPICE program, transient parameters are acceptable for resistor elements. This
allows simulation of different kinds of time-dependent resistors and switches in a simple way.

Example
R12 4 5 100k
S-Voltage-Controlled Switch
Syntax
Sxxx n+ n- nc+ nc- <mname> <on|or> <roff=val> <ron=val> <vt=val>
<vh=val>

n+, n- Positive and negative terminal nodes.
value DC value of the source in units of volts.
AC Keyword for the AC source value.
acmag AC magnitude.
transient_parameters The transient parameters are described in Section 12.4.3 “Transient
Parameters”.

Example
VCC 5 0 10.5
VIN 2 4 5.5 AC 1

650 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

X - Subcircuit Call
Syntax
Xxxx n1 <n2 n3 ...> subcktname <<PARAMS:> parname=val ...>

Description

Xxxx Subcircuit names must begin with X.

n1, n2, ... Node names for external reference.
subcktname Subcircuit definition.
PARAMS The optional parameter preceding the list of parameters and their values.
parname The parameter name whose value is set to val, which will be local to the
subcircuit. valis either a numerical value, an expression enclosed in single
quotes containing previously defined parameters, or a string, which defines a
model’s name, that can be transmitted in subcircuit. If the same parameter is
assigned a value on more than one statement in the input deck (.PARAM, .SUBCKT,
and X statements), then will be one of the following:
• a value assigned in a global .PARAM statement (outside
subcircuits) is used if it exists.
• a value assigned in a corresponding X statement is used if it
exists.
• a value assigned in a corresponding .SUBCKT statement is used if
it exists.
• a value assigned in a local .PARAM statement (inside the
subcircuit) is used if it exists.

This statement creates instance Xxxx of subcircuit subcktname. There must be the same
number of nodes n1 ... as in the subcircuit definition (see .SUBCKT statement). These nodes
replace the formal node names in the .SUBCKT statement.
Subcircuits can be nested. The number of nesting levels is not limited but nesting must not be
recursive. When the circuit is loaded, all subcircuit device names are expanded to the form
devtype.subcktname.devname, where devtype is the first letter of devname. All local
node names in the subcircuit are expanded to the form subcktname.nodename.
For nested subcircuits, expanded names have multiple dot-separated qualifiers. For example,
if a circuit has a call to subcircuit subcktname1 and the subcktname1 definition contains a
call to subcircuit subcircuit2, then expanded names will have the following format:
devtype.subcktname1.subcktname2.devname
subcktname1.subcktname2.nodename
subcktname1.subcktname2.modelname

651 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

YVLG-User-defined Verilog-A Element

Syntax
YVLGxxx n1 n2 ... mname <user_params=param-values>
Description

YVLGxxx Name of the voltage-controlled switch element.

n1, n2, .. Terminal node names.
mname Verilog-A module name or VLG model name.
user_params Optional list of user-defined Verilog-A parameters.

The following restrictions apply during a YVLGXXX statement:

• A .verilog statement that specifies the file containing the model referenced by mname
must be present in the netlist.
• The number of terminal nodes must match the number of nodes declared in the Verilog-A
module.

652 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

Z – MESFET
Syntax
Zxxx nd ng ns [nb] mname [area] [M=val] [L=val] [W=val] [OFF]
[IC=vds,vgs]
[TEMP=val] [DTEMP=val]

Description

Jxxx Name of the MESFET element. It must begin with Z.

nd, ng, ns, nb Drain, gate, source and bulk terminal nodes. The bulk node doesn’t need to be
specified. If the bulk node it not specified, then the bulk is connected to the
source node.
mname Model name. It must refer to a MESFET model.
area Area factor. The default is 1.0.
M Multiplier used to describe multiple parallel MESFETs.
L Length of the gate in meters.
W Width of the gate in meters.
OFF Sets ON/OFF startup condition for DC analysis. Default is ON.
IC Initial condition specification for vds and vgs.
TEMP Device operating temperature (oC).
DTEMP Difference (in oC) between the device operating temperature and the circuit
temperature. Default value is 0.

Example
Z44 1 4 6 jmodel

Note: See SMARTSPICE MODELING MANUAL, VOLUME 2 for a complete description of the MESFET
models.

653 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

12.4.2 Control and Analysis Statements

.AC
.AC performs an AC linear small-signal analysis on the circuit. MIXEDMODE first creates a
linearized small-signal model at the operating point of the circuit and the computes the
frequency response over a user-specified range of frequencies.
Syntax
.AC DEC|OCT|LIN nump fstart fstop/ sweep source_name start stop
step

Description

DEC Sweep frequency by decades.

OCT Sweep frequency by octaves.
LIN Linear frequency sweep. This is default.
nump Total number of points per decade or per octave, or the total number of points of the linear
sweep.
fstart Starting frequency (Hz).
fstop Final frequency (Hz).

Several.AC statements can be specified in the same command file. In this case, they will be
executed sequentially. Before executing the first.AC statement, the program will execute
all.DC statements (if any), regardless of the order of the.AC and.DC statements in the
command file. At least one independent voltage or current source must have a AC
specification.

source_name Name of the independent voltage, current, or optical source to be swept.

start Starting value of the sweep.
step Increment value of the sweep.
stop Final value of the sweep.
sweep Performs a DC sweep at every frequency point.

Example
.AC DEC 3 1.e3 1.e12
.AC LIN 20 1.e5 2.e6

654 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

.BEGIN
.BEGIN indicates the start of the circuit part of a MIXEDMODE command file. The synonym
for this parameter is START.
.BEGIN 3D

Description

3D Specifies that MIXEDMODE3D will be used instead of MIXEDMODE.

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

.LOAD
.LOAD loads a solution file.

Syntax
.LOAD INFILE=filename

Description

INFILE Name of a file (filename) to be loaded as an initial guess for further simulation. This file
must have been saved during a previous run of MIXEDMODE using the .SAVE
statement.

Example
.LOAD INFILE=pdsave

Note: This statement is not used to load SSF format solution files from ATLAS (see .OPTIONS
LOADSOLUTIONS).

OUTPORT Output port description. It should be in one of the following formats.

• V(n+,n-): two nodes (positive (n+) and negative (n-)).
• Vxxxx: where Vxxxx is the name of an existing voltage source. The
positive terminal of the source becomes the positive output port node
and the negative terminal becomes the negative output node.
• Ixxxx: where Ixxxx is the name of an existing current source. The
positive terminal of the source becomes the positive output port node
and the negative terminal becomes the negative output node.

Note: If the nodes specified as the output port are the same nodes as an existing current or voltage source, then
the name of the source must be specified as outport.

DEC Sweep frequency by decades.

OCT Sweep frequency by octaves.
LIN Linear frequency sweep. This is default.

660 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

nump Total number of points per decade or per octave, or the total number of points of the
linear sweep.
fstart Starting frequency (Hz).
fstop Final frequency (Hz).

Additional optional parameters may also be specified on the .NET statement.

Z0 Matching Impedance (default = 50 Ohms).

INDIN Inductance through which the DC voltage source is connected to the input source
(only if INPORT is given as Vxxxx).
RSIN Series resistance of INDIN.
INDOUT Inductance through which the DC voltage source is connected to the output source
(only if OUTPORT is given as Vxxxx).
RSOUT Series resistance of INDOUT.
CIN Capacitance through which the S-parameter test circuit is connected to the input port.
COUT Capacitance through which the S-parameter test circuit is connected to the output port.
SWEEP Performs a DC sweep at every frequency point.
source_name Name of the independent voltage, current, or optical source to be swept.
start Starting value of the sweep.

stop Final value of the sweep.

step Increment value of the sweep.

Note: The S-parameters will be automatically saved to the log file. The Z, Y, H, ABCD, and gain small-signal
parameters can also be written to the log file. These are selected through the.OPTIONS statement. You
can also view the default values if PRINT is specified in the.OPTIONS statement

Examples
.NET V1 V2 DEC 10 1e6 1e10
.NET I1 V2 DEC 10 1e6 1e10 Z0=75 RSOUT=100
.NET V(1,0) V(2,3) DEC 10 1e6 1e10

661 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

.NODESET
.NODESET sets initial values for circuit node voltages.

Syntax
.NODESET [V(I)=VAL_I ...]

Description
This statement specifies the initial values for circuit node voltages. If a node voltage is not
specified, the program will try to find a solution using zero as an initial guess for this node.
This statement can significantly reduce the CPU time needed to calculate the initial condition.
Example
.NODESET V(1)=50 V(2)=49.4 V(3)=10 V(5)=-1.5
.NODESET V(in1)=0 V(2)=2 V(out1)=-1

662 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

.NUMERIC
.NUMERIC specifies special numeric parameters for the circuit analysis.

Syntax
.NUMERIC [parameters]

Parameter Type Default Units

DTMAX Real 1*10100 s

DTMIN Real 1*10-12 s

FILL.RATIO Real 5.5

IMAXDC Integer 25
IMAXTR Integer 15
LTE Real 0.1
TCOMP Real K
TOLDC Real 1*10-4
TOLTR Real 1*10-4
VCHANGE Real 5*107 V

VMAX Real 5*107 V

VMIN Real -5*107 V

Description

DTMAX Specifies the maximum time step.

DTMIN Minimum time step value for transient analysis.
FILL.RATIO Specifies the fill ratio used when calculating the conductivity matrix using the
iterative solver.
IMAXDC Maximum number of mixed circuit-device iterations to be performed during
steady-state analysis.
IMAXTR Maximum number of mixed circuit-device iterations to be performed during
transient analysis.
LTE Local truncation error for transient analysis.
TCOMP Temperature compliance value.

663 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

TOLDC Relative accuracy to be achieved during steady-state analysis for the

calculation of voltages in circuit nodes.
TOLTR Relative accuracy to be achieved during transient analysis for the calculation of
voltages in circuit nodes.
VCHANGE Maximum allowable change in circuit node voltages between two mixed
circuit-device iterations. This parameter can be useful for reaching steady-state
convergence with a bad initial guess.
VMAX Maximum value for circuit node voltages.
VMIN Minimum value for circuit node voltages.

Example
.NUMERIC LTE=0.05 TOLDC=1.*10-8 DTMIN=1ns
.OPTIONS
.OPTIONS specifies various circuit simulation options.

Syntax
.OPTIONS [parameters]

Parameter Type Default Units

ABCD.PARAM Logical False

CNODE Real 1*10-16 F

CYLINDR Logical False

DC.WRITE Real 1
DT.NOCHECK Logical False
DCGMIN Real 1*10-12 1/ohms

FULLN Logical True

GAINS Logical False
GMIN Real 1*10-12 1/ohms

H.PARAM Logical False

LOADFLOATING Logical False
LOADSOLUTIONS Logical False
M2LN Logical False
M2LN.TR Logical False
NOPROJ Logical False

664 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

Parameter Type Default Units

NOSHIFT Logical False

PRINT Logical False
RELPOT Logical False
RV Real 1*10-4 W

STR.XOFFSET Real 1 m
STR.XSHIFT Logical
STR.YOFFSET Real 1 m
STR.YSHIFT Logical
STR.ZOFFSET Real 1 m
STR.ZSHIFT Logical
TEMP Real 300 K
TNOM Real 300 K
WRITE Integer 1
Y.PARAM Logical False
Z.PARAM Logical False
ZIP.SOLVER Logical False

Description

ABCD.PARAM ABCD parameters will be written to the log file. This is used in conjunction with
the.NET statement.
CNODE A very small capacitance, which for algorithmic reasons automatically connected
from each circuit node to ground. This value can be set to 0.
CYLINDR Cylindrical coordinate system for all ATLAS devices.
DC.WRITE Specifies how often a structure file (as specified by the MASTER parameter on the
.SAVE statement) will be written during .DC solutions. If this parameter is set to
0, then no .DC structure files will be written.
DT.NOCHECK Specifies that the transient pulse and table definitions will not be taken into
account when calculating the time step.
DCGMIN Specifies the minimum conductance used in SPICE models during DC analysis.
FULLN Full Newton solution method is used during steady-state simulation.

665 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

GAIN Stability factor (K), unilateral power gain (GU), maximum unilateral transducer
power gain (GTUmax) and |H21|2 are written to the LOG file. This is used in
conjunction with the .NET statement.
GMIN Specifies the minimum conductance used in SPICE models during transient
analysis.
H.PARAM H-parameters will be written to the LOG file. This is used in conjunction with the
.NET statement.

LOADFLOATING Specifies that the charge bias value for floating electrodes will be loaded.
LOADSOLUTIONS Solutions, structures, doping distributions, and meshes, are to be loaded from
standard structure files. The solutions are used as the initial guess or initial
conditions for subsequent MIXEDMODE simulation. To use this feature, you must:
• Calculate a solution for each ATLAS device and save each
solution in a separate standard structure format file using SAVE or
SOLVE.... MASTER.
• For each ATLAS device, use the A statement in the MIXEDMODE
command file to specify the associated standard structure format
file
• Set node voltages to appropriate values with the.NODESET
statement
• Specify LOADSOLUTIONS in the.OPTIONS statement

Note: If using this feature, specify solutions for all ATLAS devices.

The.NODESET statement must always be used when LOADSOLUTIONS is used. The .NODESET
statement is used to make the initial circuit voltages match those device solutions that were
obtained. You may also need to specify the NOSHIFT parameter of the OPTIONS statement. By
default, MIXEDMODE shifts device terminal voltages with respect to the voltage on the first
terminal that is specified in the A statement. You must either prepare initial solutions with this
terminal grounded, or specify NOSHIFT in the OPTIONS statement.

M2LN Uses the modified two-level Newton solution method during steady-state
simulation. The full NEWTON method provides faster solution than the modified two-
level Newton method when a good initial guess is available. The modified two-level
method is more reliable when the initial guess is far from the solution. The default is
the full NEWTON method.
M2LN.TR Uses the modified two-level Newton solution method during transient simulations.
NOPROJ Disables the initial guess project method for the ATLAS iterations. MIXEDMODE
attempts to extrapolate the values of the ATLAS device variables (such as potential and
carrier concentration) for the each iteration. Specifying NOPROJ disables the
extrapolation and the previous values of potential and carrier concentration are used
instead.

666 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

NOSHIFT Disables the shift of voltages for ATLAS device models. MIXEDMODE normally shifts
the voltages on ATLAS device terminals to be referenced to the voltage on the first
terminal. From the physical point of view, the state of the p-n diode is the same for
voltages of 0V and 0.5V on the diode terminals with 1000V and 1000.5V, but the first
situation is better for numerical simulation.
PRINT Enables printing of circuit nodes voltages after the calculation for each bias point (DC
analysis) or time step (transient the analysis).
RELPOT Enables the use of relative convergence criteria for potential for ATLAS models. By
default, ALTAS models use absolute convergence criteria for potential. When bias
voltages are large (a common situation for power devices), then absolute
convergence criteria are not appropriate and this parameter should be specified.
RV Defines the ohmic resistance that MIXEDMODE associates with all voltage sources and
all inductances. This value should never be set to 0. The default value is small enough
to avoid errors due to the influence of the internal resistance. Usually, extremely small
values of this parameters can cause convergence problems. It is usually acceptable to
decrease this parameter to the range of 1*10-6-1*10-7. This parameter should not be
varied unless there is a compelling reason to do so.
STR.XOFFSET Specifies the X direction shift offset for multiple ATLAS devices.
STR.XSHIFT Specifies that multiple ATLAS device will be shifted in the X direction.
STR.YOFFSET Specifies the Y direction shift offset for multiple ATLAS devices.
STR.YSHIFT Specifies that multiple ATLAS device will be shifted in the Y direction.
STR.ZOFFSET Specifies the Z direction shift offset for multiple ATLAS devices.
STR.ZSHIFT Specifies that multiple ATLAS device will be shifted in the Z direction.
TEMP Device temperature to be use during the simulation.
TNOM Circuit temperature to be use during the simulation.
WRITE How often the solution is to be saved in standard structure files during the
simulation. For example, write=3 specifies that the solution will be saved at every
third timestep. Specifying this parameter can help avoid disk overflow.
Y.PARAM Y-parameters should be written to the log file. This is used in conjunction with the
.NET statement.

Z.PARAM Z-parameters should be written to the log file. This is used in conjunction with the
.NET statement.

ZIP.SOLVER Specifies that the ZIP libary version of BICGST iternative solver will be used to
calculate the device conduction matrix instead of the CGS or BICGST solver. If you
specify the DIRECT parameter in the METHOD statement, then the direct solver will be
used for both the device conduction matrix calculation and the ATLAS solution.

Example
.OPTIONS TNOM=293 FULLN

667 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

.PRINT
.PRINT specifies which device output parameters will be printed to the log files.

Syntax
.PRINT [antype] parmeter(device_name) [parmeter2(device_name)...]

Description

antype Type of analysis for which the outputs are desired. If antype is unspecified, the outputs
of all simulation types will be printed. antype must be one of the following keywords:
• AC: AC analysis outputs
• DC: DC analysis outputs
• NET: Network analysis outputs
• TRAN: Transient analysis outputs
• parmeter: Output variables or expressions to be printed.
• device_name: The device name.

Example
.PRINT ic(q1) ib(q1) is(q1)
.PRINT AC ic(q1) ib(q2)
.PRINT DC ic(q1) ib(q2) i(d1)
.PRINT TRAN cd(m1) cg(m1) cs(m1) cb(m1)
.PARAM
.PARAM specifies a parameter definition.

Syntax
.PARAM parname1=val1 [parname2=val2 ...]
The .PARAM statement defines one or more parameter labels by assigning the names
parname1, parname2 to constant values val1, val2. Parameter labels must begin with
alphabetic character(s) or the underscore character(s).
The .PARAM statement can be globally and locally (i.e., inside a subcircuit definition)
specified. Once defined, a global parameter label or an expression containing global
parameter labels can be used in the input deck when a numerical value is expected. A local
parameter label can be used only inside the subcircuit.

668 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

.SAVE
.SAVE saves simulation results into files for visualization or for future use as an initial guess.

Syntax
.SAVE OUTFILE=name [MASTER=mname][TSAVE=timepts]

Description

MASTER This will cause MIXEDMODE to output standard structure files for visualization in
TONYPLOT. These files, with the base name, mname, will be written after the
calculation of each bias point during DC simulation, and after of each time step during
transient simulation. For example, if you wish to save a structure file for each DC bias
solution, you would write:
.save master=my_structure.str
MIXEDMODE will then output structure files at each DC bias point. MIXEDMODE will also
automatically append a number to the end of the file name indicating which bias point the
file is for.

Example
.SAVE MASTER=my_filename.str

OUTFILE Specifies that after the simulation is finished the solution is to be written to a file
called my_filename.str. In other words, only one structure file will output after the
simulation has finished. Even if you have several .dc solutions, MIXEDMODE will
only output one structure file if for example:
.save outfile=filename.str
has been used. The ATLAS model solutions will be written to the file (name) and the
circuit solution will be written to the file, name.cir. These files can be used later for
loading solutions to be used as an initial guess (see .LOAD statement).

Example
.save outfile=my_filename.str

TSAVE Specifies the transient time points at which a MASTER file will be written. The time
points should be specified as a comma seperated list.

The program will automatically add the following suffixes to mname:

669 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

• During DC simulation: _dc_number, where number is the number of the DC point.

• During transient simulation: _tr_number, where number is the number of the time step.
Example
.SAVE OUTFILE=pdsave MASTER=pd
.SUBCKT
.SUBCKT specifies a subcircuit defintion.

Syntax
.SUBCKT subcktname <n1 n2 ...>
+ <OPTIONAL: optn1=defval1 <optn2=defval2 ...>>
+ <<PARAMS|PARAM:> parname1=val1 <parname2=val2 ...>>

Description

subcktname The subcircuit name. It is used by an X element statement to

reference the subcircuit.
n1 ... The optional list of external nodes. Ground nodes (zero) are not
allowed.
PARAMS|PARAM The parameter preceding the list of parameters and their values.
parname1, parname2, ... The name of a parameter with value set to val1, val2, ... that
will be local to the subcircuit. val is a numerical value. If the same
parameter is assigned, a value on more than one statement in the
input deck (.PARAM, .SUBCKT, and X statements) will be one of the
following:
• a value assigned in a global .PARAM statement
(outside subcircuits) is used if it exists.
• a value assigned in a corresponding X statement is
used if it exists.
• a value assigned in a corresponding .SUBCKT
statement is used if it exists.
• a value assigned in a local .PARAM statement
(inside the subcircuit) is used if it exists.

A subcircuit definition is initiated by a .SUBCKT statement. The group of element statements,

which immediately follow the .SUBCKT statement define the subcircuit. The last statement in
a subcircuit definition is the .ENDS statement. Control statements cannot appear within a
subcircuit definition; however, subcircuit definitions may contain anything else, including
other subcircuit definitions, device models, and subcircuit calls.
All internal nodes (if not included on the .SUBCKT statement) are local with the exception of
the global node 0.

670 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

.TRAN
.TRAN specifies that a transient analysis is to be performed.

Syntax
.TRAN tstep tstop DTCONST DTMAX DTMIN

Description

tstep Time interval in seconds.

tstop Final time value for which the simulation is to be performed.
DTCONST Specifies that a constant time step is to be used for the current .TRAN statement.
DTMAX Specifies the maximum time step for the current .TRAN statement.
DTMIN Specifies the minimum time step for the current .TRAN statement.

Transient analysis is performed only after the execution of all.DC statements. If no DC

statements are used, the transient starts after the calculation of the initial circuit state with the
values of the independent sources given in the descriptions of those sources.
Multiple.TRAN statements are supported. The TSTOP parameter is not reset between
each.TRAN statement.
Example
.TRAN 1ns 100ns
.VERILOG
.VERILOG specifies a Verilog-A file to be included.
Syntax
.VERILOG filename <module_name> <compile-options>
Description

filename The name of a VerilogA file.

module_name Optional Verilog-A module name.
compile_options Optional Verilog-A Compilation options.

Example
.verilog "resistor.va"

671 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

12.4.3 Transient Parameters

MIXEDMODE allows you to specify transient parameters for voltage sources (Vxxx), current
sources (Ixxx), and resistors (Rxxx). These parameters describe the time development
behavior of the source.
EXP
EXP is used to define an exponential waveform. The waveform is specified as follows:
EXP i1 i2 td1 tau1 td2 tau2
where:
• i1 is the initial value.
• i2 is the pulsed value.
• td1 is the rise delay time.
• td2 is the fall delay time.
• tau1 is the rise time constant.
• tau2 is the fall time constant.
The transient behavior is shown in Table 12-2.

Table 12-2 Transient behavior for EXP

Time Value

0 < t < td1 i1

td1 t < td2 i1+(i2-i1).(1-exp[-(t-td1)/tau1])
td2t i1+(i2-i1).(1-exp[-(t-td1)/tau1])
+(i1-i2).(1-exp[-(t-td2)/tau2])

672 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

GAUSS
GAUSS is used to define a Gaussian waveform. The waveform is specified as follows:
GAUSS i1 i2 td1 tau1 td2 tau2
where:
• i1 is the initial value.
• i2 is the pulsed value.
• td1 is the rise delay time.
• td2 is the fall delay time.
• tau1 is the rise time constant.
• tau2 is the fall time constant.
The transient behavior is shown in Table 12-3.

Table 12-3 The Transient behavior for GAUSS

Time Value

0 < t < td1 i1

td1t < td2 i1+(i2-i1).(1-exp[-((t-td1)/tau1)2]
td2t i1+(i2-i1).(1-exp[-((t-td1)/tau1)2]
+(i1-i2).(1-exp[-((t-td2)/tau2)2]

PULSE
PULSE is used to define a pulse waveform. The waveform is specified as follows:
PULSE i1 i2 td tr tf pw per
where:
• i1 is the initial value.
• i2 is the pulsed value.
• td is the delay time before the pulse is started.
• tr is the rise time of the pulse.
• tf is the fall time of the pulse.
• pw is the pulse length per period
• per is the period.

673 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

The transient behavior is described in Table 12-4. Intermediate points are found by linear
interpolation.

Table 12-4 Transient behavior for PULSE

Time Value

0 i1
td i1
td + tr i2
td + tr + pw i2
td + tr + pw + tf i1
td + per i1
td + per + tr i2 (second period)

PWL
PWL is used to define a piece-wise linear waveform. The waveform is specified as a list of
time and value point with the time values in ascending order. The waveform is specified as
follows:
PWL t1, v1 [t2,v2 ....][R=tval]
where:
• t1 is the first time value.
• v1 is the bias value at time=t1.
• t2 is the second time value (t2>t1).
• v2 is the bias value at time=t2.\
• R is the repeat time point. This specifies the time point from which the PWL waveform
should be repeated. The section of the waveform between tval and the end of the PWL
will be repeated until the transient analysis is completed.
PWLFILE
PWLFILE is used to define a file name containing a piece-wise linear waveform. The
waveform is specified as a list of time and value points with time values in ascending order.
The file is specified as follows:
PWLFILE filename [R=tval]
where:
• filename is the PWL file.
• R is the repeat time point. This specifies the time point from which the PWLFILE
waveform should be repeated. The section of the waveform between tval and the end of
the PWL will be repeated until the transient analysis is completed.

674 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

SFFM
SFFM is used to define a modulated sinusoidal waveform. The waveform is specified as
follows:
SFFM io ia fc mdi fs
where:
• io is the DC offset.
• ia is the amplitude.
• fc is the carrier frequency.
• mdi is the modulation index.
• fs is the signal frequency.
The transient behavior will be:
value(t)=io + ia · sin[· fc · t + mdi · sin(2· fs · t)]
SIN
SIN is used to define a sinusoidal waveform. The waveform is specified as follows:
SIN io ia freq td theta
where:
• io is the offset.
• ia is the amplitude.
• freq is the frequency.
• td is the delay.
• theta is the damping factor.
The transient behavior is shown in Table 12-5.

Table 12-5 Transient Behavior for SIN

Time Value

t < td value(t) = io
t  td value(t) = i0+ia.exp[-(t-td)/THETA]
.sin[2.freq.(t-td)]

675 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

TABLE
TABLE is used to define a waveform using a table of values. This parameter is used as follows:
TABLE infile=<table_file_name>[R=tval]
where table_file_name is an ASCII text file that contains the tabulated time-dependence
of a variable in the following format:
t1 v1
t2 v2
t3 v3
...
tN vN
end
Each line contains two numbers. The first number is the time in seconds. The second number
is the time-dependent variable) voltage in volts, the current in amps, or the resistance in
ohms). Up to 1000 lines can be used. Input is terminated by the word end.
If during the simulation the transient time becomes larger than the last value in the table, then
the last value will be used for the remainder of the simulation.
R is the repeat time point. This specifies the time point from which the TABLE waveform
should be repeated. The section of the waveform between tval and the end of the PWL will
be repeated until the transient analysis is completed.

676 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

12.4.4 Expressions
Expressions can be used at input deck statements to specify (by means of calculations) the
value of a parameter.
Functions
MIXEDMODE supports the functions listed in Table 12-6. The table gives the definition for
each function, and shows the types of function arguments.

Table 12-6: Functions

Function Scalar Vector Description Notes

abs(x) + + Absolute value x

acos(x) + + Arc cosine x
acosh(x) + + Hyperbolic arc cosine x
asin(x) + + Arc sine x
asinh(x) + + Hyperbolic arc sine x
arctan(x) + - Arc tangent x
atan(x) + + Arc tangent x
atanh(x) + + Hyperbolic arc tangent x
cos(x) + + Cosine of x
cosh(x) + + Hyperbolic cosine of x
db(x) - + Magnitude of the complex vector
in decibels:
db  x  = 20  log  mag  x  

erf(x) + + Error function

x
2
2
erf  x  = -------   e dt
–t

0

erfc(x) + + Error function complement

erfc(x) = 1 - erf(x)
exp(x) + + Exponential: e raised to power x
int(x) or + + Returns the integer part of x
trunc(x)

log10(x) + + Logarithm (base 10) of x

log(x) or ln(x) + + Natural logarithm (base e) of x

677 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

Table 12-6: Functions

Function Scalar Vector Description Notes

max(x,y) + + Maximum of two values,

if x>y then max(x,y)=x; otherwise
=y
min(x,y) + + Minimum of two values,
if x<y then min(x,y)=x; otherwise
=y
pow(x,y) + - Absolute power: absolute value of
x raised to integer part of y:
pow(x,y) = abs(x)^int(y)
rnd(x) + + Returns a random integer number
(or the vector with each
component a random integer
number) between 0 and absolute
value x (absolute value of the
vector x components)
sgn(x) + + Sign of value,
if x<0 then sgn(x) = -1;
if x=0 then sgn(x) = 0;
if x>0 then sgn(x) = 1
sin(x) + + Sine of x
sinh(x) + + Hyperbolic sine of x
sqrt(x) + + Square root of x
tan(x) + + Tangent of x
tanh(x) + + Hyperbolic tangent of x
valif(log_cond,x,y) + - Logical expression: log_cond is the a
or if(log_cond) then x else y logical condition,
if (log_cond, x,y)- which consist of
in ’-pspice’ mode two expressions
compared using
operators: ==; !=;
<; >; <=; >= or
their alphabetical
equivalents: eq, ne,
lt, gt, le,ge.

678 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

Operators
Table 12-7 lists the operations supported by MIXEDMODE.
The results of the modulo operation is the remainder when the first number is divided by the
second number. Both modulo operation arguments are rounded down to the nearest integer
before the operation is performed.

Table 12-7: Operations

Operation Type Operator Meaning Example Synonym Notes

Algebraic + Plus a+b

+ Unary Plus +a
- Minus a-b
- Unary Minus -a
* Multiply a*b
/ Divide a/b
^ Power a^b
% Modulo a%b
, Comma a,b
Logical & And a&b and
| Or a|b or
&& And a&&b and
|| Or a||b or
! Unary negation !a not
[cond] ? Ternary conditional See Example
x:y expression below
Relational < Less than a<b lt
> Greater than a>b gt
<= Less than or equal a<=b le
>= Greater than or equal a>=b ge
== Equal a==b eq
!= Not equal a!=b ne

679 ATLAS User’s Manual

MixedMode Syntax MixedMode: Mixed Circuit and Device Simulator

Example
.param Param1=5 Param2=10
+ Test=’(Param>Param2) ? 1 : 2’
+ Test2=’5*((Param1<Param2) ? (Param+1) : (Param2+2))’

680 ATLAS User’s Manual

Chapter 13
Quantum: Quantum Effect
Simulator
Quantum Effects Modeling Quantum: Quantum Effect Simulator

13.1 Quantum Effects Modeling

There are several quantum effects models that work with ATLAS to simulate various effects
of quantum mechanical confinement. These models are generally independent and should not
be used simultaneously in any given simulation. These models are described in the following
sections.
In Section 13.2 “Self-Consistent Coupled Schrodinger Poisson Model”, we discuss the self-
consistent Schrodinger-Poisson model. This model self-consistently solves Poisson's equation
(for potential) and Schrodinger's equation (for bound state energies and carrier
wavefunctions). This model should not be used to look at carrier transport problems.
In Section 13.3 “Density Gradient (Quantum Moments Model)”, we describe the density
gradient model. This model is based on the moments of the Wigner function equations of
motion and calculates a quantum correction to the carrier temperatures in the transport
equations. This model can accurately reproduce the carrier concentration predicted by the
Schrodinger-Poisson model but can also predict transport properties. This model cannot,
however, predict the bound state energies or wave functions given by the Schrodinger-
Poisson model.
In Section 13.4 “Bohm Quantum Potential (BQP)”, we describe the Bohm quantum potential
model. This model can be used to address a similar set or problems as the density gradient
model. The Bohm quantum potential model has two advantages over the density gradient
model: 1) better convergence and 2) better calibration to the Schrodinger-Poisson model.
In Section 13.5 “Quantum Correction Models”, we describe two quantum correction models.
These models by Van Dort and Hansch are both phenomenological models that give
corrections for the effects of quantum confinement in the inversion layer under the gate of
MOSFET devices.
In Section 13.6 “General Quantum Well Model”, we describe a general quantum well model
used to predict the gain and spontaneous recombination in quantum well light emitting
devices. This model solves Schrodinger equation to calculate the bound state energies and
wave functions. The bound state energies are used to predict the gain and spontaneous
recombination rates and the wave functions can be used to predict the overlap integral. The
capture-escape model is used together with Quantum Well Model to predict position of Fermi
levels of quantum wells.
In Section 13.7 “Quantum Transport: Non-Equilibrium Green’s Function Approach”, we
describe a fully quantum non-equilibrium Green’s function approach to modeling transport in
nanoscale devices with strong transverse confinement. This models predicts eigen energies
and eigen functions, quantum electron and curent density, and ballistic current-voltage
characteristics.
In Section 13.8 “Drift-Diffusion Mode-Space Method (DD_MS)”, we describe a
semiclassical approach to modeling transport in nanoscale devices with strong transverse
confinement. This model is coupled to ATLAS models for mobility, generation, and
recombination. It also predicts eigen energies and eigen functions, quantum electron and
curent density, and drift-diffusion current-voltage characteristics.

682 ATLAS User’s Manual

Self-Consistent Coupled Schrodinger Poisson Model Quantum: Quantum Effect Simulator

13.2 Self-Consistent Coupled Schrodinger Poisson Model

The solution of Schrodinger's equation gives a quantized description of the density of states in
the presence of quantum mechanical confining potential variations.
The N.SCHRO parameter of the MODEL statement enables the self-consistent coupled
Schrodinger-Poisson model for electrons. Set the P.SCHRO parameter in the MODEL statement
to enable the Schrodinger-Poisson model for holes. You can specify both SCHRO and P.SCHRO
on the same statement to model two carrier types simultaneously. ATLAS2D can solve
Schrodinger equation in 1D slices or 2D plane. In the case of cylindrical coordinates,
Schrodinger equation is solved in radial direction for different orbital quantum numbers and
for all slices perpendicular to the axis. ATLAS3D solves a 1D Schrodinger equation along
one of the axes at all in-plane nodes or 2D Schrodinger equation in plane slices. To set up
dimensionality and direction of the solver, use the SP.GEOM parameter on the MODELS
statement. Table 13-1 shows the possible values of the SP.GEOM parameter. If the parameter
is not specified, ATLAS will use a default value, depending on the situation. Note that the
2DXY option in cylindrical coordinates corresponds to a case of a 3D quantum confinement in
radial, axial, and azimuthal directions.

Table 13-1 Geometry of Schrodinger solver

SP.GEOM Geometry Comments

1DX 1D in X direction default for cylindrical mesh in ATLAS2D.

1DY 1D in Y direction default for non-cylindrical mesh in ATLAS2D.
1DZ 1D in Z direction Unavailable in ATLAS2D.
2DXY 2D in XY plane Default in ATLAS2D if option 2DXY.SCHRO is on
Default in ATLAS3D for cylindrical or non Atlas mesh or if
option 2DXY.SCHRO is on
2DXZ 2D in XZ plane Unavailable in ATLAS2D.
2DYZ 2D in YZ plane Default for ATLAS mesh in ATLAS3D.
Unavailable in ATLAS2D.

When the quantum confinement is in one dimension (along Y axis), the calculation of the
quantum electron density relies upon a solution of a 1D Schrodinger equation solved for
eigen state energies Ei(x) and wavefunctions i(x,y) at each slice perpendicular to the X axis
and for each electron valley (or hole band) 

h 
2
1  i
– ----- -----  -------------------- ------------ + E C  x y  i = E i  i 13-1
2 y  m   x y  y 
y


Here, m y (x,y) is a spatially dependent effective mass in Y direction for the -th valley and
EC(x,y) is a conduction band edge. The equation for holes is obtained by substituting hole
effective masses instead of electron ones and valence band edge -EV(x,y) instead of EC(x,y).

683 ATLAS User’s Manual

Self-Consistent Coupled Schrodinger Poisson Model Quantum: Quantum Effect Simulator

Analogously, in cylindrical coordinates, the equation for the radial part of the wavefunction
Rim(r) for each orbital quantum number m reads

h 1 1
2 R im 1 m
2
– ----- --- -----  ------------------- r --------------- – ------------------- ------- + E C  r z R im = E im R im 13-2
2 r r  m   r z  r  m   r z  r 2
r r

Finally, a 2D Schrodinger equation reads:

h 
2
1  i   1  i
– ----- -----  -------------------- ------------ + -----  -------------------- ------------ + E C  x y  i = E i  i 13-3
2 x  m   x y  x  y  m   x y  y 
x y

In order to take into account exchange-correlation effects, you can specify the SP.EXCORR
parameter on the MODELS statement. The exchange-correlation, treated within local density
approximation, is an additional negative potential energy term in the Hamiltonian, which
depends on the local carrier density and is given by
2
q - 2 13
V exc = – -----------------
2
 3 n  x y z    13-4
4  0

The number of hole bands is set by NUM.BAND parameter on the MODELS statement. If
NUM.BAND is set to 1, a solution for only one valence band MV (on the MATERIAL statement) is
obtained. If NUM.BAND is set to 3, a solution for heavy holes, light holes, and split-off band
holes (MHH, MLH, and MSO on the MATERIAL statement) is obtained. If NUM.BAND is set to 2,
only heavy holes and light holes are taken into account. Hole effective mass is assumed to be
isotropic.
To choose the number of electron valleys and anisotropy of effective mass, specify the
NUM.DIRECT on the MODELS statement. If NUM.DIRECT is set to 1, a solution for one -valley
with isotropic effective mass MC (on the MATERIAL statement) is obtained. If NUM.DIRECT is
set to 3, a solution for three doubly-degenerate X-valleys with anisotropic effective mass is
obtained. If NUM.DIRECT is set to 4, a solution for four L-valley with anisotropic effective
mass is obtained.
In case of anisotropic electron effective mass, when NUM.DIRECT is set to 3 or 4, the value of
effective mass in the quantization and the perpendicular directions in each valley depends on
the crystallographic orientation, which is set by X.ORIENT, Y.ORIENT, and Z.ORIENT on
the REGION statement. For example:
region num=1 y.min=0 y.max=1 mat=silicon y.orient=’<1,1,1>’
region num=1 y.min=0 y.max=1 mat=germanium y.orient=’<1,1,0>’
x.orient=’<1,-1,0>’
The default orientation is <1,0,0> along X, <0,1,0> along Y and <0,0,1> along Z axis. If one
or two axes are not specified, they are chosen as perpendiculars to the specified axes.
Effective mass tensor i,j, whose principal values are t1=MT1-1,t2= MT2-1, and l=ML-1,
where MT1, MT2, and ML are set on the MATERIAL statement is rotated in accordance with the
relative orientation of X, Y, and Z axes of the device and the principal axes of each valley. In
case of 1D Schrodinger equation, to get effective masses in the quantization (q) and the two
perependicular directions (p1,p2), the following expressions are used [262]:
m q = 1   q q 13-5

684 ATLAS User’s Manual

Self-Consistent Coupled Schrodinger Poisson Model Quantum: Quantum Effect Simulator

2
m p1 = 1    p1 p2 –  p1 q   q ,q  13-6

2
m p2 = 1    p2 p2 –  p2 q   q ,q  13-7

In case of 2D Schrodinger equation, to get effective masses in two quantization (q1,q2) and
the one perependicular directions (p), the following expressions are used [28]:
m q1 = 1   q1 q1 13-8

m q1 = 1   q1 q2 13-9

m p = 1    q1 q1  q2 q2 –  q1 q2    t1  t2  l 13-10

In case of cylindrical Schrodinger equation, to get effective masses in radial (q1,q2) and axial
directions (r), the following expressions are used:
m rad = 2    q1 q1 +  q2 q2  13-11

m axis = 1    q1 q1  q2 q2 –  q1 q2    t1  t2  l 13-12

The electron valleys are denoted in TONYPLOT as Valley #1, Valley #2, Valley #3 and Valley
#4.
Schrodinger equation uses a conduction or valence band edge of the bulk material. You have
an option to set an additional band off-set manually without getting into details of its origin.
To do so, use the following parameters from the MODELS or REGION or MATERIAL statements.

Table 13-2 Band off-set

Valley/Band Band off-set

Longitudinal DEC.C1

Transverse1 DEC.C2

Transverse2 DEC.C3

Isotropic Electron Band DEC.ISO (when NUM.DIRECT=1)

Heavy Holes DEV.HH

Light Holes DEV.LH

Split-off Holes DEV.SO

Isotropic Hole band DEV.ISO(when NUM.BAND=1)

685 ATLAS User’s Manual

Self-Consistent Coupled Schrodinger Poisson Model Quantum: Quantum Effect Simulator

The expression for the electron concentration is obtained by using Fermi statistics and eigen
anergies and wave functions:

kB T   E i – E F
ln 1 + exp  – --------------------
  i  x y 
2
n  x y  = 2 --------- m x  x y m z  x y 13-13
2  k T 
h B
 i=0

for 1D confinement and


2k B T  E i – E F
 i  x y  F –1  2  – --------------------
 
2
n  x y  = 2 ----------------- m z  x y 13-14
h  k T B
 i=0

for a 2D confinement case and cylindrical case, where F-1/2 is the Fermi-Dirac integral of
order -1/2.
ATLAS will search for a minimum number of filled eigen states in the range between
 –  ; 20kT  . You can also use EIGEN=X when a certain fixed number X of eigen states is
required. In the case of cylindrical mesh in ATLAS2D, Schrodinger equation has two
quantum numbers: radial number nr and orbital number m. All states with non zero m are
doubly degenerate, which corresponds to +m and -m or to clockwise and counter-clockwise
polarization of electron wave function. Therefore, when you set the EIGEN parameter to X,
ATLAS will store X states for m=0 and X states for each of 2*(X-1) orbital numbers
m =  1  2   X . This will add up to a total of X*(2X-1) states. In TONYPLOT, the states are
sorted in ascending order with respect to their eigen energies, while the information about
values of radial and orbital numbers (nr,m) is not stored. You can still determine nr and m of a
state, using a general property of wave fuctions Rnr,m(r). nr is always equal to the number of
times the wave function changes sign +1. For m=0, Rnr,m(r) is always non zero and has zero
first derivative at r=0. For m=1, Rnr,m(r) is always zero and has zero second derivative at r=0.
For m>1, Rnr,m(r) is always zero and has zero first derivative at r=0.
By default, electron penetration into insulator regions is not allowed. Use the OX.SCHRO
parameter on the MODELS statement to allow electron penetration and OX.MARGIN= to specify
the maximum penetration distance (default is 0.0003 microns). By default, ATLAS attempts
to find a solution in all semiconducting regions. If a solution is unnecessary in a particular
region, switch off SCHRO parameter on the REGION statement for that particular region.
When Schrodinger equation is solved in 1D or 2D slices, it is required that all nodes are
aligned (1D) or lie in one plane (2D). In case of unstructured or non-ATLAS mesh, ATLAS
can find a solution on an additional rectangular mesh and then interpolate results onto original
mesh. To set an additional rectangular mesh, use SPX.MESH, SPY.MESH, and SPZ.MESH
statements in a way similar to X.MESH, Y.MESH, and Z.MESH. In ATLAS3D, if Z.MESH is not
specified, Z planes of original mesh will be used.
Note that 2D Schrodinger does not require rectangular mesh within the 2D slice and can
handle meshes created outside of ATLAS (e.g., DEVEDIT or ATHENA). In devices with
crossection close to rectangular, ATLAS gives an option to speed up a solution by using the
SP.FAST parameter on the MODELS statement. When this parameter is on, a 2D solution is
sought in the form of linear combination of products on 1D solutions in the middle of the
solution domain along perpendicular axes. The solution will be exact for rectangular structure
and deviate slightly for non rectangular. SP.FAST parameter requires a rectangular mesh. If it

686 ATLAS User’s Manual

Self-Consistent Coupled Schrodinger Poisson Model Quantum: Quantum Effect Simulator

is unavailable, use SPX.MESH, SPY.MESH, and SPZ.MESH to set up an additional rectangular

mesh.
Since the wavefunctions diminish rapidly from the confining potential barriers in the
Schrodinger solutions, the carrier concentrations become small and noisy. You can refine
these carrier concentrations by setting a minimum carrier concentration using the QMINCONC
parameter on the MODELS statement. This parameter sets the minimum carrier concentration
passed along to the Poisson solver and the output to the structure files. The transition between
the Schrodinger solution and the minimum concentration is refined between 10QMINCONC
and QMINCONC so that it is continuous in the first derivative.
Once the carrier concentration is calculated using Equations 13-13 and 13-14, it is substituted
into the charge part of Poisson’s Equation. The potential derived from solution of Poisson's
equation is substituted back into Schrodinger's equation. This solution process (alternating
between Schrodinger’s and Poisson’s equations) continues until convergence and a self-
consistent solution of Schrodinger's and Poisson's equations is reached. A default predictor-
corrector scheme is used to avoid instability and oscillations of Poisson convergence. In this
scheme, after wave functions are found and fixed, eigen energies are changed locally and
plugged into Poisson equation (predictor). After several predictor iterations, a new set of
wave functions and eigen energies is computed (corrector). You can control the number of
Predictor iterations by a parameter NPRED.NEGF= (default is 7) on the MODELS statement.
The convergence criterion for potential for Schrodinger-Poisson as well as for NEGF and
DDMS models is given by the QCRIT.NEGF parameter on the MODELS statement with a
default of 0.001 eV. Maximum number of iterations after which Schrodinger-Poisson
proceeds to the next bias is set by SP.NUMITER with a default of 30. ATLAS also checks for
oscillatory convergence behavior for the last SP.NUMOSC (default 30) iterations starting from
iteration number SP.STAOSC (default 3). If logarithm of potential update or residual is
smaller than SP.MINDIF (default 0.001), then ATLAS will proceed to the next bias point.
For modeling of devices with transverse confinement, it is sometimes more preferable to set
von Neumann Boundary Conditions (BC) for potential in a contact. Von Neumann BC will
be applied if you set the REFLECT parameter on the CONTACT statement and if SCHRO or
P.SCHRO parameter is present on the MODELS statement.
To set number of eigen states stored in a structure file, use the EIGENS parameter on the
OUTPUT statement. Use the SAVE statement or the OUTFILE parameter on the SOLVE statement
to write the solutions of the self-consistent system into a structure file.
In obtaining self-consistent solutions for the Schrodinger's Equation, an assumption is made
about the location of the electron or hole quasi-fermi level. Normally, ATLAS tries to set the
quasi-fermi to that of a contact attached to the region where solution is obtained. To ensure
that the fermi level is zero, set FIXED.FERMI parameter on the MODELS statement.
If you want to solve Schrodinger equation as a postprocessing step under non-equilibrium,
specify CARRIERS=2 or CARRIERS=1 HOLES or CARRIERS=1 ELECTRONS on the METHOD
statement. In this case, a classical quasi Fermi level will be computed by solving drift-
diffusion equations. It is then used to find quantum electron density self-consistently with
Poisson equation. Flag FIXED.FERMI will be ignored. In case of bias sweep, ATLAS will
solve Schrodinger for each bias point. The SP.LAST parameter on the SOLVE statement
allows to solve the Schrodinger equation with classically computed Fermi levels only for the
last bias point.

687 ATLAS User’s Manual

Density Gradient (Quantum Moments Model) Quantum: Quantum Effect Simulator

13.3 Density Gradient (Quantum Moments Model)

The quantum models apply to several different types of problems. These problems are HEMT
channel confinement simulation, thin gate oxide MOS capacitors and transistors, and other
problems such as small geometry MESFETs and heterojunction diodes.
The effects due to confinement of carriers associated with variations of local potential on the
scale of the electron wave functions (i.e., quantum effects) can be modeled in ATLAS using a
density gradient. This model is based on the moments of the Wigner function equations-of-
motion [334, 335, 316, 305], which consists of quantum correction to the carrier temperatures
in the carrier current and energy flux equations (see Equations 3-20 to 3-23).
In the density gradient model, the expression for electron current given in Equation 3-11 is
replaced by the expression given in Equation 13-15.

J n = qD n n – qn n    –   –  n n  kT L   ln n ie   13-15

Here,  is a quantum correction potential.

Equation 13-16 is the density gradient model for holes.

J p = – q D p p – qp p    –   +  p p  kT L   ln n ie   13-16

The quantum potential can be calculated using either of the two expressions given in
Equations 13-17 and 13-18.
2
h 2 1 2
 = – -----------  log n + ---  log n  13-17
12m 2

2 2
h  n
 = – -------- -------------- 13-18
6m n

where  is a fit factor, m is the carrier effective mass, and n represents electron or hole
concentration as appropriate. To choose between these expressions, specify DGLOG for
Equation 13-17 or DGROOT for Equation 13-18 in the MODELS statement. By default, DGLOG is
used. The fit factors for electrons and holes can be specified independently using the
DGN.GAMMA or the DGP.GAMMA parameters of the MODELS statement respectively.
These expressions and particularly their derivatives can be seen to be sensitive to low carrier
concentrations. You can specify a minimum concentration to use in these calculations using
the QMINCONC parameter in the MODELS statement. This parameter specifies the minimum
concentration used in the calculations of Equations 13-17 and 13-18 in cm-3.
You can activate the model for electrons and holes independently. To activate the quantum
model with the quantum moments equation for electrons, use the quantum switch, MODELS
QUANTUM, in the MODELS statement. To activate the quantum moments equation for holes, use
the quantum switch, MODELS P.QUANTUM.
Specify T.QUANTUM in the OUTPUT statement to write the quantum temperatures in the
standard structure file. Once written you can examine the quantum temperature distribution
using TONYPLOT.

688 ATLAS User’s Manual

Density Gradient (Quantum Moments Model) Quantum: Quantum Effect Simulator

Since the distributions of carriers given by the density gradient can vary greatly from the
distributions predicted by the standard drift-diffusion or energy-balance models, the standard
initial guess strategies (e.g., INIT) aren’t usually suitable for obtaining solutions for quantum
moments. Until more suitable initial guess strategies can be devised, we’ve included a
damping factor to gradually apply to the density gradient.

This damping factor is specified by the QFACTOR parameter in the SOLVE statement. The
QFACTOR is implemented as a pre-factor to the expression for the quantum correction
potential, , in Equations 13-17 and 13-18. As such, a value of QFACTOR of 0.0 implies that
the density gradient is either turned off or not applied. A value of QFACTOR of 1.0 implies that
the Quantum Moment Model is turned on and applied.

You can vary the value of QFACTOR between 0.0 and 1.0 to overcome the problems of initial
guess. During this ramping of QFACTOR, PREVIOUS should be used as an initial guess. Also,
while varying the QFACTOR, trapping is available if a given solution doesn’t converge. In such
cases, the QFACTOR is reduced and solutions proceed until convergence is obtained or the
maximum number of trap steps is exceeded.
The following is a typical fragment from an input file that ramps the QFACTOR parameter
from 0 to 1:
MODEL NUMCARR=1 QUANTUM FLDMOB CONMOB SRH PRINT
OUTPUT CON.BAND VAL.BAND BAND.PARAM T.QUANTUM
SOLVE INIT
SOLVE QFACTOR=0.0
SOLVE QFACTOR=0.0001
SOLVE QFACTOR=0.001
SOLVE QFACTOR=0.01
SOLVE QFACTOR=0.1
SOLVE QFACTOR=1.0
LOG OUTF=test.log
SOLVE VDRAIN=0.01

689 ATLAS User’s Manual

Bohm Quantum Potential (BQP) Quantum: Quantum Effect Simulator

13.4 Bohm Quantum Potential (BQP)

This model was developed for SILVACO by the University of Pisa and has been implemented
into ATLAS with the collaboration of the University of Pisa. This is an alternative to the
Density Gradient method and can be applied to a similar range of problems. There are two
advantages to using Bohm Quantum Potential (BQP) over the density gradient method. First,
it has better convergence properties in many situations. Second, you can calibrate it against
results from the Schrodinger-Poisson equation under conditions of negligible current flow.
The model introduces a position dependent quantum potential, Q, which is added to the
Potential Energy of a given carrier type. This quantum potential is derived using the Bohm
interpretation of quantum mechanics [117] and takes the following form
–1 
– h   M   n  
2
Q = --------- -------------------------------------- 13-19
2 
n

where  and  are two adjustable parameters, M-1 is the inverse effective mass tensor and n is
the electron (or hole) density. This result is similar to the expression for the quantum
potential in the density gradient model with  = 0.5, but there are some differences about
how they are implemented. Q is added to the continuity equations the same way  is for the
density gradient method as shown in Equations 13-15 and 13-16.
The Bohm Quantum Potential (BQP) method can also be used for the Energy balance and
hydrodynamic models, where the semi-classical potential is modified by the quantum
potential the same way as for the continuity equations.
The default iterative scheme used to solve the non-linear BQP equation along with a set of
semi-classical equations is as follows. After an initial semi-classical solution has been
obtained, the BQP equation is solved on its own Gummel iteration to give Q at every node in
the device. The semi-classical potential is modified by the value of Q at every node and the
set of semi-classical equations is then solved to convergence as usual (using a Newton or
block iterative scheme). Then, the BQP equation is solved to convergence again and the
process is repeated until self-consistency is achieved between the solution of the BQP
equation and the set of semi-classical equations. The set of semi-classical equations solved
can be any of the combinations usually permitted by ATLAS.
You can enable an alternative iteration scheme by specifying the BQP.NEWTON flag on the
METHOD statement. This causes the BQP equation to be included in the fully coupled Newton
iteration. This can only be used with either BQP.N or BQP.P models seperately, and not with
both enabled, and is only available in QUANTUM2D. Additionally, this model also modifies
the intrinsic carrier density, ni, by the Bohm quantum potential. This ensures that under
2
equilibrium conditions np – n i = 0.0 . On subsequent SOLVE statements, the FREEZEBQPNI
flag should be enabled to ensure that the value of ni used is the equilibrium one
To use the BQP model for electrons (or holes), specify BQP.N (BQP.P) in the MODELS
statement. You can also set the parameter values ( and ) and the direction of the
quantization (confinement). Tables 13-3 and 13-4 show the parameters to use for the MODEL
statement.

690 ATLAS User’s Manual

Bohm Quantum Potential (BQP) Quantum: Quantum Effect Simulator

Table 13-3 MODEL Statement Parameters for Electrons

Parameter Type Default Units

BQP.N Logical false

BQP.NGAMMA Real 1.2 --
BQP.NALPHA Real 0.5 --
BQP.QDIR Integer 2 --

Table 13-4 MODEL Statement Parameters for Holes

Parameter Type Default Units

BQP.P Logical false

BQP.PGAMMA Real 1.0 --
BQP.PALPHA Real 0.5 --
BQP.QDIR Integer 2 --

BQP.N switches on the model for electrons. BQP.P switches it on for holes. The BQP.NGAMMA
and BQP.PGAMMA allow you to set the  parameter for electrons and holes respectively.
BQP.NALPHA and BQP.PALPHA allow you to set the  parameter for electrons and holes
respectively.
You can specify BQP.N and BQP.P separately. But if quantum effects for electrons and holes
occur in the same device, then you can specify them together in the same MODELS statement.
The BQP.QDIR parameter specifies the principal quantization direction.
BQP.QDIR = 1 means the X direction
BQP.QDIR = 2 means the Y direction
BQP.QDIR = 3 means the Z direction
For example, if a MOSFET channel is in the XY plane, the direction of quantization
(quantum confinement) is the Z direction you would set BQP.QDIR=3. For semiconductors
with spherical bands, BQP.QDIR will have no effect.

691 ATLAS User’s Manual

Bohm Quantum Potential (BQP) Quantum: Quantum Effect Simulator

13.4.1 Calibration against Schrodinger-Poisson Model

You can obtain close agreement between BQP and the results of Schrodinger-Poisson (S-P)
calculations for any given class of device. To obtain comparisons with S-P results, we
recommend to use either the new quasistatic capacitance-voltage profile feature or compare
charge-voltage curves. This will ensure similar charge control properties between the two
models.
The first part of the calibration is to choose a suitable biasing for the device. There should be
negligible current flow and quantum confinement effects that manifest at the chosen biases.
The second part of the calibration is to set the appropriate BQP parameters in the MATERIAL
or MODELS statements, and to set CARRIERS=0 in the METHOD statement.
This will cause the BQP equation to be coupled with Poisson's equation using the charge
density terms.
 E c + qQ 
n = N c exp  – ------------------------ 13-20
 kT 
L

 qQ – E v 
p = N v exp  – -----------------------
- 13-21
 kT 
L

This gives the same results as solving the current continuity equations with the constraint of
zero current density.
The third part of calibration is to choose the quantity to compare with S-P results.
For example, for a MOSFET holding the drain and source voltages at the same bias and
ramping the gate bias will give us a bias dependent capacitance with negligible current flow.
So for an NMOS, you may have the statement
SOLVE VGATE=0.0 NAME=GATE VSTEP=0.01 VFINAL=2.0 QSCV
to give us the quasi-static C-V curve as it is biased into inversion. It is best to use a fine
voltage step with QSCV to give good resolution. The process can be repeated by setting the
S-P model in the MODELS statement instead of BQP to obtain the same set of curves for the S-
P model.
The BQP model is then rerun with different sets of parameters until an acceptable agreement
with the curves produced by the S-P model is achieved.

692 ATLAS User’s Manual

Bohm Quantum Potential (BQP) Quantum: Quantum Effect Simulator

13.4.2 Post Calibration Runs

After obtaining the parameters for BQP, either the Drift-Diffusion or energy balance
(hydrodynamic) equations can be solved as usual. For cases where Lattice Heating is
important then LAT.TEMP can be enabled at the same time.
The iteration scheme uses a modified version of BLOCK. Set BLOCK in the METHOD statement
(although NEWTON and GUMMEL are ignored, BLOCK is always used if the BQP model is set). If
an energy balance model is chosen (HCTE.EL or HCTE.HO on the MODELS statement), then an
alternative iteration scheme will become available by specifying BQP.ALTEB in the METHOD
statement. This method is slower and is only available in ATLAS2D but may have better
convergence properties.
By using BQP.NOFERMI, the BQP equation will only use its Boltzmann statistics form.
Without this parameter, the statistics used are those specified for the other equations. With
fermi statistics, the convergence can be poor for very high carrier densities, and this
parameter can circumvent some convergence properties.
Re-calibrate the BQP parameters if you set BQP.NOFERMI.

Table 13-5 METHOD Statement Parameter

Parameter Type Default Units

BQP.ALTEB Logical false

BQP.NEWTON Logical false
BQP.NOFERMI Logical false
BQPX.TOL Real 2.510-7 BQPX.TOL

BQPR.TOL Real 1.010-26 (2D) BQPR.TOL

1.010-18 (3D)
BQP.NMIX Real 1 BQP.NMIX

BQP.PMIX Real 1 BQP.PMIX

NBLOCKIT Integer 15 NBLOCKIT

ITLIMIT Integer 25 ITLIMIT

GUMITS Integer 100 GUMITS

To speed up convergence, specify the NOCURRENT parameter on the first SOLVE statement
after SOLVE INIT. It should prevent the need to use the QFACTOR parameter as was necessary
for the density gradient method. QSCV enables the quasistatic capacitance calculation and
output.

693 ATLAS User’s Manual

Bohm Quantum Potential (BQP) Quantum: Quantum Effect Simulator

Table 13-6 SOLVE Statement Parameters

Parameter Type Default Units

FREEZEBQPNI Logical false

NOCURRENT Logical false
QSCV Logical false
QFACTOR Real 1.0

Use the parameters in Table 13-5 to control solution convergence behavior.

The solution comprises of OUTER iterations (which end when self-consistency occurs) and
within these are solutions of groups of equations or single equations. The BQP equation is
implemented so that both LHS and RHS convergence are required. You cannot override this
default behavior. Tests carried out have shown it gives the best performance. ITLIMIT
controls how many iterations can occur for the BQP equation in SOLVE INIT. This is 10 *
ITLIMIT. For CARRIERS=0 solutions, the maximum number of iterations for the individual
equation solvers is GUMITS. The maximum number of OUTER iterations is ITLIMIT.
NBLOCKIT controls the number of OUTER cycles in the drift-diffusion and energy-balance
cases. the maximum number of drift-diffusion solution and carrier-energy solution iterations
is ITLIMIT, and the maximum number of BQP solutions is GUMITS. The convergence may
become unsteady when the carrier density is very high and Fermi statistics is used. In this
case, you may decrease BQP.NMIX or BQP.PMIX parameters or both to a value between 0 and
1. This will cause relaxation of Fermi factors between two consecutive iterations and result in
steadier convergence.

Note: The Bohm Quantum potential is stored in ATLAS as an energy and so the convergence criteria are actually
of order KT less than those for scaled electrostatic potential. In ATLAS 3D, the BQP equation had an extra
scaling factor applied and so the RHS norms are larger than in 2D. LHS norms, however, are the same.

The Bohm Quantum Potential is automatically calculated in insulators. If you specify the
SEMICONDUCTOR parameter in the MATERIAL statement for the insulator, the drift-diffusion
equations will be solved in the insulator and will include the BQP corrections. It is also
recommended to set the densities of states of the insulator to be the same as the
semiconductor to which they interface. For example for Silicon/SiO2 system, set
MATERIAL REGION=N NC300=2.8E19 NV300=1.04E19
where N is region number of SiO2.
You can also set the BQP parameters on a material by material basis, which may be necessary
for heterojunction based devices. The BQP model is applied globally and cannot be set on a
region by region basis.

694 ATLAS User’s Manual

Bohm Quantum Potential (BQP) Quantum: Quantum Effect Simulator

Table 13-7 shows the material parameters that have a direct impact on the BQP equation.

Table 13-7 MATERIAL Statement

Parameter Type Default

SEMICONDUCTOR Logical false

EG300 Real
AFFINITY Real
NC300 Real
NV300 Real
BQP.NALPHA Real 0.5
BQP.NGAMMA Real 1.2
BQP.PALPHA Real 0.5
BQP.PGAMMA Real 1.0

There is one more pertinent parameter, BQP.NEUMANN, which can be set on the MODELS
statement.
BQP.NEUMANN sets the normal gradient of classical+quantum potential explicitly to zero on
non-contact boundaries. If cleared using MODELS ^BQP.NEUMANN, the Neumann conditions
will then be applied implicitly. Clearing this flag may give a slight speed increase and slight
differences in solution.

Table 13-8 OUTPUT Statement Parameters

Parameter Type Default

P.QUANTUM Logical False

Setting the P.Quantum parameter in the OUTPUT statement will cause the Bohm Quantum
Potential to output to a structure file. The Electron Quantum Potential and Hole Quantum
Potential will both output.

695 ATLAS User’s Manual

Quantum Correction Models Quantum: Quantum Effect Simulator

13.5 Quantum Correction Models

In deep submicron MOS devices, quantum effects in the channel can have significant effects
on the device characteristics. These effects are directly due to the increased doping levels and
thinner gate oxide. In the channel of such devices, the potential well formed during inversion
where the effects of quantum confinement must be considered. The direct effect of such
quantum confinement is that the peak of the carrier concentration is shifted away from the
interface and the thinner gate oxide can cause a marked difference in gate capacitance.
Consideration of the effects may be essential for accurate prediction of the device turn-on
voltage.
To fully treat quantum effects, solving the Schrodinger’s equation, isn’t always desirable or
necessary.

13.5.1 Hansch’s Model

The quantum mechanical correction given by Hansch [97] is suitable for accurate simulation
of the effects of quantum mechanical confinement near the gate oxide interface in MOSFETs.
To enable the model, specify the HANSCHQM parameter of the MODEL statement. This
correction model is a modification of the density of states as a function of depth below the Si/
SiO2 interface, which is given by Equation 13-22.
2
N C = N C 1 – exp –  ----------------------
* z
13-22
 LAMBDA

Here, NC is the standard density of states, z is the depth below the interface. LAMBDA is a user-
definable parameter in the MODEL statement.

Table 13-9 User-definable parameter for Equation 13-22

Statement Parameter Default Units
MODELS LAMBDA 1.010-3 µm

696 ATLAS User’s Manual

Quantum Correction Models Quantum: Quantum Effect Simulator

13.5.2 Van Dort’s Model

In the Van Dort’s model [283], the effects of the quantum confinement are modeled by
broadening the bandgap near the surface, which makes a function of a perpendicular electric
field and distance from the surface. In Van Dort’s model, the change in bandgap is given by
the expression in Equation 13-23.
 si 13

Eg =  B DORT    -------------------  E – Ec  g  y  13-23
· 4qk B T L

Here, B.DORT is a user-definable parameter in the MODEL statement, E  is the perpendicular

electric field, and g(y) is a function to restrict the application of the model to the channel
region, given by Equation 13-24.
2
–  Y   D · DORT  
2e
g  y  = --------------------------------------------------------2- 13-24
– 2  Y   D · DORT  
1+e
The default parameter values for , Ec, and  are dependent on technology as described in
Table 13-10.

Table 13-10 Technology Dependent Parameter Values of General Van Dort Model

Technology Ec  

electron inversion NINV.EC NINV.EXP NINV.BETA

hole inversion PINV.EC PINV.EXP PINV.BETA

electron accumulation NACC.EC NACC.EXP NACC.BETA/(1+Nd/NACC.N)

hole accumulation PACC.EC PACC.EXP PACC.BETA/(1+Na/PACC.N)

The default values for the various parameter of the MODELS statement are given in Table 13-
11.

Table 13-11 Default Parameter Values for the Van Dort Model

Parameter Type Default Units Reference

NINV.EC Real 0 V/cm [283]

NINV.BETA Real 5.92E-8 (1)
NINV.EXP Real 0.666666 (1)
PINV.EC Real 100000 V/cm [99]
PINV.BETA Real 6.10E-8 (2)
PINV.EXP Real 0.666666 (2)
NACC.EC Real 0 V/cm [100]
NACC.BETA Real 8.45E-8 (3)

697 ATLAS User’s Manual

Quantum Correction Models Quantum: Quantum Effect Simulator

Table 13-11 Default Parameter Values for the Van Dort Model
NACC.EXP Real 0.666666 (3)
NACC.N Real 2.0E19 cm-3 [100]
PACC.EC Real 0 V/cm [100]
PACC.BETA Real 4.97E-8 (3)
PACC.EXP Real 0.666666 (3)
PACC.N Real 3.6E19 cm-3 [100]

The new models are enabled for the various condicitons are enabled by the flags NINV.DORT,
NACC.DORT, PINV.DORT, and PACC.DORT. The flags N.DORT is aliased to NINV.DORT and
P.DORT is aliased to PINV.DORT.

698 ATLAS User’s Manual

General Quantum Well Model Quantum: Quantum Effect Simulator

13.6 General Quantum Well Model

The general quantum well model is used to predict bound state energies in a region for
subsequent use in modelling optoelectronic gain, radiative recombination and absorption.
Figure 13-1 shows the simulation flow for such modeling.

band edges
STRAIN BAND eff. masses SCHRODINGER
CALCULATION
momentum
band edges matrix elem bound state energies
eff. dens. of state wave functions

POLARIZATION DRIFT-DIFFUSION GAIN AND SPONTANEOUS

POISSON EMISSION

radiative recomb
photo-generation stimulated recomb

LUMINOUS LASER LED

VCSEL

Figure 13-1: Simulation Flow For Physically Based Optoelectronic Models

The solution for the bound state energies is done by solving Schrodinger equation (see
Equation 13-1) along discrete slices in the quantization direction. Effective masses and band-
edge parameters are taken from multi-band k.p-based models set by ZB.ONE, ZB.TWO,
ZB.THREE, and WZ.THREE parameters on the MODELS statement, which stand for zincblende
or wurtzite type of material and the number of valence bands used in the calculation.
You can enable the quantum well model by specifying QWELL in the REGION or MODELS
statement. The orientation and dimensionality of Schrodinger solver is set by SP.GEOMETRY
parameter on the MODELS statement with a default of 1DY. Alternatively, you may solve 1DX
and 1DZ cases. Regions with QWELL parameter are treated as independent quantum wells,
unless they are adjacent to each other and have the same orientation, in which case they are
merged into one quantum well region. This means that a single quantum well region can
include different materials. Merging of adjacent regions into a single quantum well can be
overridden by assigning different numbers to different wells, using QWNUM parameter on the
REGION or the MODELS statement (e.g., QWNUM=1).
Whenever a quantum well region is specified, the solution domain for Schrodinger equation
will actually be expanded into neighboring regions to account for wavefunction penetration
into the barriers. The maximum penetration length is set by WELL.MARGIN parameter on the
MODELS statement with a default value of 0.01 micron. It is perfectly admissible to have
wavefunctions of two different wells to overlap.

By default, ATLAS mesh is used for discretization. If the ATLAS mesh too coarse, you can
specify WELL.NX or WELL.NY or WELL.NZ or all three in the REGION or the MODELS statement

699 ATLAS User’s Manual

General Quantum Well Model Quantum: Quantum Effect Simulator

to use an auxiliary quantum well mesh. If an auxiliary mesh is used, the sampling is uniform
and bound state energy values are interpolated back onto the device mesh for the optical
modelling. Generally, you should select WELL.NX, WELL.NY, and WELL.NZ to adequately
resolve the geometry but there is a trade-off between computational accuracy and speed.
When Schrodinger equation is discretized, the conduction and valence bands are modified in
the barriers to avoid bound states at the edges. The schematics below shows the original
conduction band Ec(x) and the modified band E/c(x) as seen by confined carriers.

E/c(x)
Ec(x)

Figure 13-2: Modification of conduction band Ec(x), as seen by confined carriers

The WELL.CNBS and WELL.VNBS parameters of the REGION or the MODELS statement specify
the maximum number of bound state energies to resolve for each conduction and valence
band with default values of 1. An output structure file of the QWELL model contains bound
state energies and wavefunctions for each well, band, and sub-band. The actual number of
bound states found may be less than or equal to the value specified by WELL.CNBS and
WELL.VNBS. Additionally, the output contains band edge energies for each band and bound
carrier densities.
You can also specify regions for quantum well simulation in the SUPERLATTICE (DBR)
statement. To enable the model, specify the QWELL1 and QWELL2 parameters. These enable
quantum modeling for the first and second superlattice cycles respectively. The sampling of
the quantum wells are controlled by the WELL1.NX, WELL2.NX, WELL1.NY and WELL2.NY
parameters. The numbers of bound states are controlled by the WELL1.CNBS, WELL2.CNBS,
WELL1.VNBS and WELL2.VNBS parameters.

13.6.1 Superlattice Model

In some applications, you may want to model a long set of independent quantum wells. In
other cases, you may be interseted in collective bound and traveling states of a set of strongly
coupled quantum wells.
In the first case, you can specify regions for quantum well simulation in the SUPERLATTICE
(DBR) statement. To enable the model, specify the QWELL1 and QWELL2 parameters. These
enable quantum modeling for the first and second superlattice cycles respectively. The
sampling of the quantum wells are controlled by the WELL1.NX, WELL2.NX, WELL1.NY, and
WELL2.NY parameters. The numbers of bound states are controlled by the WELL1.CNBS,
WELL2.CNBS, WELL1.VNBS, and WELL2.VNBS parameters. The rest of the model is identical
to QWELL model.
In the second case, when a set of quantum wells is strongly coupled, you may set SLATT
parameter on the REGION statement. All adjacent regions with the SLATT attribute will be

700 ATLAS User’s Manual

General Quantum Well Model Quantum: Quantum Effect Simulator

lumped into the same superlattice domain. To assign adjacent regions to different
superlattice domains, use the SL.NUMBER parameter. You also need to set direction of
Schrodinger solver using SL.GEOM = 1DX or 1DY (default). The SLATT model uses the same
band structure as the QWELL model. It will also reuse WELL.NX and WELL.NY parameters, if
specified. Calculation of collective states of a superlattice domain is done using Quantum
Transmitting Boundary Method (QTBM), which is a non-linear Schrodinger equation with
open-boundary contacts. Calculated eigen states are sorted into bound and traveling based on
the location of the wavefunction and eigen energy with respect to band edge. SLATT model
can coexist with QWELL model or used independently.

13.6.2 Radiative Recombination in Quantum Wells

Knowledge of bound states allows ATLAS to calculate spontaneous recombination and gain
due to interband transitions. To calculate spontaneous recombination and include it into the
continuity transport equations, use SPONTANEOUS parameter on the MODELS statement. If the
QWELL region is also assigned LED parameter, then luminous power from this region will be
computed and stored in the IV log file.
Spontaneous emission rate per unit energy and gain are given as functions of photon energy
h and polarization TE, TM by the following expressions:

  nr e2      2  n m 2
- f  1 – f  I m n M dE
2D 2
r spon  h  =  ----------------------------
  hc 3  0 m 20
   r  E  ---------------------------------------------------
E – h +   2
2 2 c v  b
13-25
n m

  e2     2  n m 2
-  f – f  I m n M dE
2
  r
2D
g  h   =  ----------------------------
2
 ---------------------------------------------------
2 2 c 
13-26
 n r c  m 0  0 E – h +   2 v b
n m

where
2D
• r = mr   h 2 Lz is the two-dimensional density of states.
• mr is a reduced electron-hole effective mass.
• nr is a material refractive index.
• I m n =  mn is an overlap integral of electron and hole wavefunctions.

• M is a polarization- dependent bulk momentum matrix element.
b
•  is the line-width due to Lorenzian broadening.
• The sum is taken over all transitions between electron sub-bands with energies En and
hole sub-bands with energies Em of all valence bands.

Fermi-Dirac distribution functions have the following form ( E mn  E n – E m ):

n 1
f c  E  = ---------------------------------------------------------------------------------------------------------------------------------
- 13-27
1 + exp   E n +  m r  m e   E – E mn  – E Fc   kT  

m 1
f v  E  = -------------------------------------------------------------------------------------------------------------------------------
- 13-28
1 + exp   E m –  m r  m h   E – E mn  – E Fv   kT  

Here, we assume that bound carriers are in the detailed balance with bulk carriers. Therefore,
we use local Fermi levels of bulk carriers to compute the recombination rate at each node.

701 ATLAS User’s Manual

General Quantum Well Model Quantum: Quantum Effect Simulator

See Section 13.6.4 “Capture-Escape Model” for more realistic treatment of bound carrier
dynamics.
In order to compute total radiative recombination rate, we perform averaging over
polarizations and integration over the whole emission spectrum:

  2 r TE  h  + r TM  h 
 spon spon 
R sp  x y z  =  ------------------------------------------------------------------------ d h  
3
13-29
0

The integration over the spectrum is done using Gauss-Laguere quadrature rule. The number
of energy points is set by WELL.NERSP (default is 32) parameter on the METHOD statement. If
the parameter is set to zero (not recommended), a trapezoidal integration will be used with a
spacing set by WELL.DENERGY parameter on the MODELS or MATERIAL statement.
Spectrum of spontaneous recombination and gain can be stored in a separate file using
SPECTRUM=’filename’ parameter on the SAVE statement. The spectra are averaged over all
slices of the quantum well.

13.6.3 Intersubband Radiative Transitions

In order to compute intersubband gain and spontaneous emission, specify INTERSUB.SPONT
on the MODELS statement. In order to save spectrum of intersubband transitions, set
SPEC.INTERSUB=’filename.log’ parameter on the SAVE statement.
Intersubband spontaneous emission rate per unit energy and gain are given as functions of
photon energy h by the following expressions (TE light does not couple to intesubband
transitions):
3
 nr e2     2 
 d m n f
TM 2 2D n m
r spon  h  =  ------------------
-  r ---------------------------------------------------------
-  1 – f  dE t 13-30
  hc 3  0 n m  E mn – h   +    2 
2 2

 e 2    2 
 h   =  ---------------  2
TM 2 2D n m
g d m n  r ---------------------------------------------------------
-  f – f  dE t 13-31
nr c 0  E mn – h   +    2 
2
n m

where
• d m n =  mx  n is a dipole moment between initial and final states.
•  is the line-width due to Lorenzian broadening.
• The sum is taken over all transitions between sub-bands with energies En and Em of the
same carrier type.
• Integral is taken over transverse energy Et
Fermi-Dirac distribution functions have the following form:
n 1
f  E  = -----------------------------------------------------------------------------
- 13-32
1 + exp   E n + E t – E F   kT 

702 ATLAS User’s Manual

General Quantum Well Model Quantum: Quantum Effect Simulator

13.6.4 Capture-Escape Model

The model is launched by adding WELL.CAPT parameter to the MODELS statement. Capture-
escape model treats the dynamics of bound carriers by breaking carriers into bulk and bound
subsystems, coupled to each other by a capture-escape rate. The capture-escape rate is added
to the continuity equation to remove bulk carriers. The same rate serves as a generation
mechanism for bound carriers and the following additional equations are solved for each slice
of each quantum well to find quantum well Fermi energies:
2D
 ñ ˜ ˜ 1 ˜ 2D
------------- = R capt n – R recomb – ---  J n 13-33
t e

2D
p̃ ˜ ˜ 1 ˜ 2D
------------ = R capt p – R recomb – ---  J p 13-34
t e

where ˜ means that the integration of the variable is performed over the box, belonging to the
slice of the quantum well. The Fermi energies thus found are used in the expressions for
radiative transitions and all other important phenomena, such as in-plane transport, inter-well
tunneling and inter-well spontaneous emission.
The capture-escape rate, which is added to the continuity equation for bulk carriers is given
by

  E 2D – E 3D 
n
3D
2D   F  F 
R capt = ----------   1 – f  E  E    1 – exp  ----------------------------------  13-35
   F      kT 
   

where n3D is a bulk carrier density, is a capture time, EF,2D is a fermi level of sub-band ,
EF3D is a local bulk fermi level, and E is the sub-band energy. The summation is performed
over all sub-bands of the same carrier type. In the first bracket under the summation, f is a
Fermi-Dirac function, which reduces capture rate as the sub-band is getting filled. Carrier
capture is a multi-phonon process, whose detailed physics is not the objective of this model.
Rather, we use capture time  as a parameter set by WELL.TAUN (for electrons) and
WELL.TAUP (for holes) parametrs on the MATERIAL statement.
As seen from the expression above, capture rate may become negative when Fermi level for
bound carriers exeeds that for bulk carriers. In this case, it will act as an escape mechanism of
2D carriers from the well into the bulk.
For LED and laser applications, it is usually sufficient to have one Fermi level for all sub-
bands in the same slice of the same carrier type. However, if an individual equation for each
sub-band of each slice is required, specify WELL.SEPARATE on the MODELS statement.
In the simplest form, only the first term on the right hand side of the is used, which in steady
state, means that bound carriers are in the detailed balance with bulk carriers.
If SPONTANEOUS parameter is specified on the MODELS statement, a second term will be
added, which will depopulate 2D sub-bands. When capture-escape model is active,
spontaneous recombination due to quantum wells is no longer a part of bulk continuity
equations.

703 ATLAS User’s Manual

General Quantum Well Model Quantum: Quantum Effect Simulator

In order to take into account spontaneous emission between bound states in the neighboring
wells, specify the WELL.WWSPONT parameter on the MODELS statement. This mechanism is a
must for type-II quantum wells and may be important when the wave function overlap of
neighboring wells is large.
The third term, switched on by WELL.INPLANE parameter on the MODELS statement, models
the in-plane transport of 2D carriers in the well, which becomes significant when there is a
voltage drop along the well. Electron and hole current for 2D carriers are given by
2D 2D
Jn = – e   n n   E Fn 
2D
and J 2D
p
2D 2D
= – e   p p   E Fp  .
 

Parameters on the METHOD statement regulate the convergence of the model. WELL.ERROR
(default 1e-5 V) is the largerst relative error of bound carrier density. WELL.ITERMAX
(default is 20) sets the maximum number of Newton iterations, after which the 2D carrier
densities are assumed to be converged. WELL.PROJ (default is true) sets the projection
method as initial guess. At each Newton iteration, the solution of the coupled system of
Poisson, bulk continuity and quantum well rate equations is done using the Schur
Complement method by default. A direct method, although less effective, can still be called
by switching off the SCHUR parameter on the METHOD statement.
The output structure file will contain Fermi levels for each sub-band, capture rates and in-
plane current densities.

13.6.5 Charge Self-Consistency

In order to take into account confined carriers in the Poisson equation, set WELL.SELFCON
parameter on the MODELS statement. This parameter should be used only with the capture-
escape model. The parameter will trigger the following changes to the normal computation.
At first, the conduction and valence bands in the well are modified before plugging into
continuity equation, so that the well is effectively removed for bulk carriers. The schematics
below shows the original conduction band Ec(x) and the modified band E/c(x), as seen by
bulk carriers. The value of the conduction band in the well is given by the energy of the
lowest of the left and rifgt barriers. This reduces bulk carrier density in the well by not letting
the bulk carriers to fill it. Then, the 2D carrier density is computed with Equation 13-13
added to the bulk carrier density and used in Poisson equation n tot = n 3D + n 2D .

E/c(x)
Ec(x)

Figure 13-3: Modification of conduction band Ec(x), as seen by bulk carriers

704 ATLAS User’s Manual

Quantum Transport: Non-Equilibrium Green’s Function Approach Quantum: Quantum Effect Simulator

13.7 Quantum Transport: Non-Equilibrium Green’s Function Approach

This section describes a Non Equilibrium Green’s Function approach (NEGF). This fully
quantum method treats such effects as source-to-drain tunneling, ballistic transport, quantum
confinement on equal footing. In the first part, we discuss a mode space approach, where
NEGF formalism in transport direction is coupled with Schrodinger equation in transverse
plane. This situation is common to double gate transistors, FinFETs, nanowire FETS, and so
on. In the second part, we discuss a planar NEGF approach, where the transverse plane is
assumed to be infinite. This approach is more applicable to multiple barrier structures, such a
resonant tunneling diode (RTD).

13.7.1 Mode Space NEGF Approach

By specifying the NEGF_MS and SCHRODINGER options on the MODELS statement, you can
launch an NEGF solver to model ballistic quantum transport in such devices as double gate or
surround gate MOSFET in ATLAS2D or ATLAS3D. All the parameters described in Section
13.2 “Self-Consistent Coupled Schrodinger Poisson Model” have the same meaning in the
context of NEGF.
An effective-mass Hamiltonian Ho of a two-dimensional device, given by

h     
2
1 1
H o = – ----- -----  -------------------- ----- + -----  -------------------- ----- 13-36
2 x  m  x y  x y  m  x y  y
 
x y

is discretized in real space using a finite volume method. A corresponding expression in

cylindrical coordinates is

h 1 1   1 
2 2
1 m
H o = – ----- --- -----  ------------------- r ----- – ------------------- ------- + -----  ------------------- ----- 13-37
2 r r  m   r z  r m   r z  r 2 z  m   r z  z
r r z

Instead of solving a 2D or 3D problem, which may take too much computational time, a
Mode Space (MS) approach is used. A Schrodinger equation is first solved in each slice of the
device to find eigen energies and eigen functions. Then, a transport equation of electrons
moving in the sub-bands is solved. Because only a few lowest eigen sub-bands are occupied
and the upper sub-bands can be safely neglected, the size of the problem is reduced.
Moreover, in the devices where the cross-section does not change (e.g., a nanowire
transistor), the sub-bands are not quantum-mechanically coupled to each other and the
transport equations become essentially 1D for each sub-band. Therefore, you can further
divide the method into Coupled (CMS) or Uncoupled Mode Space (UMS) approaches.
ATLAS will automatically decide on the minimum number of sub-bands required and the
method to be used. It is possible, however, to set the number of sub-bands by using the EIGEN
parameter on the MODELS statement. To enforce either CMS or UMS approaches, you can use
NEGF_CMS or NEGF_UMS instead of NEGF_MS on the MODELS statement. The transformation of
a real space Hamiltonian Ho to a mode space is done by taking a matrix element between m-th
an n-th wave functions of k-th and l-th slices:
MS k l
H mnkl =  m  y H o  n  y  13-38

705 ATLAS User’s Manual

Quantum Transport: Non-Equilibrium Green’s Function Approach Quantum: Quantum Effect Simulator

Note, that in cylindrical geometry, different orbital quantum numbers will result in zero
matrix element and are therefore uncoupled. The quantum transport equations in the mode
space read:

 E – H MS R R R
 mnkl –  S –  D G = I 13-39

 E – H MS mnkl –  R –  R  G < =   < +  < G A 13-40

 S D  S D

Here, E is an electron (hole) energy. S,DR are retarded self-energies, carrying information
about the density of states of source and drain semi-infinite contacts. S,D< are less-than self-
energies, carrying information about the Fermi distribution function in contacts.
The solution of these equations are a retarded Green’s function GR(E), whose diagonal
elements have a meaning of the local density of states as a function of energy and a less-than
Green’s function G <  E  , whose diagonal elements have a meaning of carrier density as a
†
function of energy. GA  E  =  G R  E   is an advanced Green’s function. A non-uniform energy
grid will be generated internally, while you can control the number of energy grid points
using a parameter ESIZE.NEGF= on the MODELS statement. After solving the above equations
at each energy, following quantities of interest are computed by ATLAS and stored in a
structure file (expressions are given for rectangular grid):
• carrier density

   G mnii  E  m  yj 
i < i *i dE
n  x i y j  = – ----- n  y j  ------ 13-41
Lz 2
k z  mn

• x-component of current density

   e  ti + 1ijj G mnii + 1  E   m  yj 
2e < i *i + 1 dE
J x  x i y j  = – --------------- n  y j  ------ 13-42
hL z  x 2
k z  mn

• y-component of current density

   e  tiijj + 1 G mnii  E   m  yj 
2e < i *i dE
J y  x i y j  = – --------------- n  y j + 1  ------ 13-43
hL z  y 2
k z  mn

• total current density

2 2 12
J = J x + J y  13-44

Here, tijkl is an off-diagonal element of real space Hamiltonian Ho, which couples nodes
(xi,yk) and (xj,yl). The summation over kz is absent in the cylindrical geometry. Instead, a
summation over orbital quantum numbers is performed.

706 ATLAS User’s Manual

Quantum Transport: Non-Equilibrium Green’s Function Approach Quantum: Quantum Effect Simulator

Additionally, you can store energy dependent quantities as a spectrum at a certain coordinate
along transport direction or as a contour plot of energy and coordinate. There are two ways to
save spectrum or contour file :
1. Specify FILENAME= parameter on PROBE statement and set NEGF.LOG on the SAVE or
SOLVE statement. An incremental number will be added to the file name, each time a new
spectrum is saved. To save contour, specify FILENAME parameter on PROBE statement
and set NEGF.CONT on the SAVE or SOLVE statement.
2. To save spectrum log file, specify NEGF.LOGFILE= parameter on the SAVE or SOLVE
statement. To save contour, specify NEGF.CONTFILE= parameter on the SAVE or SOLVE
statement
The following quantities can be stored into spectrum or contour file:
• Transmission coefficient for m-th sub-band (only for spectrum file)

 R A   R 
T m  E  = tr  G  S G  D ,  S D = 2m   S D  13-45
   

You must set a TRANSMISSION parameter on the PROBE statement. It is followed by BAND=
and STATE= parameters to specify the effective mass band (valley) and the number of sub-
band respectively. If BAND= or STATE= is missing, a summation over this index is performed.
• Density of states in the m-th sub-band

 R 
DOS m  E x i  = 2m  G mmii  E   13-46
 

To save DOS into spectrum file, you must set a DOSVSE parameter on the PROBE statement. To
save DOS into a contour file, set DOS.CONT on the PROBE statement. It is followed by BAND=
, STATE= , and (only for spectrum file) X= parameters to specify the effective mass band
(valley), the number of sub-band, and the location along transport direction respectively. If
BAND= or STATE= is missing, a summation over this index is performed.
• Electron density per energy in m-th sub-band

 < 
N m  E x i  = m  G mmii  E   13-47
 

To save density into spectrum file, you must set a parameter DENSVSE on the PROBE
statement. To save density into a contour file, set DENS.CONT on the PROBE statement. It is
followed by BAND=, STATE=, and (only for spectrum file) X= parameters to specify the
effective mass band (valley), the number of sub-band, and the location along transport
direction respectively. If BAND= or STATE= is missing, a summation over this index is
performed.
• Current density per energy in the m-th sub-band

2e  < 
J m  E x i  = ------ e  h mmi + 1i G mmii + 1  E   13-48
h  

707 ATLAS User’s Manual

Quantum Transport: Non-Equilibrium Green’s Function Approach Quantum: Quantum Effect Simulator

To save current density into spectrum file, you must set a parameter CURDVSE on the PROBE
statement. To save density into a contour file, set CURD.CONT on PROBE statement. It is
followed by BAND=, STATE=, and (only for spectrum file) X= parameters to specify the
effective mass band (valley), the number of sub-band, and the location along transport
direction respectively.
The energy dependent quantities are stored on a uniform energy grid. You can specify the size
of the grid by setting a parameter ESIZEOUT.NEGF= on OUTPUT statement. If it is not
specified, the size will be the same as ESIZE.NEGF on the MODELS statement
A solution of NEGF equations is performed self-consistently with Poisson equation using a
predictor-corrector scheme. You can specify the number of predictor iterations using a
NPRED.NEGF= parameter on the MODELS statement. The criterion for convergence of Poisson
iterations is set by a QCRIT.NEGF= parameter on the MODELS statement.
By default, Dirichlet boundary conditions (BC) for potential and open-boundary condition for
electrons are used in the contacts. You can also use von Neumann boundary conditions for
potential by specifying REFLECT parameter on the CONTACT statement. This type of BC is
useful because ballistic carriers do not create voltage drop near the source and drain contacts.
Moreover, the potential in the source and drain must be allowed to float because the number
of electrons reflected backward and thus total carrier concentration in contacts can change
with bias. The third type of BC is applicable to Schottky contacts, where potential is fixed by
Dirichlet BC, but carriers are injected at all energies, mimicking a near-constant density of
states of metal contact near the Fermi level. To set Schottky BC, set NEGF.TUNNEL on the
CONTACT statement and also specify metal density of states by DOS.NEGF with the default of
0.1 1/eV. In the case of Schottky contacts, contact self-energy is given by
R 2
 c = i  t  DOS.NEGF 13-49

Finally, a few comments has to be made about NEGF model:

• NEGF is a fully quantum approach to electron transport, which takes into account the
wave nature of electron in both transport and transverse directions. This is the main
difference of this method and drift-diffusion or Boltzmann Transport equation, which
treat electron classically.
• In the present version, the transport is ballistic (i.e., no scattering occurs). Currently, this
model gives an upper limit of current. In future versions, electron-phonon scattering may
be added.
• The model is robust to predict subthreshold current, threshold voltage. The Poisson
convergence, however, may be unstable when a device in completely open because the
assumption of equilibrium contacts breaks down. In the long run, adding scattering will
mitigate this problem.
• A discretized effective-mass Hamiltonian has a finite bandwidth, which depends on the
2 2
mesh spacing in the transport direction and is given by h   2m  x  . A spacing in the
transport direction should be fine enough, so that a half-bandwidth of discretized effective
mass Hamiltonian is larger than applied bias. ATLAS will give a warning if the grid is
not fine enough.

708 ATLAS User’s Manual

Quantum Transport: Non-Equilibrium Green’s Function Approach Quantum: Quantum Effect Simulator

13.7.2 Planar NEGF approach

Most of the equations and syntax of the mode-space approach are also applicable to planar
devices. Here, we discuss only a few new parameters specific to planar approach. Planar
NEGF solver is based on 1D NEGF with effective mass Hamiltonian discretized in real space.
Current and carrier density are obtained by analytically integrating Green’s function over 2D
transverse k-space. The solution is done for each electron or hole band independently of all
other bands. Ballistic transport is assumed. In real devices, such as resonant tunneling diodes
(RTD), injection into the double barrier structure occurs from emitter quasi-bound states. To
model this effect without including inelastic phonon scattering, we treat emitter and collector
as quasi-equilibrium regions, while the central region with double barrier structure is treated
as non-equilibrium.
To launch the model, specify N.NEGF_PL1D or P.NEGF_PL1D or both on the MODELS
statement for electrons and holes respectively. This will tell ATLAS to solve 1D NEGF
equation in the first slice in transport direction and copy carrier and current density to all
other slices. If the slices are not equivalent, use N.NEGF_PL or P.NEGF_PL or both instead to
solve 1D NEGF equation for each slice separately.
To specify a region as quasi-equilibrium, use EQUIL.NEGF on the MODELS or the REGION
statements. It is a good idea to specify all regions in front of the first barrier and behind the
last barrier as quasi-equilibrium, so that all possible quasi-bound states are filled. Quasi-
equilibrium regions are characterized by a broadening, which can be set by ETA.NEGF
parameter on the MODELS or the REGION statements, with a default of 0.0066 eV.
The output quantities of interest are the IV characteristics, spectrum of transmission, DOS,
density and current, and also spatial profiles of carrier density and conduction or valence band
edge.
The solution of 1D NEGF equations starts with resonance finding in the whole device using
Quantum Transmitting Boundary Method (QTBM) and a construction of the energy grid. It is
the resolution of these resonances that is critical for correct simulation. The resonances can be
seen as extremely narrow peaks in the transmission vs. energy plot. Use log scale in
TONYPLOT to get a better understanding of the position of the transmission and DOS
resonances.
In case of poor convergence, energy grid size (ESIZE.NEGF on the MODELS statement) and
magnitude of the broadening in quasi-equilibrium regions have to be increased.

709 ATLAS User’s Manual

Drift-Diffusion Mode-Space Method (DD_MS) Quantum: Quantum Effect Simulator

13.8 Drift-Diffusion Mode-Space Method (DD_MS)

This model is a semiclassical approach to transport in devices with strong transverse
confinement and is a simpler alternative to mode-space NEGF approach (NEGF_MS) as
described earlier. Similar to NEGF_MS, the solution is decoupled into Schrodinger equation in
transverse direction and 1D transport equations in each sub-band. In this model, however, a
classical drift-diffusion equation is solved instead of a quantum transport equation. Thus, the
model captures quantum effects in transverse direction and yet inherits all familiar ATLAS
models for mobility, recombination, impact ionization, and band-to band tunneling. Unlike
NEGF_MS, the model can handle only devices with uniform cross-section because quantum
mechanical coupling between electron sub-bands is neglected. To activate the model, use
DD_MS together with SCHRO or P.SCHRO parameters or both on the MODELS statement. Note
that CARRIERS parameter on the METHOD statement should be set to 0 because multi-
dimensional ATLAS continuity equation solvers are not used in this method.
One-dimensional drift-diffusion transport in each sub-band, characterized by sub-band index
 and effective mass index b is described by

1 J
--- b = R b – G b 13-50
q x

E b n b
J b = q b n b – q Db 13-51
x x

where Jb(x) is sub-band current, nb(x) is sub-band carrier density to be found, and Eb(x) is
sub-band eigen energy as found by Schrodinger solver in transverse direction. The 1D
transport equation is discretized using Scharfetter-Gummel scheme. Mobility b(x) is
defined by ATLAS mobility models and is in general dependent on material, doping, local
longitudinal electric field, and temperature. Diffusion coeffiecient Db(x) is related to
mobility by
–1
 n b 
Db =  b   13-52
  E b

As Fermi-Dirac statistics is always assumed, the relation between sub-band carrier densities
(cm-3), quasi-fermi levels and eigen energies is given by the following relations:
kB T E b – E F b
n b = 2 ------------2-  m x m z ln 1 + exp  – ----------------------------
b b
- 13-53
Ah  kB T

in case of 1D confinement in y direction and

2k B T E b – E F b
n b = 2 -----------------  m x F –1  2  – ----------------------------
b

-

13-54
Ah kB T


in case of 2D confinement in y-z plane, where A is a normalizing area.

710 ATLAS User’s Manual

Drift-Diffusion Mode-Space Method (DD_MS) Quantum: Quantum Effect Simulator

Recombination rates Rb(x) are computed by taking into account recombination of a

particular electron (hole) sub-band with all hole (electron) sub-bands:
2
n b p b' – n i
R b =  Rbb'   ------------------------------
 bb'
- 13-55
b' b'

Here, Rbb’(x) are computed using ATLAS bulk recombination models with a substitution of
sub-band carrier densities for bulk densities, sub-band eigen energies for conduction and
valence bands, and sub-band fermi levels for bulk fermi levels. Unlike the rest of ATLAS,
recombination rates are set to 0 when only one carrier type is solved for.
Total impact ionization generation rates are computed by taking a sum over all electron and
hole sub-bands:
b b'
G total =   n J b +  p J b'  13-56
bb'

Impact ionization coefficients n and p are computed using ATLAS bulk impact ionization
models. Then, the generation rate in each sub-band is given by Gtotal, weighted by the carrier
population in the sub-band:
n b
G b = G total  ------------- 13-57
n total

p b'
G b' = G total  ------------- 13-58
p total

Unlike the rest of ATLAS, it is allowed to use impact ionization model with only one carrier
type. While one carrier type calculation may give incorrect answer, it is sometimes very
useful to understand the effect of electrons and holes separately.
When generation-recombination mechanisms are present, you can iterate between carrier
density and G-R rates before solving the Poisson equation. The self-consistency between
carrier density and G-R rates, achieved in this inner iteration procedure, will result in a more
stable Poisson convergence. To control the number of inner iterations, use the
RGITER.DDMS=N and the RGCONV.DDMS=X parameters on the METHOD statement. Here, N is the
maximum number of inner iterations (default is 0) and X is minimum error between 0 and 1
(default is 1e-5).
After 1D transport equations are solved in each sub-band and quasi-fermi levels are extracted,
bulk (3D) carrier densities and currents are computed by summing over all sub-bands and
weighting with corresponding wave function squared as done in Equation 13-13 or Equation
13-14. The same procedure is used to compute total generation and recombination rates,
which are stored in output structure file.
A usual predictor-corrector scheme is used to solve Poisson equation self-consistently.
Similar to NEGF_MS, Dirichlet boundary conditions for potential in contacts are default. Von
Neumann boundary conditions can be used by specifying REFLECT on the CONTACT
statement. Schottky contacts boundary conditions can be set by DDMS.TUNNEL parameter on
the CONTACT statement. In the case of Schottky contacts, potential is fixed but new generation

711 ATLAS User’s Manual

Drift-Diffusion Mode-Space Method (DD_MS) Quantum: Quantum Effect Simulator

terms will be added to the drift-diffusion equations. The local tunneling generation rate is
given by
* 1 + exp  E   x  – E F   x  
G TUN  x  = --------- F  x    x  ln  ---------------------------------------------------------------------
A T
 1 + exp  E   x  – E F contact 
, 13-59
k

where A* is Richardson constant, F(x) is the local electric field, and (x) is the tunneling
probability, computed within WKB approximation:

 x

   x  = exp  – ------  2m  E   0  – E   x'   dx'
 4
13-60
h
 0 

An output structure file contains quantities summed over all sub-bands. If you want detailed
information about 1D sub-band-resolved eigen energies, carrier densities, currents, quasi
fermi levels or generation-recombination rates, you can save an additional log file along with
master structure file by specifying the DDMS.LOG option on the SAVE or SOLVE statements. If
you specify DDMS.LOG on the SOLVE statement, then you must also set MASTER and OUTF. The
name of the extra file is the same as the corresponding structure file but ending with
"_ddms.log".

712 ATLAS User’s Manual

Chapter 14
TFT: Thin-Film Transistor
Simulator
Polycrystalline and Amorphous Semiconductor Models TFT: Thin-Film Transistor Simulator

14.1 Polycrystalline and Amorphous Semiconductor Models

TFT is an ATLAS module that simulates disordered material systems. TFT doesn’t contain
material models so you need to combine either S-PISCES or BLAZE with TFT to simulate these
material systems.
TFT enables you to define an energy distribution of defect states in the bandgap of
semiconductor materials. This is necessary for the accurate treatment of the electrical
properties of such materials as polysilicon and amorphous silicon.
The syntax used by TFT is a part of the ATLAS syntax so there’s no need for you to learn how
to use a new simulator to run TFT simulations.
Before continuing with this chapter, you should be familiar with ATLAS physics. If not, read
Chapters 2 “Getting Started with ATLAS” and 3 “Physics” before proceeding further. For
more information on S-PISCES and BLAZE, see Chapters 4 “S-Pisces: Silicon Based 2D
Simulator” and 5 “Blaze: Compound Material 2D Simulator”.

714 ATLAS User’s Manual

Simulating TFT Devices TFT: Thin-Film Transistor Simulator

14.2 Simulating TFT Devices

This section is intended to illustrate the basic building blocks for a thin-film transistor
simulation.
To use TFT, specify the addition of defect states into the bandgap of a previously defined
crystalline material. Throughout this section the example of polysilicon is used. Amorphous
silicon and other disordered materials are handled in a similar manner.

14.2.1 Defining The Materials

Use the ATLAS command syntax to define a simple polysilicon thin-film transistor structure.
The MESH, X.MESH, and Y.MESH statements can be used to construct a mesh as described in
Section 2.6 “Defining A Structure”. For more information on MESH statements, Sections
21.34 “MESH” and 21.2 “A.MESH, R.MESH, X.MESH, Y.MESH, Z.MESH”.
When defining the material regions, use the following syntax.
REGION Y.MIN=-0.05 Y.MAX=0 OXIDE
REGION Y.MIN=0 Y.MAX=0.2 SILICON
REGION Y.MIN=0.2 Y.MAX=2 OXIDE
Note that the region is defined as silicon. You can also define the material as polysilicon. But
note that it is the defect distribution rather than this initial material definition that will
determine the electrical characteristics.

14.2.2 Defining The Defect States

Disordered materials contain a large number of defect states within the band gap of the
material. To accurately model devices made of polycrystalline or amorphous materials, use a
continuous density of states. The DEFECT statement is used to specify the density of defect
states (DOS) as a combination of exponentially decaying band tail states and Gaussian
distributions of mid-gap states [133, 94]. In addition, you may need to model the grain-grain
boundary interface as a thermionic field emission boundary [136].

Note: The INTDEFECTS, INTERFACE S.S and TRAP.COULOMBIC models are avaiable for grain-grain boundary
interface modeling. These models have been developed in collaboration with Epson and Cambridge
University. For more information, see Sections 21.23 “INTDEFECTS”, 21.24 “INTERFACE”, and 21.37
“MODELS”.

14.2.3 Density of States Model

It is assumed that the total density of states (DOS) and g  E  , is composed of four bands: two
tail bands (a donor-like valence band and an acceptor-like conduction band) and two deep
level bands (one acceptor-like and the other donor-like) which are modeled using a Gaussian
distribution.

g  E  = g TA  E  + g TD  E  + g GA  E  + g GD  E  14-1

715 ATLAS User’s Manual

Simulating TFT Devices TFT: Thin-Film Transistor Simulator

Here, E is the trap energy, EC is the conduction band energy, EV is the valence band energy and
the subscripts (T, G,A, D) stand for tail, Gaussian (deep level), acceptor and donor states
respectively.
E – Ec
g TA  E  = NTA exp --------------- 14-2
WTA

Ev – E
g TD  E  = NTD exp --------------- 14-3
WTD

EGA – E 2
g GA  E  = NGA exp – -------------------- 14-4
WGA

E – EGD 2
g GD  E  = NGD exp – -------------------- 14-5
WGD

For an exponential tail distribution, the DOS is described by its conduction and valence band
edge intercept densities (NTA and NTD), and by its characteristic decay energy (WTA and WTD).
For Gaussian distributions, the DOS is described by its total density of states (NGA and NGD),
its characteristic decay energy (WGA and WGD), and its peak energy/peak distribution (EGA and
EGD). Table 14-1 shows the user-specifiable parameters for the density of defect states.

Table 14-1 User-Specifiable Parameters for Equations 14-2 to 14-5

Statement Parameter Default Units

DEFECTS NTA 1.121021 cm-3/eV

DEFECTS NTD 4.01020 cm-3/eV
DEFECTS NGA 5.01017 cm-3/eV
DEFECTS NGD 1.51018 cm-3/eV
DEFECTS EGA 0.4 eV
DEFECTS EGD 0.4 eV
DEFECTS WTA 0.025 eV
DEFECTS WTD 0.05 eV
DEFECTS WGA 0.1 eV
DEFECTS WGD 0.1 eV

716 ATLAS User’s Manual

Simulating TFT Devices TFT: Thin-Film Transistor Simulator

14.2.4 Trapped Carrier Density

The ionized densities of acceptor and donor like states (nT and pT respectively) are given by:
p T = p TA + p GA 14-6

n T = n TD + n GD 14-7

where nTA, nGA, pTD, and pGD are given below.

p TA =
 gTA  E   f t  E n p  dE
TA
14-8
EV

p GA =
 gGA  E   f t  E  n  p  dE
GA
14-9
EV

n TD =  gTD  E   f t  E n p  dE
TD
14-10
EV

n GD =  gGD  E   ft
GD
 E n p  dE 14-11
EV

f t  E  n p  and f tGA  E n p  are the ionization probabilities for the tail and Gaussian acceptor
TA

DOS, while ftTD  E n p  and ftGD  E n p  are the ionization probabilities for the donors.

In the steady-state case, the probability of occupation of a trap level at energy E for the tail
and Gaussian acceptor and donor states are given by Equations 14-12 through 14-15.
Ei – E
v n SIGTAE n + v p SIGTAH n i exp --------------
kT
f t  E n p  = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 14-12
TA E – Ei Ei – E
v n SIGTAE  n + n i exp --------------  + v p SIGTAH  p + n i exp -------------- 
 kT   kT 

Ei – E
v n SIGGAE n + v p SIGGAH n i exp --------------
kT
f t  E n p  = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 14-13
GA E – Ei Ei – E
v n SIGGAE  n + n i exp --------------  + v p SIGGAH  p + n i exp -------------- 
 kT   kT 

717 ATLAS User’s Manual

Simulating TFT Devices TFT: Thin-Film Transistor Simulator

E – E
v p SIGTDH p + v n SIGTDE n i exp ---------------
kT
f t  E n p  = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 14-14
TD E – Ei Ei – E
v n SIGTDE  n + n i exp --------------  + v p SIGTDH  p + n i exp -------------- 
 kT   kT 

E – Ei
v p SIGGDH p + v n SIGGDE n i exp --------------
kT
f t  E n p  = --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 14-15
GD E – Ei Ei – E
v n SIGGDE  n + n i exp --------------  + v p SIGGDH  p + n i exp -------------- 
 kT   kT 

where vn is the electron thermal velocity and vp is the hole thermal velocity, ni is the intrinsic
carrier concentration. SIGTAE and SIGGAE are the electron capture cross-section for the
acceptor tail and Gaussian states respectively. SIGTAH and SIGGAH are the hole capture cross-
sections for the acceptor tail and Gaussian states respectively and SIGTDE, SIGGDE, SIGGDH,
and SIGGDH are the equivalents for donors states.

Table 14-2 User-Specifiable Parameters for Equations 14-12 to 14-15

Statement Parameter Default Units

DEFECT SIGTAE 1.010-16 cm2

DEFECT SIGTDE 1.010-14 cm2
DEFECT SIGGAE 1.010-16 cm2
DEFECT SIGGDE 1.010-14 cm2
DEFECT SIGTAH 1.010-14 cm2
DEFECT SIGTDH 1.010-16 cm2
DEFECT SIGGAH 1.010-14 cm2
DEFECT SIGGDH 1.010-16 cm2

718 ATLAS User’s Manual

Simulating TFT Devices TFT: Thin-Film Transistor Simulator

14.2.5 Steady-State Trap Recombination

For steady-state conditions, the net recombination/generation rate is identical for electrons
(Rn) and holes (Rp) (i.e., instantaneous equilibrium). Using Equations 14-12 through 14-15 to
give the values of ft and following the derivation by Shockley and Read [258] and Hall [96],
the Shockley-Read-Hall recombination/generation rate due to the defect states is given by:
EC 
2
 v n v p SIGTAE SIGTAH  n p – n i  g TA  E 
R n p =
  ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
 
E – Ei
 
Ei – E

E V  v n SIGTAE  n + n i exp --------------  + v p SIGTAH  p + n i exp --------------  14-16
kT kT
2
v n v p SIGTGAE SIGGAH  n p – n i  g GA  E 
+ -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
E – Ei Ei – E

v n SIGGAE n + n i exp -------------- + v p SIGGAH p + n i exp --------------   
 kT   kT 
2
v n v p SIGTDE SIGTDH  n p – n i  g T'D  E 
+ ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
E – Ei Ei – E

v n SIGTDE n + n i exp -------------- + v p SIGTDH p + n i exp --------------   
 kT   kT 
2 
v n v p SIGGDE SIGGDH  n p – n i  g GD  E  
+ ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ dE
E – Ei Ei – E 
v n SIGGDE  n + n i exp --------------  + v p SIGGDH  p + n i exp --------------  
 kT   kT 

14.2.6 Transient Traps

For the transient case, time is required for carriers to be emitted or captured and therefore
instantaneous equilibrium cannot be assumed. This means that Equation 14-16 is no longer
valid for transient simulations. Instead, the total recombination/generation rate for electrons
(which is equal to electron recombination rate minus the generation rate for electrons) is
calculated using the transient probabilities of occupation for acceptors (ftTA and ftGA). These
are calculated by solving additional rate equations (Equations 14-17 and 14-18).
EC
E – Ei
g TA  E  v n SIGTAE  n  1 – f t  E   – f t  E  n i exp -------------- 

d-
---- p  = 14-17
dt TA TA TA kT
EV

Ei – E
– v p SIGTAH  p f t  E  –  1 – f t  E   n i exp --------------  dE
TA TA kT

EC
E – Ei
SIGGAE  n  1 – f t  E   – f t  E  n i exp -------------- 
 gGA E  vn
d
-----  p GA  = 14-18
dt  GA GA kT 
EV

Ei – E
– v p SIGGAH  p f t  E  –  1 – f t  E   n i exp --------------  dE
 GA GA kT 

719 ATLAS User’s Manual

Simulating TFT Devices TFT: Thin-Film Transistor Simulator

The total hole recombination/generation rate can also be determined from the transient values
of ftTD and ftGD (see Equations 14-19 and 14-20).
EC
Ei – E
g TD  E  v n SIGTDH  p  1 – f t  E   – f t  E  n i exp -------------- 

d
-----  n TD  = 14-19
dt  TD TD kT 
EV

E – Ei
– v n SIGTDE  n f t  E  –  1 – f t  E   n i exp --------------  dE
 TD TD kT 

EC
Ei – E
 gGD  E  vp SIGGDH  p  1 – f t  E   – f t  E  n i exp -------------- 
d
-----  n GD  = 14-20
dt  GD GD kT 
EV

E – Ei
– v n SIGGDE  n f t  E  –  1 – f t  E   n i exp --------------  dE
 GD GD kT 

A transient trap simulation using this model is more time consuming than using the static
model but gives a much more accurate description of the device physics. It may sometimes be
acceptable to perform transient calculations using the static trap distribution and assume that
traps reach equilibrium instantaneously. Specifying FAST on the DEFECTS statement will
neglect the trap rate equation from the simulation.
Trap-Assisted Tunneling
Trap-Assisted Tunneling models the trap-to-band phonon-assisted tunneling effects for Dirac
wells. At high electric fields, tunneling of electrons from the valence band to the conduction
band through trap or defect states can have an important effect on the current.
Trap-assisted tunneling is modeled by including field-effect enhancement terms [115]
(nDIRAC and pDIRAC) in the trap lifetimes in capture cross-sections. This model is enabled by
specifying TRAP.TUNNEL in the MODELS statement.
The electron capture cross-section (SIGTAE, SIGGAE, and SIGTDE, and SIGGDE) are modified
by including the electron field-effect term (nDIRAC). For example, the electron capture cross-
section for acceptor tail states (SIGTAE) becomes:
The field-effect enhancement term for electrons is given by:
DIRAC
SIGTAE   1 +  n  14-21

SIGGAE, SIGTDE, and SIGGDE are also modified this way.

1
En En 32
= ----------- exp  ----------- u – k n u  du

DIRAC
n 14-22
kT L  KT L 
0

720 ATLAS User’s Manual

Simulating TFT Devices TFT: Thin-Film Transistor Simulator

While the field-effect enhancement term for hole is:

1
Ep En 32
= ----------- exp  ----------- u – k p u  du

DIRAC
p 14-23
kT L  KT L 
0

The hole capture cross-sections (SIGTAH, SIGGAH, SIGTDH, and SIGGDH) are modified by
including the hole field effect term (nDIRAC). For example, SIGTAH now becomes:
DIRAC
SIGTAH  1 +  n 14-24

SIGGAH, SIGTDH, and SIGGDH are also modified this way.

where u is the integration variable, En is the energy range where tunneling can occur for
electrons, Ep is the tunneling energy range for holes, and kn and kp are given by:
3
2m 0 MASS · TUNNELEn
4
k n = --- ---------------------------------------------------------------------- 14-25
3 3qh E

3
2m 0 MASS · TUNNELEp
4
k p = --- ---------------------------------------------------------------------- 14-26
3 3qh E

 is the reduced Planck's constant, m0 is the rest mass of an electron, and MASS.TUNNEL is the
relative effective mass (You can specify MASS.TUNNEL by setting the MASS.TUNNEL
parameter in the MODELS statement).
Poole-Frenkel Barrier Lowering
The Poole Frenkel barrier lowering effect enhances the emission rate for trap-to-band
phonon-assisted tunneling and pure thermal emissions at low electric fields. The Poole
Frenkel effect occurs when the Coulombic potential barrier is lowered due to the electric field
and only occurs in traps with a Coulombic potential (such as traps that are neutral when
filled).
The Poole-Frenkel effect is modeled by including field-effect enhancement terms for
Coulombic wells (nCOUL and pCOUL) and thermal emission (F) [167] in the capture cross-
sections. The model also includes the trap-assisted tunneling effects in the Dirac well. To
enable this model, specify TRAP.COULOMBIC in the MODELS statement.
In the electron recombination/generation term (RN), the Coulombic barrier lowering term
(cF), and the electron Coulombic field-effect enhancement term (nCOUL) are applied to the
electron capture cross-sections.
For example, SIGTAE now becomes:
COUL
SIGTAE    F +  n  14-27

SIGGAE, SIGTDE, and SIGGDE are modified in the same manner.

721 ATLAS User’s Manual

Simulating TFT Devices TFT: Thin-Film Transistor Simulator

The hole capture cross-section are modified by including the Dirac field-effect enhancement
term for holes (nDIRAC).
For example, SIGTAH now becomes:
COUL
SIGTAE    F +  n  14-28

SIGGAH, SIGTDH, and SIGGDH are modified in the same manner.

In the hole recombination/generation term (Rp), the Coulombic terms are applied to the hole
capture cross-sections and the Dirac terms applied to the electron capture cross-sections.
The Poole-Frenkel thermal emission enhancement factor, F, is defined as:
E fp
 F = A · TRAPCOULOMBIC exp  ------------ 14-29
 kT  L

Efp is the barrier lowering term for a Coulombic well (see Equation 14-30).

3
B · TRAPCOULOMBICq E
E fp = --------------------------------------------------------------------- 14-30


The Coulombic field-enhancement terms, nCOUL and pCOUL, are defined as:
1
En En 32 E fp 5  3
exp  ----------- u – k p u 1 –  -------------  du

COUL
n = ----------- 14-31
kT L  KT L  uE n 
E fp
-----------
E n

1
E p E p 32 E fp 5  3
exp  ---------- u – k p u 1 –  -------------  du

COUL
p = ---------- 14-32
kT L  KT L  uE p 
E fp
-----------
E n

Table 14-3 User-Specificable Parameters for Equations 14-29 and 14-30

Statement Parameter Default Units

MATERIAL A.TRAPCOULOMBIC 1.0

MATERIAL B.TRAPCOULOMBIC 1.0

722 ATLAS User’s Manual

Simulating TFT Devices TFT: Thin-Film Transistor Simulator

14.2.7 Continuous Defects

If CONTINUOUS is specified in the DEFECTS statement, then the integral equations for the
charge and recombination are evaluated using a numerical integral scheme [252]. In this case,
the NUMA and NUMD (DEFECTS statement) parameters correspond to the number of acceptor
and donor energy level intervals used in the integral.

14.2.8 Discrete Defects

If CONTINUOUS is not specified on the DEFECTS statement, the equation is modeled with
discrete energy levels. The integrals terms in Equations 14-8 to 14-11 are replaced by
summations, which run over the number of discrete energy levels (NUMA and NUMD). The
acceptor and donor density of states terms are integrated separately. For example, the
equation for the electron trap concentration (Equation 14-6) is replaced by:
NUMA + +
 
 f  E  n p   
nT =  t TA i

gTA  E  dE + f 
t GA iE  n  p   g GA  E  d E


14-33
i=0 – –

Table 14-4 Additional Parameters for the DEFECTS Statement

Statement Parameter Default Units

DEFECT FAST FALSE

DEFECT CONTINUOUS FALSE

DEFECT NUMA 12

DEFECT NUMD 12

Syntax for a typical defect states definition is given below.

DEFECTS NTA=1.12E21 NTD=4.E20 WTA=0.025 WTD=0.05 \
NGA=5.E17 NGD=1.5E18 EGA=0.4 EGD=0.4 \
GA=0.1 WGD=0.1 SIGTAE=1.E-16 \
SIGTAH=1.E-14 SIGTDE=1.E-14 \
SIGTDH=1.E-16 SIGGAE=1.E-16 SIGGAH=1.E-14 \
SIGGDE=1.E-14 SIGGDH=1.E-16
Figure 14-1 shows how the syntax is used to define the peak density of states and distribution
widths for the two tail states and two Gaussian distributions.

723 ATLAS User’s Manual

Simulating TFT Devices TFT: Thin-Film Transistor Simulator

14.2.9 Amphoteric Defects

In hydrogenated amorphous silicon (a-Si:H), the effect of dangling-bonds states on
recombination can be significant. The dangling-bond states are amphoteric and located
around the middle of the bandgap. If you specify the AMPHOTERIC parameter on the DEFECTS
or INTDEFECTS statements, the dangling-bond density of states for a-Si:H will be calculated
using the defect-pool model [220, 221] as will the amphoteric defect density and
recombination [280].
The amphoteric defects are given by:
+ +
n AMP = D  E f AMP  E  14-34

0 0
n AMP = D  E f AMP  E  14-35

– –
n AMP = D  E f AMP  E  14-36

where:
+
• n AMP  E  is the occupied density of the positive amphoteric traps.
0
• n AMP  E  is the occupied density of the neutral amphoteric traps.
–
• n AMP  E  is the occupied density of the negative amphoteric traps.
+
• f AMP  E  is the probability of occupation of the positive amphoteric traps,
0
• f AMP  E  is the probability of occupation of the neutral amphoteric traps,
–
• f AMP  E  is the probability of occupation of the negative amphoteric traps.
• D(E) is the dangling-bond density of states given by Equation 14-38.
The total charge caused by the presence of amphoteric traps is subtracted from the right hand
side of Poisson’s equation. The total charge value is defined by:
EC

  nAMP  E  – nAMP  E   dE
+ –
Q TAMP = q 14-37
EV

724 ATLAS User’s Manual

Simulating TFT Devices TFT: Thin-Film Transistor Simulator

kT0 AMPHOTERIC ·
---------------------------------------------------------------------
-
 HCONC AMPHOTERIC 4EV0 · AMPHOTERIC
D  E  = NV0 · AMPHOTERIC  -----------------------------------------------------------
·  14-38
 NSISI · AMPHOTERIC

 2EV0 · AMPHOTERIC
2 
  --------------------------
---------------------------------------
 2EV0 · AMPHOTERIC – kT0 · AMPHOTERIC

2
1  SIGMA · AMPHOTERIC 
exp – --------------------------------------------------------  EP · AMPHOTERIC – E v – --------------------------------------------------------------
2EV0 · AMPHOTERIC  4EV0 · AMPHOTERIC 

kT0 AMPHOTERIC ·
---------------------------------------------------------------------
-
 2  2EV0 · AMPHOTERIC 1
  -----------------------  --------------------------------------------------------------------------
0
 f AMP  E  2  SIGMA · AMPHOTERIC
2
0

2 2
 E + SIGMA · AMPHOTERIC – EP AMPHOTERIC
--------------------------------------------------------------
 · 
 exp –  -------------------------------------------------------------------------------------------------------------------------------
-
2EV0 · AMPHOTERIC
 2 
 2SIGMA · AMPHOTERIC 
 

where:
• EP.AMPHOTERIC is the most probable potential defect energy.
• EV is the valence band edge.
• EV0.AMPHOTERIC is the characteristic energy.
• HCONC.AMPHOTERIC is the density of hydrogen.
• NSISI.AMPHOTERIC is the density of Si-Si bonds.
• NV0.AMPHOTERIC is the density of states of the valance-band tail exponential region
extrapolated to the valence-band edge.
• SIGMA.AMPHOTERIC is the defect pool width.
• T0.AMPHOTERIC is the freeze-in temperature.
0
• f AMP  E  is the equilibrium probability of occupation of the neutral amphoteric traps.
0

The probabilities of occupation can be calculated from the equilibrium probabilities of

+ 0 –
occupation ( f AMP  E  , f AMP  E  , and f AMP  E  ) and the capture and emission rates:
0 0 0

+ 1
f AMP  E  = ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 14-39
+ 0
e p + n n SIGNP · AMPHOTERIC e p + n n SIGN0 · AMPHOTERIC
1 + ---------------------------------------------------------------------------------------- 1 + ---------------------------------------------------------------------------------- -
0 –
e n + p p SIGP0 · AMPHOTERIC e n + p p SIGPN · AMPHOTERIC

725 ATLAS User’s Manual

Simulating TFT Devices TFT: Thin-Film Transistor Simulator

0 1
f AMP  E  = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 14-40
0 0
e n + p p SIGP0 · AMPHOTERIC e p + n n SIGN0 · AMPHOTERIC
1 + ---------------------------------------------------------------------------------------- + -----------------------------------------------------------------------------------
+ –
e p + n n SIGNP · AMPHOTERIC e n + p p SIGPN · AMPHOTERIC

– 0 +
f AMP  E  = 1 – f AMP  E  – f AMP  E  14-41

where:
• SIGN0.AMPHOTERIC is the electron capture cross-section for neutral defects.
• SIGNP.AMPHOTERIC is the electron capture cross-section for positive defects.
• SIGP0.AMPHOTERIC is the hole capture cross-section for neutral defects.
• SIGPN.AMPHOTERIC is the hole capture cross-section for negative defects.
0
• e n is the electron emission coefficient for neutral defects.
–
• e n is the electron emission coefficient for negative defects.
+
• e p is the hole emission coefficient for positive defects.
0
• e p is the hole emission coefficient for neutral defects.

+
0
n 0 f AMP  E  n SIGNP · AMPHOTERIC
0
en = -----------------------------------------------------------------------------------------------
- 14-42
0
f AMP  E 
0

0
–
n 0 f AMP  E  n SIGN0 · AMPHOTERIC
0
en = -----------------------------------------------------------------------------------------------
- 14-43
–
f AMP  E 
0

–
0
p 0 f AMP  E   p SIGPN · AMPHOTERIC
0
ep = ---------------------------------------------------------------------------------------------------
- 14-44
0
f AMP  E 
0

0
+
p 0 f AMP  E  p SIGP0 · AMPHOTERIC
0
ep = -----------------------------------------------------------------------------------------------
- 14-45
+
f AMP  E 
0

+ 1
f AMP  E  = ------------------------------------------------------------------------------------------------------------------------------------------------------
- 14-46
0
EF – ET 2E F – 2E T – EU · AMPHOTERIC
1 + 2 exp ------------------- + exp ---------------------------------------------------------------------------------
kT kT

EF – ET
2 exp -------------------
0 kT
f AMP  E  = ------------------------------------------------------------------------------------------------------------------------------------------------------- 14-47
0
EF – ET 2E F – 2E T – EU · AMPHOTERIC
1 + 2 exp ------------------- + exp ---------------------------------------------------------------------------------
kT kT

726 ATLAS User’s Manual

Simulating TFT Devices TFT: Thin-Film Transistor Simulator

– + 0
f AMP  E  = 1 – f AMP  E  – f AMP  E  14-48
0 0 0

where:
• n0 is the equilibrium electron concentration.
• p0 is the equilibrium hole0 concentration.
• EF is the Fermi energy.
• ET is the trap energy.
• EU.AMPHOTERIC is the defect electron correlation energy.
The recombination terms are given by:
+ + 0 0
R AMP = D  E   n n SIGNP · AMPHOTERICf AMP  E  – e n f AMP  E   14-49

– 0 – –
R AMP = D  E   n n SIGN0 · AMPHOTERICf AMP  E  – e n f AMP  E   14-50

Table 14-5 User-Specifiable Parameters for Equations 14-34-14-50

Statement Parameter Default Units

DEFECTS (INTDEFECTS) EP.AMPHOTERIC 1.27 eV

DEFECTS (INTDEFECTS) EU.AMPHOTERIC 0.2 eV
DEFECTS (INTDEFECTS) EV0.AMPHOTERIC 0.056 eV
DEFECTS (INTDEFECTS) HCONC.AMPHOTERIC 51021 (51017) cm-3 (cm-2)
DEFECTS (INTDEFECTS) NSISI.AMPHOTERIC 21023 (21019) cm-3 (cm-2)
DEFECTS (INTDEFECTS) NV0.AMPHOTERIC NV300 cm-3 (cm-2)
DEFECTS (INTDEFECTS) SIGMA.AMPHOTERIC 0.19 eV
DEFECTS (INTDEFECTS) SIGN0.AMPHOTERIC 1e-16 cm2
DEFECTS (INTDEFECTS) SIGNP.AMPHOTERIC 1e-16 cm2
DEFECTS (INTDEFECTS) SIGP0.AMPHOTERIC 1e-16 cm2
DEFECTS (INTDEFECTS) SIGPN.AMPHOTERIC 1e-16 cm2
DEFECTS (INTDEFECTS) T0.AMPHOTERIC 300 K

727 ATLAS User’s Manual

Simulating TFT Devices TFT: Thin-Film Transistor Simulator

14.2.10 Amphoteric Defect Generation

This model calculates the dangling bond density of amphoteric defects as a function of bias
stress time. If the parameter TIME.EP on the MODELS statement is specified, then the value of
EP.AMPHOTERIC in the amphoteric defect model will be replaced by Ep:

   EPBETA · AMPHOTERIC
Ep = EP · AMPHOTERIC exp  –  -------------------------------------------------------
TIME · EP
- 14-51
  EPT0 AMPHOTERIC 
 · 

where Ep is the stress time dependent most probable potential defect energy.

Table 14-6 User-Specifiable Parameters for Equation 14-51

Statement Parameter Type Default Units

MODELS TIME.EP Real 0 s

DEFECTS EPT0.AMPHOTERIC Real 1.3e6 s
DEFECTS EPBETA.AMPHOTERIC Real 0.5

14.2.11 Light Induced Defect Generation

This model calculates the dangling bond density of amphoteric defects as a function of light
[186]. The model is enabled by specifying TRAP.LID on the MODELS statement. The total
dangling bond density of amphoteric defects (NAMP) is given by:

2  KD · LID – 2KH · LID

N AMP = 3G TIME · LID  ----------------------------------------------------
- + N AMP 14-52
2
 SIG · LID  0

where
• N AMP is the initial total dangling bond density.
0

• G is the photogeneration rate.

• KD.LID is the light induced SiHD creation constant.
• KH.LID is the dissociation constant of a hydrogen atom from a SiHD defect.
• SIG.LID is the capture coefficient of free carriers by dangling bonds.
• TIME.LID is the time values used in the light induced defect generation model.
By default, the photogeneration rate (G) is calculated from all active BEAM statements. If you
specify the BEAM.LID parameter on the MODELS statement, then only active BEAM statements
that were defined with the parameter BEAM.LID will be used in the photogeneration rate
calculation. If you specify the parameter GENRATE.LID parameter on the MODELS statement,
then this value will be used as the photogeneration rate in Equation 14-52.

728 ATLAS User’s Manual

Simulating TFT Devices TFT: Thin-Film Transistor Simulator

Table 14-7 User-Specifiable Parameters for Equation 14-52

Statement Parameter Type Default Units

MODELS TIME.LID Real 0 s

MATERIAL KD.LID Real 5.0e-15 cm3 s-1
MATERIAL KH.LID Real 1.0e-18 cm3 s-1
MATERIAL SIG.LID Real 1.0e-8 cm3 s-1

14.2.12 Plotting The Density Of States Versus Energy

The DFILE and AFILE parameters were used to allow you to specify output file names for
capture of defect densities as a function of energy for donor states and acceptor states,
respectively. For example, if you want to look at the donor and acceptor defect distributions,
specify the following line:
DEFECTS DFILE=donors AFILE=acceptors
The files, donors and acceptors, could then be loaded into TONYPLOT to look at the
distributions of donor and acceptor defects as a function of energy.

14.2.13 Using the C-Interpreter to define DEFECTS

The C-INTERPRETER can be used to define the defect states in the bandgap. The F.TFTDON
and F.TFTACC parameters of the DEFECT statement indicate the filenames containing the C
functions. For more information on using the C-INTERPRETER, See Appendix A “C-
Interpreter Functions”.
DEFECTS F.TFTDON=mydefects.c F.TFTACC=mydefects.c
The file, mydefects.c, will contain C functions for donor and acceptor defect densities as a
function of energy. These user defined defects are added to the existing defect distribution. If
you want to use only your own function, set the gaussian and tail functions to zero. The
following example defect states are defined in the file, tft.lib. These are added to a zero
background set using the tail and gaussian state syntax. The resultant distribution of defects
versus energy can be plotted in the files, don.dat and acc.dat.
DEFECTS F.TFTDON=tft.lib F.TFTACC=tft.lib DFILE=don.dat
AFILE=acc.dat \
NTA=0 NTD=0 WTA=1.0 WTD=1.0 \
NGA=0 NGD=0 EGA=0.6 EGD=0.6 WGA=1 WGD=1 \
SIGTAE=1.E-16 SIGTAH=1.E-14 SIGTDE=1.E-14 SIGTDH=1.E-16
SIGGAE=1.E-16 SIGGAH=1.E-14 SIGGDE=1.E-14 SIGGDH=1.E-16

729 ATLAS User’s Manual

Simulating TFT Devices TFT: Thin-Film Transistor Simulator

14.2.14 Setting Mobility and Other Models

TFT uses adaptations of the standard models of S-PISCES or BLAZE. The example below
shows how to select the models and material parameters for polysilicon.
MATERIAL MUN=300 MUP=30
MODELS SRH
Typical mobility values for amorphous silicon can be set by:
MATERIAL MUN=20 MUP=1.5
Other models are also available in TFT. These include impact ionization and tunneling. Set
them by using:
MODELS BBT.STD IMPACT

Note: Don’t use concentration dependent mobility models (CONMOB, ANALYTIC, ARORA, KLA) because
they overwrite the low field mobilities set in the MATERIAL statement.

Figure 14-1: Syntax Guide to Define Two Tail States and Two Gaussian Distributions. NGA and NDG are the
integrated values of the Gaussian distributions. Gaussians are entered on energies EGA and EGD
respectively.

730 ATLAS User’s Manual

Chapter 15
Organic Display and Organic
Solar: Organic Simulators
Organic Device Simulation Organic Display and Organic Solar: Organic Simulators

15.1 Organic Device Simulation

Organic device simulation in the ATLAS framework is supported by two products: ORGANIC
DISPLAY and ORGANIC SOLAR. As implied by the names, these simulators are targeted at
specific applications requiring specific physical features. In the case of ORGANIC DISPLAY,
the simulator is set up to perform analysis of organic based active displays. This entails
simulation of principally organic FET devices, OFETs, and organic light emitting devices,
OLEDs. The ORGANIC SOLAR simulator is specific to light sensitive devices, particularly
solar cells. Organic based imaging sensors and the like may also be simulated using Organic
Solar.
The following sections describe each of the simulators in more detail. This is followed by
details of the physical models used in each case.

15.1.1 Organic Display

ORGANIC DISPLAY allows simulation of complex organic light emitting devices and circuits.
Much of this capability is discussed elsewhere in this manual. Before reading this section,
you should already be familiar with the general concepts discussed in Chapters 2 “Getting
Started with ATLAS” and 3 “Physics”. Chapter 2 “Getting Started with ATLAS” gives an
overview of the general concepts of building a simulation deck and using the simulator.
Chapter 3 “Physics” gives a fairly detailed discussion of the basic physical concepts and
models that define device modeling.
If you are interested in simulation of organic thin film transistors, OTFTs, then also read
Chapter 14 “TFT: Thin-Film Transistor Simulator”. The drift diffusion concepts discussed in
that chapter are completely anologous to those involved in simulation of OTFTs. Also if you
are interested in embedding physical devices in lumped element (SPICE) circuits, then read
Chapter 12 “MixedMode: Mixed Circuit and Device Simulator”. If you are interested in
simulating organic light emitting devices, OLEDs, then also read Chapter 11 “LED: Light
Emitting Diode Simulator”. This chapter discusses the physical concepts for simulation of
light emitting devices in general including OLEDs. Finally, those interested may also want to
look over Chapter 20 “Numerical Techniques” where they will be given a description of the
workings of the numerics of the simulator including information on convergence.
In this section, we will lead you through the important differences that you will need to know
about when simulating organic display devices. We will do this in a sequential manner similar
to the description of deck construction presented in Chapter 2 “Getting Started with ATLAS”.
First, with respect to definition of the device structure, there is little difference with what is
described in Chapter 2 “Getting Started with ATLAS”. The main thing you should be aware
of is that there are several default material names that are recognized by ORGANIC DISPLAY.
These materials are Organic, Pentacene, Alq3, TPD, PPV, Tetracene, and NPB. For
insulating dielectrics and conductive materials, see Appendix B “Material Systems”.
If you want to simulate a material not included in the default set recognized by ORGANIC
DISPLAY, you should follow the steps described in Section B.2.5 “Specifying Unknown or
User-Defined Materials in ATLAS”.
After you have specified the device structure, you may decide how to handle interfaces
between homogenous layers. These interfaces are analogous to heterojunctions in the drift
diffusion formulation where the band edges are replaced by the HOMO and LUMO.

732 ATLAS User’s Manual

Organic Device Simulation Organic Display and Organic Solar: Organic Simulators

You may choose to make the interfaces obey the physics of thermionic emission and
tunneling. To do this, specify the INTERFACE statement with the parameters THERMIONIC
and TUNNEL set. There are various ways in which characteristics may be assigned to a
particular interface geometrically using principally the X.MIN, X.MAX, Y.MIN, and Y.MAX
parameters. These types of interfaces are discussed in more detail in Section 5.2.2 “The
Thermionic Emission and Field Emission Transport Model”.
Next, you may want to fix the workfunction of electrodes to a certain value relative to the
LUMO and HOMO. To do this, use the CONTACT statement where the workfunction is
specified with the WORKF parameter. A particular CONTACT statement is associated with an
electrode using the NAME or NUMBER parameters. Also, you can assign various Schottky
characteristics to the electrode using, for example, the SURF.REC parameter to enable surface
recombination or BARRIER to enable barrier lowering. For more details, see Section 21.6
“CONTACT”.
Key parameters in the simulation of organic devices are the energies of the LUMO and
HOMO. In ORGANIC DISPLAY, these are set (indirectly) on the MATERIAL statement using the
EG300 and AFFINITY parameters. The LUMO energy coresponds to the negative of the value
specified by the AFFINITY parameter. The HOMO energy coresponds to the negative of the
sum of the values of the AFFINITY and EG300 parameters.
Defect states can be specified in organic materials using the ODEFECT and OINTDEFECT
statements. These statements are described in Section 15.2.2 “Organic Defects Model”. The
organic defects are important to correctly model defect recombination (Equation 15-28),
hopping mobility (Equations 15-31 and 15-32), and dopant light emission rates (see Section
15.3.1 “Dopants”).
You should also enable Poole-Frenkel mobility modeling in organic devices by specifying
PFMOB on the MODEL statement. This should be done in conjunction with the Langevin
recombination model enabled by the LANGEVIN parameter of the MODEL statement. Langevin
recombination is needed to enable exchange between charged carriers and singlet and triplet
excitons as described in Equations 15-49 and 15-76. To enable the singlet and triplet exciton
rate equations, specify SINGLET and TRIPLET on the MODELS statement.
Radiative emission can come from two sources: the host material singlets and the dopants.
The dopant rate equation is given in Equation 15-79 and is enabled by specifying DOPANT on
the ODEFECT statement. ORGANIC DISPLAY cannot predict the output spectra so you must
supply it. These are in the form of tabular descriptions of intensity vs. wavelength as
described in Section 11.7.1 “User-Definable Emission Spectra”. To connect materials and
their emission spectra, assign the USER.SPECT and DOPE.SPECT parameters of the MATERIAL
statement to the file names containing the appropriate tables. The spectra need not be
normalized as this is done automatically. To output the spectra to TONYPLOT compatible log
files, specify OUT.USPEC and OUT.DSPEC.
The normalized emission spectra are scaled by the radiative emission rates given in Equations
15-49 and 15-79. The PHEFF.EXCITON and DPHEFF.EXCITON parameters are quantum
efficiencies for host and dopant emission and act to scale the emission rates. The emission
spectra are critical in calculation through reverse ray tracing (see Section 11.5 “Reverse Ray-
Tracing”) of output coupling and output emisson spectrum. Also for reverse ray tracing, you
should accurately specify the complex indices of refraction for the various materials. This can
be done as described in Section 10.8 “Defining Optical Properties of Materials”.

733 ATLAS User’s Manual

Organic Device Simulation Organic Display and Organic Solar: Organic Simulators

Finally, a variety of useful information can be output after each solution using the LED
statement (see Section 11.5 “Reverse Ray-Tracing” and Section 21.27 “LED”). These
information include the angular distribution of emission spectra and output coupling.

15.1.2 Organic Solar

ORGANIC SOLAR allows simulation of complex organic light sensitive devices such as solar
cells and imaging sensors. Much of this capability is discussed elsewhere in this manual.
Before reading this section, you should already be familiar with the general concepts
discussed in Chapters 2 “Getting Started with ATLAS” and 3 “Physics”. Chapter 2 “Getting
Started with ATLAS” gives an overview of the general concepts of building a simulation deck
and using the simulator. Chapter 3 “Physics” gives a fairly detailed discussion of the base
physical concepts and models that define device modeling.
To understand the general concepts of photdetection, read Chapter 10 “Luminous:
Optoelectronic Simulator”. If you are interested in embedding physical devices in lumped
element (SPICE) circuits, then read Chapter 12 “MixedMode: Mixed Circuit and Device
Simulator”. Those interested may also want to look over Chapter 20 “Numerical Techniques”
where they will be given a description of the workings of the numerics of the simulator
including information on convergence.
In this section, we will lead you through the important differences you will need to know
about when simulating organic display devices. We will do this in a sequential manner similar
to the description of deck construction presented in Chapter 2 “Getting Started with ATLAS”.
First with respect to definition of the device structure, there is little difference with what is
described in Chapter 2 “Getting Started with ATLAS”. You should be aware of is that there
are several default material names that are recognized by ORGANIC DISPLAY. These materials
are Organic, Pentacene, Alq3, TPD, PPV, Tetracene, and NPB. For insulating dielectrics
and conductive materials, see Appendix Appendix B, “Material Systems,”.
If you want to simulate a material not included in the default set recognized by ORGANIC
SOLAR, you should follow the steps described in Section B.2.5 “Specifying Unknown or
User-Defined Materials in ATLAS”.
After you have specified the device structure, you may decide how to handle interfaces
between homogenous layers. These interfaces are analogous to heterojunctions in the drift
diffusion formulation where the band edges are replaced by HOMO and LUMO.
As such, you may choose to make the interfaces obey the physics of thermionic emission and
tunneling. To do this, specify the INTERFACE statement with the parameters THERMIONIC
and TUNNEL set. There are various ways in which interfaces may be specified to a particular
interface using principally the X.MIN, X.MAX, Y.MIN, and Y.MAX parameters. These types of
interfaces are discussed in more detail in Section 5.2.2 “The Thermionic Emission and Field
Emission Transport Model”.
Next, you may want to fix the workfunction of electrodes to a certain value relative to the
LUMO and HOMO. To do this, use the CONTACT statement where the workfunction is
specified with the WORKF parameter. A particular CONTACT statement is associated with an
electrode using the NAME or NUMBER parameters. Also, you can assign various Schottky
characteristics to the electrode using, for example, the SURF.REC parameter to enable surface
recombination or BARRIER to enable barrier lowering. For more details, see Section 21.6
“CONTACT”.

734 ATLAS User’s Manual

Organic Device Simulation Organic Display and Organic Solar: Organic Simulators

Key parameters in the simulation of organic devices are the energies of the LUMO and
HOMO. In ORGANIC SOLAR, these are set (indirectly) on the MATERIAL statement using the
EG300 and AFFINITY parameters. The LUMO energy coresponds to the negative of the value
specified by the AFFINITY parameter. The HOMO energy coresponds to the negative of the
sum of the values of the AFFINITY and EG300 parameters.
Defect states can be specified in organic materials using the ODEFECT and OINTDEFECT
statements. These statements are described in Section 15.2.2 “Organic Defects Model”. The
organic defects are important to correctly model defect recombination (Equation 15-28),
hopping mobility (Equations 15-31 and 15-32), and dopant light emission rates (see Section
15.3.1 “Dopants”).
You should also enable Poole-Frenkel mobility modeling in organic devices by specifying
PFMOB on the MODEL stateement. This should be done in conjunction with the Langevin
recombination model enabled by the LANGEVIN parameter of the MODEL statement. Langevin
recombination is needed to enable exchange between charged carriers and singlet and triplet
excitons as described in Equations 15-49 and 15-76. To enable the singlet and triplet exciton
rate equations, specify SINGLET and TRIPLET on the MODELS statement.
Photodetection in organic photo detectors is a two part process. First, absorbed photons
generate singlet excitons. The QE.EXCITON parameter of the MATERIAL statement (see
Section 15.3.2 “Light Generation of Excitons”) specifies the efficiency of this conversion.
Then, excitons are generated and they diffuse to interfaces in the "blend" material where they
experience dissociation as described in Section 15.3.3 “Exciton Dissociation”. Once electrons
and holes dissociate from the singlets, they experience built-in electric fields. They are then
separated and detected at the electrodes. To enble dissociation, specify S.DISSOC on the
MODELS statement.

735 ATLAS User’s Manual

Organic Materials Physical Models Organic Display and Organic Solar: Organic Simulators

15.2 Organic Materials Physical Models

This section describes the model definition required for the simulation of organic devices.

15.2.1 Gaussian Band Structure

In organic semiconductors the density of states in the conduction (LUMO) and valence
(HOMO) levels are often described by a Gaussian function [286].
2
NTC.GAUSS   E – EC  
g C  E – E C  = ----------------------------------------------- exp  – -----------------------------------------
- 15-1
2SIGC.GAUSS  2SIGC.GAUSS2

2
NTV.GAUSS   EV – E  
g V  E V – E  = ----------------------------------------------- exp  – -----------------------------------------
- 15-2
2SIGV.GAUSS  2SIGV.GAUSS  2

EC and EV are, respectively, the centers of the conduction and valence Gaussian DOS. Using
the Fermi-Dirac occupation probability, Equation 3-25, the electron density in the conduction
band and the hole density in the valence band are given by

n  EF  =
– gC  E – EC f  E – EF  dE 15-3


p  EF  =
– gV  EV – E   1 –f  E – EF   dE 15-4

These equations are written in the same form as Equations 3-27 and 3-28.
EF – EC
n  E F  = N C G C  ------------------- 15-5
 k T  B

EV – EF
p  E F  = N V G V  ------------------- 15-6
 k T  B

Here, NC and NV are effective band-edge density of states and GC and GV are Gaussian
equivalents of the Fermi-Dirac integral. GC and GV have to be evaluated numerically.
An equilibrium intrinsic carrier density and intrinsic Fermi level can be found for the
Gaussian band structure where
n  E Fi  = p  E Fi  = n i 15-7

The effective band-edge density of states are chosen so that Equations 3-34 and 3-35 are still
valid. The parameters to define a Gaussian DOS are in Table 15-1.

Table 15-1 Guassian DOS Parameters

Statement Parameter Default Units

MATERIAL NTC.GAUSS NTV.GAUSS/1.0e21 cm-3

MATERIAL SIGC.GAUSS SIGV.GAUSS/0.1 eV

736 ATLAS User’s Manual

Organic Materials Physical Models Organic Display and Organic Solar: Organic Simulators

Table 15-1 Guassian DOS Parameters

MATERIAL NTV.GAUSS NTC.GAUSS/1.0e21 cm-3
MATERIAL SIGV.GAUSS SIGC.GAUSS/0.1 eV
MATERIAL CRELMIN.N 1e-10
MATERIAL CRELMIN.P 1e-10
MATERIAL CRELMAX.N 0.5
MATERIAL CRELMAX.P 0.5
MATERIAL G3MAX.N 100.0
MATERIAL G3MAX.P 100.0

To use a Gaussian band structure, you need to specify at least one of NTC.GAUSS,
SIGC.GAUSS, NTV.GAUSS, or SIGC.GAUSS on a MATERIAL statement.
If a Gaussian DOS is used for one band, then it will automatically be used for the other band.
If a parameter is defined for one band, then that value becomes the default, if the
corresponding parameter is not explicitly set, for the other band. For example, setting the
conduction band to
MATERIAL NTC.GAUSS = 2.0e22 SIGC.GAUSS = 0.15
will have Nt = 2.0e22 and  = 0.15 for a Gaussian valence band as well.
Only a single parameter is required to activate a Gaussian band structure. For example
MATERIAL NTC.GAUSS = 2.0e22
The numerical defaults are used if a parameter is not explicitly given for either band.
In the Drift-Diffusion model, the diffusion coefficient is
 n
D = --- -------------------- 15-8
q dn  dE F

With Parabolic band structure and Boltzmann statistics

dn n-
---------- = -------- 15-9
dE F kB T

and we get the Einstein diffusion coefficient

k B T
D Einstein = ------------
- 15-10
q
With Gaussian bands, we write the diffusion coefficent as
D = g 3 D Einstein 15-11

with
1 n
g 3 n = --------- -------------------- 15-12
k B T dn  dE F

737 ATLAS User’s Manual

Organic Materials Physical Models Organic Display and Organic Solar: Organic Simulators

and
1 p
g 3 p = --------- -------------------- 15-13
k B T dp  dE F

These equations hold for

n
CRELMIN.N  ------------------------------  CRELMAX.N 15-14
NTC.GAUSS
and
p
CRELMIN.P  ------------------------------  CRELMAX.P 15-15
NTV.GAUSS
If the normalized carrier density is outside these limits, then g3 is evaluated at the appropriate
limit. G3MAX.[NP] is the maximum value that g3 is allowed to have.

15.2.2 Organic Defects Model

Density of States
The ODEFECTS statement is used to specify the energy distribution of defect states. The total
density of states (DOS) is composed of an acceptor-like conduction band DOS (gA(E)) and a
donor-like valence band DOS (gD(E)). There are two defect distribution models available.
The first model defines the defects based on a characteristic temperature and density [253].
E –Ec
g A  E  = -------------- exp  ------------------
HA
15-16
kTCA   kTCA 

Ev –E
g D  E  = -------------- exp  ------------------
HD
15-17
kTCD   kTCD 

HA is the acceptor traps density of states, HD is the donor traps density of states, TCA is the
acceptor traps characteristic temperature, TCD is the donor traps characteristic temperature,
and E is the energy.

Table 15-2 User-Specifiable Parameters for Equations 15-16 and 15-17

Statement Parameter Default Units

ODEFECTS HA 0 cm-3
ODEFECTS HD 0 cm-3
ODEFECTS TCA 0 K
ODEFECTS TCD 0 K

The trapped state concentrations for electrons (nA) and holes (pD) are given by:

738 ATLAS User’s Manual

Organic Materials Physical Models Organic Display and Organic Solar: Organic Simulators

nA =
 gA  E   ftA  E n p  dE 15-18
Ev

pD =
 gD  E   ftD  E n p  dE 15-19
Ev

where ftA(E,n,p) is the acceptor probability of occupation and ftD(E,n,p) is the donor
probability of occupation.
The second model is the effective transport hopping energy model and defines the defects
using a double peak Gaussian distribution [14],[15].
2 2
NIA   E – Ec   NA   E – E c + EA  
g A  E  = ----------------------------------------- exp  --------------------------------+ ------------------------------------- exp  ------------------------------------- 15-20
2
2 SIGMAIA  2SIGMAIA  2 SIGMAA  2SIGMAA2 

2 2
NID   Ev – E   ND   E v – E + ED  
g D  E  = ----------------------------------------- exp  --------------------------------+ ------------------------------------- exp  ------------------------------------- 15-21
2
2 SIGMAID  2SIGMAID  2 SIGMAD  2SIGMAD2 

NIA is the total intrinsic density for acceptor-like traps. NID is the total intrinsic density for
donor-like traps. SIGMAIA is the Gaussian width for the intrinsic acceptor-like traps. SIGMAID
is the Gaussian width for the intrinsic donor-like traps. NA is the total doping density for
acceptor-like traps. ND is the total doping density for donor-like traps. SIGMAA is the Gaussian
width for the doping acceptor-like traps. SIGMAD specifies the Gaussian width for the doping
donor-like traps. EA is the energy shift between the intrinsic and doping states for acceptor-
like traps. ED is the energy shift between the intrinsic and doping states for donor-like traps.
The trapped state concentrations are given by:
E tr
n

nA =
 gA  E FtA  E n p dE 15-22
Ev

pD =
 gD  E FtD  E n p  dE 15-23
E tr
p

where E tr is the effective transport energy for electrons and E tr is the effective transport
n p

energy for holes. You must specify the HOPPING parameter on the ODEFECTS statement to use
the effective transport hopping energy model. The effective transport energy for electrons and
holes is calculate by solving Equations 15-20 and 15-21 along with Equations 15-22 and 15-
23.

739 ATLAS User’s Manual

Organic Materials Physical Models Organic Display and Organic Solar: Organic Simulators

E tr
n

 gA  E   Etr
3 6HOPN · BETA 3
– E  dE = ------------------------------------
-  HOPN · GAMMAkT  15-24
n 
–

E tr
p

 gD  E   Etr
3 6HOPP · BETA 3
– E  dE = ------------------------------------
-  HOPP · GAMMAkT  15-25
p 
–

where HOPN.BETA is the electron percolation constant, HOPP.BETA is the hole percolation
constant, HOPN.GAMMA is 1/carrier localization radius for electrons and HOPP.GAMMA is 1/
carrier localization radius for holes.

Table 15-3 User-Specifiable Parameters for Equations 15-20-15-25

Statement Parameter Value Units

ODEFECTS EA 0.0 eV
ODEFECTS ED 0.0 eV
MATERIAL HOPN.BETA 1.5
MATERIAL HOPN.GAMMA 5.0107 cm-1
MATERIAL HOPP.BETA 1.5
MATERIAL HOPP.GAMMA 5.0107 cm-1
ODEFECTS HOPPING False
ODEFECTS NA 0.0 cm-3
ODEFECTS ND 0.0 cm-3
ODEFECTS NIA 0.0 cm-3
ODEFECTS NID 0.0 cm-3
ODEFECTS SIGMAA 0.0 eV
ODEFECTS SIGMAD 0.0 eV
ODEFECTS SIGMAIA 0.0 eV
ODEFECTS SIGMAID 0.0 eV

In steady-state, the probabilities of occupation are given by:

740 ATLAS User’s Manual

Organic Materials Physical Models Organic Display and Organic Solar: Organic Simulators

Ei – E
v n SIGAEn + v p SIGAHn i exp --------------
kT L
f tA = -----------------------------------------------------------------------------------------------------------------------------------------------------------
- 15-26
 E – Ei  Ei – E
v n SIGAE n + n i exp --------------  + v p SIGAH  p + n i exp -------------- 
 kT L   kT L 

Ei – E
v n SIGDEn + v p SIGDHn i exp --------------
kT L
f tD = ----------------------------------------------------------------------------------------------------------------------------------------------------------
- 15-27
E – Ei Ei – E
v n SIGDE n + n i exp  --------------  + v p SIGDH p + n i exp  -------------- 
 kT L   kT L 

where vn is the electron thermal velocity, vp is the hole thermal velocity, ni is the intrinsic
carrier concentration, SIGAE and SIGAH are the acceptor electron and hole capture cross-
sections respectively and SIGDE and SIGDH are the donor electron and hole capture cross-
sections respectively.

Table 15-4 User-Specifiable Parameters for Equations 15-26 and 15-27

Statement Parameter Default Units

ODEFECTS SIGAE 10-20 cm2

ODEFECTS SIGAH 10-20 cm2
ODEFECTS SIGDE 10-20 cm2
ODEFECTS SIGDH 10-20 cm2

Steady-State and Transient Recombination

The steady-state net recombination/generation rate is identical for electrons (Rn) and holes
(Rp) as equilibrum is assumed to be instantaneous. The Schockley-Read-Hall recombination/
generation rate due to the defect states is given by:

Ec 2
v n v p SIGAE SIGAH  np – n i g A  E 
R n p =
 -----------------------------------------------------------------------------------------------------------------------------------------------------------
 E – Ei  Ei – E
- 15-28
Ev v n SIGAE  n + n exp --------------  + v p SIGAH  p + n i exp -------------- 
 i kT L   kT L 

2
v n v p SIGDE SIGDH  np – n i g D  E 
+ ----------------------------------------------------------------------------------------------------------------------------------------------------------
- dE
E – Ei Ei – E
v n SIGDE  n + n i exp  --------------  + v p SIGDH  p + n i exp  -------------- 
kT  kT 

741 ATLAS User’s Manual

Organic Materials Physical Models Organic Display and Organic Solar: Organic Simulators

In transient simulations, the instantaneous equilibrium can no longer be assumed. Instead, the
recombination/generation rate is calculated using the transient acceptor and donor
probabilities of occupation. These are calculated by solving additional rate equations:
Ec
E – Ei
 gA  E  vn SIGAE n  1 – ftA  E   –ftA  E ni exp -------------- 
d
-----  n A  =
dt kT L 
Ev 15-29
E – Ei
– v p SIGAH  pf tA  E  –  1 – f tA  E  n i exp --------------  dE
 kT L 

Ec
E – Ei
 gD  E  vn SIGDE n  1 – ftD  E   –ftD  E ni exp -------------- 
d
-----  p D  = 15-30
dt kT L 
Ev
E – Ei
– v p SIGDH  pf tD  E  –  1 – f tD  E  n i exp --------------  dE
 kT L 

Additional Parameters
If the parameter FAST is specified in the ODEFECTS statement, then the traps are assumed to
reach equilibrum instantaneously. Therefore, Equations 15-29 and 15-30 will not be solved.
In this case, the steady-state probabilities of a occupation will be used in transient
simulations.
You can use the C-Interpreter to define the defect states in the bandgap. The F.OTFTDON and
F.OTFTACC parameters of the ODEFECTS statement indicate the filenames containing the C
functions. For more information on using the C-Interpreter, see Appendix A “C-Interpreter
Functions”.
ODEFECTS F.OTFTDON=mydefects.c F.OTFTACC=mydefects.c
The file, mydefects.c, will contain C functions for donor and acceptor defect densities as a
function of energy. The following example defect states are defined in the file, otft.lib.
The resultant distribution of defects versus energy can be plotted in the files, odon.dat and
oacc.dat.
ODEFECTS F.OTFTDON=otft.lib F.OTFTACC=otft.lib DFILE=odon.dat
AFILE=oacc.dat \
SIGAE=1.E-16 SIGAH=1.E-14 SIGDE=1.E-14 SIGDH=1.E-16
If you specify the CONTINUOUS parameter in the ODEFECTS statement, then a numerical
integral will be used to calculate the charge and recombination integrals. In this case, you can
use the parameters NUMA and NUMD to specify the number of acceptor and donor energy levels
intervals used in the integral. If CONTINUOUS is not specified, then the integrals will be
modelled with discrete energy levels. The integral terms are replaced by summations, which
run over the number of discrete energy levels (NUMA for acceptors and NUMD for donors).

742 ATLAS User’s Manual

Organic Materials Physical Models Organic Display and Organic Solar: Organic Simulators

Table 15-5 Additional Parameters for the ODEFECTS statement

Statement Parameter Default Units

ODEFECTS AFILE

ODEFECTS CONTINUOUS False

ODEFECTS DFILE

ODEFECTS F.OTFTACC

ODEFECTS F.OTFTDON

ODEFECTS FAST False

ODEFECTS NUMA 12
ODEFECTS NUMD 12

15.2.3 Hopping Mobility Model

If the effective transport hopping energy model is specified (HOPPING on the ODEFECTS
statement), then the field-independent hopping mobility can be used (HOPMOB on the MODELS
statement). This is given by [15]:
E tr
n
–2  3 E tr
n
–1  3
3 HOPN · BETA- 1  3
 g A  E  dE exp –2  ------------------------------------
 gA  E  dE
q HOPN · V0-
 n0 = ---------------------------- 
HOPN GAMMA
·
15-31
kT 4
– –

E tr
p
–2  3 E tr
p
–1  3
3 HOPP · BETA- 1  3
 g D  E  dE exp – 2  ------------------------------------
 g D  E  dE
q HOPP · V0-
 p0 = ----------------------------   HOPP GAMMA
·
15-32
kT 4
– –

where HOPN.V0 is the attempt-to-jump frequency for electrons and HOPP.V0 is the attempt-
to-jump frequency for holes.

Table 15-6 User-Specifiable Parameters for Equations 15-31 and 15-32

Statement Parameter Value Units

MODEL HOPMOB False

MOBILITY HOPMOB.N False
MOBILITY HOPMOB.P False
MATERIAL HOPN.V0 1.0e11 Hz
MATERIAL HOPP.V0 1.0e11 Hz

743 ATLAS User’s Manual

Organic Materials Physical Models Organic Display and Organic Solar: Organic Simulators

15.2.4 Poole-Frenkel Mobility Model

The Poole-Frenkel mobility model is given by [85][111]:

 n  E  =  n0 exp  – -----------------------------------------------
DELTAEN · PFMOB  BETAN · PFMOB
- + ---------------------------------------- – GAMMAN · PFMOB E 15-33
PF  kT neff  kT neff  

 p  E  =  p0 exp  – -----------------------------------------------
DELTAEP · PFMOB  BETAP · PFMOB
- + ---------------------------------------- – GAMMAP · PFMOB E 15-34
PF  kT peff  kT peff  

where  n  E  and  p  E  are the Poole-Frenkel mobilities for electrons and holes
PF PF

respectively, n0 and n0are the zero field mobilities for electrons and holes respectively, and
E is the electric field. DELTAEN.PFMOB and DELTAEP.PFMOB are the activation energy at zero
electric field for electrons and holes respectively. BETAN.PFMOB is the electron Poole-Frenkel
factor, and BETAP.PFMOB is the hole Poole-Frenkel factor.
Tneff is the effective temperature for electrons and is defined as:

T0N · PFMOB T
T neff = ------------------------------------------L- 15-35
T0N · PFMOB – T L

Tpeff is the effective temperature for holes and is defined as:

T0P · PFMOB T
T peff = ------------------------------------------L- 15-36
T0P · PFMOB – T L

If you specify the BETANAUTO.PFMOB parameter on the MOBILITY statement, then

BETAN.PFMOB in Equation 15-33 will be calculated from the permittivity:

q
BETAN · PFMOB = q ------------ 15-37
0

Similarly, if you specify BETAPAUTO.PFMOB, then BETAP.PFMOB in Equation 15-34 will be

calculated from the permittivity.
If you use E0N.PFMOB, E0P.PFMOB, E0NAUTO.PFMOB or E0PAUTO.PFMOB instead of
BETAN.PFMOB, BETAP.PFMOB, BETANAUTO.PFMOB and BETAPAUTO.PFMOB, you will then use
a simplified form of the Poole-Frenkel mobility [111][239]:

 E 
 n  E  =  n0 exp  – DELTAEN · PFMOB- + --------------------------------
----------------------------------------------- - 15-38
PF
 kT neff E0N · PFMOB

 DELTAEP PFMOB E 
 p  E  =  p0 exp  – -----------------------------------------------
· -
- + -------------------------------- 15-39
PF
 kT peff E0P · PFMOB

where E0 is the characteristic field. You can use the parameters E0NAUTO.PFMOB and
E0PAUTO.PFMOB in the MOBILITY statement to calculate the electron and hole characteristic
field values from the permittivity:

744 ATLAS User’s Manual

Organic Materials Physical Models Organic Display and Organic Solar: Organic Simulators

kT L 2 0
E0 =  ---------  ------------ 15-40
 q   q 

If you specify the PFMOB parameter on the MODELS statement, the Poole-Frenkel mobility
model will be used for both electrons and holes. You can specify this model individually for
electrons and holes by using the PFMOB.N and PFMOD.P parameters in the MOBILITY
statement.

Table 15-7 User-Specifiable Parameters for Poole-Frenkel Mobility Model

Statement Parameter Default Units

MOBILITY BETAN.PFMOB 0 eV(cm/V)1/2

MOBILITY BETAP.PFMOB 0 eV(cm/V)1/2
MOBILITY BETANAUTO.PFMOB False
MOBILITY BETAPAUTO.PFMOB False
MOBILITY DELTAEN.PFMOB 0 eV
MOBILITY DELTAEP.PFMOB 0 eV
MOBILITY EON.PFMOB 1956 V/cm
MOBILITY EOP.PFMOB 1956 V/cm
MOBILITY EONAUTO.PFMOB False
MOBILITY EOPAUTO.PFMOB False
MOBILITY GAMMAN.PFMOB 0 (cm/V)1/2
MOBILITY GAMMAP.PFMOB 0 (cm/V)1/2
MOBILITY PFMOB False
MOBILITY PFMOB.N False
MOBILITY PFMOB.P False
MOBILITY TON.PFMOB 0 K
MOBILITY TOP.PFMOB 0 K
MOBILITY VTHN.PFMOB cm/s
MOBILITY VTHP.PFMOB cm/s

Due to the strong dependence on the electric field, the Poole-Frenkel mobility model can
cause convergence problems. To increase the stability of the Poole-Frenkel mobility model, it
is combined with a limiting mobility (  n  E  and  p  E  ) calculated from the thermal
lim lim

velocities (n and p).

745 ATLAS User’s Manual

Organic Materials Physical Models Organic Display and Organic Solar: Organic Simulators

1
 n  E  = ----------------------------------------------- 15-41
1 1
------------------- + --------------------
n  E  n  E 
PF lim

1
 p  E  = ----------------------------------------------- 15-42
1 1
------------------- + --------------------
p  E  p  E 
PF lim

where:

 n  E  = ----n- 15-43
lim E


p  E  = ----p- 15-44
lim E

If you specify VTHN.PFMOB on the MOBILITY statement, then Equation 15-44 becomes:

 n  E  = VTHN · PFMOB-
------------------------------------ 15-45
lim E

Similarly, if you specify VTHP.PFMOB on the MOBILITY statement, then Equation 15-45
becomes:

 p  E  = VTHP · PFMOB-
------------------------------------ 15-46
lim E

15.2.5 Bimolecular Langevin Recombination Model

To enable this model, specify the parameter LANGEVIN in the MODELS statement. The
Langevin recombination rate coefficient is given by [30][248]:
q  n  E  + p  E  
r L  x y t  = A · LANGEVIN -------------------------------------------- 15-47
  r 0

If you specify the logical parameter KOSTER on the MODEL statement, the Langevin rate is
given by:
qmin   n  p 
r L  x y t  = A.LANGEVIN --------------------------------- 15-48
r o

Table 15-8 User-Specifiable Parameters for Poole-Frenkel Mobility Model

Statement Parameter Default Units

MATERIAL A.LANGEVIN 1

746 ATLAS User’s Manual

Organic Materials Physical Models Organic Display and Organic Solar: Organic Simulators

The Langevin recombination rate is given by:

2
R L n p = r L  x y t   np – ni  15-49

When enabled, the Langevin recombination rate is included in recombination terms in the
carrier continuity equations (Equations 3-3 and 3-4). This rate is also included in the exciton
singled and triplet continuity equations described in Section 15.3 “Singlet and Triplet
Excitons”.

15.2.6 Juska Two-Dimensional Langevin Recombination Model.

This is a 2D Langevin recombination model that only considers the drift current. To enable
this model, specify the DR.2D.LANGEVIN parameter on the MODELS statement. The Langevin
recombination rate is given by [127]:
32 12
qL · 2D · LANGEVIN c -    E  +   E  np
R L ,n ,p = A · 2D · LANGEVIN  ----------------------------------------------------------------------- 15-50
r 0 n p

with

c = np
L.2D.LANGEVIN is the thickness of the 2D lamellas in the organic structure. In [127], the
value of the prefactor A.2D.LANGEVIN is 3/4. Another paper [192] has a similar expression
with a prefactor of 15/8.

Table 15-9 User-Specifiable Parameters for Juska Langevin Recombination Model

Statement Parameter Default Units

MATERIAL A.2D.LANGEVIN 0.75

MATERIAL L.2D.LANGEVIN 1.6 nm

When enabled, the Langevin recombination rate is included in recombination terms in the
carrier continuity equations (Equations 3-3 and 3-4). This rate is also included in the exciton
singlet and triplet continuity equations described in Section 15.3 “Singlet and Triplet
Excitons”.

15.2.7 Nenashev Two-Dimensional Langevin Recombination Model

This is a 2D Langevin recombination model that considers both the drift and the diffusion
current. To enable this model, specify the DRDI.2D.LANGEVIN parameter on the MODELS
statement. The Langevin recombination rate is given by [192]:

4kTL · 2D · LANGEVIN
R L ,n ,p = -----------------------------------------------------------------------------    n  E  +  p  E  np  15-51
 ALPHA · 2D · LANGEVIN
q ln  ---------------------------------------------------------------
 a 2 L · 2D · LANGEVINc 

with

c = np

747 ATLAS User’s Manual

Organic Materials Physical Models Organic Display and Organic Solar: Organic Simulators

and the Onsager radius

2
q
a = ------------------------
4 r  0 kT

L.2D.LANGEVIN is the thickness of the 2D lamellas in the organic structure. The

ALPHA.2D.LANGEVIN parameter can either be specified directly or by providing
G.2D.LANGEVIN for the equation

1
ALPHA · 2D · LANGEVIN = ----------------------------------------------------------------------- 15-52
 exp  2 + 2D · LANGEVIN 

where γ is Euler's constant. If ALPHA.2D.LANGEVIN is provided then any value given for
G.2D.LANGEVIN is ignored.

Table 15-10 User-Specifiable Parameters for Nenashev Langevin Recombination Model

Statement Parameter Default Units

MATERIAL L.2D.LANGEVIN 1.6 nm

MATERIAL ALPHA.2D.LANGEVIN 0.037
MATERIAL G.2D.LANGEVIN 1

When enabled, the Langevin recombination rate is included in recombination terms in the
carrier continuity equations (Equations 3-3 and 3-4). This rate is also included in the exciton
singlet and triplet continuity equations described in Section 15.3 “Singlet and Triplet
Excitons”.

15.2.8 Generalized Gaussian Disorder Transport Model

To activate this model for electrons, specify the PASVEER.N flag on the MOBILITY statement.
To activate this model for holes, specify the PASVEER.P flag.
The model described in [286] assumes a Gaussian density of states
2
N t n  En 
N C  E  = ---------------------------------------------------  exp  – ------------------------------------------
- 15-53
2SIG · PASV · N
2  2SIG · PASV · N  2

2
N t p  Ep 
N V  E  = ---------------------------------------------------  exp  – ------------------------------------------
- 15-54
2SIG · PASV · P
2  2SIG · PASV · P  2

NC(E) and NV(E) are the distributions for modeling electrons and holes respectively. En and
Ep are the energies relative to the center of the appropriate Gaussian distribution.
The densities of sites, Nt, are described by the average distance, a, between sites

748 ATLAS User’s Manual

Organic Materials Physical Models Organic Display and Organic Solar: Organic Simulators

1
N t n = -------------------------------- 15-55
3
A · PASV · N

1
N t p = -------------------------------- 15-56
3
A · PASV · P

The organic semiconductor is disordered so the carrier wave-functions are localized with an
inverse length of . The inverse of the electron localization length is ALP.PASV.N and the
inverse of the hole localization length is ALP.PASV.P.

The expressions in [286] are for the case where -1=0.1×a. If ALP.PASV.N or ALP.PASV.P
are defined, then the associated A.PASV.N or A.PASV.P are calculated as
10
A · PASV · N = ------------------------------------ 15-57
ALP · PASV · N

10
A · PASV · P = -----------------------------------
- 15-58
ALP · PASV · P

If both A.PASV.N and ALP.PASV.N are defined, then the explicit value of A.PASV.N is used.
Similarly, if both A.PASV.P and ALP.PASV.P are defined, then the explicit value of
A.PASV.P is used.
The Pasveer model [209] of the enhancement to the mobility is
  T n E  =  300  g 0  T   g 1  n T   g 2  E T  15-59
The temperature dependent terms, g0(T), are

 TC2 · PASV · N  SIG · PASV · N2  1 1- 

g 0 n = exp  ---------------------------------------------------------------------------------
-  ------------ – ----- 
2
15-60
2
 k  300 2
T 

 TC2 · PASV · P  SIG · PASV · P2  1 1- 

g 0 p = exp  ---------------------------------------------------------------------------------
-  ------------ – -----  15-61
2
 k  300 2
T 
2

The carrier density dependent enhancement to the mobilities, g1, are

n n
g 1 n  n T  = exp ---   -----------------------------------
1 SIG · PASV · N-  SIG · PASV · N
 ------------------------------------ – 1   2 ----- 15-62
2  kB T   kB T   N t

p p
g 1 p  p T  = exp ---   -----------------------------------
1 SIG · PASV · P-  SIG · PASV · P
 ------------------------------------ – 1   2 ----- 15-63
2  kB T   kB T   N t

where

kB T 2  
 n = 2   ------------------------------------   ln  -----------------------------------
SIG · PASV · N  SIG · PASV · N
-  ------------------------------------ – 1 – ln  ln 4   15-64
 SIG PASV N  k T   
· ·  B kB T 

749 ATLAS User’s Manual

Organic Materials Physical Models Organic Display and Organic Solar: Organic Simulators

kB T 2  
 p = 2   ------------------------------------   ln  -----------------------------------
SIG · PASV · P-  SIG · PASV · P
 ------------------------------------ – 1 – ln  ln 4   15-65
 SIG PASV P  k T   
· ·  B k B T 

These equations hold for

n
----------  CCUTOFF · N 15-66
N t n

n
----------  CCUTOFF · P 15-67
N t p

If the normalized carrier concentration is higher than the cutoff, the enhancement g1 is
calculated at the cutoff concentration.
The electric field dependent enhancement to the mobilities, g2, are

 2 
 32   
 SIG · PASV N
g 2 n = exp  0.44  -----------------------------------
· -  q A · PASV N E
– 2.2  1 + 0.8  ------------------------------------------
· - – 1  15-68
 k T   SIG · PASV · N 
 B 
 

 2 
  SIG · PASV · P 3  2  q A · PASV · P E 
g 2 p = exp  0.44  ------------------------------------ – 2.2  1 + 0.8  ------------------------------------------- – 1  15-69
 kB T   SIG · PASV · P 
 
 

These equations hold for

q------------------------------------------
A · PASV · N E-
 FCUTOFF · N 15-70
SIG · PASV · N

q A · PASV · P E
-------------------------------------------  FCUTOFF · P 15-71
SIG · PASV · P

If the normalized electric field is higher than the cutoff, the enhancement g2 is calculated at
the cutoff electric field.
If a Gaussian band structure (Section 15.2.1 “Gaussian Band Structure”) is being used then
the default parameter values for the Pasveer model are extracted from this band structure.
–1  3
A.PASV.N = NTC.GAUSS 15-72

SIG.PASV.N = SIGC.GAUSS 15-73

–1  3
A.PASV.P = NTV.GAUSS 15-74

SIG.PASV.P = SIGV.GAUSS 15-75

Any parameters explicitly given on a MOBILITY statement overload these defaults (but do not
change the equivalent parameters in the Gaussian band structure).

750 ATLAS User’s Manual

Organic Materials Physical Models Organic Display and Organic Solar: Organic Simulators

Table 15-11 User-Specifiable Parameters for Pasveer Mobility Model

Statement Parameter Value Units

MOBILITY PASVEER.N FALSE

MOBILITY PASVEER.P FALSE

MOBILITY SIG.PASV.N 0.1 eV

MOBILITY SIG.PASV.P 0.1 eV
MOBILITY A.PASV.N 0.1 nm
MOBILITY A.PASV.P 0.1 nm
MOBILITY ALP.PASV.N nm-1
MOBILITY ALP.PASV.P nm-1
MOBILITY CCUTOFF.N 0.1
MOBILITY CCUTOFF.P 0.1
MOBILITY FCUTOFF.N 2
MOBILITY FCUTOFF.P 2
MOBILITY TC2.PASV.N 0.42
MOBILITY TC2.PASV.P 0.42

751 ATLAS User’s Manual

Singlet and Triplet Excitons Organic Display and Organic Solar: Organic Simulators

15.3 Singlet and Triplet Excitons

The distribution of singlet or triplet excitons can be used to infer the radiative rates for
luminescense or phosphorescence in organic material LEDs. In ATLAS, you can self-
consistently solve the singlet and triplet exciton continuity equations along with the electron
and hole drift diffusion equations. The singlet exciton continuity equation is given by [239]
and [291]:

dS  x y t  =
------------------------ RST · EXCITONR Ln p + F.GENSINGLET  x y z t n p  15-76
dt
RST · TT · EXCITON 2
- KTT EXCITON T  x y t 
+ ------------------------------------------------------------ ·
1 + RST · TT · EXCITON
– KISC · EXCITON S  x y t  – KST · EXCITON S  x y t  T  x y t  – KDD · EXCITONS  x y t 

KSS · EXCITON 2
– K SP · EXCITON S  x y t   n  x y t  + p  x y t   – ---------------------------------------- S  x y t 
2
PHEFF · EXCITON-
– K NRS · EXCITON S  x y t  – ----------------------------------------------- S  x y t  + D s S  x y t  – Q  x y S  x y z 
TAUS · EXCITON
– R D n p S  x y t  + G ph QE · EXCITON

while the triplet exciton continuity equation is given by:

dT  x y t  =  1 – R ST EXCITON R
------------------------ + F.GENTRIPLET  x y z t n p  15-77
dt · Ln p

+ KISC · EXCITON S  x y t  – KTP · EXCITON T  x y t   n  x y t  + p  x y t  

2
– KTT · EXCITON T  x y t  – KNRT · EXCITON T  x y t 
T  x y t 
– -------------------------------------------- +   D T T  x y t  
TAUT · EXCITON

where:
• S(x,y,t) is the spatial and temporal singlet exciton density,
• T(x,y,t) is the spatial and temporal triplet exciton density,
• RST.EXCITON is the fraction of singlets formed during Langevin recombination,
• RST.TT.EXCITON is the fraction of singlets formed during triplet-triplet annihilation,
• RLn,p is the Langevin recombination rate,
• n(x,y,t) is the electron concentration,
• p(x,y,t) is the hole concentration,
• TAUS.EXCITON is the singlet radiative decay lifetime,
• TAUT.EXCITON is the triplet radiative decay lifetime,
• KISC.EXCITON is the intersystem crossing constant,
• KST.EXCITON is the singlet-triplet constant,
• KSP.EXCITON is the singlet-polaron constant,
• KTP.EXCITON is the triplet-polaron constant,
• KSS.EXCITON is the singlet-singlet constant,

752 ATLAS User’s Manual

Singlet and Triplet Excitons Organic Display and Organic Solar: Organic Simulators

• KTT.EXCITON is the triplet-triplet constant,

• KNRS.EXCITON is the singlet non-radiative decay constant
• KNRT.EXCITON is the triplet non-radiative decay constant.
• KDD.EXCITON is the dipole-dipole exciton transfer rate (Forster).
• Q(x,y) is the metal quenching rate.
• PHEFF.EXCITON is the host photoluminescence quantum efficiency.
• RDn,p is the excition dissociation rate described in Section 15.3.3 “Exciton Dissociation”.
• Gph is the photoabsorption rate.
• QE.EXCITION is the proportion of absorbed photons that is responsible for generating
singlets.
• F.GENSINGLET gives an independent generation of singlet excitons as a function of
position and time via a c-interpreter function.
• F.GENTRIPLET gives an independent generation of triplet excitons as a function of
position and time via a c-interpreter function.
The singlet diffusion constant, DS, can be expressed as:
2
LDS · EXCITON
D S = -------------------------------------------
- 15-78
TAUS · EXCITON

where LDS.EXCITON is the singlet diffusion length.

The triplet diffusion constant, DT, can be expressed as:
2
LDT · EXCITON
D T = -------------------------------------------
- 15-79
TAUT · EXCITON

where LDT.EXCITON is the triplet diffusion length.

Table 15-12 User-Specifiable Parameters for Equations 15-76, 15-77, 15-78, and 15-79

Statement Parameter Default Units

MATERIAL KISC.EXCITON 0 s-1

MATERIAL KNRS.EXCITON 0 s-1
MATERIAL KNRT.EXCITON 0 s-1
MATERIAL KSP.EXCITON 0 cm3 s-1
MATERIAL KST.EXCITON 0 cm3 s-1
MATERIAL KTP.EXCITON 0 cm3 s-1
MATERIAL KSS.EXCITON 0 cm3 s-1
MATERIAL KTT.EXCITON 0 cm3 s-1

753 ATLAS User’s Manual

Singlet and Triplet Excitons Organic Display and Organic Solar: Organic Simulators

Table 15-12 User-Specifiable Parameters for Equations 15-76, 15-77, 15-78, and 15-79
MATERIAL LDS.EXCITON 0.01 microns
MATERIAL LDT.EXCITON 0.0632 microns
MATERIAL RST.EXCITON 0.25
MATERIAL RST.TT.EXCITON RST.EXCITON

MATERIAL TAUS.EXCITON 110-9 s

MATERIAL TAUT.EXCITON 110-4 s

MATERIAL KDD.EXCITON 0 s-1

MATERIAL PHEFF.EXCITON 1.0

Table 15-13 User-Specifiable Parameters for Tetracene

Statement Parameter Default Units

MATERIAL KISC.EXCITON 0 s-1

MATERIAL KNRS.EXCITON 0 s-1
MATERIAL KNRT.EXCITON 0 s-1
MATERIAL KSP.EXCITON 410-8 cm3 s-1
MATERIAL KST.EXCITON 210-7 cm3 s-1
MATERIAL KTP.EXCITON 210-9 cm3 s-1
MATERIAL KSS.EXCITON 510-8 cm3 s-1
MATERIAL KTT.EXCITON 110-9 cm3 s-1
MATERIAL LDS.EXCITON 0.01 microns
MATERIAL LDT.EXCITON 0.0632 microns
MATERIAL RST.EXCITON 0.25
MATERIAL TAUS.EXCITON 110-7 s

MATERIAL TAUT.EXCITON 110-4 s

To enable the self-constant simulation of both singlet and triplet excitons, specify EXCITONS
in the MODEL statement. Specifying SINGLET instead of EXCITONS will enable the solution of
the singlet exciton equations, while specifying TRIPLET will enable the triplet exciton
equation.

754 ATLAS User’s Manual

Singlet and Triplet Excitons Organic Display and Organic Solar: Organic Simulators

Table 15-14 Exciton Model Flags

Statement Parameter Default

MODELS EXCITONS False

MODELS SINGLET False
MODELS TRIPLET False
MODELS F.GENSINGLET

MODELS F.GENTRIPLET

When solving for excitons, you can examine the density distribution of the singlet and triplet
excitons in TONYPLOT in files saved in the standard structure file format. You can also
specify EXCITON in the PROBE statement to save the singlet exciton density to log files. For
further analysis of LED devices, see Chapter 11 “LED: Light Emitting Diode Simulator”.

15.3.1 Dopants
The singlet exciton density for dopants [160] can also be calculated. If you specify the
parameter DOPANT on the ODEFECTS statement, then the trap densities defined using the NA,
SIGMAA, EA, ND, SIGMAD, and ED paramaters will be used in dopant singlet exciton
calculation.

Note: Only one dopant is allowed per region.

The equation for dopant singlet excitons [160] is given by:

dS d  x y t  1 2
--------------------------- = G d + KDD · EXCITON S  x y t  – --- DKSS · EXCITON S d  x y t  15-80
dt 2

 
–  DKNRS · EXCITON + DPHEFF · EXCITON- + DKCQ EXCITON N
--------------------------------------------------- · d TOT + Q  x y  S d  x y t 
 DTAUS · EXCITION 

where:
• Sd(x,y,t) is the spatial and temporal singlet exciton density for dopants.
• DKNRS.EXCITON is the dopant nonradiative decay rate.
• DTAUS.EXCITON is the dopant radiative decay lifetime.
• DKCQ.EXCITON is the dopant concentration quenching rate.
• Nd is the dopant concentration as calculated from the ODEFECTS statement.
• DKSS.EXCITON is the bimolecular annililation constant.
• KDD.EXCITON is the dipole-dipole exciton transfer rate (Forster).
• Q(x,y) is the metal quenching rate.
• DPHEFF.EXCITON is the dopant photoluminescent quantum efficiency.
• Gd is the dopant Langevin recombination rate given by:

755 ATLAS User’s Manual

Singlet and Triplet Excitons Organic Display and Organic Solar: Organic Simulators

qDRST · EXCITION
G d = ---------------------------------------------------  nN dd  n  E  + pN da  p  E   15-81
 o

where:
• DRST.EXCITON is the fraction of singlets formed.
• Ndd is the density of occupied dopant donor traps.
• Nda is the density of occupied dopant acceptor traps.
An additional Langevin recombination term [160] is also added to the electron (Rdn) and hole
(Rdp) current continuity equations:
nN dd  n  E 
R dn = ---------------------------
- 15-82
 o

nN da  p  E 
R dp = ---------------------------
- 15-83
 o

Table 15-15 User-Specifiable Parameters for Equations 15-80, 15-81, 15-82, and 15-83

Statement Parameter Default Units

MATERIAL DRST.EXCITON 0.25

MATERIAL DTAUS.EXCITON 5.010-9 s

MATERIAL DKSS.EXCITON 3.010-8 cm3 s-1

MATERIAL DKCQ.EXCITON 1.110-8 s-1
MATERIAL DKNRS.EXCITON 0.0 s-1
MATERIAL KDD.EXCITON 0.0 s-1
MATERIAL DPHEFF.EXCITON 1.0

15.3.2 Light Generation of Excitons

Optical generation of of excitons is available in ORGANIC SOLAR. All of the capabilities
available in LUMINOUS2D are also available in ORGANIC SOLAR for the description of the
propogation and absorption of light. If you are interested in photogeneration in organic
materials, then read Chapter 10 “Luminous: Optoelectronic Simulator”.
The additional parameter QE.EXCITON of the MATERIAL statement specifies the fraction of
the absorbed photons that generate singlet excitons.
The fraction 1-QE.EXCITON of the absorbed photons leads to the generation of electron hole
pairs. By default, the value of QE.EXCITON is 0. For ORGANIC SOLAR analysis, a value of one
is more recommended. The actual detection mechanism requires exciton dissociation as
described in the next section.

756 ATLAS User’s Manual

Singlet and Triplet Excitons Organic Display and Organic Solar: Organic Simulators

15.3.3 Exciton Dissociation

Another source of carrier generation in the carrier continuity equations (Equations 3-3 and 3-
4) is due to exciton dissociation. Exciton dissociation is the dissociation of a neutral singlet
exciton into a free electron hole pair. By default, the dissociation rate is given by:

3r L  x y t  J 1  2 – 2b 
R D n p = ------------------------------------------- exp  – --------------------------------
S · BINDING --------------------------- - S  x y t 
- 15-84
3  kT  – 2b
4A · SINGLET

Here, rL is the Langevin recombination rate constant (Equation 15-47). A.SINGLET and
S.BINDING are user-specified parameters as described in Table 15-16 representing the
electron and hole separation distance and the singlet exciton binding energy. J1 is the first
order Bessel function and S(x,y,t) is the local density of singlet excitons.
The parameter b is given by:
3
q E
b = ------------------------------ 15-85
2 2
8 r  0 k T

where E is the local electric field and r is the relative permittivity.

Generally, the geminate pair distance can be described by a distribution function specified by
SDISS.TYPE on the MODELS statement as follows [87]:

SDISS · TYPE = 1 fr = r – a

2 2
2 –r  a
SDISS · TYPE = 2 fr = r e
2
15-86
–r  a
SDISS · TYPE = 3 f r = e
2 –r  a
SDISS · TYPE = 4 fr = r e

The dissociation rate is also included in the singlet continuity equation described in Section
15.3 “Singlet and Triplet Excitons”. To include the singlet dissociation rate, you must specify
S.DISSOC on the MODELS statement and also enable Langevin recombination.

Table 15-16 User Specifiable Parameters of the Singlet Dissociation Model

and Metal Quenching

Statement Parameter Default Units

MATERIAL A.SINGLET 1.8 nm

MATERIAL S.BINDING 0.1 eV
MATERIAL QR.EXCITON 0 cm3/s

In the geminate pair dissociation model, the geminate pair separation distance A.SINGLET
and the geminate pair binding energy S.BINDING are independent fitting parameters. This
can be slightly non-physical. One could suggest that we apply Coulomb's law to link the two
parameters as if they have the same dependence as for ion pairs.

757 ATLAS User’s Manual

Singlet and Triplet Excitons Organic Display and Organic Solar: Organic Simulators

This correction can be expressed using a new parameter K.SINGLET in Equation 15-87.
2
q
S · BINDING = K · SINGLET ------------------------------------------------------ 15-87
A · SINGLET 4 r  o

Here, K.SINGLET is a user-specifiable parameter expressing the relationship between

S.BINDING and A.SINGLET. Ideally, K.SINGLET should have a value of one. You can
independently express any two of the parameters and the third will be automatically
calculated. The values of S.BINDING and A.SINGLET are eventually used in Equation 15-85.

15.3.4 Metal Quenching

The metal quenching rate is given by:

Q  x y  = QR · EXCITON  M  x y  15-88

where M(x,y) is the metal atomic density. QR.EXCITION is the quenching constant specifiable
on the MATERIAL statement. To specify the metal atomic density, use the METAL parameter on
the DOPING statement. The defaults for QR.EXCITON are given in Table 15-16.

758 ATLAS User’s Manual

Chapter 16
NOISE: Electronic Noise
Simulator
Introduction NOISE: Electronic Noise Simulator

16.1 Introduction
NOISE is an ATLAS based product that simulates the small-signal noise generated by
devices. Electronic noise results in an unavoidable degradation of a circuit’s performance. It
is important to understand the properties of noise to minimize its effect.
You should already be familiar with ATLAS before using NOISE. If not, see Chapter 2
“Getting Started with ATLAS”.

760 ATLAS User’s Manual

Simulating Noise in ATLAS NOISE: Electronic Noise Simulator

16.2 Simulating Noise in ATLAS

ATLAS models the noise of a device by calculating the statistical behavior of equivalent
random voltage sources at its ports. You can use the results of this calculation in a circuit
simulator (e.g., SMARTSPICE) to minimize the disruption caused by the (inevitable) presence
of noise. The following shows the minimum needed to simulate noise in ATLAS.
1. Add the noise parameter to a log command. For example:
log outfile=noise.log inport=gate outport=drain noise
2. Use the noise parameter on the subsequent solve command. For example:
solve noise frequency=100.0
NOISE is an extension of the AC analysis of a device (see Section 2.9.3 “Small-Signal AC
Solutions” for more information on AC analysis). You can perform a noise simulation on any
device (one-port or two-port) where small-signal AC analysis can be performed.
A direct AC analysis is automatically performed on the device before the noise simulation
(i.e., you don’t need to perform the analysis).
For more information about the LOG and SOLVE commands, see Sections 21.30 “LOG” and
21.57 “SOLVE”.

761 ATLAS User’s Manual

Circuit Level Description of Noise NOISE: Electronic Noise Simulator

16.3 Circuit Level Description of Noise

Noise is the name given to the random fluctuations in the currents and voltages of real
devices.
The simple Drude model of an n-type resistor gives
 J =    E = q   n   v 16-1

where  J is the mean current density, q is the electronic charge,  n is the mean electron
density, and  v is the mean electron drift velocity. Time dependent fluctuations in the
instantaneous electron density and drift velocity will cause fluctuations in the resultant
current density.
The current at a contact of the resistor would be
i  t  =  i + i  t  16-2

where the  i  t  term is the noise generated by the resistor.

A noisy device can be represented by external current sources added to the terminals of an
ideal (noiseless) device (see Figure 16-1).

Figure 16-1: A real (noisy) device is modelled as an ideal (noiseless) device with random current sources
attached to the ports.
These current sources are small and random. The description of noise is the statistical
properties of these current sources. Normally, you describe the behavior of the frequency
domain representation of the current sources.
The means of the current sources are zero. For example:
 i 1    = 0 and  i 2    = 0 16-3

The noise is described by the auto-correlation:

2 2
 i 1     and  i     16-4
2

and the cross-correlation:

 i 1     i 2    and  i 2     i 1    16-5

762 ATLAS User’s Manual

Circuit Level Description of Noise NOISE: Electronic Noise Simulator

It may be more convenient to describe the noise in terms of the behavior of voltage sources
instead of current sources. The noise current can be easily translated into a noise voltage. For
example:

 v 1     Z 11    Z 12     i 1   
  =     16-6
 v 2     Z 21    Z 22     i 2   

Figure 16-2 shows that both representations are the same.

Figure 16-2: The noise can be modelled with random current sources or random voltage sources.

16.3.1 Noise “Figures of Merit” for a Two-Port Device

Figure 16-3 shows a simple two-port circuit.

Figure 16-3: A block diagram of a simple circuit. We are interested in how much noise reaches the load.
The noise figure, F, is a measure of the increased amount of noise power delivered to the load
because of the device’s noise.
P S D
F = -----------
- 16-7
PS

Here:
• PS is the noise power that would be delivered to the load if the device was noiseless (i.e.,
only the noise from the source is transferred to the load).
• PS,D is the noise power that is delivered to the load from both the source and the device.

763 ATLAS User’s Manual

Circuit Level Description of Noise NOISE: Electronic Noise Simulator

The standard definition of noise figure assumes the source is generating thermal noise at the
temperature of the device:
2
 v s  = 4kTR s 16-8

Here, we have for the two-port circuit:

gn 2
F = F min + ----- Z S – Z 0 16-9
Rs

where the traditional two-port figures of merit for noise are as follows:
• the minimum noise figure, Fmin ;
• the best source impedance, Z0 ;
• the noise conductance, gn.
The best source impedance gives the minimum noise figure (i.e., F=Fmin when ZS=Z0). The
noise conductance is a measure of how much the noise figure increases as the source
impedance moves away from Z0.

764 ATLAS User’s Manual

Noise Calculation NOISE: Electronic Noise Simulator

16.4 Noise Calculation

ATLAS does the following to calculate the noise of a device.
1. Takes a small volume of a device and calculates the random current flucluations in that
volume.
2. Uses the impedence field to calculate the resultant voltage on a contact.
3. Repeats the above steps until the noise from each part of the device has been calculated,
which is then the total noise of a device.

16.4.1 The Impedance Field

The impedance field is a transfer function relating current (injected at some point in the
device) to the resultant contact voltage.
Suppose a current is injected into a point r of a device (see Figure 16-4).

Figure 16-4: We calculate the impedance field at point r by injecting a unit current at the point getting the
resultant voltages V1 and V2. (This is impossible to do in practice but is easy for ATLAS to simulate.)
This will generate a voltage on the open circuit contacts. The device is assumed to be linear so
that the voltage will be at the same frequency as the input current. Here, we get
v 1    = Z 1  r ;   i    16-10

and

765 ATLAS User’s Manual

Noise Calculation NOISE: Electronic Noise Simulator

v 2    = Z 2  r ;   i    16-11

The Zi(r;) parameters, when taken over the whole device, are called the impedance fields.
There are separate impedance fields for when the injected current is electrons or holes. In
other words, a two-port device has four associated impedance fields: Z1n  r ;  , Z2n  r ;  , Z1p  r ;  ,
and Z2p  r ; .

16.4.2 Microscopic Noise Source

We assume the statistical behavior of the noise from one point in the device is totally
uncorrelated with the statistical behavior of the noise at all other points. So we describe the
correlation of a parameter, x, as:
 x  x  r r ;  = k x x  r ;     r – r  16-12

where the term k x x  r ;  is called the microscopic noise source.

The theory of Generation-Recombination noise gives the auto-correlation of the current:
k i i  r ;  16-13

The theory of Diffusion noise gives the auto-correlation of the current density:
k J J  r ;  16-14

16.4.3 Local Noise Source

The local noise source is the effect that a microscopic noise source has on the noise behavior
of the device.
For k i i  r ;  we have

 
 v  ,v      =
 Z  r ; ki i  r ;   Z  r ;   dr 16-15


where the integral is over the device volume. The integrand function is called the local noise
source.
For k J J  r ;  we have

 
 v  ,v      =
 Z  r ;  kJ J  r ;   Z  r ;   dr 16-16


Again, the integral is over the device volume and the integrand function is called the local
noise source.
The vector impedance field is defined as
Z  r ;  = Z  r ;  16-17

766 ATLAS User’s Manual

ATLAS Models NOISE: Electronic Noise Simulator

16.5 ATLAS Models

ATLAS has models for three types of microscopic noise source: Diffusion Noise, Generation-
Recombination Noise, and Flicker Noise.
If noise is being calculated then diffusion noise is on by default. Generation-recombination
noise is on if any Generation-Recombination (GR) or Impact Ionization (II) models are
defined. Flicker noise is off by default.
The default noise models for diffusion and GR/II have no user-definable parameters They get
all the needed information from the simulation.

16.5.1 Diffusion Noise

Diffusion noise is caused by variations in the velocity of the carriers.
Einstein Diffusion Equation
There are no independent user-definable parameters for this model. You can set the mobility
model in a MATERIAL statement (See Section 21.32 “MATERIAL”) but that will affect the
whole simulation not just the noise calculation. It is on by default and can only be turned off
by choosing a different model for diffusion noise.
The equation for the microscopic noise source is
2
k J J  r ;  = 4q n  r D    16-18

where:
• q is the fundamental electronic charge.
• n(r) is the carrier concentration.
• D() is the diffusion coefficient.
The diffusion coefficient is given by Einstein's relationship, which is:
kT
D    = ------  16-19
q

where:
• k is Boltzmann's constant.
• T is the temperature.
• µ is the carrier mobility.

767 ATLAS User’s Manual

ATLAS Models NOISE: Electronic Noise Simulator

C-Interpreter
You can define a C-Interpreter function to calculate the microscopic noise source for diffusion
noise. If a C-Interpreter function is defined, then the default model is turned off.
For a C-Interpreter function to calculate the electron microscopic noise source, use the
MATERIAL statement:
material F.MNSNDIFF=<filename>
The header for this C-Interpreter function is
/*
* Microscopic Noise Source for electron diffusion noise.
* Statement: MATERIAL
* Parameter: F.MNSNDIFF
* Arguments:
* xcomp composition fraction x
* ycomp composition fraction y
* temp lattice temperature (K)

* n electron concentration (cm-3)

* mun electron mobility (cm2/V s)

* e electric field (V/cm)

* *mns microscopic noise source (s A2 / cm)

*/
int mnsndiff(double xcomp, double ycomp, double temp, double n,
double mun, double e, double *mns);
For a C-Interpreter function to calculate the hole microscopic noise source, use the MATERIAL
statement:
material F.MNSPDIFF=<filename>
The header for this C-Interpreter function is
/*
* Microscopic Noise Source for hole diffusion noise.
* Statement: MATERIAL
* Parameter: F.MNSPDIFF
* Arguments:
* xcomp composition fraction x
* ycomp composition fraction y
* temp lattice temperature (K)

* p hole concentration (cm-3)

* mup hole mobility (cm2 / V s)

768 ATLAS User’s Manual

ATLAS Models NOISE: Electronic Noise Simulator

* e electric field (V / cm)

* *mns microscopic noise source (s A2 / cm)

*/
int mnspdiff(double xcomp, double ycomp, double temp, double p,
double mup, double e, double *mns);

16.5.2 Generation-Recombination Noise

Generation-Recombination noise is caused by variations in the number of the carriers.
There are no user-definable parameters for GR noise. You can, however, set the parameters
for the various GR models in the MATERIAL statement (see Section 21.32 “MATERIAL”) and
the various II models in the IMPACT statement (Section 21.22 “IMPACT”). Setting these
parameters will affect the whole simulation, not just the noise calculation.
There are two families of GR in ATLAS: direct and trap assisted. Direct GR is when the
electron travels directly from the conduction band to the valence band (or vice versa). Trap
assisted GR is when the electron travels from a band to a trap level. If more than one trap
level is defined, then they are treated as independent trap assisted GR sites (i.e., electrons are
not assumed to move between different trap levels).
Generation-Recombination is defined by four parameters:
• Gn
• Gp
• Rn
• Rp
G is the generation rate, R is the recombination rate, n is for electrons, and p is for holes. The
generation rate of electrons is the number of electrons that appears (per unit time, per unit
volume) in the conduction band. The recombination rate is the number of electrons that
disappears from the conduction band.
Direct GR
Since there are no trap levels, all electrons leaving the conduction band end up in the valence
band. All electrons leaving the valence band end up in the conduction band.
Therefore:
G n = G p and R n = R p 16-20

The GR parameters are calculated with

2
G = C  ni
16-21
R = C  np

For optical GR (OPTR in the MODELS statement), the C parameter is COPT in the MATERIALS
statement. For Auger GR (AUGER in the MODELS statement), the C parameter is a function of
the carrier density.

769 ATLAS User’s Manual

ATLAS Models NOISE: Electronic Noise Simulator

The microscopic noise sources are

k n  n = k p  p = k n  p = 2  G + R  16-22

Trap Assisted GR
There are assumed to be no direct transitions in trap assisted GR. Therefore, an electron
leaving the conduction band ends up at the trap level. An electron entering the conduction
band comes from the trap level. Unlike direct GR, the following four parameters are
independent.
n1 nt
G n = ----- -----
n Nt
p1 n
G p = -----  1 – -----t 
p  N t
n
R n = -----  1 – -----t 
n
n Nt 16-23

p nt
R p = ----- -----
p Nt

The ratio nt/Nt is the fraction of ionized traps. The concentration n1 is the electron
concentration if the Fermi level were at the trap level. p1 is the hole concentration if the Fermi
level were at the trap level.
You can define the lifetimes n and p using the TAUN0 and TAUP0 parameters from the
MATERIALS statement.
The microscopic noise sources are
k n  n = 2  G n + R n  16-24

There are no direct transitions between the conduction and valence band. Therefore:
k p  p = 2  G p + R p  16-25

Impact Ionization
Impact ionization is a pure generation term. For each electron created in the conduction band,
a corresponding hole is also created in the valence band. This is similar to direct GR without
a recombination term.
The microscopic noise sources are
k n  n = k p  p = k p = 2G 16-26

770 ATLAS User’s Manual

ATLAS Models NOISE: Electronic Noise Simulator

16.5.3 Flicker Noise

Flicker noise is observed experimentally, however, a complete theory about what causes
flicker noise isn’t known [32].
Hooge
The Hooge model, which is shown below, is a phenomenological microscopic noise source.
2 2
 Hn J n  r   Hp J p  r 
K Jn Jn  r ;  = ---------- ------------------ and K Jp  Jp  r ;  = ---------- ------------------ 16-27
f nr f pr

In this model:
• Hn is the electron Hooge constant.
• Hp is the hole Hooge constant.
• f is the frequency.
• Jn(r) is the electron current density.
• Jp(r) is the hole current density.
• n(r) is the electron concentration.
• p(r) is the hole concentration.
The Hooge constants are defined in the MATERIAL statement:
material HOOGEN=<value> HOOGEP=<value>
The default values for these parameters is zero (which means the flicker noise is turned off).
C-Interpreter
You can define a C-Interpreter function to calculate the microscopic noise source for flicker
noise. If a C-Interpreter function is defined, then the default model is turned off.
For a C-Interpreter function to calculate the electron microscopic noise source, use the
MATERIAL statement:
material F.MNSNFLICKER=<filename>
The header for this C-Interpreter function is:
/*
* Microscopic Noise Source for electron flicker (1/f) noise.
* Statement: MATERIAL
* Parameter: F.MNSNFLICKER
* Arguments:
* x location x (microns)
* y location y (microns)
* temp lattice temperature (K)
* n electron concentration (cm^-3)
* jn electron current density (A / cm^2)
* f frequency (Hz)

771 ATLAS User’s Manual

ATLAS Models NOISE: Electronic Noise Simulator

* e electric field (V / cm)

* *mns microscopic noise source (s A^2 / cm)
*/
int mnsnflicker(double x, double y, double temp, double n,
double jn, double f, double e, double *mns);
For a C-Interpreter function to calculate the hole microscopic noise source, use the MATERIAL
statement:
material F.MNSPFLICKER=<filename>
The header for this C-Interpreter function is
/*
* Microscopic Noise Source for hole flicker (1/f) noise.
* Statement: MATERIAL
* Parameter: F.MNSPFLICKER
* Arguments:
* x location x (microns)
* y location y (microns)
* temp lattice temperature (K)

* p hole concentration (cm-3)

* jp hole current density (A / cm2)

* f frequency (Hz)
* e electric field (V/cm)

* *mns microscopic noise source (s A2/cm)

*/
int mnspflicker(double x, double y, double temp, double p,
double jp, double f, double e, double *mns);

772 ATLAS User’s Manual

Output NOISE: Electronic Noise Simulator

16.6 Output
The noise simulation results can output to the log file. The intermediate values in the
calculation can output to the structure file.

16.6.1 Log Files

There are five groups of data that can output to a log file. They are as follows:
• The two-port figures of merit (FOM): Fmin, Zopt, gn.
• The correlation of the total noise voltage sources.
• The correlation of the total noise current sources.
• The correlation of the individual noise voltage sources (GR, II, electron and hole
diffusion; electron and hole flicker).
• The correlation of the individual noise current sources (GR, II, electron and hole
diffusion; electron and hole flicker).
Table 16-1 shows the logical parameters in the LOG command (see Section 21.30 “LOG” for
more information) that cause these groups to output. The columns in this table correspond to
the groups above.

Table 16-1 Logical Parameters in the LOG command

two-port FOM Total V Total I Individual V Individual I

(a) correlation correlation correlation correlation

NOISE X -(b) - - -

NOISE.V X X - - -
NOISE.I X - X - -
NOISE.IV X X X - -
NOISE.VI X X X - -
NOISE.V.ALL X X - X -
NOISE.I.ALL X - X - X
NOISE.ALL X X X X X

Note a: If the device is a one-port, then this data will not be output: the two-port figures of merit have no meaning for a
one-port device.

Note b: If the device is a one-port and NOISE is the only noise output in the LOG command, then the total noise voltage
sources will output.

773 ATLAS User’s Manual

Output NOISE: Electronic Noise Simulator

16.6.2 Structure Files

There are five groups of data that can output to a structure file. They are as follows:
• Total Local Noise Source (LNS)
• Impedance fields
• Short-circuit Green's function
• Individual Microscopic Noise Sources (MNS)
• Individual local noise sources
Table 16-2 shows the logical parameters from the OUTPUT command (see Section 21.43
“OUTPUT” for more information) that cause these groups to output. The columns in this
table correspond to the groups above.

Table 16-2 Logical Parameters in the OUTPUT command

Short-circuit
Impedance Individual
Total LNS Green’s Individual LNS
Fields MNS
function

NOISE X - - - -
NOISE.IMP - X - - -
NOISE.ALL X X X X X

774 ATLAS User’s Manual

Chapter 17
Thermal 3D: Thermal Packaging
Simulator
Overview Thermal 3D: Thermal Packaging Simulator

17.1 Overview
THERMAL3D solves the steady-state heat equation to find the equilibrium temperature
distribution in planar and non-planar three-dimensional structures. Specify the heat sinks and
sources and choose from several temperature-dependent models for thermal conductivity
within each region. A typical application for THERMAL3D is a package simulation for a power
circuit or III-V ICs.
If you’re unfamiliar with THERMAL3D, see also Chapter 6 “3D Device Simulator”. If you’re
unfamiliar with ATLAS, see Chapter 2 “Getting Started with ATLAS”.

17.1.1 3D Structure Generation

THERMAL3D supports structure defined on 3D prismatic meshes. Structures may have
arbitrary geometries in two dimensions and consist of multiple slices in the third dimension.
There are two methods for creating a 3D structure that can be used with THERMAL3D. One
method is through the command syntax of ATLAS. The other method is through an interface
to DEVEDIT. See Section 6.2 “3D Structure Generation” for more information on both
methods.
Defining Heat Sources
Heat sources are identified with regions in the 3D structure. Regions are defined in the
manner documented in “Specifying Regions And Materials” on page 38 and “ATLAS Syntax
For 3D Structure Generation” on page 448. A region has a unique number, which is used to
identify the region on the MATERIAL statement. The POWER parameter of the MATERIAL
statement is used to set the power of the heat source in watts.
MATERIAL REGION=2 POWER=0.35
Set the POWER parameter in the SOLVE statement if you want to step the power through a range
of values. See Section 17.4 “Obtaining Solutions In THERMAL3D” for the proper syntax.
Defining Heat Sinks
Heat sinks are identified as electrodes in the 3D structure. Heat sink areas should be defined
as electrodes in the manner documented in “Specifying Electrodes” on page 39 and “ATLAS
Syntax For 3D Structure Generation” on page 448.
Each electrode (heat sink) has a unique number, which is used to set the temperature on the
heat sink during simulation. For more information on setting the temperature of the heat sink,
see Section 17.4 “Obtaining Solutions In THERMAL3D”.

776 ATLAS User’s Manual

Model and Material Parameter Selection Thermal 3D: Thermal Packaging Simulator

17.2 Model and Material Parameter Selection

17.2.1 Thermal Simulation Model
To obtain steady-state solutions for the temperature distribution, THERMAL3D solves
Poisson’s equation for temperature:
kT T  = q 17-1

where T represents the steady state temperature, k the temperature-dependent thermal

conductivity, and q the power generation per unit volume in the medium (heat sources).
The prescribed temperatures at the heat sinks form boundary conditions for Equation 17-1.
The solution of the heat equation by THERMAL3D is invoked by the following syntax:
MODELS THERMAL

17.2.2 Setting Thermal Conductivity

To set the thermal conductivities, see Section 7.2.2 “Specifying Thermal Conductivity”.

777 ATLAS User’s Manual

Numerical Methods Thermal 3D: Thermal Packaging Simulator

17.3 Numerical Methods

No special numerical methods are required for thermal simulation. A METHOD statement with
no parameters is assumed by default.

778 ATLAS User’s Manual

Obtaining Solutions In THERMAL3D Thermal 3D: Thermal Packaging Simulator

17.4 Obtaining Solutions In THERMAL3D

The SOLVE statement in used in THERMAL3D for heat-flow solutions much the same as it is
used in other ATLAS simulations involving electrical biases. The temperature in kelvin on
each heat sink is used to prescribe the boundary condition temperatures. For example:
SOLVE T1=300 T2=500
sets the temperature at 300 K and 500 K on heat sink #1 and #2 respectively.
Multiple SOLVE statements are allowed in THERMAL3D. This is useful for obtaining solutions
for several combinations of heat sinks and thermal power sources. A range of solutions can
also be obtained by stepping the value of a heat sink or power source. For example:
SOLVE T1=300 POWER2=0.35 POWER3=0.4 NSTEPS=5 STEPREGION=3 \
POWERFINAL=0.8 OUTFILE=thermal_out0
increments the thermal power source in region #3 from 0.4 watts to 0.8 watts in 5 steps and:
SOLVE T1=300 POWER2=0.35 NSTEPS=3 ELECTRODE=1 \
TEMPFINAL=600 OUTFILE=thermal_out0
increments the temperature on electrode #1 from 300 K to 600 K in 3 steps. Thermal power
and temperature can be simultaneously sweep.
If more than one region power is specified during a sweep, the region to be stepped must be
specified by STEPREGION=# as shown in the first example above. If the STEPREGION
parameter is not specified, the smallest numerical value of POWER# is stepped.
During temperature and power sweeps, the output filename is modified according to the
following rule. The rightmost character is incremented using the sequence: 0-9, A-Z, a-z.
When a character is incremented beyond z, the character is set to 0> and the character to the
left (smaller than ‘z’) is incremented.
Complete syntax information for the SOLVE statement can be found in Section 21.57
“SOLVE”.
The SOLVE statement is used in THERMAL3D for heat flow solutions much as it is used in
other ATLAS simulations for electrical biases. The temperature in Kelvin on each heat sink in
the device must be set in the SOLVE statement. The Tn parameter, where n is the number of the
heat sink, is used to set the temperature. For example:
SOLVE T1=300 T2=500
This sets 300K on heat sink #1 and 500K on heat sink #2. Only one SOLVE statement is
allowed in any THERMAL3D input file.

779 ATLAS User’s Manual

Interpreting The Results From THERMAL3D Thermal 3D: Thermal Packaging Simulator

17.5 Interpreting The Results From THERMAL3D

The output of thermal simulation consists of the minimum and maximum calculated
temperature of each region and its location. The three-dimensional temperature distribution
can be saved in an output structure file by setting the OUTFILE parameter in the SOLVE
statement and visualized using TONYPLOT3D.

780 ATLAS User’s Manual

More Information Thermal 3D: Thermal Packaging Simulator

17.6 More Information

Many examples using THERMAL3D have been installed on your distribution tape or CDROM.
More information about the use of THERMAL3D can be found by reading the text associated
with each example.
You can also find some information at www.silvaco.com.

781 ATLAS User’s Manual

Chapter 18
Mercury: Fast FET Simulator
Introduction Mercury: Fast FET Simulator

18.1 Introduction
MERCURY is an ATLAS module for the fast simulation of FET devices. The MERCURY command
syntax is similar to the ATLAS command syntax. See Chapter 2 “Getting Started with ATLAS”
if you are unfamilar with the ATLAS command syntax.

18.1.1 Accessing Mercury

To access the MERCURY module, use the MERCURY command or the MERCURY parameter on
the MESH command. This command must be specified immediately after the GO ATLAS
command, so you can activate the MERCURY module either with
GO ATLAS
MERCURY
MESH WIDTH=100
or with
GO ATLAS
MESH MERCURY WIDTH=100

783 ATLAS User’s Manual

External Circuit Mercury: Fast FET Simulator

18.2 External Circuit

The MERCURY module can use a Harmonic Balance simulation to calculate the large signal
behavior of a FET. As the large signal behavior of a device depends upon the circuit around
the device, the MERCURY module always embeds the FET under simulation in an external
circuit. The system that MERCURY simulates is shown in Figure 18-1.

Figure 18-1: FET Embedded In An External Circuit

The input matching network is described by a set of G two-port parameters

 Ii   G 11 G 12  V i
  =     18-1
 V g  G 21 G 22  I g

The output matching network is described by a set of H two-port parameters

 V d  H 11 H 12  I d 
  =     18-2
 Io   H 21 H 22  V o

The contact elements Zi, Zo, Zs, Zg, and Zd are simple impedances. The input and output
matching circuits, Gi and Ho, are defined as two-port networks. At DC, all elements are
assumed to be real. At RF, all elements can be complex.
The NETWORK command is used to define the characteristics of the external circuit.
For a DC or small-signal AC simulation, the external circuit is usually not wanted so the
CLEAR parameter will remove the external circuit. This will set all the contact impedances to
0. The input matching network becomes

 Ii  0 – 1  V i
  =    18-3
 V g 1 0   I g

which just applies the input voltage to the gate and passes the gate current to the output. The
output matching network becomes

 V d 0 1   Id 
  =    18-4
 Io  – 1 0  V o

784 ATLAS User’s Manual

External Circuit Mercury: Fast FET Simulator

which just applies the output voltage to the drain and passes the drain current directly to the
output.
For a harmonic balance simulation, you must define the external circuit. The circuit should
be defined at DC, the fundamental frequency, and at all the harmonics that are being modeled.
At DC, the contact impedances are set as resistances. You can define the input matching
network as a set of G-parameters, S-parameters, or Z-parameters. You can define the output
matching network as a set of H-parameters, S-parameters, or Z-parameters.
The RF circuits must be defined at a range of frequencies. There are several ways to do this.
The first way is to define the impedance as an RLC triplet. This calculates the impedance as

Z = R + j · · L – j  · C  18-5

The default capacitance is infinite, so the 1/(.C) term goes to zero if the capacitance is not
set. The second way is to set an explicit value that is the same for all frequencies. The third
way is to use the HARMONIC=<n> parameter and set a value for a single frequency (where
HARMONIC=1 is the fundamental, f, HARMONIC=2 is 2.f, HARMONIC=3 is 3.f, and so on).
The RF contact impedances can be defined as RLC triplets, as admittances, or as impedances.
The RF input matching network can be defined as a G-parameter network, an S-parameter
network, or as a Z-parameter network. If it is defined as a Z-parameter network, then you can
set the z-parameters or you can define them as an RLC triplet.
The RF output matching network can be defined as a H-parameter network, an S-parameter
network, or a Z-parameter network. If it is defined as a Z-parameter network, then you can set
the Z-parameters or you can define them as an RLC triplet.
Example
NETWORK CLEAR
This removes the network and provides a clean slate to work with
NETWORK ZI.RE.RF=50.0 ZI.IM.RF=0.0 ZO.RE.RF=50.0 ZO.IM.RF=0.0
This sets the input and output impedance to 50 Ohms at all frequencies.
NETWORK RG.RF=0.01 LG.RF=1e-13
This sets a resistance and inductance on the gate contact, the impedance will be calculated at
each frequency.
NETWORK GI.11.RE.RF=10.0 GI.11.IM.RF=5.0 GI.12.RE.RF=7.0 ...
NETWORK HARMONIC=1 ZI.11.RE.RF=...
This sets an input network at all harmonics and then changes the network at the fundamental.

785 ATLAS User’s Manual

Device Mercury: Fast FET Simulator

18.3 Device
The FET in Figure 18-1 consists of an intrinsic FET and some parasitics as shown in Figure
18-2.

Figure 18-2: Intrinsic FET With Parasitics

The MERCURY module imposes some restrictions on device structures and requires some
assumptions of device geometry to use the quasi-2D method to speed up the simulation (see
Figure 18-3). All material layers and doping profiles are planar: they have no variation in the
x-directon. The only exceptions to this rule are the gate and a possible recess in the surface.
The source and drain contacts are not considered part of the intrinsic device and are modeled
as perfectly ohmic contacts to the channel on the left and right of the device.
The method for defining a device in the MERCURY module is similar to the method used in
ATLAS. First, the MESH statements are used to set the z-depth of the device and to define the
physical extent of the device in the xy-pane
MESH WIDTH=100
X.MESH LOCATION=0.0 SPACING=0.1
X.MESH LOCATION=1.1 SPACING=0.2
X.MESH LOCATION=2.5 SPACING=0.1
Y.MESH LOCATION=0.0 SPACING=0.01
Y.MESH LOCATION=3.0 SPACING=0.01

786 ATLAS User’s Manual

Device Mercury: Fast FET Simulator

But the MERCURY module only uses the extreme values of x and y to define the extent of the
device, any locations between these two extremes (such as the "X.MESH LOCATION=1.1"
command) are ignored.
Then REGION commands are used to define the epilayers. Only the y-extent of the regions are
recognized by MERCURY: each layer must cover the whole x-range of the device
REGION MATERIAL=GaAs THICKNESS=0.2
REGION MATERIAL=AlAs THICKNESS=0.3
REGION MATERIAL=GaAs THICKNESS=2.0
Then, DOPING commands are used to define the doping. Again, these commands can only
have a y-variation. Any x-variation in the doping will be ignored.
DOPING UNIFORM N.TYPE CONCENTRATION=1e17 Y.MIN=0.0 Y.MAX=0.2
The only non-uniformity in the X direction is a recess in the surface of the device
SURFACE X=0.50 Y=0.00
SURFACE X=0.65 Y=0.15
SURFACE X=1.35 Y=0.15
SURFACE Y=1.50 Y=0.00
and the position of the gate
ELECTRODE NAME=gate X.MIN=0.7 X.MAX=1.3

787 ATLAS User’s Manual

Device Mercury: Fast FET Simulator

Figure 18-3: Typical Planar Device for a MERCURY Simulation

788 ATLAS User’s Manual

Quasi-2D Simulation Mercury: Fast FET Simulator

18.4 Quasi-2D Simulation

The quasi-2D method has two parts. The first calculates Poisson's equation in the Y direction.
The second calculates the transport equations in the X direction.

18.4.1 Poisson Equation

The first part of the simulation calculates Poisson's equation in the Y direction of the device
over a range of surface conditions. This calculates the carrier density as a function of depth
into the device (see Figure 18-4). The program then converts this profile into a "channel"
with a width and a constant 3D density (see Figure 18-5).

Figure 18-4: Carrier Density as a Function of Depth

789 ATLAS User’s Manual

Quasi-2D Simulation Mercury: Fast FET Simulator

Figure 18-5: Channel with a Width and a Constant 3D Density

Two look-up tables are generated. One from the results of Poisson's equation for an ungated
slice over the range of recess depths of the device (if the device is planar this is a single
point). The other from the results of Poisson's equation for a gated slice over the range of
contact voltages expected in the simulation, see Figure 18-6.

790 ATLAS User’s Manual

Quasi-2D Simulation Mercury: Fast FET Simulator

Figure 18-6: Channel Height and Electron Volume Density

The look-up tables are generated with the command
POISSON LOOK_UP VG_MIN=-2.5 VG_MAX=0.0 VD_MIN=0.0 VD_MAX=1.0
The parameters VG_MIN and VG_MAX are used to define the range of gate voltages. The
parameters VD_MIN and VD_MAX are used to define the range of drain voltages. These ranges
are needed to ensure that the gate look-up table is calculated over a sufficient range.

791 ATLAS User’s Manual

Quasi-2D Simulation Mercury: Fast FET Simulator

18.4.2 Channel Simulation

The main channel simulation is essentially one-dimensional. The device is divided into slices
from the source to the drain. The transport equations are derived around the parallelogram
surrounding a grid point (see Figure 18-7).

Figure 18-7: The section of the channel around the grid point xi.
Here, ya and yb are the heights of the channel (rather than the device). The volume of this
parallelogram is
 xi + 1 – xi – 1   ya + yb 
V = ----------------------------------  ---------------------  z = 0.25   x i + 1 – x i – 1    y a + y b   z 18-6
2 2

The resultant equations are similar to the standard set of equations used in the drift diffusion
model.

792 ATLAS User’s Manual

Quasi-2D Simulation Mercury: Fast FET Simulator

Poisson’s Equation
MERCURY’s version of Poisson’s equation is derived by analyzing the electric fields on the
surface of a section of the channel.

Figure 18-8: The electric fields around a section of the channel.

Start with Gauss’s law

   dA = ----
Q
18-7

Q is the total charge contained within the surface. Initially, xa=0 and xb=0.
Therefore
Q
–  y  x  z = ------0- 18-8


The relationship between Q0 and y is calculated in the charge-control simulation. Therefore,

when you add the X direction electric fields you have

 xb  y b  z –  xa  y a  z –  y  x  z = Q
---- 18-9


Replace the y term with the values from the charge control
Q – Q0
 xb  y b  z –  xa  y a  z = ------------------ 18-10


Write the charge in terms of the carrier density (where q is the charge on a hole, V is the
volume of the parallelogram, and  is the density of carriers).
Q = qV 18-11

793 ATLAS User’s Manual

Quasi-2D Simulation Mercury: Fast FET Simulator

Therefore, p is the density of holes, n is the density of electrons, ND+ is the density of ionized
donors, and NA- is the density of ionized acceptors.
+ –
 = p – n + ND – NA 18-12

If the ionization density is a constant, then

 – 0 =  p – p0  –  n – n0  18-13

Therefore
qV
 xb  y b  z –  xa  y a  z =  ------------    p – p 0  –  n – n 0   18-14
  

In the limit x0 (ya=yb), this becomes Poisson’s equation. MERCURY, however, implements
the discrete version.
Continuity Equation
MERCURY’s version of the continuity equation is derived by analyzing the currents on the
surface of a section of the channel.

Figure 18-9: The currents around a section of the channel.

A current is a flow of positive charge. Therefore, the change in the number of carriers in a
small time t due to a current i is (s is the sign of the carriers)
1
C = s  ---  i  t 18-15
q

So in a small time t, the change in the number of carriers in the parallelogram is
1 1 1
C = s  ---  i xa  t – s  ---  i xb  t – s  ---  i g  t + genTerm 18-16
q q q

794 ATLAS User’s Manual

Quasi-2D Simulation Mercury: Fast FET Simulator

Here, genTerm is the number of carriers generated in the volume in the small time t.
Dividing through by t
C s s s genTerm
------- = ---  i xa – ---  i xb – ---  i g + ----------------------- 18-17
t q q q t

Writing in terms of the current density rather than the current and dividing by the volume

C s s s xi + 1 – xx – 1 genTerm
------------- = -----------  J xa  y a  z – -----------  J xb  y b  z – -----------  J g  ------------------------------  z + ----------------------- 18-18
V  t Vq Vq Vq 2 V  t

Take the limit as t0 and get

s
dc = ---------- s s xi + 1 – xi – 1
------ -  J xa  y a  z – -----------  J xb  y b  z – -----------  J g  ----------------------------
-  z + Uc 18-19
dt Vq Vq Vq 2

Here, c is now the density of the carriers and Uc is the generation rate of the carrier. In the
limit as x0, this becomes the standard continuity equation.
Channel Current
The channel current is described by the standard drift diffusion model written using the
Bernoulli function
s  k  T  c
J = ---------------------------   B  –  i i + 1   c i – B  –  i i + 1   c i + 1  18-20
xi + 1 – xi

with
sq
 i i + 1 = ----------   v i – v v + 1  18-21
kT

795 ATLAS User’s Manual

DC and Small-Signal AC Mercury: Fast FET Simulator

18.5 DC and Small-Signal AC

As mentioned in Section 18.2 “External Circuit”, the FET under simulation is embedded in an
external circuit. But when investigating the DC or small-signal AC behavior, you will
normally be interested in the characteristics of the bare FET. The NETWORK CLEAR command
sets the external circuit so that the input voltage is applied directly to the gate and the output
voltage is applied directly to the drain.
To calculate a set of DC-IV curves, you need to issue a set of commands like this (the DC-IV
data would be saved to the file DC.LOG).
LOG OUTFILE=DC.LOG
SOLVE VINPUT=-0.0 VOUTPUT=0.0 VSTEP=0.5 VFINAL=5.0 NAME=output
SOLVE VINPUT=-0.5 VOUTPUT=0.0 VSTEP=0.5 VFINAL=5.0 NAME=output
SOLVE VINPUT=-1.0 VOUTPUT=0.0 VSTEP=0.5 VFINAL=5.0 NAME=output
SOLVE VINPUT=-1.5 VOUTPUT=0.0 VSTEP=0.5 VFINAL=5.0 NAME=output
SOLVE VINPUT=-2.0 VOUTPUT=0.0 VSTEP=0.5 VFINAL=5.0 NAME=output
In order to make a MERCURY input deck as close to an ATLAS input deck as possible, GATE is
recognized as an alias for INPUT and DRAIN is recognized as an alias for OUTPUT. So the
previous set of commands can been written as:
LOG OUTFILE=DC.LOG
SOLVE VGATE=-0.0 VDRAIN=0.0 VSTEP=0.5 VFINAL=5.0 NAME=drain
SOLVE VGATE=-0.5 VDRAIN=0.0 VSTEP=0.5 VFINAL=5.0 NAME=drain
SOLVE VGATE=-1.0 VDRAIN=0.0 VSTEP=0.5 VFINAL=5.0 NAME=drain
SOLVE VGATE=-1.5 VDRAIN=0.0 VSTEP=0.5 VFINAL=5.0 NAME=drain
SOLVE VGATE=-2.0 VDRAIN=0.0 VSTEP=0.5 VFINAL=5.0 NAME=drain
But be careful, VGATE is just an alias for VINPUT and VDRAIN is just an alias for VOUTPUT.
Therefore, the command VGATE sets the voltage Vi (rather than the voltage Vg). The command
VDRAIN sets the voltage Vo (rather than the voltage Vd). But if the external network has been
cleared (as it usually will during a DC or small-signal AC simulation), setting Vi is equivalent
to setting Vg and setting Vo is equivalent to setting Vd.

18.5.1 Small-Signal AC
In MERCURY, the FET is always arranged in a common source configuration, port 1 is always
the gate-source port, and port 2 is always the drain-source port. A typical set of commands
for a small-signal AC simulation will be
SOLVE VGATE=-1.0 VDRAIN=3.0
LOG S.PARAM GAINS NOISE.V OUTFILE=AC.LOG
SOLVE AC NOISE FREQUENCY=1.0E6 FSTEP=1.2589 NFSTEP=40 MULT.FREQ
MERCURY can calculate the small-signal noise of the device in addition to the small-signal
AC behavior.

796 ATLAS User’s Manual

Harmonic Balance Mercury: Fast FET Simulator

18.6 Harmonic Balance

Harmonic Balance simulates the behavior of a device when its input is large enough to cause
a non-linear response.
The following shows the minimum needed to simulate the large-signal response in MERCURY.
1. Solve at a dc bias point
SOLVE vinput=-0.5 voutput=2.0
2. Open a file to store the results.
log num.harmonic=3 outfile=hb.log
The harmonic balance simulation is slow so it’s a good idea to record a result when it
calculates.
3. Use the SOLVE command to simulate the large-signal behavior. You define the frequency
and the range of the size of the input signal at this frequency.
SOLVE hb vinput=1e-4 vfinal=1.0 nvstep=20 name=input mult.v
For more infomation about the LOG and SOLVE commands see Chapter 21 “Statements”.
The SOLVE Command
The NUM_HARMONICS parameter on the SOLVE command defines the number of harmonics
used in the harmonic balance simulation. That is, the number of frequency components used
to describe the time domain signal. The number of harmonics should be high enough to model
the time domain signal with sufficient fidelity. But there should be no more harmonics than
necessary: the more harmonics the slower the simulation.
The LOG Command
The following data is stored in the harmonic balance log file:
• the DC bias point,
• a list of the frequency components of the input and output voltages and currents,
• the input power at the fundamental,
• the output power at the fundamental and each harmonic.
The NUM_HARMONICS parameter on the LOG command defines how many harmonics to output
the data for. The complex numbers can output as a real/imaginary pair or a magnitude/phase
pair.

797 ATLAS User’s Manual

Harmonic Balance Mercury: Fast FET Simulator

18.6.1 Non-Linearity
Figure 18-10 shows a schematic of a two-port circuit.

Figure 18-10: Two-Port Circuit

A signal is input into port 1 of the circuit. It is then transmitted through the two-port to the
output impedance on port 2. In reality, the input signal will be irregular. But for analysis, the
input signal is usually assumed to be sinusoidal at a frequency of . The frequency is called
the fundamental frequency.
This section focuses on the steady-state output from the circuit to a periodic input signal. It
does not focus on the transient response of the circuit to non-periodic changes to the input.
A Linear Circuit
In a linear circuit, the shape of the output signal is a perfect reproduction of the shape of the
input signal. (The amplitude of the output signal is likely to be different from the input:
almost all circuits amplify or attenuate. And there may be a delay between a signal appearing
at the input and being transmitted to the output.)
If the input is a sinusoid at a frequency of , then the output will be a sinusoid at a frequency
of . But probably with a different amplitude and phase than the amplitude and phase of the
input sinusoid.
The behavior of this circuit can be summarized by a set of two-port parameters.

 v 1     Z 11    Z 12     i 1   
  =     18-22
 v 2     Z 21    Z 21      i 2   

The Z-parameters depend only upon the frequency of the input signal. They don't depend
upon the magnitude of the input signal.

798 ATLAS User’s Manual

Harmonic Balance Mercury: Fast FET Simulator

A Non-linear Circuit
In a non-linear circuit, the shape of the output signal is not the same as the shape of the input
signal.
If the input is a sinusoid at a fundamental frequency of , then the output is a signal that is a
sum of a component at the fundamental frequency and additional components at the integer
harmonics of this fundamental frequency.
v 0  v 0 1    + v 0 2  2    + v 0 3  3    +  18-23

In principle, this series is infinite but in practice we need only the first few terms. The bigger
the magnitude of the input signal the more terms you need to include in the calculation of the
output signal.
Real Devices
In reality, all devices and components are non-linear. If the input signal is small enough, only
the first term in the expansion of the output is required and the circuit will be effectively
linear.

18.6.2 Harmonic Balance

Time Domain and Frequency Domain
Any measured signal is real (rather than complex). A periodic real signal that is described by
n harmonics can be written as
v  t  = V 0 + V 1  exp  j    t  + V 2  exp  2  j    t  +  + V n  exp  n  j    t  18-24
+ V 1  exp  – j    t  + V 2  exp  – 2  j    t  + V n  exp  – n  j    t 

The "positive frequencies" correspond to sinusoidal waves moving in the negative direction.
The "negative frequencies" correspond to sinusoidal waves moving in the positive direction.
The signal v(t) is completely described by the set of coefficients V0 to Vn. This is 2.n+1 real
numbers (V0 is real and the coefficients V1 to Vn are complex). This representation of v(t) is
called the frequency domain description.
Correspondingly, the signal v(t) is completely described by a set of 2.n+1 values equally
spaced over one period. The period of v(t) is dictated by the period of the fundamental. The
period T is such that
v t = v t + T 18-25
Therefore
2
T = ----------- 18-26


We define the time step ts as

T
t s = -------------------- 18-27
2n+1

799 ATLAS User’s Manual

Harmonic Balance Mercury: Fast FET Simulator

The time domain representation of v(t) is the set of points

v 0 = v  0  v 1 = v  t s  v 2 = v  2  t s  v 2  n = v  2  n  t s  18-28
These two representations of the signal v(t) are equivalent. The discrete Fourier transform is
used to translate between them.
The Problem With Large-Signal Simulation
The equations that describe the behavior of non-linear components are mostly written in the
time domain. Therefore, they must be solved in the time domain. The time domain behavior
must be solved at 2.n+1 points over one period of the signal.
A linear circuit almost certainly contains resistance, capacitance, and inductance and has an
RCL time constant. The output from a time domain simulation of a linear circuit will initially
contain a transient part in addition to the steady state part. For the transient part of the result to
decay and leave only the steady state part, the simulation must continue for several multiples
of the RCL time constant. If the time constant is longer than the period of the input signal, a
time domain simulation may have to run for many hundreds of periods before reaching the
steady state solution.
This is a problem when trying to solve a circuit that contains both linear and non-linear
elements. The non-linear elements must be solved in the time domain. Solving the linear
elements, however, in the time domain is very inefficient.
Harmonic Balance Method
While the steady state response of a linear circuit is difficult to calculate in the time domain,
this response can be directly calculated in the frequency domain. This is the key to the
Harmonic Balance method.
The circuit is divided into the linear parts and the non-linear parts (see Figure 18-11).
Attention is focused on the ports where the two circuits connect.

Figure 18-11: A Circuit divided into Linear and Non-linear Parts

800 ATLAS User’s Manual

Harmonic Balance Mercury: Fast FET Simulator

The Harmonic Balance algorithm is as follows:

1. MERCURY makes an initial guess of the time domain voltage at the ports.
2. MERCURY uses the non-linear circuit to calculate the resultant time domain current at the
ports.
3. MERCURY uses the Fourier transform to translate each time domain current into a
frequency domain current.
4. MERCURY uses the linear circuit to calculate the resultant frequency domain voltage at the
ports.
5. MERCURY uses the Fourier transform to translate each frequency domain voltage into a
time domain voltage.
6. If the new time domain voltage is close enough to the time domain voltage from the
previous iteration, MERCURY has solved the response of the circuit. If the new time
domain voltage is significantly different from the time domain voltage in the previous
iteration, MERCURY returns to step 2 and performs another iteration.

801 ATLAS User’s Manual

NETWORK Mercury: Fast FET Simulator

18.7 NETWORK
NETWORK allows you to define the external network surrounding the FET in a MERCURY
simulation. This command is valid only when using the MERCURY module.
Syntax
NETWORK [CLEAR] [HARMONIC=<n>] [dc_circuit] [rf_circuit]

Parameter Type Default Units

CLEAR Logical False

HARMONIC Integer

CD.RF Real 0.0 F

CG.RF Real 0.0 F

CI.11.RF Real 0.0 F

CI.12.RF Real 0.0 F

CI.21.RF Real 0.0 F

CI.22.RF Real 0.0 F

CI.RF Real 0.0 F

CO.11.RF Real 0.0 F

CO.12.RF Real 0.0 F

CO.21.RF Real 0.0 F

CO.22.RF Real 0.0 F

CO.RF Real 0.0 F

CS.RF Real 0.0 F

GI.11.DC Real 0.0 Mho

GI.11.IM.RF Real 0.0 Mho

GI.11.RE.RF Real 0.0 Mho

GI.12.DC Real 0.0 Mho

GI.12.IM.RF Real 0.0

GI.12.RE.RF Real 0.0

GI.21.DC Real 0.0

GI.21.IM.RF Real 0.0

GI.21.RE.RF Real 0.0

GI.22.DC Real 0.0 Ohm

802 ATLAS User’s Manual

NETWORK Mercury: Fast FET Simulator

GI.22.IM.RF Real 0.0 Ohm

GI.22.RE.RF Real 0.0 Ohm

HO.11.DC Real 0.0 Ohm

HO.11.IM.RF Real 0.0 Ohm

HO.11.RE.RF Real 0.0 Ohm

HO.12.DC Real 0.0

HO.12.IM.RF Real 0.0

HO.12.RE.RF Real 0.0

HO.21.DC Real 0.0

HO.21.IM.RF Real 0.0

HO.21.RE.RF Real 0.0

HO.22.DC Real 0.0 Mho

HO.22.IM.RF Real 0.0 Mho

HO.22.RE.RF Real 0.0 Mho

LD.RF Real 0.0 H

LG.RF Real 0.0 H

LI.11.RF Real 0.0 H

LI.12.RF Real 0.0 H

LI.21.RF Real 0.0 H

LI.22.RF Real 0.0 H

LI.RF Real 0.0 H

LO.11.RF Real 0.0 H

LO.12.RF Real 0.0 H

LO.21.RF Real 0.0 H

LO.22.RF Real 0.0 H

LO.RF Real 0.0 H

LS.RF Real 0.0 H

RD.DC Real 0.0 Ohm

RD.RF Real 0.0 Ohm

RG.DC Real 0.0 Ohm

RG.RF Real 0.0 Ohm

RI.11.RF Real 0.0 Ohm

803 ATLAS User’s Manual

NETWORK Mercury: Fast FET Simulator

RI.12.RF Real 0.0 Ohm

RI.21.RF Real 0.0 Ohm

RI.22.RF Real 0.0 Ohm

RI.DC Real 0.0 Ohm

RI.RF Real 0.0 Ohm

RO.11.RF Real 0.0 Ohm

RO.12.RF Real 0.0 Ohm

RO.21.RF Real 0.0 Ohm

RO.22.RF Real 0.0 Ohm

RO.DC Real 0.0 Ohm

RO.RF Real 0.0 Ohm

RS.DC Real 0.0 Ohm

RS.RF Real 0.0 Ohm

SI.11.DC Real 0.0

SI.11.IM.RF Real 0.0

SI.11.RE.RF Real 0.0

SI.12.DC Real 0.0

SI.12.IM.RF Real 0.0

SI.12.RE.RF Real 0.0

SI.21.DC Real 0.0

SI.21.IM.RF Real 0.0

SI.21.RE.RF Real 0.0

SI.22.DC Real 0.0

SI.22.IM.RF Real 0.0

SI.22.RE.RF Real 0.0

SO.11.DC Real 0.0

SO.11.IM.RF Real 0.0

SO.11.RE.RF Real 0.0

SO.12.DC Real 0.0

SO.12.IM.RF Real 0.0

SO.12.RE.RF Real 0.0

SO.21.DC Real 0.0

804 ATLAS User’s Manual

NETWORK Mercury: Fast FET Simulator

SO.21.IM.RF Real 0.0

SO.21.RE.RF Real 0.0

SO.22.DC Real 0.0

SO.22.IM.RF Real 0.0

SO.22.RE.RF Real 0.0

YD.IM.RF Real 0.0 Mho

YD.RE.RF Real 0.0 Mho

YG.IM.RF Real 0.0 Mho

YG.RE.RF Real 0.0 Mho

YI.IM.RF Real 0.0 Mho

YI.RE.RF Real 0.0 Mho

YO.IM.RF Real 0.0 Mho

YO.RE.RF Real 0.0 Mho

YS.IM.RF Real 0.0 Mho

YS.RE.RF Real 0.0 Mho

ZD.IM.RF Real 0.0 Ohm

ZD.RE.RF Real 0.0 Ohm

ZG.IM.RF Real 0.0 Ohm

ZG.RE.RF Real 0.0 Ohm

ZI.11.DC Real 0.0 Ohm

ZI.11.IM.RF Real 0.0 Ohm

ZI.11.RE.RF Real 0.0 Ohm

ZI.12.DC Real 0.0 Ohm

ZI.12.IM.RF Real 0.0 Ohm

ZI.12.RE.RF Real 0.0 Ohm

ZI.21.DC Real 0.0 Ohm

ZI.21.IM.RF Real 0.0 Ohm

ZI.21.RE.RF Real 0.0 Ohm

ZI.22.DC Real 0.0 Ohm

ZI.22.IM.RF Real 0.0 Ohm

ZI.22.RE.RF Real 0.0 Ohm

ZI.IM.RF Real 0.0 Ohm

805 ATLAS User’s Manual

NETWORK Mercury: Fast FET Simulator

ZI.RE.RF Real 0.0 Ohm

ZO.11.DC Real 0.0 Ohm

ZO.11.IM.RF Real 0.0 Ohm

ZO.11.RE.RF Real 0.0 Ohm

ZO.12.DC Real 0.0 Ohm

ZO.12.IM.RF Real 0.0 Ohm

ZO.12.RE.RF Real 0.0 Ohm

ZO.21.DC Real 0.0 Ohm

ZO.21.IM.RF Real 0.0 Ohm

ZO.21.RE.RF Real 0.0 Ohm

ZO.22.DC Real 0.0 Ohm

ZO.22.IM.RF Real 0.0 Ohm

ZO.22.RE.RF Real 0.0 Ohm

ZO.IM.RF Real 0.0 Ohm

ZO.RE.RF Real 0.0 Ohm

ZS.IM.RF Real 0.0 Ohm

ZS.RE.RF Real 0.0 Ohm

Description

CLEAR Sets all contact impedances to 0, sets the input matching network to G11=0, G12=-1,
G21=1, and G22=0, and sets the output matching network to H11=0, H12=1, H21=-1,
and H22=0.
HARMONIC Specifies which harmonic, in a harmonic balance simulation, the RF circuit is applied
to. If HARMONIC is not set, then the RF circuit will be applied to all harmonics.

DC Contact Resistances

Note: At DC all circuits are assumed to be real.

RD.DC Defines the DC drain resistance.

RG.DC Defines the DC gate resistance.
RI.DC Defines the DC input resistance.
RO.DC Defines the DC output resistance.
RS.DC Defines the DC source resistance.

806 ATLAS User’s Manual

NETWORK Mercury: Fast FET Simulator

DC Input Matching Network

GI.11.DC, GI.12.DC, GI.21.DC, Defines the DC input matching network as a set of

GI.22.DC g-parameters.
SI.11.DC, SI.12.DC, SI.21.DC, Defines the DC input matching network as a set of s-
SI.22.DC parameters.
ZI.11.DC, ZI.12.DC, ZI.21.DC, Defined the DC input matching network as a set of
ZI.22.DC z-parameters.

DC Output Matching Network

HO.11.DC, HO.12.DC, HO.21.DC, Defines the DC output matching network as a set of h-

HO.22.DC parameters.
SO.11.DC, SO.12.DC, SO.21.DC, Defines the DC output matching network as a set of s-
SO.22.DC parameters.
ZO.11.DC, ZO.12.DC, ZO.21.DC, Defines the DC output matching network as a set of z-
ZO.22.DC parameters.

RF Contact Impedances

RD.RF, CD.RF, LD.RF Defines the RF drain impedance as a series RCL circuit.
RG.RF, CG.RF, LG.RF Defines the RF gate impedance as a series RCL circuit.
RI.RF, CI.RF, LI.RF Defines the RF input impedance as a series RCL circuit.
RO.RF, CO.RF, LO.RF Defines the RF output impedance as a series RCL circuit.
RS.RF, CS.RF, LS.RF Defines the RF source impedance as a series RCL circuit.
YD.RE.RF, YD.IM.RF Defines the RF drain impedance as an admittance.
YG.RE.RF, YG.IM.RF Defines the RF gate impedance as an admittance.
YI.RE.RF, YI.IM.RF Defines the RF input impedance as an admittance.
YO.RE.RF, YO.IM.RF Defines the RF output impedance as an admittance.
YS.RE.RF, YS.IM.RF Defines the RF source impedance as an admittance.
ZD.RE.RF, ZD.IM.RF Defines the RF drain impedance as an impedance.
ZG.RE.RF, ZG.IM.RF Defines the RF gate impedance as an impedance.

807 ATLAS User’s Manual

NETWORK Mercury: Fast FET Simulator

ZI.RE.RF, ZI.IM.RF Defines the RF input impedance as an impedance.

ZO.RE.RF, ZO.IM.RF Defines the RF output impedance as an impedance.
ZS.RE.RF, ZS.IM.RF Defines the RF source impedance as an impedance.

RF Input Matching Network

RI.11.RF, CI.11.RF, Defines the RF input matching network, each parameter of the
LI.11.RF, RI.12.RF, CI.12.RF, z-parameter two-port is defined as a series RCL circuit.
LI.12.RF, RI.21.RF, CI.21.RF,
LI.21.RF, RI.22.RF, CI.22.RF,
LI.22.RF

GI.11.RE.RF, GI.11.IM.RF, Defines the RF input matching network as a set of g-parameters.

GI.12.RE.RF, GI.12.IM.RF,
GI.21.RE.RF, GI.21.IM.RF,
GI.22.RE.RF, GI.22.IM.RF
SI.11.RE.RF, SI.11.IM.RF, Defines the RF input matching network as a set of s-parameters.
SI.12.RE.RF, SI.12.IM.RF,
SI.21.RE.RF, SI.21.IM.RF,
SI.22.RE.RF, SI.22.IM.RF
ZI.11.RE.RF, ZI.11.IM.RF, Defines the RF input matching network as a set of z-parameters.
ZI.12.RE.RF, ZI.12.IM.RF,
ZI.21.RE.RF, ZI.21.IM.RF,
ZI.22.RE.RF, ZI.22.IM.RF

RF Output Matching Network

RO.11.RF, CO.11.RF, Defines the RF output matching network, each parameter of the
LO.11.RF, RO.12.RF, z-parameter two-port is defined as a series RCL circuit.
CO.12.RF, LO.12.RF,
RO.21.RF, CO.21.RF,
LO.21.RF, RO.22.RF,
CO.22.RF, LO.22.RF,

808 ATLAS User’s Manual

NETWORK Mercury: Fast FET Simulator

HO.11.RE.RF, HO.11.IM.RF, Defines the RF output matching network as a set of h-parameters.

HO.12.RE.RF, HO.12.IM.RF,
HO.21.RE.RF, HO.21.IM.RF,
HO.22.RE.RF, HO.22.IM.RF

SO.11.RE.RF, SO.11.IM.RF, Defines the RF output matching network as a set of s-parameters.

SO.12.RE.RF, SO.12.IM.RF,
SO.21.RE.RF, SO.21.IM.RF,
SO.22.RE.RF, SO.22.IM.RF

ZO.11.RE.RF, ZO.11.IM.RF, Defines the RF output matching network as a set of z-parameters.

ZO.12.RE.RF, ZO.12.IM.RF,
ZO.21.RE.RF, ZO.21.IM.RF,
ZO.22.RE.RF, ZO.22.IM.RF

Examples
NETWORK CLEAR
Do this before DC and small-signal AC simulations to remove the input and output matching
networks.
network zi.re.rf=50.0 zi.im.rf=0.0 zo.re.rf=50.0 zo.im.rf=0.0
This sets the input and output impedance to 50 Ohms at all harmonics.

809 ATLAS User’s Manual

POISSON Mercury: Fast FET Simulator

18.8 POISSON
POISSON allows you to solve Poisson's equation in the Y direction of a FET in a MERCURY
simulation. This command is valid only when using the MERCURY module.
Syntax
POISSON <cap simulation> <gate simulation> <look-up table
calculation>

Parameter Type Default Units

ASCII Logical False

BINARY Logical False

CAP Logical False

COMPRESSED Logical False

DE.DX Real 0.0 V/cm2

DEBUG.CAP Character

DEBUG.GATE Character

E.SURF Real 0.0 V/cm

GATE Logical False

IN.FILE Character

LOOK.UP.TABLE Logical False

NUM.RECESS.POINTS Integer

OUT.FILE Character

QUICK Logical False

RECESS Real 0.0 um

SAFE Logical False

STANDARD Logical True

V.GATE Real V

V.SUBSTRATE Real V

V.TYPICAL Real 0.0 V

VD.MAX Real 0.0 V

VD.MIN Real 0.0 V

VG.MAX Real 0.0 V

VG.MIN Real 0.0 V

VG.NUM Real

810 ATLAS User’s Manual

POISSON Mercury: Fast FET Simulator

VG.START Real V

VG.STOP Real V

VG.TYPICAL Real 0.0 V

VI.MAX Real 0.0 V

VI.MIN Real 0.0 V

VO.MAX Real 0.0 V

VO.MIN Real 0.0 V

Description

ASCII Saves the data file in a standard, uncompressed format.

BINARY Saves the data file in a binary format.
COMPRESSED Saves the data file in a compressed ASCII format.
OUT.FILE Specifies the name of the output file.
V.SUBSTRATE Specifies the potential on the substrate contact.

Cap Poisson Solution

CAP Defines that Poisson's equation is solved once, over an ungated slice of the
device.
DE.DX Specifies the gradient of the X direction electric field.
E.SURF Specifies the Y direction electric field at the surface of the slice.
RECESS Specifies the position of the surface of the slice.

Gate Poisson Solution

DE.DX Specifies the gradient of the X direction electric field.

GATE Defines that Poisson's equation is solved once, over a gated slice of the device. The
recess depth is automatically set to the position of the gate specified in the device
construction.
V.GATE Specifies the potential on the gate contact.

811 ATLAS User’s Manual

POISSON Mercury: Fast FET Simulator

Poisson Look-Up Table Calculation

DEBUG.CAP Specifies a file to save the results of the cap look-up table calculation.
This file is suitable for viewing in TONYPLOT but has no other purpose in a
device simulation.
DEBUG.GATE Specifies a file to save the results of the gate look-up table calculation.
This file is suitable for viewing in TONYPLOT but has no other purpose in a
device simulation.
IN.FILE Specifies the file name of a previously calculated Poisson look-up table.
If a table cannot be read, then you must calculate it.
LOOK.UP.TABLE Defines that the Poisson look-up table must be calculated.
NUM.RECESS.POINTS Specifies the number of points to use when calculating the cap look-up
table.
QUICK Specifies that a sparse look-up table should be calculated (16 points in the
cap look-up table and 128 points in the gate look-up table).
SAFE Specifies that a dense look-up table should be calculated (48 points in the
cap look-up table and 384 points in the gate look-up table).
STANDARD Specifies that a normal look-up table should be calculated (32 points in the
cap look-up table and 256 points in the gate look-up table)
V.TYPICAL Specifies a typical voltage that is used to initialize the Poisson
solution. This should be a gate voltage that generates a noticable 2D sheet
carrier density. But it doesn't generate too much additional charge at the
surface.
VD.MAX Specifies the maximum expected bias on the drain contact. This is used to
calculate the range over which to solve the gate look-up table.
VD.MIN Specifies the minimum expected bias on the drain contact. This is used to
calculate the range over which to solve the gate look-up table.
VG.MAX Specifies the maximum expected bias on the gate contact. This is used to
calculate the range over which to solve the gate look-up table.
VG.MIN Specifies the maximum expected bias on the drain contact. This is used to
calculate the range over which to solve the gate look-up table.
VG.NUM Specifies the number of points to use when calculating the gate look-up
table.
VG.START Sets the lower value of gate bias to use when calculating the gate
look-up table.
VG.STOP Sets the upper value of gate bias to use when calculating the gate look-up
table.
VG.TYPICAL This is a synonym for V.TYPICAL.

812 ATLAS User’s Manual

POISSON Mercury: Fast FET Simulator

VI.MAX This is a synonym for VG.MAX.

VI.MIN This is a synonym for VG.MIN.
VO.MAX This is a synonym for VD.MAX.
VO.MIN This is a synonym for VD.MIN.

Examples
poisson cap recess=0.0 out.file=cap.str
This solves Poisson's equation for an ungated slice with the surface at y=0. The results are
stored in a file cap.str.
poisson gate v.gate=0.0 out.file=gate.str
This solves Poisson's equation for a gated slice with a bias of Vg=0. The results are stored in
a file gate.str.
poisson look.up vg.min=-1.5 vd.max=5.0 out.file=look_up.dat
The Poisson look-up table must be calculated before a channel simulation is attemped
(basically before a SOLVE command). The simulation is going to have gate biases in the range
of 0 to -1.5 V and drain biases in the range of 0 to 5 V.

813 ATLAS User’s Manual

SURFACE Mercury: Fast FET Simulator

18.9 SURFACE
SURFACE allows you to define the surface topography of a FET in a MERCURY simulation.
This command is valid only when using the MERCURY module.
Syntax
SURFACE X=<x> Y=<y>

Parameter Type Default Units

X Real 0.0 um

Y Real 0.0 um

Description

X Gives the X coordinate of the point being defined.

Y Gives the Y coordinate of the point being defined.

Examples
surface x=0.5 y=0.0
surface x=0.6 y=0.1
surface x=1.4 y=0.1
surface x=1.5 y=0.0
These commands define a 0.1 um deep recess on the surface of the device. The bottom of the
recess is 0.8 um long and centered around x=1 um. The sides of the recess slope at 45°.

814 ATLAS User’s Manual

Chapter 19
MC Device
What is MC Device MC Device

19.1 What is MC Device

MC DEVICE is a module for performing full-band Monte Carlo simulation of silicon and
strained silicon devices. MC DEVICE is based on MOCA. The MOCA software was
developed by the Department of Electrical and Computer Engineering at the University of
Illinois at Urbana-Champaign.
MC DEVICE has two operating modes:
• A bulk mode, which is for fitting material parameters to experimental data. This mode
provides a unipolar, space-homogeneous transport model that can be used to investigate
femtosecond transients, impact-ionization coefficients, doping, field or stress-dependent
mobilities, and many other bulk transport properties. This mode is intrinsically unipolar
(i.e., only the behavior of electrons or holes can be simulated).
• A 2D mode, which is for performing device MC simulation with regions of different
materials, ohmic and blocking contacts, and so on. A non-linear Poisson solver and
Schroedinger solver are included. This mode shares the band structure and scattering
models with the bulk mode.
Providing the two modes above in a single module has advantages of maintainability and
consistency. For example, a scattering mechanism added to the bulk mode is available in the
2D mode. The complication of sharing some models but not other models between modes is
tolerable. For example, position-dependent impurity scattering is necessary in the 2D mode
because the impurity concentration is not uniform over the device. A different model that
considers only one doping level is used for the bulk mode.

816 ATLAS User’s Manual

Using MC Device MC Device

19.2 Using MC Device

The MC DEVICE module can be used in all modes of operations of ATLAS (see Section 2.3
“Modes of Operation”). We recommend that you run MC DEVICE in the DECKBUILD’s run-
time environment as described in Sections 2.3.1 “Interactive Mode With DeckBuild” through
2.3.4 “Running ATLAS inside Deckbuild”.
The MC DEVICE module can also be used in batch mode without DECKBUILD, although this is
not recommended by Silvaco. Running in batch mode without DECKBUILD can be used for
reading input files in MOCA input format (see Section 19.3.3 “Migrating From MOCA Input
Format”). If you don’t want the overhead of a DeckBuild window, then use DECKBUILD with
the no-windows mode (see Section 2.3.3 “No Windows Batch Mode With DeckBuild”).
Many important features such as variable substitution, automatic interfacing to process
simulation, and parameter extraction are only available inside the DECKBUILD run-time
environment.
To use the MC DEVICE module in batch mode without DECKBUILD, use the statement:
atlas <input file>
<input file> is the name of an input file that contains the simulation parameters. The
input file can have any extension. Here, we will use.in. The following statement line runs
MC DEVICE using the input file mcdeviceex01.in:
atlas mcdeviceex01.in
The file defaults.in contains default values for all parameters; the input file overrides
those values that need to be changed. For example, the default lattice temperature for the
material is 300 K. So you do not need to set the TEMP parameter to run a room-temperature
simulation. Section 19.3 “Input Language and Syntax” describes the format of the input file.
Appendix D: “MC Device Files” gives the complete default configuration file.
At startup, MC DEVICE reads the defaults.in file (i.e., $<S_INSTALL>/lib/atlas/
<version>/common/mcdevice/defaults.in). Next, MC DEVICE reads your input file
(i.e., mcdeviceex01.in). Simulation parameters defined in your input file override those in
the defaults.in file. Next, upon processing a SOLVE statement, MC DEVICE loads database
files (i.e., band structure or scattering rate data or both) from a database directory (i.e.,
$<INSTALL>/lib/mcdevice_db/<version>).
You can modify the file MC DEVICE defaults.in file in the directory where this file is
installed (namely, $<S_INSTALL>/lib/atlas/<version>/common/mcdevice). But, we
do not recommend this. Instead, if you need to modify the defaults.in file, we
recommend that you set the MCDEVICEDEFDIR environment variable to a new directory. Copy
$<S_INSTALL>/lib/atlas/<version>/common/mcdevice/defaults.in to that new
directory. Then, make whatever changes you want to the copied defaults.in file. If the
MCDEVICEDEFDIR environment variable is unset, it is automatically set in the start-up ATLAS
script to $<S_INSTALL>/lib/atlas/<version>/common/mcdevice. On Windows
platforms, MCDEVICEDEFDIR begins with the driver letter and a colon (i.e,. C:/).

817 ATLAS User’s Manual

Using MC Device MC Device

You can modify the MC DEVICE database files in the directory where they are installed
(namely, $<S_INSTALL>/lib/mcdeivce_db/<version>). But, we do not recommended
this. Instead, if you need to modify the database files, we recommend that you set the
MCDEVICEDIR environment variable to a new directory. Copy $<S_INSTALL>/lib/
mcdeivce_db/<version> to that new directory. Then, make whatever changes you want to
the copied database files. If the MCDEVICEDIR environment variable is unset, it is
automatically set in the start-up ATLAS script to $<S_INSTALL>/lib/mcdevice_db/
<version>. On Windows platforms, MCDEVICEDIR begins with the driver letter and a colon
(i.e,. C:/).
Upon processing SOLVE statements, MC DEVICE writes a summary file that reports all the
user-definable parameters that have been used for the solve. The name of this file is by default
summary.out. You can change this name in the input file. See the description of the OUTPUT
statement in Section 19.3.4 “Commonly-Used Statements”. You can permanently change this
name by editing defaults.in. Most of the simulation parameters affect the physical
models (e.g., optical-phonon energies) and will not be changed often. Others, like the number
of iterations or the simulation mode, will be redefined at every run. There are also some
numerical parameters that cannot be changed (e.g., , , and the unstrained lattice constant of
silicon).

818 ATLAS User’s Manual

Input Language and Syntax MC Device

19.3 Input Language and Syntax

19.3.1 The Input Language
MC DEVICE reads input files in MC DEVICE input format with data grouped on to statements.
Each statement can have parameter settings or no parameter settings.
For example, the BULK statement initializes the bulk simulation parameters and looks like
this:
BULK FIELD=(100e3,5e3) DOPING=1e17 \
CONC=0.0
The body of the statement follows its name and is continued by using the backslash (\) at end
of each line except the last line of the statement. The body of the statement may be empty or
it may contain one or more settings of simulation parameters. For example, the (electric)
field, the doping, and the (carrier) concentration are parameters on the BULK statement.
A parameter setting has the form:
<parameter>=value
The value can be an integer, a floating-point number, a string, or a vector of one of the three
simple types. A vector is a list of values separated by commas and enclosed in parentheses.
For example, the FIELD parameter on the bulk statement has a list of two floating-point
numbers as its value: the first is the value for TIME<0; the second is the value for TIME>=0. In
the example input above, we are simulating the sudden reduction of the electric field from
100e3 V/cm to 5e3 V/cm. Normally, you should provide the value in cgs units although
lengths are given in um by default (see the UNITS statement in Section 19.5.1 “Device
Geometry”).

Note: Field values for TIME<0 and TIME>=0 field may be set equal. In this case, TIME<0 is used to improve
the quality of the initial conditions used at TIME=0. Output averaging in MC DEVICE is done for
TIME>=0.

A vector should include values of the same type. MC DEVICE can make conversions between
real and integer types. For clarity, we recommend you use values with the same type as the
parameter. All example input files and the defaults.in files demonstrate this approach. MC
DEVICE stops if you set a parameter using an incompatible type (e.g., a string for a floating-
point parameter).

819 ATLAS User’s Manual

Input Language and Syntax MC Device

19.3.2 Statement allocation and usage

Memory associated with MC DEVICE statements is either allocated once or is allocated for
each occurrence. The allocation behavior of each statement relates to its usage. For example,
memory associated with the BULK statement, shown above, is allocated once. This means that
if the BULK statement occurs more than once in the input file before a SOLVE statement, then
each occurrence of the BULK statement changes parameter settings associated with prior BULK
statements beginning with the BULK statement in the defaults.in file. All MC DEVICE
statements that appear in the defaults.in file (except for the MATDEF statement) are
allocated once. The SOLVE statement is also allocated once, although it does not appear in the
defaults.in file.
The MC DEVICE statements associated with mesh, regions, materials, and doping (see Section
19.5.1 “Device Geometry”) are allocated for each occurrence. All statements involving other
types of regions (e.g., current regions and Schrödinger regions) are also allocated for each
occurrence. The parameters on statements that are allocated for each occurrence are
initialized to zero (or null). Therefore, non-zero settings from prior occurrences are not
retained. All MC DEVICE statements that are allocated for each occurrence (except the
MATDEF) do not appear in the defaults.in file.

19.3.3 Migrating From MOCA Input Format

Although MC DEVICE can read input files in MOCA input format as an aid to users migrating
from MOCA to MC DEVICE (e-mail [email protected] for details), Silvaco strongly
recommends that MOCA users transform their input files to MC DEVICE input format.
Files in MOCA input format look like this:
bulk {
field = {100e3,5e3};
doping = 1e17;
conc =0.0;
}
If you compare this bulk statement to the same statement definition (in MC DEVICE input
format) in Section 19.3.1 “The Input Language”, you can see the differences. Semicolons
terminate parameter assignments rather than white space. Curly braces enclose statement
bodies rather than using continuation characters. Curly braces enclose vectors rather than
using parentheses.
To transform a file from MOCA input format into MC DEVICE input format, you will need to
make the following changes:
1. Add the line “MCDEVICE” as the first uncommented line of the input file.
2. Delete any semicolons.
3. Delete any curly braces that surround the body of a statement.
4. Use a continuation character (\) to continue statements to the next line. Move any
comments that were inside a statement to outside of that statement.
5. Replace curly braces that surround parameter vectors with parentheses.
6. Remove any braces that surround the value of the n parameter on the PARTICLE
statement. For example, particle {n={10000};}; becomes PARTICLE N=10000. Do

820 ATLAS User’s Manual

Input Language and Syntax MC Device

likewise, for any other scalar parameter which unnecessarily used braces surrounding its
value. Using array syntax on scalar parameters is allowed in the MOCA input format but
is not allowed in the MC DEVICE input format.
7. Replace any MOCA’s 1d statements with MC DEVICE’s oned statements. In MC DEVICE
input format, statements names and parameter names may not begin with a number, so the
1d statement has been changed to ONED.
8. Replace any use of the p parameter on an IMPACT statement with the PARAM parameter.
Both MC DEVICE input format and MOCA input format have changed from using the p
parameter to using the PARAM parameter to distinguish it from the symbol for phosphorus,
P.
9. Insert the contents of the $MCDEVICEDIR/“bands2d.hole.in” file into an input file that
uses carrier=h (for MC holes) on the ALGO statement. (The “bands2d.*.in” files are
not used by MC DEVICE, so this input must now reside in your input file.)
10. Insert the contents of the $MCDEVICEDIR/“bands2d.elec.ssi.in” file into an input
file that uses TYPE=1 (for MC electrons in strained silicon) on the mmat statement. (The
“bands2d.*.in” files are not used by MC DEVICE, so this input must now reside in your
input file.)
11. If you used the IMPORT statement, change the TYPE and filename parameters as needed.
(See Section 19.5.8 “Importing Data”.) The default in MC DEVICE is TYPE=8 for Silvaco
structure files.
12. If you have meshing or boundary specifications in centimeters in your input file, add the
statement units length=cm. By default, MC DEVICE uses units length=um.
Alternatively, you may want to change values in cm to um by multiplying them by 1e4.
13. Add a SOLVE statement at the end of your input file. The SOLVE statement is not required
in MOCA input format. In MOCA input format, all explicit SOLVE statements are ignored
and a single implicit SOLVE statement is performed after processing all input. The SOLVE
statement is required in MC DEVICE input format. In MC DEVICE input format, SOLVE
statements are processed explicitly at the point at which they appear in the input file.
You can add the GO ATLAS and QUIT statements to the start and end of your input file. These
statements tell DECKBUILD to start and stop ATLAS. These statements are ignored when you
use them outside DECKBUILD.
You not required to convert statement and parameter keywords to uppercase when converting
an input file in MOCA input format to MC DEVICE input format. Our convention in the body
of this manual is to use uppercase for statement and parameter keywords to make these stand
out in the input file. (Our defaults and examples are provided, however, using lowercase
keywords.) Our convention is to use lowercase for keywords in MOCA input format, since
UIUC MOCA only supported lowercase keywords. Uppercase, lowercase, and even
mixedcase keywords are supported in MC DEVICE. In MC DEVICE, arbitaryly mixedcase
keywords are supported in both the MC DEVICE input format and the MOCA input format.
Abbreviated keywords are supported in exclusively in MC DEVICE input format.

821 ATLAS User’s Manual

Input Language and Syntax MC Device

19.3.4 Commonly-Used Statements

Many input statements are used to establish MC DEVICE’s default configuration and are not
commonly used in input files. As mentioned in Section 19.2 “Using MC Device”, the default
values of most parameters are set in the defaults.in file. Any parameters that do not
appear in the defaults file take the value 0|FALSE|NO, 0.0, or a null string, depending on the
type of the parameter.

Note: MC DEVICE uses 0, FALSE, and NO interchangeably. MC DEVICE also uses 1, TRUE, and YES
interchangeably. As part of MC DEVICE input format, MC DEVICE also supports the use of bare
parameter names (i.e. ALGO RESTART) to set a boolean variable to true, and bare parameter names
preceeded by the negation symbol, ^, (i.e. ALGO ^RESTART) to set a boolean variable to false.

The following is a description of the most commonly used input statements and their
parameters. For a more complete list of input statements, see Section 19.3.5 “List of
Statements”.
The MCDEVICE statement below must be included at the start of every MC DEVICE input file
or interactive input session.
MCDEVICE
The statement above must be the first non-comment line and indicates that the input that
follows pertains to the MC DEVICE module.
The ALGO statement contains the fundamental Monte Carlo algorithm parameters:
ALGO MODE=2 CARRIER=E DT=1e-15 ITER=6000 TRANS=4000 \
RESTART=YES SEED=65428
The MODE parameter controls the type of simulation to be run. Set MODE=0 perform a bulk
simulation. Set MODE=2 to perform a 2D simulation. CARRIER is the Monte Carlo carrier type,
Set CARRIER=E to simulate electrons with an MC transport model. Set CARRIER=H to
simulate holes with an MC transport model. When MODE=0 (bulk simulation), only one carrier
type is possible and it is set by the CARRIER parameter. When MODE=2 (2D), both carrier
types are possible. The carrier type not specified by the carrier parameter will be treated with
the constant quasi-Fermi level model (see “Regions” on page 834 and Section 19.7.1 “Self-
Consistent Simulation”). The DT parameter is the time step for the free-flights. The default is
21015 s. Use a lower value for high fields (100 kV/cm and up) or for relatively more
accurate free-flight calculations. The ITER parameter is the number of iteration steps in a
simulation and relates to the simulation time as TIME=ITER*DT. The TRANS parameter is the
number of time steps (DT) taken as an initial transient. During the initial transient, no
estimator averaging and estimator output is performed.

Note: MC DEVICE uses parameters which begin with the letter D (DT, DX, and DE) to denote differential
intervals of quantities (here time, space, and energy). Similarly, MC DEVICE uses parameters, which end
with the word STEP (TSTEP, XSTEP, and ESTEP) to denote integer multiples of the corresponding
differential intervals. For example, TSTEP is an integer parameter, DT is a floating-point parameter (in
seconds), TSTEP denotes an integer multiple of DT, and a time of TSTEP*DT seconds.

If RESTART is 0, as it is by default, a new simulation will be started. Otherwise, the carrier

states (positions and momenta) are read from the "status.in" file.

822 ATLAS User’s Manual

Input Language and Syntax MC Device

SEED is the initial seed for the random-number generator. Since the outcome of the MC
simulation depends on a sequence of random numbers, you can use the variable SEED to
perform different simulations of the same physical system.
PARTICLE N=20000 NNORM=10000 CONC=0 DZ=-1.0
The PARTICLE statement defines parameters for the particle (electron or hole) ensemble. N is
the maximum number of particles. If NNORM>0, then NNORM sets the number of particles to use
at the beginning of the simulation to normalize the charge. NNORM must be less than or equal
to N. If NNORM=0 (as it is by default), then NNORM=N is assumed.
If the device width, DZ, is greater than 0.0, then DZ is the device width in cm. If CONC>Ntot,
where Ntot is the internally-calculated initial integrated particle concentration (cm-1) and
where CONC is a user-specified initial integrated particle concentration (cm-1), then the device
width is NNORM/CONC. (Ntot and CONC have the units cm-1 since they are volumetric
concentrations integrated over x and y.) If neither of the two previous conditions are true, then
the device width is determined as N/1.5/Ntot. If NNORM, CONC, and DZ are left at their
default values of 0, 0.0, and -1.0 respectively, then the choice above for the device width
means that the initial particle number is (2/3)*N. This is normally the case. The fraction of 2/3
represents the fraction of initially-used particle memory compared to allocated particle
memory. This choice allows the particle number to vary during the simulation.

Note: For the mcdevicex02.in example (Section D.3.2 “A 25-nm n-MOSFET: mcdeviceex02”), the device
width and initial particle number are determined as they normally are where NNORM, CONC, and DZ are left
at their default values of 0. In this example, the device width is N/1.5/Ntot=40,000/1.5/(1.76182E+09
cm-1)=1.51359E-05 cm and the initial particle number is (2/3)*N=(2/3)(40,000)=26,666.

You can use CONC to set a total charge in the device. This is useful for strong-inversion
devices, where you need to help MC DEVICE establish the amount of stored charge in the
device. This may be necessary because the transient time of the charging is much longer than
can realistically be simulated with MC DEVICE.
MAT TEMP=77.0 CUT=(1.0,0.0,0.0)
This MAT statement defines material parameters (i.e., temperature and wafer cut or crystal
orientation. TEMP is the lattice temperature in Kelvin. CUT is a triplet of real numbers that give
the orientation of k-space with respect to real space. When properly normalized, they are the
direction cosines of the X in the kxkykz space. For example, 1.0, 1.0, 0.0 means that the
X-axis of real space (i.e., the direction of the electric field) is in the 110 direction of the
crystal.
OUTPUT SUMMARYOUTFILE="summary.out" DIFFLOGFILE=”diff.log” \
CURRENTLOGFILE=”current.log” SOLSTRFILE=”sol.str”
SCATOUTFILE="scat.out" \
TSTEP=100 XSTEP=(1,1,1) ESTEP=50 RESTART=500 \
WPSTEP=PINF INIT=NO HYDRO=NO HIST=YES QETA=NO OUTFILES=NO

823 ATLAS User’s Manual

Input Language and Syntax MC Device

The OUTPUT statement has parameter settings associated with output. The SUMMARYOUTFILE
parameter sets the name of a file written for each SOLVE statement, which holds the input
related to each solve statement. If you don’t change the value of SUMMARYOUTFILE from its
default value of summary.out, then after your simulation is complete, summary.out will
hold the input used for your last solve statement. The DIFFLOGFILE parameter sets the name
of the file used to hold time-averaged quantities for a bulk simulation (see diff.log in
Section D.2.1 “Silvaco Output Files”). The CURRENTLOGFILE parameter sets the name of the
file used to hold currents for a 2D simulation (see current.log in Section D.2.1 “Silvaco
Output Files”). The SOLSTRFILE parameter sets name of the file used to hold the time-
averaged and instanteous quantities related to a 2D solution (see sol.str in Section D.2.1
“Silvaco Output Files”).
The SCATOUTFILE parameter sets the name of the file used to hold the scattering rates that
were calculated and used during your simulation (see scat.out in Section D.2.4 “Scattering
Output Files”).

Note: The aliases SUMMARYFILE (for SUMMARYOUTFILE) and SCATFILE (for SCATOUTFILE) have
been provided on the OUTPUT statement for backward compatibility with MOCA.

The TSTEP parameter is the number of time steps (DT on the ALGO statement) between output
of observables (or estimators). The XSTEP parameter is the size of spatial bins (DX, DY, DZ) in
units of mesh cells used for output. The ESTEP parameter is the size of energy bins in meV
used for output. The RESTART parameter is the number of time steps between dumps of the
particle status to the restart.out file. The WPSTEP parameter is the number of time steps
between successive writes of the instantaneous electric potential (voltage) and the
instantaneous MC carrier concentration. The init parameter indicates MC DEVICE will
write out the initial electrostatic potential (voltage) and initial MC carrier concentration at 5
points during the initialization process. The HYDRO parameter indicates MC DEVICE should
write out 2D estimators for hydrodynamic parameter extraction. The HIST parameter
indicates that MC DEVICE should write the energy histograms. The QETA parameter indicates
that MC DEVICE should write the quantum potential.

Note: The OUTFILES parameter on the OUTPUT statement is an important control variable in MC DEVICE.
The OUTFILES parameter enables/disables the writing of almost all *.out files generated by MC
DEVICE. By default, the outfiles parameter is set to 0|FALSE|NO so that *.out files are not
written. (The summaryfile is treated as an exception and is always created.) Setting
OUTFILES=FALSE results in no *.out files in the directory where the test is run and makes it easy to
see the output files that can be used with other Silvaco tools, especially TONYPLOT. To enable the
generation of all *.out files, set the OUTFILES parameter to 1|TRUE|YES. If other boolean
parameters are specifically associated with a *.out file (like the init parameter above), then you must
also set this parameter to 1|TRUE|YES if it is not true by default.

DEBUG NPICK=0
The DEBUG statement defines debugging parameters like a particle to track and holds the
output parameter, which controls the generation of virtually all non-Silvaco-standard file
formats.

824 ATLAS User’s Manual

Input Language and Syntax MC Device

If NPICK>0, MC DEVICE tracks the corresponding particle index, dumping debugging data to
various files. If NPICK<0, a large amount of debugging data is stored for the entire simulation
(enough to fill up a 1 GB disk in a few hours). Also when NPICK0, many internal self-
consistency checks are performed to catch bugs. The status of subroutine calls are then
printed out at the console for each iteration.
SOLVE
If the SOLVE statement includes no parameters settings, it indicates that an MC solve should
be performed for the biases specified by the input that precedes it (see “Biasing” on page
838).
SOLVE VGATE=1.0 VDRAIN=1.0
The SOLVE statement can also include one or more parameter settings of the form:
V<NAME>=#. For these settings, # is a the applied voltage in V. <NAME> must be from the list:
GATE, GG, DRAIN, DD, SOURCE, BULK, SUBSTRATE, EMITTER, EE, COLLECTOR, CC, BASE, BB,
ANODE, CATHODDE, FGATE, CGATE, NGATE, PGATE, WELL, NWELL, PWELL, CHANNEL, and
GROUND. Multiple SOLVE statements may be used in an input file. The V<NAME> settings are
preserved from one SOLVE statement to the next. The V<NAME> settings are updated after the
ramping SOLVE (see example below) so that they can be used in subsequent solves. The
V<NAME> parameters default to 0 on the first SOLVE, so you only have to set non-zero biases.
You can generate all output files when MC DEVICE processes the SOLVE statement. To retain
output files from one SOLVE after executing a second SOLVE, change the output files using an
OUTPUT statement between the SOLVE statements.
SOLVE NAME=”drain” VSTEP=1.0 VFINAL=2.0
If you set VSTEP!=0 or NSTEPS>0 on the SOLVE statement, then MC DEVICE performs a
voltage ramp in 2D simulations. The voltage ramp is done on the contact specified by the
NAME parameter. The voltage ramp completes after solving at a voltage of VFINAL. When
USELOG=0|false|no (as it is by default), MC DEVICE will use VSTEP for uniform spacing
on a linear scale. When USELOG=1|true|yes, MC DEVICE will use VSTEP for uniform
spacing on a logarithmic scale. The USELOG parameter defaults to 0 on each SOLVE statement.
Transient solves are performed for each bias condition beginning with those specified on the
same or prior SOLVE statements.
The *current.log and *sol.str files produced during each voltage ramp step are saved by
inserting the contact voltage after the base file name. For example, if the voltage on the drain
were 1.0 V prior ot the solve above, then current_vdrain=1.log will pertain to the first
ramp step. In this case, only one step is performed. The currents after this voltage step are
saved in current_vdrain=2.log. Correspondingly, for this example, the solution structure
files are saved as sol_vdrain=1.str and sol_vdrain=2.str. The NAME parameter must
match a contact name as specified by the NAME parameter on a REGION statement (with
TYPE=CONTACT). After the ramping SOLVE is complete, the corresponding V<GATE>
parameter is automatically set to VFINAL so that this value will be used in subsequent solves.
The VSTEP and VFINAL parameters are in V. The values of the voltage included in the
filenames are also in V. The VSTEP, VFINAL, NAME, and NSTEPS parameters default to 0 or
null on each SOLVE statement.

Note: MC DEVICE checks that VSTEP takesw the voltage evenly to VFINAL. It stops with an error if it does
not. Please choose VSTEP and VFINAL consistently or use NSTEPS instead of VFINAL or VSTEP.

825 ATLAS User’s Manual

Input Language and Syntax MC Device

The results in *current.log pertaining to the last simulation time for each voltage in the
ramp are saved into the *current_ramp.log file. To specify the file name, use
CURRENTRAMPLOGFILE parameter on the OUTPUT statement. Since this file contains both the
voltage at the contact and the currents, you may use it to plot a current-voltage (I-V) curve.
SOLVE NSTEPS=3 FFINAL=1e6 USELOG=yes
If you set FSTEP!=0 or NSTEPS>0 on the SOLVE statement, then MC DEVICE performs an
electric field ramp in bulk simulations. The field ramp completes after solving with the field
at FFINAL. When USELOG=0|false|no (as it is by default), MC DEVICE will use FSTEP for
uniform spacing on a linear scale. When USELOG=1|true|yes, MC DEVICE will use FSTEP
for uniform spacing on a logarithmic scale. The USELOG parameter defaults to 0 on each
SOLVE statement. Transient solves are performed for each ramp step beginning with those
specified by the FIELD parameter on the BULK statements (see Section 19.5.1 “Device
Geometry”). To simplify solves during field ramping, FSTEP is added to both components of
the FIELD vector (namely both the value for time<0 and the value for time>=0). If you want
to control each component of the FIELD parameter on the BULK statement separately, perform
the ramp steps as a sequence of separate solves by setting the FIELD parameter on the ALGO
statement as desired. Then, change the DIFFLOG file parameter on the OUTPUT statement
between the SOLVE statements.

Note: MC DEVICE checks that FSTEP takes the field evenly to FFINAL. It stops with an error if it does not.
Please choose FSTEP and FFINAL consistently or use NSTEPS instead of FFINAL or FSTEP.

The results in *diff.log files produced during each field ramp step are saved by inserting
the field after the base file name. For example, if the field were 1000 V/cm prior to the solve
above, then diff_field=1000.log will pertain to the first ramp step. In this case, then only
two more iterations are performed at 1e4 V/cm and 1e5 V/cm. The results for these ierations
are saved in diff_field=1e4.log and diff_field=1e5.log respectively. The FSTEP and
FFINAL parameters are fields in V/cm. The values of the fields included in the filenames are
also in V/cm. The FSTEP, FFINAL, and NSTEPS parameters default to 0 on each SOLVE
statement.
The results in the *diff.log file pertaining to the last simulation time for each field ramp
step will be saved into the the *diff_ramp.log file. To specify the file name, use the
DIFFRAMPLOGFILE parameter on the OUTPUT statement. Since this file contains both the field
and the bulk results (like mobilities), you may use this file to plot a mobility-field curve.
When you use the drift diffusion transport model to obtain the solution for a particular DC
bias point, it is reasonable to think that you get the same solution for the same bias conditions
even when the DC bias points that preceed the final point differ. This is normally not true
(because normally each solution uses the prior solution as an initial conditions), but the
solutions are close enough to each other that you can think of them as being the same.
When you use the Monte Carlo transport model (MC DEVICE), small variations in the initial
conditions can have a relatively larger effect on the solutions than they do for the drift
diffusion model. To keep this behavior similar to what you are accustom to with drift
diffusion, we have implemented the SOLVE statement in MC DEVICE to exactly reproduce the
results (on a particular hardware platform) whenever the bias conditions are exactly the same
regardless what the preceeding bias point was. For example, this means that the solution you
obtain as the last step of a ramp will be identical to the solution you obtain by a single solve
for the final condition of the ramp. This means that you can add solution points to your

826 ATLAS User’s Manual

Input Language and Syntax MC Device

results without re-running the entire ramp. The results you obtain for the additional single
solves are identical to those you would obtain by re-running the entire ramp.
The QUIT statement stops execution at the point in the input where the QUIT statement
appears. All input lines after the QUIT statement are ignored. The QUIT statement should be
the last statement in your input file. The QUIT statement must be included when you run
DECKBUILD in batch mode and use DECKBUILD’s outfile command-line option. Otherwise,
your output file will finish with an ATLAS> prompt and the END-OF-RUN message of MC
DEVICE will be omitted.

19.3.5 List of Statements

Table 19-1. Lists of Statements

Statement Description

AC Acoustic phonon scattering. See Section 19.7.5 “Phonon Scattering”.

ALGO Simulation mode parameters. See Section 19.3 “Input Language and Syntax”.
BANDS Band structure information. See Section 19.7.2 “Band Structure”.
BIPOLAR Pipe for interprocess communication. See Section 19.5.5 “Bipolar”.
BULK Bulk simulation parameters. See Section 19.4 “Bulk Simulation”.
CREGION Current estimation region. See Section 19.5.3 “Estimators”.
CROSSEST Boundary crossing estimators. See Section 19.5.6 “Track Boundary Crossings”.
DEBUG Toggle debugging on/off. See Section 19.3 “Input Language and Syntax”.
DEVICE Device applied voltages. See Section 19.5.1 “Device Geometry”.
DOPING Doping profile definition. See Section 19.5.1 “Device Geometry”.
DTREGION Fine time-step region. See Section 19.5.7 “Time Step Regions”.
EBREGION Eta barrier region definition. See Section 19.7.8 “Size Quantization”.
EEREGION Energy estimator region definition. See “Energy Regions” on page 842.
ENHAN Statistical enhancement. See Section 19.6 “Statistical Enhancement”.
EREGION Statistical enhancement region. See Section 19.6 “Statistical Enhancement”.
ESTIM1D 1D averaging. See Section 19.6.6 “Statistical Enhancement Statements”.
FIELD Initial field definition. See “Biasing” on page 838.
FFREGION Fixed-field region definition. See Section 19.5.4 “Initial Conditions”.
FINAL Final state selection. See Section 19.7.4 “Coulomb Interaction”.
GO DECKBUILD control. See Section 19.3.3 “Migrating From MOCA Input Format”.
ICLREGION Initial concentration limit region. See Section 19.5.4 “Initial Conditions”.

827 ATLAS User’s Manual

Input Language and Syntax MC Device

Table 19-1. Lists of Statements

Statement Description

IMPACT Impact ionization parameters. See Section 19.7.6 “Impurity Scattering”.

IMPORT Import initial guess. See Section 19.5.8 “Importing Data”.
INJECT 2D injection parameters. See Section 19.5.2 “Ohmic Contacts”.
LINBIAS Additional linear biasing. See “Biasing” on page 838.
MAT Material parameters. See Section 19.3 “Input Language and Syntax”.
MATDEF Material definition. See Section 19.5.1 “Device Geometry”.
MCDEVICE Module specification. See Section 19.3.4 “Commonly-Used Statements”.
OUTPUT File outputs. See Section 19.3 “Input Language and Syntax”.
PARTICLE Number of particles. See Section 19.3 “Input Language and Syntax”.
PHON High-energy phonon scattering. See Section 19.7.5 “Phonon Scattering”.
POISSON Poisson solver parameters. See Section 19.7.1 “Self-Consistent Simulation”.
QREGION Quantum region definition. See Section 19.7.8 “Size Quantization”.
QUANTUM Quantum correction parameters. See Section 19.7.8 “Size Quantization”.
QUIT Stop execution. Section 19.3.4 “Commonly-Used Statements”.
QSS Surface charge parameters. See Section 19.5.4 “Initial Conditions”.
REFLECT Properties for reflection. See Section 19.5.2 “Ohmic Contacts”.
REGION Region definition. See Section 19.5.1 “Device Geometry”.
RESIST Contact resistance information. See Section 19.5.2 “Ohmic Contacts”.
RIDLEY Ridley model for impurity scattering. See Section 19.7.6 “Impurity Scattering”.
SCHRMESH Schrödinger solver mesh definition. See Section 19.7.8 “Size Quantization”.
SCHRREGION Schrödinger region definition. See Section 19.7.8 “Size Quantization”.
SEREGION Scattering estimation region. See Section 19.7.4 “Coulomb Interaction”.
SMREGION Smoothing region definition. See Section 19.6 “Statistical Enhancement”.
SOLVE Perform MC solve. See Section 19.3.4 “Commonly-Used Statements”.
SSPARAM Surface scattering parameters. See Section 19.7.6 “Impurity Scattering”.
SSREGION Surface scattering region. See Section 19.7.6 “Impurity Scattering”.
TRACKSCAT Track scattering region definition. See Section 19.7.6 “Impurity Scattering”.
TRAJ Track particle trajectories. See Section 19.7.6 “Impurity Scattering”.

828 ATLAS User’s Manual

Input Language and Syntax MC Device

Table 19-1. Lists of Statements

Statement Description

TTEMP Transverse temperature calculation region. See Section 19.7.8 “Size Quantization”.
TUNEL Gate tunneling current calculator. See Section 19.7.7 “Tunneling”.
UNITS Set units used in other statements. See Section 19.5.1 “Device Geometry”.
XL X-L phonon scattering. See Section 19.7.5 “Phonon Scattering”.
XMESH X-mesh definition. See Section 19.5.1 “Device Geometry”.
XXF X-X f-type phonon scattering. See Section 19.7.5 “Phonon Scattering”.
XXG X-X g-type phonon scattering. See Section 19.7.5 “Phonon Scattering”.
YMESH Y-mesh definition. See Section 19.5.1 “Device Geometry”.
ZSREGION Zero-scattering region definition. See Section 19.7.6 “Impurity Scattering”.

829 ATLAS User’s Manual

Input Language and Syntax MC Device

19.3.6 Accessing The Examples

MC DEVICE is supplied with standard examples that demonstrate the tool. We recommended
that new users study the examples and use them as a starting point for creating their own sim-
ulations. Please learn how to access, load, plot, and run these examples by following the pro-
cedure below.
The examples are accessed from the menu system in DECKBUILD.
1. Start DECKBUILD with ATLAS as the simulator as described in Section 19.2 “Using MC
Device”.
2. Click the MainControl button to open this pull-down menu.
3. Click Examples... button to open the DECKBUILD: Examples window (see Figure 19-1).

Figure 19-1: DeckBuild: Examples Window

4. Scroll down to show MCDEVICE: Monte Carlo Appliation Examples by clicking on

the bottom arrow of the slider at the right of the window.
5. Choose MC DEVICE by double clicking the MC DEVICE line of the index. A list of
examples will appear (see Figure 19-2). These examples illustrate different types of
devices or different applications of MC DEVICE.

830 ATLAS User’s Manual

Input Language and Syntax MC Device

Figure 19-2: MC Device Examples Window

6. Choose each example by double clicking that item in the list. A description of the
example will appear in the window. This description gives important details about how to
use MC DEVICE. Pleaes read this before proceeding. We recommend that you begin with
example 1 and go in order since important concepts and statements are described only in
the first example in which they appear.
7. Click on the Load example button. All files associated with that example are copied into
the current working directory. The input command (*.in) file is loaded into DECKBUILD.
The sample output (*_diff.log or *_current.log and *_sol.str) files produced on
x86-64-linux are copied to the working directory. TONYPLOT set (*.set) files associated
with each output file are also copied into the local working directory. DECKBUILD will
overwrite exiting files with fresh copies when you click the load button and click to
confirm.
8. Look over the input file in DECKBUILD. One or more tonyplot statements are included
near the end of the file to plot some of the simulation results. You may view the sample
results without running the example. To do this, highlighting the first tonyplot
statement and click the line button to have DECKBUILD to prepare to start at this line.
Then, click the next button to execute the next line or click the cont button to continue
with all remaining lines (to generate all the plots).
9. To run the entire example, click on the run button. The examples will finish by
generating the plots in Tonyplot. Each tonyplot statement near the end of the input file
will produce a separate TONYPLOT window.

831 ATLAS User’s Manual

Bulk Simulation MC Device

19.4 Bulk Simulation

Bulk simulation parameters are specified on the BULK statement.
BULK FIELD=(10e3,200e3) DOPING=1e18 CONC=4e17 MEMORY=NO MMFRAC=0.0
As mentioned in Section 19.3.1 “The Input Language”, the FIELD parameter sets the values
of the electric field for TIME<0 and TIME>=0 in V/cm. Splitting the simulation into two parts
serves two purposes. First, waiting for a transient time before computing time averages.
Second, allowing the time-dependent study of carrier dynamics on the femtosecond or
picosecond scale. The first case obviously applies only when the electric field before and after
TIME=0 is the same. The second case applies when the field changes abruptly.

The DOPING parameter is the background doping level in cm-3.

The CONC parameter allows you to specify the carrier concentration. This is relevant for a
bulk simulation since it affects the screening of impurities and the scattering of carrier with
ionized-impurities.
If you set the MEMORY parameter to 1|TRUE|YES, the status of the carrier ensemble will be
maintained during the simulation. The energy and other quantities associated with MC
carriers are recorded at time steps determined by the step field of the OUTPUT statement.
This slows down MC DEVICE but allows the time-dependent study of the microscopic
histories of the particles. The file memory.out is printed out at the end of the run with
statistics on impact ionization events.
For modeling the bulk behavior of strained silicon, the MMFRAC parameter sets the effective
strain level in the bulk from 0.0 to 1.0. When modeling the bulk behavior of strained silicon
layers on a relaxed Si1-xGex substrate, MMFRAC refers to x, the fraction of Germanium in the
substrate. When modeling the bulk behavior of strain silicon on a silicon substrate, MMFRAC
refers to an effective strain which can vary from 0.0 (unstrained) to 1.0 (max strain of the
fitting). For a full description of the strain model in MC DEVICE, see Section 19.7.13
“Strain”.

832 ATLAS User’s Manual

837 ATLAS User’s Manual

Device Simulation MC Device

Biasing
Assign any non-zero applied voltage biases using the SOLVE statement (see “Commonly-Used
Statements” on page 822).
For backward compatibility with MOCA, you may assign the voltage for every contact defined
using the VBIAS parameter on the DEVICE statement as follows:
DEVICE VBIAS=(Vi,Vj,Vk,Vl)
where each value of the VBIAS parameter is specified in Volts. The subscripts i, j, k, and l
refer to the contacts in the order they are defined. In this case, first suppose the source and
drain contacts are defined, followed by the gate and finally a substrate contact. Then, the
statement above will assign voltages as follows:
DEVICE VBIAS=(Vs=Vi Vd=Vj Vg=Vk Vsub=Vl)
The DEVICE statement supports up to 16 contact definitions or 16 values on the VBIAS
parameter.
Defining contacts and then assigning each one a potential is essentially setting the boundary
conditions for the simulation. There are some cases where you must be careful. For example,
if you define a MOS device with a very shallow body, but in reality when the device is
fabricated, then there is a much larger distance to the substrate grounding potential. The
simulation results then may not be accurate.
The preferred method is to set the applied biases with the SOLVE statement. MC DEVICE will
stop with an error if you use both the DEVICE statement and the SOLVE statement to set non-
zero biases. Instead, you may only use the DEVICE statement or the SOLVE statement to do
this.
MC DEVICE provides you ways to define fields over chosen regions. The FIELD statement
lets you do this.
FIELD MAXFIELD=(start,end) RAMP=(start,end) NEGFIELD=0|1
USEBARR=0|1
MAXFIELD defines the starting and ending maximum electric fields in [V/cm]. The RAMP
parameter sets the iteration (ITER) on which to start ramping the field and the iteration on
which to end ramping the field. NEGFIELD sets whether to use the negative of the field.
USEBARR sets whether to use the material-dependent band-offset barriers in the field
calculation.
The material-dependent band-offset barriers are determined by the AFFINITY parameter (for
the conduction band barriers) and the AFFINITY and EGAP parameters (for the valence band
barriers) on the MATDEF statements, which appear at the end of the defaults.in file. The
material-dependent band-offset barriers are only used when using the multi-material model
(when TYPE=1 on the MMAT statement).
MC DEVICE also allows you to setup bias conditions that vary linearly between them. The
LINBIAS statement allows you to choose a specific REGION and specify the boundary
voltages of that region.
LINBIAS LBREGION=n LBX1=Vx1 LBX2=Vx2 LBY1=Vy LBY2=Vy2
The LBREGION parameter chooses which region this LINBIAS statement applies to. The
remaining fields set the voltage for each of the 4 boundaries of the chosen region. LBX1 sets
the potential at the left-hand x boundary. LBX2 sets the potential at the right-hand x boundary.

838 ATLAS User’s Manual

Device Simulation MC Device

LBY1 sets the potential at the top y boundary. LBY2 sets the potential at the bottom y
boundary.

19.5.2 Ohmic Contacts

For metal-semiconductor contacts, the barrier height,  B , represents the difference between
the metal work function and the electron affinity of the semiconductor,  , namely
B = m –  19-4
and is taken from the material library. For silicon contacts, the barrier height is found using
the surface electron concentration, n s  x  (where x is the position along the boundary), and the
conduction band density of states, Nc (where n s is found by assuming charge neutrality at
each node on the contact). For the non-degenerate case:
kB TL Nc
 B  x  = ------------ ln  ------------ 19-5
q  n s  x 

For the degenerate case, the logarithm above is replaced by an inverse Fermi-Dirac integral.
In all cases, the surface potential,  s , is determined by the barrier height,  B , and the applied
bias, Va , as:

s  x  = Va – B  x  19-6

Regarding the MC simulation, ohmic contacts must obviously force a quasi-equilibrium

concentration. We adopt the same algorithm as used by DAMOCLES [110].
The total charge along a contact is computed, and particles are injected to the mesh nodes
where the depletion is maximum. It must be observed that, for the case of non-uniform
doping or mesh spacing or both, this method can easily lead to unphysical non-uniformities of
concentration. The only solution to this problem is to use small, uniform contacts.
In MC DEVICE, the INJECT statement controls injection from the contacts and RESIST
controls the resistance of the contact region.
INJECT DN=0.0 DP=0.0 HEMIMAX=NO OVERINJ=0.0 TSTEP=1
DN and DP allow the selection of the fraction of electron and holes injected. This is more
useful in bipolar mode when both particle types are actually simulated simultaneously.
HEMIMAX sets injection to be a hemi-Maxwellian distribution (or not). OVERINJ sets the
allowed fraction of over-injection. So OVERINJ=0.01 allows 1% over-injection. TSTEP sets
the injection time step relative to the simulation time step. The default in MOCA is to inject
every time Poisson is solved.
RESIST OHM=0.0 CRNUM=-1 DIR=1 CONTNUM=-1 TSTEP=PINF
OHM defined the ohmic value of the contact in Omega-cm. CRNUM is the index of the region in
the input file from which to take material properties. DIR is the direction primary injection
direction for the contact, where 1 is for the x-direction and 2 is for the y-direction. CONTNUM
is the index of the actual contact in the .in input file. TSTEP is the simulation step interval on
which to update values.

839 ATLAS User’s Manual

Device Simulation MC Device

You can also control reflection from boundaries and contacts in MOCA. The REFLECT
statement is used to set the minimum distance travelled after reflection. By default, this value
is set to -1 to turn off the function.
REFLECT MINDIST=-1E0
The value of MINDIST is given in cm.

Note: Unlike other regions, the information about a contact is not overwritten by subsequent regions. Once you
specify a grid point to be a contact, it will remain that way regardless what follows in the input file.

19.5.3 Estimators
Current Regions
MC DEVICE allows you to define one or more rectangular current regions (C-Regions) where
you can calculate the current. A CREGION statement takes the form:
CREGION BOUNDP=(0.0275 0.0675 0.0 0.08)
The BOUNDP parameter contains the boundaries of the region as on the REGION statement (see
“Regions” on page 834). Again, a BOUND parameter is provided for backward compatibility.
The total current is the sum of the conduction current and the displacement current. The
conduction current is computed by assuming that carriers flow either along the X or along the
Y direction (or through two opposite sides of each rectangular region). The displacement
current is computed by considering the current across all four edges of each rectangular
region.
Consider the case of the X-oriented current, flowing in the negative X-direction (the normal
case for an n-MOSFET). Case a in Figure 19-4 illustrates how to calculate the current exiting
from the left-hand side of a rectangle, such as at the source of an n-MOSFET. Case b in
Figure 19-4 illustrates how to calculate the current entering from the right-hand-side of a
rectangle, such as at the drain of an n-MOSFET.

Figure 19-4: Rectangular regions are employed to compute the total current I for cases a and b, which pertain
to the source and drain of an n-MOSFET.

840 ATLAS User’s Manual

Device Simulation MC Device

Since the sum of the currents into any region is 0, the current I can be written, in both cases,
as:
I = I1 + I2 + I3 19-7

where I, I1, I2, and I3 refer to either to case a or case b collectively.

We will use the equation
L L

 0  D· + J  dl =
0  E + J  dl
I = J dl = · 19-8
tot

where the total current density, J tot , is written as the sum of the displacement current density,
D· , and the conduction current density, J , and where I represents a current per unit length in
a 2D domain.
The source current I in case a of Figure 19-4 is then:
x Y x

0 E· y  x' 0 dx' – 0  E· x  x y  + J  x y   dy – 0 Ey  x' Y  dx'

I = · 19-9

We now consider the average current within the C-Region by integrating the RHS of Equation
19-9 from x0 to xX and by dividing by X, namely:
X x X Y X Y

  
1
I = ---   E· y  x' 0  – E· y  x' Y   dx' dx – J  x y dy dx –  E· x  x y  dy dx 19-10
X 0 0 0 0 0 0

The first integral in Equation 19-10 is evaluated as:

X x
·
·E  x'   dx' dx =
X X 
   
·
 y   E y  x'  H  x – x'  dx' dx 19-11
0 0  0 0 

X X
=
0 E· y  x'   dx'
0 H  x – x'  dx' 19-12

0 Ey  x    X – x  dx
= · 19-13

where H is the Heaviside or unit step function.

The expression for the source current I in case a in Figure 19-4 is then:
X Y  Y X

   
1 · ·
I = --- –  J   x y dy  dx +     X y  –   0 y  dy +   X – x   E· y  x 0  – E· y  x Y   dx 19-14
0 0 
X 0 0

where the x component of the electric field, Ex , has been expressed in terms of the
electrostatic potential,  , using the fundamental theorem of calculus.

841 ATLAS User’s Manual

Device Simulation MC Device

Similarly, the drain current I in case b in Figure 19-4 is:

X Y X

0  E· y  x' 0  dx' –
0  E· x  x y  + J  x y   dy +
x Ey  x' Y  dx'
I=– · 19-15

Again, we consider the average current in the C-Region by integrating the RHS of Equation
19-15 from x=0 to x=X and by dividing by X. The expression for the drain current I in case b
in Figure 19-4 is then:
X Y  Y X

   
1 · ·
I = --- –  J  x y dy dx +     X y  –   0 y  dy +  x  E· y  x Y  – E· y  x 0   dx 19-16
0 0 
X 0 0

In both Equations 19-14 and 19-16, the first term represents the conduction current. The
second term represents the the displacement current flowing through the left (x=0) and right
(x=X) boundaries. The third term represents the displacement current flowing through the
lower (y=0) and upper (y=Y) boundaries.
In steady state, all displacement currents average to zero, so the second and third terms in
Equations 19-14 and 19-16 drop out. So in steady state, the current in both cases is negative
the integral over y of the X-average of the conduction current density.
During the MC simulation, the current will be averaged over time to reduce the noise and
increase the accuracy of the estimates. The time derivatives of the potentials and electric
fields do not need to be time-averaged. Instead, these can be directly calculated with a
backward difference approximation to the derivatives. The conduction current density,
however, is time averaged.
The sign convention for the current was adopted from MOCA where a net flow of electrons in
the positive x direction (from source to drain in an n-MOSFET) was defined as positive
current despite the fact this is contrary to the standard definition. Conversely, a net flow of
holes in the positive x direction (from source to drain in a p-MOSFET) is defined as negative
current. Since the currents calculated by a current regions is not necessarily associated with a
terminal, the reference direction for the current is not related to the terminals. This means that
the source and drain currents from current regions (normally the x-component of the current
in each current region) will be positive for n-MOSFETs and negative for p-MOSFETs when
these devices are biased in the linear or saturations regions.
Energy Regions
MC DEVICE also allows you to define one or more rectangular regions used for the
calculating the energy estimator (EE-Regions). A EEREGION statement takes this form:
EEREGION BOUNDP=(0.0051,0.0102,-0.0015,0.0102)
The BOUNDP parameter contains the boundaries of the region as on the REGION statement (see
“Regions” on page 834). Again, a BOUND parameter is provided for backward compatibility.
The energy estimator is calculated for each EEREGION in a way similar to how the total
current is calculated (see “Current Regions” on page 840).

842 ATLAS User’s Manual

Device Simulation MC Device

19.5.4 Initial Conditions

MC DEVICE allows you to customize many quantities used in the simulation. You can do the
following:
• Defining regions over which the field is constant.
• Setting the initial particle concentration in a given region.
• Setting the particle injection rate.
• Setting a linear bias over a region.
FFREGION allows setting up a constant field region. The FFREGION statement has three
parameters: FIELDX, FIELDY, and BOUNDP. FIELDX sets the x-directed field at the four edges
of the region. FIELDY sets the y-directed field at the four edges of the region. BOUNDP defines
the boundaries of the region (see “Regions” on page 834). Again, a BOUND parameter is
provided for backward compatibility.
FFREGION \
FIELDX=(Ex1,Ex2,Ey1,Ey2) \
FIELDY=(Ex1,Ex2,Ey1,Ey2) \
BOUNDP=(x1,x2,y1,y2)
You can use the ICLREGION statement to set the initial carrier concentration. This statement is
useful for controlling the initial solution to Poisson’s equation.
ICLREGION CONC=conc BOUNDP=(x1,x2,y1,y2)
CONC sets the initial concentration. BOUNDP sets the boundary wherein this concentration is
applied (see “Regions” on page 834). Again, a BOUND parameter is provided for backward
compatibility.
You can use the QSS statement to set a fixed surface charge density. This statement is useful
for studying trapped oxide charge in MOS devices.
QSS QSCHARGE=conc BOUNDP=(x1,x2,y1,y2)
QSCHARGE sets the fixed surface charge density. BOUNDP sets the boundary wherein this
concentration is applied (see “Regions” on page 834). Again, a BOUND parameter is provided
for backward compatibility.
If you need to define a line charge, then set y1 and y2 to be equal to one another.

19.5.5 Bipolar
For devices where impact ionization is important, you can run MC DEVICE in bipolar mode.
The BIPOLAR statement sets the interprocess communication parameters.
BIPOLAR COMM=0|1|2 MSFILE=”msfile” SMFILE=”smfile” NSLAVE=n
INITNLIN=conc \
NFRACII=0.0
The comm field sets the communication method, 0 is to turn bipolar mode off, 1 is for master,
and 2 is for slave (unix pipe). If COMM is set to 1, the variable MSFILE is set to the master to
slave pipe. If COMM is set to 2, then the variable SMFILE is set to the slave to master pipe.
NSLAVE is the number of slave particles. INITNLIN is the initial slave concentration. NFRACII
is the fraction at which to increase the number of particles.

843 ATLAS User’s Manual

Device Simulation MC Device

19.5.6 Track Boundary Crossings

Along with using statements in the scattering section to track particles (see “Impact
Ionization and other topics” on page 881), there is CROSSEST to track particles crossing
boundary regions. The CROSSEST statement specifies where region you want to track
particles. Multiple CROSSEST regions may be defined.
CROSSEST CROSSREG=n

19.5.7 Time Step Regions

To precisely simulate some areas of a device where carriers move quickly, you can use
DTREGION to set a small time step. Conversely, to efficiently simulate areas of a device where
carriers move slowly, you can use DTREGION statements to set a larger time step.
Use the DTREGION statement to define the area over which to use the fine time step.
DTREGION BOUNDP=(x1 x2 y1 y2)
The region defined by DTREGION will be simulated every flight time step while the rest of the
device will only be updated every Poisson time step. This way you can accurately simulate,
say, the channel region with a flight time step of 0.01 fs while still keeping the Poisson step
around 0.1 fs. The BOUNDP parameter defined the rectangular boundary of the region (see
“Regions” on page 834). Again, a BOUND parameter is provided for backward compatibility.

19.5.8 Importing Data

MC DEVICE allows you to import data from input data files. In particular, you can import the
electric potential, the initial MC carrier concentration, doping profiles, regions and mesh. You
can import all these data or any combination of them. All imports must be done from one file
type, specified by the TYPE parameter on the IMPORT statement.
IMPORT TYPE=1 VOLT=FALSE AVGVOLT=FALSE \
NCONC=FALSE AVGNCONC=FALSE PCONC=FALSE AVGPCONC=FALSE DOP=TRUE \
MESH=FALSE REGIONS=FALSE \
DIPSH=0 OFFSET={x,y} BOUNDP=(x1,x2,y1,y2) \
DOPFILE=”ref/dop.in”
The TYPE parameter allows you select from eight different file formats:
• TYPE=1 is for UIUC MOCA files.
• TYPE=2 is for PISCES version-2 files.
• TYPE=3 is for PISCES version-3 files.
• TYPE=4 is for ISE files.
• TYPE=5 is for rectangular grid files.
• TYPE=6 is for TIF files.
• TYPE=7 is for Well-Tempered MOSFET (WTM) files.
• TYPE=8 is for Silvaco structure files.
The TYPE parameter sets the file type for all imported files. The default value of TYPE=8 is for
Silvaco Structure files.

844 ATLAS User’s Manual

Device Simulation MC Device

VOLT, AVGVOLT, NCONC, AVGNCONC, PCONC, AVGPCONC, and DOP are boolean parameters that
enable you to import the related quantities. By default, these parameters are FALSE so that no
import occurs. Setting VOLT to 1|TRUE|YES will import the “Potential” as the electric
potential. Setting AVGVOLT to 1|TRUE|YES will import the “Average Potential” as the electric
potential. You may not set both VOLT=1 and AVGVOLT=1. When CARRIER=E on the ALGO
statement, setting NCONC to 1|TRUE|YES will import the “Electrons” as the initial electron
concentration and setting AVGNCONC to 1|TRUE|YES will import the “Average Electrons” as
the initial electron concentration. You may not set both NCONC=1 and AVGNCONC=1. When
CARRIER=H on the ALGO statement, setting PCONC to 1|TRUE|YES will import the “Holes” as
the initial hole concentration and setting AVGPCONC to 1|TRUE|YES will import the “Average
Holes” as the initial hole concentration. You may not set both PCONC=1 and AVGPCONC=1.
Setting DOP=1|TRUE|YES will import the “Net Doping” if IONIZ=0|FALSE|NO on the
POISSON statement (for complete ionization of dopants). Setting DOP=1|TRUE|YES will
import “Boron”, “Arsenic”, “Indium”, and “Phosphorus” if IONIZ=0|FALSE|NO on the
POISSON statement (for incomplete ionization of dopants).
MESH and REGIONS are boolean parameters that enable you to import these structure
elements from a Silvaco structure files (TYPE=8). Setting MESH=1|TRUE|YES will import the
mesh from STRFILE and ignore XMESH and YMESH statements in your input file. Since MC
DEVICE supports only rectangular tensor product meshes, meshes of a type other than this are
transformed to a rectangular tensor mesh which includes every x position and every y
position in the imported mesh. Setting REGIONS=1|TRUE|YES will import regions from
STRFILE. You must comment out any REGION statements in your input file unless you want
to add additional regions or redefine regions which get imported. Since MC Device supports
only rectangular regions, regions with a shape different than this are expanded to a rectangle
which includes every mesh point in the imported region. You can not import contacts from a
structure file using an IMPORT statement. Instead, when REGIONS=1|TRUE|YES, MC DEVICE
will continue to process any uncommented REGION statements in your input file. Those with
TYPE=CONTACT will represent the contact regions for the device and any contacts in the
structure file will be ignored. Contacts in your input file are added to the imported structure
in the order they appear in the input file. So when you import a structure from strfile
using TYPE=8, you will still need to use REGION statements with TYPE=CONTACT in your input
file to create contacts for the imported structure.
The DIPSH parameter is set to 0 (off) by default. You may set DIPSH to 1 (on) to use ISE
DIPSH with the import.
The OFFSET parameter is an (X,Y) offset for importing quantities. The OFFSET parameter is
(0,0) by default and (X,Y) are specified in cm.
The BOUNDP=(x1 x2 y1 y2) parameter sets a rectangular boundary over which doping
imports are applied. If any of x1, x2, y2, or y2 are left as their default values of 0, then these
are set to the xmin, xmax, ymin, ymax of the device structure respectively, so that doping
imports are applied to the entire device by default. Alternatively, you can set x1, x2, y1, and
y2 to set to defined a window for importing the doping. Imports for the voltage and
concentration are always done over the whole device, so these are not affected by the BOUNDP
parameter (see “Regions” on page 834). Again, a BOUND parameter is provided for backward
compatibility.
The VOLTFILE, CONCFILE, DOPFILE, GRIDFILE, SOLFILE, VERFILE, TRIFILE, SDFILE,
RETROFILE, and STRFILE parameters provide the filenames of the data to be imported. The
particular parameters that are used depend on the TYPE parameter.

845 ATLAS User’s Manual

Device Simulation MC Device

For TYPE=1, VOLTFILE, CONCFILE, and DOPFILE are used, where VOLTFILE is a 2D voltage
file, CONCFILE is a 2D MC carrier concentration file, and DOPFILE is a 2D doping file.
For TYPE=2|3, GRIDFILE and SOLFILE are used, where GRIDFILE is a 2D grid file and
SOLFILE is a 2D solution file.
For TYPE=4|5|6, VERFILE, TRIFILE, DOPFILE, and SOLFILE are used, where VERFILE is a
2D vertex file (a 2D grid file when DIPSH=1), TRIFILE is a 2D triangle file (a 2D data file
when DIPSH=1), DOPFILE is a 2D doping file, and SOLFILE is a 2D solution file.
For TYPE=7, SDFILE, RETROFILE, and VOLTFILE are used, where SDFILE is a 2D source-
drain doping file, RETROFILE is a 1D retrograde doping file, and VOLTFILE is a 2D voltage
file.
For TYPE=8, STRFILE is used, where STRFILE is a 2D Silvaco structure file.
For example, if you set volt=1|TRUE|YES and TYPE=1 (for the UIUC MOCA file type), then
MC DEVICE will import the electric potential (voltage) from the file specified by the
VOLTFILE parameter (which is, by default, ‘volt.in’). Correspondingly, if you set
CONC=TRUE and TYPE=1, then MC DEVICE will import the MC carrier concentration from the
file specified by the CONCFILE parameter (which is, by default, ‘conc.in’). Finally, if you
set DOPING=TRUE and TYPE=1, then MC DEVICE will import the doping profiles from the file
specified by the DOPFILE parameter (which is, by default, ‘dop.in’).

846 ATLAS User’s Manual

Statistical Enhancement MC Device

19.6 Statistical Enhancement

The methods for statistical enhancement are controlled with the ENHAN statement.
ENHAN MODE=2 TSTEP=2 ESTEP=100 XSTEP=(0,0,0) \
LOTHRE=1e-20 HITHRE=1e20 RELTHRE=0.05 RATIO=2. N=10
The default value of MODE=0 indicates that statistical enhancement is turned off. MODE=1 is
deprecated and should not be used. MODE=2 selects the non-local gathering algorithm also
known as phase-space balancing [203, 204]. MODE=3 selects the comb method (for bulk
simulation only). In all cases, TSTEP is the interval in time steps between calls of the
variance-reduction subroutine. This depends on a compromise between quality of the
simulation (reduction of the noise) and CPU time spent in the balancing routine. Usually, a
balancing step of less that 100 fs is needed to achieve significant reduction of the noise.
ESTEP and XSTEP specify the partitioning of phase space in units of meV and mesh cells
respectively. Time steps between application of the enhancement algorithm.
You can also specify only a sub-region of the device to perform the statistical enhancement.
The BOUND parameter defines a rectangular boundary where to apply the statistical
enhancement.
The non-local gathering method works by a two-phase process. First, the ensemble of
particles will be scanned. A linked list is then created that contains all particles that do not
contribute to the simulation or redundant particles. A particle is marked as redundant if its
weight is less than RELTHRE times that of the largest particle in the same bin. Then, the
average number of particles per bin is computed. Particles in under-populated bins are split
until the occupation of phase space is balanced. Every splitting needs some storage for the
newly-created particles. This storage is obtained by gathering particles. You can perform
gathering on the redundant set if this is available. If more splittings are needed than redundant
particles are available, some local gatherings will be performed in bins with higher particles.
The algorithm is completely controlled by only five variables:
• LOTHRE
• HITHRE
• RELTHRE
• RATIO
• N
These will be described in the following sections.

19.6.1 LOTHRE
The LOTHRE parameter is the single most important parameter in the non-local gathering
method. It is the minimum weight allowed for particles in the simulation. That is, particles
with weight below LOTHRE cannot be split further. This minimum threshold defines the
accuracy of the simulation, since the weakest details will have to be resolved by very small
particles. Usually, a minimum weight of 1020 to 1010 is adequate to resolve most important
high-energy effects. Since this parameter is a double-precision number, you can specify a
very small value. A small value for LOTHRE uses particles to populate a large number of bins.
Therefore, it trades the overall accuracy of the simulation for a wider coverage of phase
space.

847 ATLAS User’s Manual

Statistical Enhancement MC Device

19.6.2 HITHRE
The HITHRE parameter controls the maximum allowed size for particles in the simulation. It
is used in two different ways. One way is to control the maximum total weight for particles in
a gathering set if the redundant set is used. Another way is to control the maximum size of
any particle in the gathering set if local gathering is used. A value of 1.0 will force a
maximum unit weight for all particles, so that no particle will be larger than in the
conventional situation. This will inhibit any gathering starting from a situation where all
particles have unit weight. In this case, if you apply balancing to an initial status from a non-
balanced simulation, then some extra particles will have to be allocated (NREF larger than the
initial number of particles). Therefore, the algorithm will be able to expand the ensemble
without gathering particles.

19.6.3 RELTHRE
The RELTHRE parameter defines the criterion for marking particles as redundant. A value of
103 means that a particle will be included in the redundant set only if there exists another
particle in the same bin with a weight 1,000 times larger. While a value of 103 is safe
because it only tries to “reuse” very small particles, it may be inefficient in cleaning the bins
of particles that do not contribute to the simulation. A large value like 101, however, may
mark as redundant particles that actually contribute significantly to the estimators.
The optimum value of RELTHRE depends on the dynamic range of weights in any given bin. A
simple rule of thumb can be derived from the exponential dependence of electron occupancy
on energy. We can reasonably take the thermal-equilibrium distribution as the steepest
possible distribution in a common device. If E is the width of the energy bins, we have a
possible dynamic range inside a bin of exp  EkBTL50 for 0.1 eV bins and room
temperature.
In this case, you can see a value of 102 for RELTHRE is safely low enough. Also for space
dependence, you can apply the same criterion, where the energy range is simply the
maximum voltage drop across a space bin. For a typical bin width of 10 nm and an electric
field of 200 kV/cm, we have an energy difference of 0.2 eV, which yields a dynamic range of
about 2500. For this case, a value for RELTHRE below 410-4 is suggested by our rule of
thumb.

Note: In high-field regions, you cannot assume a thermal-equilibrium distribution. But in high-field regions, you
should not use the guidelines described above.

848 ATLAS User’s Manual

Statistical Enhancement MC Device

19.6.4 RATIO, N
These RATIO and N parameters define the gathering process. RATIO is the maximum dynamic
range of weights for any gathering. n is the maximum number of particles in the gathering set.
Let us assume that n particles are being gathered and the RATIO of weights between the
largest and the smallest particle is RATIO. This means that if the smallest particle is retained,
its weight will increase by a factor between RATIO and nRATIO. If there is a great increase
in weight, it will worsen the numerical noise. To estimate the maximum tolerable increase of
weight, consider the case of gathering of redundant particles. For each redundant particle,
there is another particle in the same part of phase space with a weight RELTHRE1 times larger.
The increase will be tolerable if the retained particle remains quite smaller than the dominant
particles in the same bin. In other words, if:
–1
ratio  n « relthre 19-17
This gives us the rule of thumb:
–1
ratio « relthre
--------------------- 19-18
n

For example, with RELTHRE10-2 and n=10, a dynamic range equal to 10 can be tolerated
before an excessive increase of variance occurs. If you use a very aggressive criterion for
marking redundant particles (e.g., RELTHRE=101 or more) then rescale RATIO accordingly.
Otherwise, the sudden growth of the redundant particles will create large spikes in the
estimated distribution function. The optimum value for RATIO depends on the number of
particles used. For a large number of particles (20000 or more), a value as low as 2 allows a
very small increase in variance without impairing the performance. For small number of
particles, a value of 10 or more can be necessary to collect a sufficient number of redundant
particles for the gathering to be effective. If the constraint becomes too strict, lower RELTHRE
instead of setting RATIO to a very low value.

19.6.5 Particle Compression

MC DEVICE automatically sets a reference number of particles, Nref, as slightly less than the
number physically allocated in memory. In this way, it allows for some fluctuations of the
number of computational number of particles. If the current number n of particles is larger
than Nref, MC DEVICE can perform some gatherings with no splittings to reduce the total
number. This step is called compression. If nNref, the extra storage available is used to
perform some “free” splittings with no gatherings. Under normal conditions, at the exit of the
variance-reduction procedure the number of particles will be equal to Nref. The introduction of
Nref solves a problem intrinsic with the manipulation of the particle states. The total number of
particles in the simulation is related to the transit time of carriers through the device. If no
variance reduction is used, this will be the physical transit time, so that the number of particle
will be easily estimated and controlled. The introduction of variance reduction algorithms
changes the distribution of MC particles in phase space, so that the computational transit time
can be higher or lower than the physical one. This leads to unpredictable drifts and
fluctuations in the storage required by the program. You can only use the compression step by
a many-particle variance-reduction scheme [204]. This solves this problem by forcing an
almost-constant number Nref of particles.

849 ATLAS User’s Manual

Statistical Enhancement MC Device

19.6.6 Statistical Enhancement Statements

The following statements don’t specifically reduce the noise during the simulation. These are
only statements that smooth the output. ESTIM1D allows you to select a region and make a 1D
cut through it and study quantities of that cut. SMREGION sets a region to perform the
smoothing of the outputs.
ESTIM1D DIR=X/Y BOUNDP=(x1,x2,y1,y2)
DIR sets what direction to make the cut in. BOUNDP defines the boundary where to run the 1D
estimator (see “Regions” on page 834). Again, a BOUND parameter is provided for backward
compatibility.
SMREGION BOUNDP=(x1,x2,y1,y2) TYPE=n
The BOUNDP parameter contains the boundaries of the region, as on the REGION statement
(see “Regions” on page 834). Again, a BOUND parameter is provided for backward
compatibility. TYPE sets the quantity to smooth. Setting TYPE to 1 smooths the potential, ETA.
Setting it to 2 smooths the quantum potential, QETA.

850 ATLAS User’s Manual

Physical Models MC Device

19.7 Physical Models

19.7.1 Self-Consistent Simulation
Background
Poisson’s equation is solved on the same grid used to define the device. Assuming
CARRIER=E on the ALGO statement, the non-linear Poisson equation takes the form
+ –
–      = e  –  n MC + n     + N D    – N A    + p     . 19-19

Here,  is the permittivity.  is the electrostatic potential. e is the fundamental electron charge.
nMC is the electron concentration computed using the Monte Carlo cloud-in-cell (CIC) scheme
[110]. nMC is zero in semiconductor contact regions and non-contact gate regions and non-
zero in MC regions. n() is the electron concentration computed using  and Nc using
Equation 19-23. NC is the effective density of states of the conduction band. NC and n() are
zero in MC regions. NC and n() are non-zero in semiconductor contact regions and non-
contact gate regions and zero in MC regions. N D + is the ionized donor concentration. N A - is
the ionized acceptor concentrations. p() is the hole concentration computed using  and NV
using Equation 19-22. NV and p(are non-zero in all semiconducting regions.
All potential-dependent terms in the RHS of Equations 19-19 are determined self-consistently
using the quasi-Fermi potentials in each mesh cell and the electrostatic potential, , at each
grid point. For 1D simulation, the linear Poisson equation is solved and holes must be
considered in the MC simulation.
For CARRIER=H on the ALGO statement, -n and p are switched in Equation 19-19.

Figure 19-5: Vi is the contact bias at contact i. By default, Efp=-e*Vi is the hole quasi-Fermi energy.
 (x)=(Eg(Si)/2 - Ec(x))/e is the electrostatic potential.

851 ATLAS User’s Manual

Physical Models MC Device

Still assuming CARRIER=E on the ALGO statement, the hole density is computed as
 
p =
0 dE g v  E f D   E – E V + E fp   k B T L  =
0 dE gv  E fD   E + e  x  + Eg – eVi   kB TL  19-20

where f D    = fD   g=1  is the Fermi-Dirac distribution and generalizes for a degeneracy g

to
1
f D   g  = --------------------------------- 19-21
1 + g  exp   

This is the Fermi-Dirac distribution including degeneracy.  is a unitless factor proportional

to the energy, Eg is the bandgap energy. gvE is the energy-dependent valence-band density of
states. Efp is the hole quasi-Fermi energy, and Vi is the applied voltage at contact i.
By default, within contact region i, Efp=-eVi. Also by default, within the doping well of
contact i, Efp=-eVi, where each contact is assumed to reside in a doping well of one type (n or
p). This is the default implementation of the constant quasi-Fermi level method in MC
DEVICE.
The USEFERMI and FERMI parameters on the REGION statement provide you with a way to
modify the default behavior of the constant quasi-Fermi level method. For example, for a n-
MOSFET with an inversion channel and assuming Ohmic contacts, you can change the
default choice for the hole quasi-Fermi energy of Efp=-eVdrain in the well adjacent to the drain
contact by setting Efp using USEFERMI=1 and FERMI=<value> on the REGION statement for
the drain (see “Regions” on page 834). For example, you may wish to set Efp= -
min(Vsource,Vdrain) = -min(0.0 eV,eVdrain>0 V) = 0.0 eV. To do so, you can then use the default
value of FERMI=0.0 eV by simply including USEFERMI=1 on the REGION statement for the
drain contact. Similarly, for a p-MOSFET when Vsource=0.0 V and Vdrain<0 V and when using
CARRIER=H on the ALGO statement, you can use USEFERMI=1 on the REGION statement for the
drain contact to get Efn=-max(Vsource,Vdrain)=FERMI=0.0 eV in the drain well instead of the
default of Efn=-eVi=-e*Vdrain.
For non-MC carrier concentrations:
p    = N v F 1/2   E V – E fp   k B T L  = N v F 1/2   – e  x  – E g + eV i   k B T L  19-22

n    = N c F 1/2   E fn – E c   k B T L  = N c F 1/2   – eV i + e  x    k B T L  19-23

where


0 
1/2
F 1/2    = f D   –   d 19-24

is the Fermi-Dirac integral of order 1/2 and where fD is defined above. Efn is the electron
quasi-Fermi energy and is equal to negative the applied voltage at contact i, Vi, within each
contact region i. For incomplete ionization of donors, Efn is also assumed to be constant and
equal to the contact value within each doping well surrounding contact i. Efn is calculated

852 ATLAS User’s Manual

Physical Models MC Device

from Equation 19-23 using the average electron concentration, <n>, and the average
conduction band edge, <Ec>, and taking Efn=0 wherever <n>=0. Although MC DEVICE uses
Equation 19-23 to calculate Efn throughout semiconductor regions, MC DEVICE treats
electrons to be in quasi-equilibrium (or to have thermalized to a distribution based on the
lattice temperature). Instead, Efn, is simply an output quantity that you can plot in energy band
diagrams. Efn obtained with MC DEVICE compares closely with Efn obtained with the drift
diffusion model inside regions of your device where quasi-equilibrium holds. Conversely, Efn
obtained with MC DEVICE diverts from Efn obtained with the drift-diffusion model in regions
of your device where quasi-equilibrium does not hold.
For the fixed charge, if ED and EA are the donor and acceptor impurity energies (with respect
to each corresponding band edge), we have for donors:
+
N D    = N D f D  – E c + E D + E fn g D  = N D f D  e + E D – eV i g D  19-25
and for acceptors:
–
N A    = N A f D  E c + E A – E g – E fp g A  = N A f D  – e + E A – E g + eV i g A  19-26
where gD=2 is the donor degeneracy factor for phosphorus and arsenic. gA=4 is the acceptor
degeneracy factor for boron and indium.
The ionization energies of donors and acceptors are assumed constant. No heavy-doping
effect, such as the donor deionization effect [149], are considered.
The 1D simulation has no specific input statement. The solver is actually controlled by the
MODE variable on the ONED statement.
Most of the energies and concentrations used in the Equations 19-19 through 19-26 are saved
in *sol.str (structure) files when performing a 2D simulation. See Section D.2 “File
Formats”. You can perform 1D cuts of the 2D structures to plot energy-band diagrams
relating to your solution.
The 2D Poisson solver is controlled through the POISSON statement
POISSON TSTEP=5 ITER=100 TOL=(2e-4,1e-6) IONIZ=NO \
START=1 FREEZE=PINF LINPOI=NO CUTOFF=1e25
TSTEP sets the number of flight time steps between each Poisson solution. ITER is the
maximum number of Newton iterations. TOL is a vector of 2 values, which is the L2-norm
tolerance for the Newton and conjugate-gradient methods respectively.
IONIZ is used to switch on and off incomplete ionization of dopants. Two reasons exist to
disable incomplete ionization. First, the response time of dopant levels is much longer than
the time scale used in the MC simulation. Therefore, it is unrealistic to treat dopants in the
“adiabatic” approximation (i.e., with instantaneous response). Second, for very high doping
levels (above 10E18 cm-3), it is approximately correct to assume that all dopant atoms are
ionized [213]. In any case, the program automatically switches off the IONIZ flag whenever
the quasi-Fermi energies are not constant across a non-depleted region (e.g., in a JFET or
MESFET simulation) where a finite voltage is applied across a non-depleted channel.

853 ATLAS User’s Manual

Physical Models MC Device

The fields START and FREEZE define the time window during which the solver actually
works. The field is frozen outside that window, either using a result from a previous
simulation, or an average from the same simulation. In the example above, the field is kept
frozen during the first 1000 time steps. The Poisson solver is then used until the end of the
simulation. The keyword PINF is internally defined as the largest positive integer. The
keyword NINF is internally defined for the largest negative integer.
If the first field of FREEZE is greater than one, then an initial fixed field is needed to start the
simulation. By default, MC DEVICE reads this field from the file volt.in. This file is also
written at the end of each run for use in a subsequent simulation.
LINPOI selects whether the linear or non-linear Poisson equation is solved. If
LINPOI=1|TRUE|YES, then a linear version of the Poisson equation is solved where complete
ionization of dopants is assumed and where the opposite carrier type as set by the CARRIER
parameter on the ALGO statement is assumed to be 0. If LINPOI=0|FALSE|NO (as it is by
default), the non-linear poisson equation is solved. In this case, the IONIZ parameter may be
used to include incomplete ionization and the Poisson equation includes a expression for the
density of the opposite carrier type as set by the CARRIER parameter on the ALGO statement as
a function of the unknown potential.
CUTOFF sets an upper limit for the complete ionization of dopants. The default value of
CUTOFF is 1e25 cm-3. Because the default value of CUTOFF is high compared to typical
doping concentrations, the CUTOFF parameter typically has no effect when the default value is
used. The CUTOFF parameter will only have an effect if it is reduced to below the peak of the
net doping profile of your device and if IONIZ=0|FALSE|NO for complete ionization. Since
using the CUTOFF parameter can change the net doping profile that will be used internally by
MC DEVICE for solving Poisson’s equation and for treating impurity scattering, the effects of
using the CUTOFF parameter can be significant. The CUTOFF parameter should not be reduced
below 3e19 cm-3 since this may dramatically affect impurity scattering. The CUTOFF
parameter also affects the doping in the gate, so a value of the CUTOFF parameter below the
net doping in the gate will change the threshold voltage. Since there are many restrictions
associated with proper use of the CUTOFF parameter, we recommend that you not include the
CUTOFF parameter in your input files. Instead, we encourage you to choose doping levels at or
below the doping ionization limit when using complete ionization of dopants. Note that
CUTOFF has no effect when IONIZ=1|TRUE|YES for incomplete ionization of dopants.
In MOCA, the IPOLY parameter turned the model for the depletion of polysilicon on and off.
In MC DEVICE, the IPOLY parameter is ignored and the depletion of polysilicon gates (or of
any semiconducting region where the TYPE is not MC) is considered automatically whenever
they appear in a device. For example, in MC DEVICE, you can define the entire gate as type
CONTACT, and no depletion will occur on the gate. Instead, you can set the region immediately
above the gate oxide to a MAT=POLY (or any semiconducting material) and TYPE=BLOCK, and
you can define a contact above this semiconducting non-contact region. In this case, the non-
linear Poisson equation including potential-dependent terms for both the electron and hole
concentrations is solved within the non-contact portion of the gate to account for the
depletion effect.

854 ATLAS User’s Manual

Physical Models MC Device

Microscopic Analysis of Noise

Here is a brief analysis of voltage fluctuations in a conductive medium under the hypothesis
of weak screening. This amounts to assuming a non-collisional system because we compute
the fluctuations within a screening sphere from the generating carriers. Since the result will be
expressed as a power spectrum, the analysis is valid for frequencies above the plasma
frequency. The result will be useful when the plasma frequency is low (i.e., for low-density
regions).
Velocity Autocorrelation
We start with an elementary model of one-dimensional velocity autocorrelation in a non-
interacting gas of classical particles:
2 – |t|  
C v  t  = E  v  t' v  t' + t    v th e 19-27
where vth is the thermal velocity and  is the momentum-relaxation time. At thermal
equilibrium:

2 kB TL
v th = -----------
- 19-28
*
m

The corresponding power spectrum of the velocity is

 2
2 – 2ft – |t|   2v th 
Sv  f  =
0 v th e dt = ---------------------------
1 + (2f)
-
2
19-29

and is approximately white up to the frequency 1   2  , which is several THz. This
frequency must be compared with the plasma frequency that characterizes the dynamic
screening of the thermal oscillations by the electron gas.

In the following, we will assume a white spectrum with uniform power density v2th  for each
of the three components of the particle velocity, truncated at the autocorrelation frequency.
Although we consider the simple, constant-relaxation-time model at thermal equilibrium, the
non-equilibrium case with variable scattering rate can be analyzed by replacing vth with an
average RMS group velocity and  with an average relaxation time.
Current-Density Fluctuations
The current density at a point r in space is defined as the limit.

 evi
1
Jr = lim ---- 19-30
0 
xi  

From this definition, the power spectrum of the fluctuations of the k-th component of the
current density can be readily computed as
2

 Sv  f 
e
SJ  f  = lim ------- 19-31
k   0 2
xi  

855 ATLAS User’s Manual

Physical Models MC Device

2
2ne k B T L
S J  f  = lim --------------------------
- 19-32
k 0 m
*

It is clear that, as  tends to zero, the fluctuation diverges. It follows that the fluctuation is not
well defined for the current density but only for its average on a finite volume. This can be
verified by a simple derivation, from Equations 19-21, of the familiar formula for the Johnson
noise. Assuming for simplicity a cylindrical resistor of length L and area A, the total
(conduction plus displacement) terminal current can be computed as
L A 

 
1 3
I = ---  J d r 19-33
L 0 0 


1
I = --- e(v i ) x 19-34
L
xi  L  A

The short-current noise power spectrum is therefore

2
A 2ne k B T L
S I  f  = ---  --------------------------- 19-35
L *
m

2Ak B T L 2k B T L
S I  f  = ----------------------
- = --------------
- 19-36
L R

This is the usual short-circuit thermal-noise current formula for a resistance R, in the
relaxation time hypothesis where the conductivity  is given by ne2/m*.
Potential Fluctuations
The main reason for the divergence for the current-density fluctuation is the finite size of the
charge carriers. Let us now analyze the power spectrum of the electostatic potential. Each
carrier of charge e generates a potential
e
V i  r  = -------------------------
- 19-37
4|r – r i |

In a given point r of space, the total potential, V, will be given by the sum of all contributions
from nearby carriers:

 ---ri
e- 1
V = -------- 19-38
4
i

where the origin has been conveniently chosen at r. Deriving Equations 19-22 with respect to
time and using Equations 19-20, we obtain the power spectrum of the time-derivative of the
potential:
2
e k B T L

1
S ·  f  = ---------------------
- ---- 19-39
V 2 2 * 4
8  m i ri

856 ATLAS User’s Manual

Physical Models MC Device

The summation, for a large number of particles, can be approximated with an integral:
2
e k B T L 

2 1
S · f = ---------------------
- 4r n ---- dr 19-40
V 2 2 * 4
8  m 0 r

2
e k B T L n  1
S · f =
V
-----------------------
2 m 0 r
2 *
- ---- dr
2  19-41

Again, the integral presents a pole for vanishing distances, due to the 1/r dependence of the
potential. We may remove the singularity by assuming a minimum characteristic distance, rc.
In a physical sense, that forces a discretization cutoff in Equations 19-25. With computer
simulations, rc is the size of the discretization mesh on which Poisson’s equation is solved.
Applying this minimum-distance cutoff, we have
2
e k B T L n
S ·  f  = -----------------------
- 19-42
V 2 *
2 m r c

We have found the high-frequency behavior of the fluctuations of the potential can be useful
for estimating the oscillations of potential barriers. Unfortunately, we have assumed no
correlation between positions of the carriers, which results in a perfectly white spectrum of
the time-derivative of the potential. This implies an infinite power of the potential
fluctuations (i.e., a “random walk” of the potential). In reality, Coulomb interaction between
carriers limits the possible oscillation of the potential. You can get a rough estimate of the
total amplitude of the potential fluctuations by limiting the bandwidth of its derivative to
some characteristic frequency fc. It’s usually the inverse of the dielectric-relaxation time. The
power spectrum of the potential is then
2
e k B T L n
S V  f  = -----------------------------------------------------------------------
- 19-43
2 * 2 2
2 m r c  (2f) + (2f c ) 

Integrating over the entire spectrum, we have

2
e k B T L n
-.
P V = ---------------------------------
2 2 *
8  m r c f c

It is evident that as fc   , the fluctuations of the potential diverge without bound.

Even after removing the similarities, we see that the total power is proportional to the factor
e2n. This means that the granularity of the charge carriers affect directly the noise. Only when
the critical radius rc is scaled along with the charge n, the limit remains finite. With computer
simulations, this means the mesh size must decrease by the same factor as the weight of the
particles. In a physical system, you can obtain a rough estimate of the microscopic noise
using the dielectric relaxation frequency /2 as the cutoff frequency. If the conductivity is
approximated as
2
n
 = e----------- 19-44
m

857 ATLAS User’s Manual

Physical Models MC Device

we have
2
e k B T L kB TL e
- = ------------  --------------
P V = --------------------------- 19-45
4m r c 
* e 4r c

The standard deviation of the electrostatic potential is the geometric mean between the
thermal voltage and the potential at a distance rc from a charge e.
In Equation 19-25, the volume integral converges within a radius of a few times rc from the
center point. This means that most of the potential noise is local and thus electrostatic screen.
Entering Equation 19-25 as a convergence factor exp(-rs) does not affect it as long as rc is
smaller than the screening length 1/s.
You may want to estimate the fluctuations for the potential for the case of a two-dimensional
system of particles with charge  per unit length in the third dimension. The corresponding
potential generated form the i-th particle is
 ln r
V i = -------------i 19-46
2

Along the lines of the derivation of Equation 19-26, we have

Types of Coulomb Interaction
Modeling the intrinsically three-dimensional Coulomb interaction by a two-dimensional
simulation requires some care to conserve the main physical features of the system. Screening
must be accounted for first. In a non-degenerate situation, we can think of screening as an
exponential decay of the effective interaction potential, according to the space-dependent
screening factor:

–2   kB Te  e 
s = -------------------------
- 19-55
en

where Te is the effective temperature of the carriers and n is their concentration. This picture
holds up to plasma frequencies. For higher frequencies, the ensemble of carriers will not
follow the perturbing potential quickly enough and anti-screening or no screening will occur.
To check the validity of the screening model from a thermodynamical point of view, the
plasma frequency must be compared with the power spectrum of the carrier velocity. The
auto-correlation of velocity extends up to about the frequency S2, where S is the scattering
rate. For thermal carriers, this frequency is of the order of 1012 Hz. The plasma frequency
ranges from 21012 Hz for n1017 cm-3 to about 1014 Hz for n31020 cm-3. Within this range
of concentrations, you can assume all interactions as screened. For lower concentrations, part
of the power spectrum will remain unscreened. In these cases, however, the microscopic
Coulomb interactions can usually be neglected when compared to the macroscopic electric
field. Moreover, the screening length is so large (13 nm at n1017 cm-3) that unscreened
interaction will occur in any case over a significant distance.
The general idea is to separate Coulomb interaction into two classes:
• Short-range Coulomb interaction: This occurs within a screening length. Since the
wave packets representing the electrons can be assumed as overlapping, a simple space-
homogeneous theory based on Bloch functions can be used to estimate the scattering rate
for a given electron concentration. Given the average electron density, you can compute
the full three-dimensional scattering probability per unit time. Screening of the Coulomb
perturbation potential is included in the calculation if necessary. Electron-electron
scattering is applied to carriers within a screening length in the two-dimensional
simulation. The effect of short-range interaction is an exchange of energy between
carriers. The position of the particles is not changed. Therefore, the potential energy
remains constant and ensemble modes are not affected.
• Long-range Coulomb interaction: We can assume that beyond a few screening lengths,
no interaction between single carriers occurs. Carrier-carrier is ineffective on these length
scales because screening attenuates exponentially the mutual force. But, the perturbation
has a large effect on the entire ensemble of carriers. In other words, beyond a screening
length the individual carrier no longer sees other isolated carriers but rather a plasma. The
long-range interaction is thus basically an exchange of energy between a single carrier
and the particle-field ensemble system.

869 ATLAS User’s Manual

Physical Models MC Device

You can model the cutoff between short-range and long-range interactions by choosing a
Poisson-mesh spacing equal to the local screening length [158]. Assuming that the short-
range interactions are correctly represented by a two-particle scattering rate, there remains the
problem of representing the 3D electron-plasma interactions in a 2D model. The most
physical approach that can be followed is the conservation of collisionality in the system.
Discretization and Collisionality
The long-range interaction can be pictured as exchange of energy between the single particles
and the plasma modes. The latter are correctly modeled by a fine representation of the
continuous plasma. Particularly, the mesh spacing must be fine enough to represent plasma
modes that suffer Landau damping. The required maximum spacing is
v th T p
h Landau = ------------- 19-56
2

where vth is the thermal energy and Tp is the plasma period. For example, for vth2107 cm/s
and Tp10-13 s, we have hLandau1 nm. But, the thermal agitation of the carriers depends on the
granularity of the simulation (i.e., each computational particle represents a physical energy
E3D equal to its “computational energy” E2D) times the number of carriers per unit length s
that it represents. No perfect mapping is possible between a completely 2D model and a
completely 3D physical system as the density of states for collective modes changes also with
dimensionality. As a basic constraint, we can try to preserve the collisionality of the system.
For systems of low collisionality, the kinetic energy of carriers dominates over the potential
energy, and interaction with plasma modes does not change it very much. For systems at
thermal equilibrium, the collisionality is given by the number of carriers within a Debye
sphere. This quantity is roughly conserved in the transition between 2D and 3D if
s = s 19-57

This condition ensures that if Coulomb interaction is negligible in the physical system, it will
also be in the 2D model and vice versa. Coupling between single carriers and collective
modes will thus be qualitatively modeled. This criterion also minimizes the difference in
short-range Coulomb interaction with respect to the 3D case [158]. It must be observed that a
higher weight implies a higher physical kinetic energy, so that large particles will heat the
plasma sea more than small particles. If all particles have the same size, the system will reach
thermal equilibrium. If particles of different weights are plasma-coupled, there will be a
transfer of energy from the large particles to the small particles.
Collisionality in Silicon Devices
The critical carrier concentration above, which an uniform sample of silicon is collisional,
can be obtained as:

4 k L  e 3  2 1
n ---  s = ---   ---------------
4 –3
------- = 1 19-58
3 3  e  n

Finally, Fermi’s Golden Rule can be applied to obtain final transition probabilities that can be
applied to all scattering mechanisms.

  Nq 
 G  ij q j  i     k  –   k  
2
P  k k  = --------------   h q  19-67
V q  N q + 1

873 ATLAS User’s Manual

Physical Models MC Device

Acoustic Phonons
For a non-degenerate band at the center of the Brillouin zone, the deformation potential,  , is
diagonal and can be treated as a scalar. This simple situation, energy, and momentum
conservation imply
*
 m u
 2  k cos  – -----------l
q =  19-68
 h 

where ul is the longitudinal velocity of sound, m* is the effective of an electron,  is the angle
between k and q, and the + and - are for emission and absorption respectively. The second
term above is the correction for the phonon involved in the scattering event. Therefore, for an
elastic event, the equation will just be
q = 
 2k cos  19-69
One important observation to make when tying this into a simulation is that when collisions
are assumed elastic, the energy distribution is also assumed to be Maxwellian. Generally, this
assumption is FALSE and cannot be used in Monte Carlo simulation. But, the only case when
this assumption does not hold is for low fields and temperature. At high fields and high
temperature, acoustic phonon scattering can be assumed to be elastic and is the same as
optical phonons.
For realistic band structures, there will be separate probabilities for longitudinal and
transverse modes. But the anisotropy is small and can be neglected by averaging the velocity
of sound over the two directions as
1
u = ---  2u t + u l  19-70
3

The form of the final scattering probabilities are the same for inelastic acoustic phonon
scattering. The main addition is that the function has phonon energy in it.
Keeping the same isotropic approximation for deformation potential coupling and using the
Herring-Voigt transformation on k and q vectors, the magnitude of q becomes
2 m 2
* ml
q = q  -------  cos   + -------  sin   
* t *
19-71
 m0 m0 

m l 1/2 m l 1/2
q  -------   ------- , 19-72
 m 0  m 0

where * is the angle between k* and the principal axis of the valley. This approximation is
sufficient for Silicon.

874 ATLAS User’s Manual

Physical Models MC Device

Optical Phonons
For optical modes, the perturbation Hamiltonian to the lowest value of approximation is
assumed to be proportional to the atomic displacement cite [125]. The continuous medium
approximation from long-wavelength acoustic phonons would be inconsistent for optical
modes. The scattering probability can be written by substituting an optical coupling constant,
D, for potential. The energy associated with intravalley transitions can be assumed to be
constant and given by h op = k B  op where  op is the temperature of optical phonons. The
resulting probability becomes
2
  D t K   N op 
P  k k  = ---------------------       k  –   k  
 h op  19-73
V op  N op + 1

where Nop is replaced by Nop + 1 for emission. Note there is no angular dependence and
therefore scattering is isotropic.
For the final state, it is random with constant probability over the energy-conserving sphere of
k*.
Intervalley Scattering
Electron transitions between states in two different equivalent valleys can be induced by both
acoustic and optical phonons.
The phonon wave vector q involved in the transition remains close to the distance between
initial and final minima, even for high electron energies. So the change in momentum and
frequencies are approximately constant. Therefore, this scattering mechanism is treated
exactly like intravalley scattering by optical phonons.
2
The coupling constant,  D t K  , depends on the types of valleys involved in the scattering
event and the branch of phonons involved.
In the case that electron energies are high and the particles can move up to bands of higher
energy, the appropriate k in the first Brillouin zone must be considered along with the
change in kinetic energy of the electron.
The integrated scattering probability for intervalley scattering is then
2 3/2
 Dt K  m Zf  Ni 
P i    = ----------------------------------   h –   1/2
   19-74
3 i fi
2h  i  N i + 1

where Zf is the number of total number of final equivalent valleys for the scattering in
consideration, Ni is the phonon number, and fi is the energy difference between the bottoms
of the initial and final valleys.

875 ATLAS User’s Manual

Physical Models MC Device

Hole Intraband Scattering

There are both acoustic and optical types of this scattering for holes as there is for electrons.
Due to the complexity of the valence band structure, use three deformation potential
parameters. The characteristic wave function of holes also introduces an overlap factor.
To overcome analytical difficulties without losing the physics, use one coupling constant
instead of the three deformation parameters. For a warped band, the resulting scattering
probability is similarly to the electron acoustic scattering case.

qE  N q  1
2
 ---  1 + 3  cos       k  –   k  
2
P  k k  = -------------   hqu  19-75
Vu  N q + 1 4

Here, E is the coupling constant, and  is the angle between k and k’. A numerical method is
used to solve this integral in MC DEVICE.
For optical hole intraband scattering, the rate can be described with one deformation potential
parameter and has been found to be isotropic [125].
The integrated probability is then
2
3d 0  N q 
P  k k  = ---------------------------   h 
     k  –   k   19-76
2 op
2V op a 0  N q + 1

where d0 is the deformation potential parameter. The resulting energy dependence of the final
equation is analogous to that for optical scattering of electrons.
Hole Interband Scattering
When a multi-band model is considered, scattering rates should be used to account for the
scattering (transitions) between the bands. In MC DEVICE, only two bands (the heavy and
light hole bands) of the three-bands hole scattering model are used. Since the calculations for
the heavy-to-light and light-to-heavy interband scattering rates are somewhat complex, these
rates are approximated by setting heavy-to-light rate equal to the light-to-light acoustic rate
and by setting the light-to-heavy rate equal to the heavy-to-heavy acoustic rate. The file
scat.out contains scattering rates for all possible processes. Any rates not calculated in
MC DEVICE (such as those involving the split-off band) are set to 0.0.
Density-of-States Scaling
The calculation of density of states is a basic ingredient of any bull-band Monte Carlo
transport model to compute scattering rates and probabilities of final states. The density of
states including both spin directions is

1 
N  E  = ------ 
VR       k  – E 

19-77
n k  BZ

where n is the band index. The sum is over all cubes/tetrahedra in all bands. It is assumed that
the energy (k) is known at the corners of whatever shape is used to discretize the Brillouin
zone (BZ).

876 ATLAS User’s Manual

Physical Models MC Device

Final State Selection

In MC DEVICE, final state selection is based on the energy as opposed to being based on the
momentum (or wave vector). To accomplish this, a relation for k(E) is obtained. Since the
bandstructure is stored in a table that is ordered by the wave vectors (k values). A final state
table can be obtained by reordering the bandstructure table by energies, E.
The uncertainty principle is satisfied in MC DEVICE by broadening the energy using the
imaginary part of the self-energy (i.e.,  h   2   ) and not using a one-to-one relationship
between k and E. For each final energy, there are a range of final energies (and a
corresponding range of momenta) that may be selected. Therefore, there is uncertainty in the
momentum of the electron after a scattering event in accordance with the uncertainty
principle.
Using Scattering Parameters
At the beginning of the simulation, scattering rates are either calculated or are loaded from
data files. In either case, the scattering rate table is written to scat.out. The manner of
obtaining the scattering rates and the number and type of the mechanisms, depends on
whether the strain model is on (TYPE=1|TRUE|YES on the mmat statement) and whether the
TYPE=ANALYTIC or TYPE=FULLBAND is used on the ESCAT statement. Below is a description
of its contents of the scat.out file. Table 19-2 is for electrons and holes when the strain
model is off or when the strain model is on and TYPE=ANALYTIC. Table 19-2 is for electrons
when the strain model is on and when TYPE=FULLBAND on the escat statement.

Table 19-2. Form of scat.out (strain model off or TYPE=ANALYTIC)

Column Format when CARRIER=E Format when CARRIER=H

1 Energy Energy

2 Total Total

3 Ac abs h-h AC abs

4 Ac emi h-h AC emi

5 XX f abs (Tp=TF1) l-l AC abs

6 XX f abs (Tp=TF2) l-l AC emi

7 XX f abs (Tp=TF3) s-s AC abs

8 XX g abs (Tp=TG1) s-s AC emi

9 XX g abs (Tp=TG2) h-h OP abs

10 XX g abs (Tp=TF3) h-h OP emi

11 XX f emi (Tp=TF1) l-l OP abs

12 XX f emi (Tp=TF2) l-l OP emi

13 XX f emi (Tp=TF3) s-s OP abs

14 XX g emi (Tp=TG1) s-s OP emi

877 ATLAS User’s Manual

Physical Models MC Device

Table 19-2. Form of scat.out (strain model off or TYPE=ANALYTIC)

15 XX g emi (Tp=TG2) h-l Interband

16 XX g emi (Tp=TG3) l-h Interband

17 XL abs (Tp=TL1) h-s Interband

18 XL abs (Tp=TL2) s-h Interband

19 XL abs (Tp=TL3) l-s Interband

20 XL abs (Tp=TL4) s-l Interband

21 XL emi (Tp=TL1) Ionized impurity

22 XL emi (Tp=TL2) Impact ioniz.

23 XL emi (Tp=TL3)

24 XL emi (Tp=TL4)

25 Ionized impurity

26 Impact ioniz.

Table 19-3. Form of scat.out (strain model on and TYPE=FULLBAND)

Column Format when carrier=e

1 Energy

2 Total

3 Longitudinal AC abs

4 Longitudinal AC emi

5 Transverse AC abs

6 Transverse AC emi

7 OP abs

8 OP emi

9 Impact ioniz.

The statements relating to phonon scattering deal with both acoustic (AC) and optical (OP)
types, X-L and X-X, and f and g types of scattering events. The abbreviations abs and emi
refer to absorption and emission. The abbreviations h, l, and s refer to heavy, light, and split-
off. Tp refers to the phonon temperature, which can be one of 3 values for f-type and g-type
transitions and one of 4 values for X-L transitions.

878 ATLAS User’s Manual

Physical Models MC Device

Acoustic scattering parameters are set by the AC statement as

AC DEFO=0.0
where DEFO is the desired deformation potential in [eV]. The default value is 9.0 eV. Optical,
or high energy, phonon scattering is characterized by the PHON statement as
PHON ETH=0.0 MODEL=0 FILE=“name”
where ETH is the energy threshold, in [eV], to activate this type of scattering. MODEL chooses
which scattering model to use.
Currently, the only model implemented in MC DEVICE is model=0 (Tang model). FILE
chooses the density of states file. The default is dos_c.in, which is located in the
$MCDEVICEDIR directory.
The intervalley scattering statements XL, XXF, and XXG set phonon temperatures and
deformation potentials. The default values have been determined experimentally for silicon.
As shown above, 4 values are used for X-L transitions and 3 values are used for f and g-type
X-X transitions.
XL TEMP=(672.0,634.0,480.0,197.0) DEFO=(2e8,2e8,2e8,2e8)
XXF TEMP=(220.0,550.0,685.0) DEFO=(3e7,2e8,2e8)
XXG TEMP=(140.0,215.0,720.0) DEFO=(5e7,8e7,11e8)

19.7.6 Impurity Scattering

For all materials, the lattice sites are many thousands of times heavier than electrons.
Therefore, impurity scattering is essentially elastic in nature. Therefore, for the energy
distribution to match reality, it must be accompanied by some dissipative mechanism.
The Brooks and Herring approach from [125] is described below.
For an ionized impurity, the scattering source is a screened Coulomb potential. The potential
can be represented by a simple decaying exponential as
2
Ze
V  r  = -------- exp  –  s r  19-78
r

–1
where  s is the screening length,  the dielectric constant, and Z the number of charge units
in the impurity. The Debye formulation for nondegenerate statistics gives
2 12
 4e n 0
 s =  ------------------ 19-79
 k B T 

where n0 is the free carrier density.

879 ATLAS User’s Manual

Physical Models MC Device

The Fourier transform of a screened Coulomb potential is given by


1
H'  q  = ------------------- V  r  exp  – iq  r  dr
32
(2)
2 19-80
2Ze 1
= ---------------- -----------------
  + q 2 2
s

and leads to a scattering probability per unit time of

5 3 2 4
2  Z n Ie 1
P  k k'  = ------------------------- G -------------------------     k'  –   k   19-81
2 2 2 2
hV ( s + q )

where G is the overlap integral.

For electrons we can look at the case for spherical, parabolic bands or ellipsoidal, non-
parabolic bands. For the first case, G is 1, and multiplying by the density of states V(2)3 we
get
52 2 4 12
2 n I Z e 
P e I    = -------------------------------- ------------------------ 19-82
2 2 1  2 1 + 4  
  m 

where
2
h2 
  = -----------2 19-83
2m

For non-parabolic bands, The Herring-Voigt model is used. The results are given as
52 2 4
2 n I Z e 1 + 2
P e I    = -------------------------------- ----------------------- 19-84
2 12 12
   m d 1 + 4 --------- 
-
 '
where
2
2

h- 2
 ' = ----------- 19-85
2m d

There are also two cases to account for impurity scattering of holes. There is the spherical
parabolic band case and the warped band case. For the parabolic band, we can adapt the
general scattering probability given above by accounting for the overlap factor. This gives
5 3 2 4
2  Z n I e 0.25  1 + 3 cos 2  
P  k k'  = ----------------------------- ---------------------------------------------      k'  –   k   19-86
2 2 2 2
hV ( s + q )

880 ATLAS User’s Manual

Physical Models MC Device

and finally integrating over the solid angle we arrive at

52 2 4 12
2 n I Z e 
P h I = -------------------------------- ------------------------ F ov 19-87
2 2 1  2 1 + 4  
  m 

where Fov is the overlap factor. The overlap factor essentially works to reduce the efficiency
of the impurity scattering mechanism and can be found in [125].
We will neglect the warped band case because it can be neglected as long as the warping is
small. For treatment of the case, see [125].
Implementation of Impurity Scattering
In MOCA, the Ridley impurity scattering model is implemented. The scattering rate is
calculated as follows:

2 1
------ --- 2 *
1 m A
* ae m k B T  1 + 2E 
---------- = -------------------
-  1 – exp – ------------------------------------------------------------ 19-88
 imp 1 + 2E  Ne h 2 
2
8 h 2   1 + ------------------------- -
 8m k B T
* - 

Final state selection is for the Ridley model are calculated including the factor 2 from
Chandrashekar.
2
e md kB T
F 1 = ----------------------
2
8  h 2 N
2 2
Ne h 19-89
F 2 = ----------------------
-
8m d k B T
2 2
– 2 N
F 3 = -------------------
3

These factors are used in density of states scaling. See “Impact Ionization and other topics”
on page 881 for more information.
MOCA includes the option to modify Ridley parameters using the RIDLEY statement.
RIDLEY SCALE=1.0 TSTEP=n
Here, SCALE is a scaling factor. TSTEP is the simulation time step where to update the
concentration.
Impact Ionization and other topics
The other impurity-related scattering statements are: IMPACT, SSPARAM, SSREGION,
TRACKSCAT, TRAJ, and ZSREGION. These are described below.
The IMPACT statement is for impact ionization.
IMPACT MODEL=0|1|2|3 PARAM=(i,j,k) ETH=(E1,E2,E3) \
WEIGHT=NO FINAL=YES sec=NO CSEC=NO \
START=1000 RANDSEC=NO RESPOS=(min,max)

881 ATLAS User’s Manual

Physical Models MC Device

Monte Carlo Schrödinger Formulation

The approach used in MC DEVICE is to periodically solve the Schrödinger equation during a
simulation along the direction perpendicular to transport using the self-consistent electrostatic
potential as input. The quantum potential in this method is found using the exact energies and
wave functions. The correction to the Monte Carlo transport potential is found by linking the
quantum density to the potential through
V qc  z  = – k B Tlog(n q  z ) – V p  z  + V 0 19-104

Here, Vqc is the quantum correction. z is the direction normal to the interface. nq is the
quantum density from the Schrödinger equation or equivalently the converged Monte Carlo
concentration. Vp is the potential from the solution of Poisson. V0 is a reference potential
determined by the fact that the correction should go to zero in a bulk-like region.
In Equation 19-104, the Schrödinger concentration profile goes in an arbitrary multiplicative
constant. For the purpose of filling the energy levels, only the relative populations are
significant. The particles simulated by Monte Carlo, once the potential correction has been
added, move in a potential environment that shapes the space distribution as in the result of
the Schrödinger equation. The non-equilibrium full band Monte Carlo model determine the
magnitude of the particle distribution at different points along the transport path and
momentum space trajectories.
MC DEVICE chooses the Fermi level to fill the energy levels according to the Maxwell-
Boltzmann distribution. The setting of the Fermi level, EF, also fixes the potential V0.
Explicitly, this can be seen by
V p  z  + V qc  z  V
–  ----------------------------------- -----0-
 kT  kT
n q  z  = exp exp 19-105

V0  z  – Vp  z  Vp  z  – EF
–  --------------------------------- –  --------------------------
 kT   kT 
exp  N eff exp 19-106

V 0 = kTlog(N eff) + E F 19-107

Neff is the effective density of states and EF is the specified Fermi level. You can see that nq
nc as Vqc 0.
Using the Maxwell-Boltzmann relation in the formulation for the correction should be
questioned since degeneracy is not explicitly taken into account. Degeneracy can be
important for filling the levels in highly inverted MOS structures. But, despite using non-
degenerate statistics to relate voltage and concentration in the correction, note that sufficient
accuracy is retained even in the highly degenerate regime for all practical cases tested. This
can be explained by the similar shape of the first three sub-bands in the region where the
quantum correction is important. The correction only begins to break down at non-physical
oxide fields where the higher subbands become degenerate.

889 ATLAS User’s Manual

Physical Models MC Device

Using the Quantum Correction

Using the quantum correction proceeds in a similar fashion to using the Monte Carlo
simulation. First, quantum parameters are defined using the QUANTUM statement. Then,
simulation regions are defined using QREGION and SCHRREGION. Finally, you build a sub-
mesh for the solution of the Schrödinger equation using SCHRMESH.
Default values for the quantum parameters are set in the default parameters file. These
parameters have been optimized and you should not change them for simulation of practical
devices. The defaults for the QUANTUM statement are
QUANTUM RESTART=NO OPTION=(NO,NO) DECAY=(100e-15,2000e-15) \
MINCONC=(1e10,1e10) MAXQV=-1.0 MAXQF=-1.0 \
STRENGTH=1.0 Q1D=(-1,-1,-1) FREEZE=PINF SIGMA=8e-8 \
RADIUS=3 LADDERS=(1,0,0) ENERGY=NO
In the example, quantum-corrected device input file is given in Section D.3 “Examples”. The
QUANTUM statement is not included because the defaults are adequate. If you want to restart
from an old quantum correction solution, change RESTART to YES. DECAY is the decay in time
averaging. MINCONC is for setting the minimum Monte Carlo concentration and the minimum
quantum concentration. MAXQV sets the maximum allowed correction to the potential. MAXQF
does the same for the field. STRENGTH is only a multiplicative factor. Q1D is for 1D averaging
and is turned off if the value is -1. FREEZE is to specify a freezing iteration. An option not
included in the default configuration is SMOOTH1D and is for 1D smoothing. SIGMA is the
smoothing standard deviation, and RADIUS is the integration radius for smoothing. LADDERS
is the number of ladders to use, and ENERGY is whether quantum potential should contribute
to carrier energy. One last option that is also not included in the defaults file is NOBARR. This
is for choosing whether to perform barrier fitting.
The first statement you need in order to use the quantum correction is QREGION.
QREGION MODE=SCHROEDINGER RAMP=(0,0) \
BOUNDP=(x1,x2,y1,y2) ABOUNDP=(x1,x2,y1,y2) \
FIT=(-1,-1) FBOUNDP=(x1,x2,y1,y2) \
SHIFT=-1 SHIFTDIST=0 DIR=Y \
TSTEP=4000 XSTEP=1 ZERO=0
MODE is to choose the type of correction to use, such as FEYNMAN and WIGNER. RAMP is to
specify a transient period where to ramp up the correction added to the Monte Carlo potential.
BOUNDP is the region where to calculate the correction (see “Regions” on page 834). Again, a
BOUND parameter is provided for backward compatibility. ABOUNDP is a region where to
calculate attenuation. ABOUND is provided for backward compatibility. FIT is to specify a
fitting function to fit the correction potential to. By setting the field to -1, it turns off the
fitting function. FBOUNDP is the region where to fit the correction. FBOUND is provided for
backward compatibility. SHIFT and SHIFTDIST are for specifying a grid to shift for
interpolation and the distance to shift the grid respectively. DIR specifies the direction for the
1D correction, for a horizontal MOS device. This direction will be y. TSTEP is to specify the
simulation step interval where Schrödinger is to be solved. XSTEP is the space step where to
perform the calculation. Finally, the ZERO parameter specifies a grid line (>=1) where the
quantum correction should go to zero. Leaving the ZERO parameter set to 0 (or -1 for

890 ATLAS User’s Manual

Physical Models MC Device

backward compatibility with MOCA) lets MC DEVICE calculate this on its own positive
integral value for the grid line at which the quantum correction will go to to zero.
The SCHRREGION statement works very similar to the REGION statement. Any material chosen
in SCHRREGION must have an entry in schroedinger.material.in. A quantum region is
specified as follows.
SCHRREGION N=n MAT=AIR|SI|SIO2|SI3N4 BOUNDP=(x1,x2)
N is an integer (where N>=1) and represents the quantum region number. MAT is the material
type. BOUNDP is the minmum and maximum position along the quantization axis (the y axis).
A BOUND parameter, which uncharacteristically also uses positions rather than mesh indicies,
is provided for backward compatibility. The mesh for the quantum regions is defined using
SCHRMESH. It is used in exactly as XMESH and YMESH. See the following.
SCHRMESH MODE=n LOC=x-pos RATIO=r
The definitions for SCHRMESH are the same as for the regular mesh statements. Because the
correction is only calculated along a single direction, only a single grid needs to be defined.
To complete the quantum correction section of the input file, include the EBREGION statement
for an energy barrier.
EBREGION BOUNDP=(x1,x2,y1,y2)
The BOUNDP parameter defines a boundary and is typically set to the boundaries of the whole
device (see “Regions” on page 834). Again, a BOUND parameter is provided for backward
compatibility.
The statement associated with the Schrödinger quantum correction is the calculation of
transverse carrier temperature. You can calculate transverse temperature, Tt, with TTEMP.
TTEMP RESTART=NO SIGMA=0 SMOOTH=NO BOUNDP=(x1,x2,y1,y2)
The TTEMP statement allows you to define a region over which to calculate the transverse
carrier temperature, T t . The RESTART parameter, as with other statements, allows you to start
using a previous solution file. SIGMA is set to 0 to run in normal mode or to 1 to run in best-fit
mode. The SMOOTH parameter allows you to smooth the transverse temperature prior to
writing it out. The BOUNDP parameter sets the boundary where to calculate the transverse
temperature (see “Regions” on page 834). Again, a BOUND parameter is provided for
backward compatibility.

891 ATLAS User’s Manual

Physical Models MC Device

19.7.9 Schottky Barrier Injection

Background
The Schottky barrier is a junction between a metal and semiconductor where the work
function of the metal is greater than the electron affinity of the semiconductor. When in
contact, these two materials then form a fixed energy barrier of constant height but varying
width.
Although a MOSFET with Schottky source and drain regions looks similar to a bulk
MOSFET, the behavior between the two is very different. Carrier flow from the source
contact to the conduction channel takes place mainly by tunneling through the Schottky
barrier at the silicon/silicide interface. The width of the barrier is spatially modulated by the
gate voltage, and conduction occurs when the barrier is sufficiently thin [310]. When the gate
voltage is small, the barrier between source and channel is sufficiently wide so that
appreciable tunneling does not occur. A certain amount of leakage current due to thermionic
emission from the metal may be present, originating from the tail of the quasi-equilibrium in
the silicide. For higher gate voltages, the Schottky barrier just below the Si–oxide interface
may become sufficiently thin to allow tunneling. A Schottky barrier is also present at the
drain to channel interface and may impede the flow of carriers at low drain biases.
Transport in the channel remains semi-classical because scattering in silicon is strong enough
to break quantum coherence. This justifies a coupled quantum injection – Monte Carlo
particle transport approach.
The Airy Function Approach
The injection between contacts is handled by transmission probabilities, which are solutions
of the Schrödinger equation along paths perpendicular to the silicon/silicide interface. The
transmission probability is calculated using a standard Airy function approach based on the
1D Schrödinger equation. The Airy approach was chosen over WKB because of its accuracy
with thermionic reflection and particles with energy near the top of the tunneling barrier.
For an equation of the form:
2
----------
- = z 19-108
2
z

the solution looks like

  z  = aAi  z  + bBi  z  19-109
where Ai(z) and Bi(z) are the Airy functions given by
 3
Ai  z  = --- cos  ---- + sz ds

1 s
19-110
 0 3 

3
s
 – ---- + sz 3
+ sin  ---- + sz ds

1 3 s
Bi  z  = --- exp 19-111
 0 3 

The direct solution of the Schrödinger equation also has the added benefit of seamlessly
treating thermionic emission and tunneling over the energy range.

892 ATLAS User’s Manual

Physical Models MC Device

To compute the current injected by the silicide contacts, a transmission coefficient TEinck|| is
defined as the ratio of the transmitted and incident probability current densities.
2 *
J transmitted  k2 m1
- = ---------------  ---------------
T  E inc k ||  = ----------------------- 19-112
J incident * 2
m2  k1

Here, the subscripts 1 and 2 refer to the silicide and silicon regions respectively. ||2
represents the wave function probability density for the carriers. At each iteration, a table of
transmission probabilities is generated for each mesh location along the contact interface.
Once the quasi-equilibrium distribution in the silicide is determined in terms of incident
energy, the injected current density is obtained by integrating the product between carrier
density and transmission probability.
Usage of Schottky Contacts
Compared to typical MOS devices, simulation of Schottky devices require few particles.
Since the behavior of the silicide contacts can be shown by injection boundary conditions,
depending on a quasi-equilibrium distribution, the entire ensemble of simulated particles is
applied to transport in the conduction channel. This greatly improves performance and
resolution of the method and allows you to easily probe the subthreshold regime.
For Schottky barrier devices, you can set the value for N on the PARTICLE statement to around
10000 without losing accuracy. To create a Schottky contact, you can use the REGION
statement in the same way. Two silicides are provided, platinum silicide PTSI, and erbium
silicide ERSI. To set the barrier height for your Schottky contact, use the MATDEF statement
and modify the BARRIER field.
For simulation of Schottky barrier devices, there is also a secondary input file. The file
inj_2.in is read in by MC DEVICE to interpret the Schottky-specific statements. The only
thing is that the Poisson barrier height in the inj_2.in file must match the value for
BARRIER set in the MATDEF statement for the Schottky material.

19.7.10 Gather/Scatter Algorithm

The identification of particles that need special handling in the simulation is performed as
follows:
1. The predictor step is computed, yielding the k-vector and the position at the end of the
time step. The first-order formula is used in k-space. The second-order formula in real
space.
2. The time of the next scattering is computed for all particles as the ratio between the
remaining normalized lifetime and the total scattering rate. This can be less than zero for
particles that should have scattered in the previous time step, or greater than DT for
particles that do not scatter.
3. The k and x-vectors at the end of the free flight are computed. The mesh and region
indices are also computed. These state variables have the suffix NEW: XNEW, KNEW, IXNEW,
and so on. A forward difference is used for the k-vector, assuming that the electric field
doesn’t change much during one time step. A second-order formula is used to compute
the final position, in case the velocity varies during the free flight (e.g., for very string
electric fields, or long time steps).

893 ATLAS User’s Manual

Physical Models MC Device

4. If a particle changes region during the predicted time step, the time of boundary crossing
is computed. A straight-line approximation is used for the flight trajectory. This reduces
the calculation to a simple geometrical problem, speeding up the calculation considerably.
Since region crossings are relatively rare events, this approximation does not affect
accuracy very much. If the time of crossing is less than the scattering time, the particle is
added to the appropriate reflection-handling list. The time of crossing and the side of the
rectangular region where the crossing occurs are stored.
5. Particles that scatter are moved to their scattering position XNEW. A final state is selected
and used as a new initial state for the remainder of the time step. The new predictor step is
computed.
6. For each scattered particle, the region index is recomputed at the end of the time step, and
compared with that at the beginning of the time step. The intermediate region index is not
used in case the particle touches the corner of a region while scattering. If they differ, the
particle is added to the reflection list.
7. For each rectangular region, the reflection is performed for all particles in the associated
list. Boundary check is performed again and particles are added to the reflection lists. The
step is repeated until all particles maintain the same region index during the last subflight
in the time step.

19.7.11 Interpolation Schemes

Trilinear Interpolation
Consider a trilinear function in the form:
f  x y z  = a 0 + a 1 x + a 2 y + a 3 z + a 4 yz + a 5 xz + a 6 xy + a 7 xyz 19-113
where the following definitions are made:
f  0 0 0  = f 000 19-114

f  1 0 0  = f 100 19-115

f  0 1 0  = f 010 19-116

f  1 1 0  = f 110 19-117

f  0 0 1  = f 001 19-118

f  1 0 1  = f 101 19-119

f  0 1 1  = f 011 19-120

f  1 1 1  = f 111 19-121

894 ATLAS User’s Manual

Physical Models MC Device

Equations 19-113, 19-114, 19-115, 19-116, and 19-118 give:

a 0 = f 000 19-122

a 1 = f 100 – a 0 19-123

a 2 = f 010 – a 0 19-124

a 3 = f 001 – a 0 19-125

Equations 19-113, 19-116, and 19-120 yield:

f 011 = a 0 + a 2 + a 3 + a 4 = f 010 + a 3 + a 4 19-126

Therefore, by symmetry:
a 4 = f 011 – f 010 – a 3 19-127

a 5 = f 101 – f 001 – a 1 19-128

a 6 = f 110 – f 100 – a 2 19-129

Equations 19-113 and 19-121 yield:

f 111 =
 ai = f011 + a1 + a5 + a6 + a7 19-130
j=0

and therefore:
a 7 = f 111 – f 011 – a 1 – a 5 – a 6 19-131

An efficient implementation of the interpolation algorithm can be achieved by factoring

Equation 19-113.
f  x y z  = a 0 + x  a1 + a 5 z + y  a 6 + a 7 z   + y  a 2 + a 4 z  + a 3 z 19-132

895 ATLAS User’s Manual

Physical Models MC Device

Figure 19-17: Tricubic interpolation scheme. The letters denote the vertices of the cube and lower-order
interpolation points. The numbers denote the edges of the cube and higher-order (cubic) interpolation points.
Tricubic Interpolation
The tricubic interpolation is achieved by a sequence of 7 simple cubic interpolations (see
Figure 19-17). The 1D cubic interpolation consists in finding a function of the form:
2 3
f  x  = a0 + a1 x + a2 x + a3 x 19-133
where the following definitions are made:
f  0  = f0 19-134

f'  0  = f' 0 19-135

f  1  = f1 19-136

f'  1  = f' 1 19-137

Substituting them, we have:

a0 = f0 19-138

a 1 = f' 0 19-139

a 3 = f' 1 – 2f 1 + f' 0 + 2f 0 19-140

896 ATLAS User’s Manual

Physical Models MC Device

a2 = f1 – a3 – a1 – a0 19-141

The complete evaluation of one tricubic point requires 73 = 21 multiplications.

19.7.12 WKB Formula

The Schottky-lowered barrier can be described by the formula:
e
E  x  = W – Fx – --------------- 19-142
16x

Here, W is an effective barrier (i.e., the ideal (non-lowered) band-edge discontinuity, minus
the normal component of the kinetic inside the oxide). F is the electric field and  is the
effective image-force dielectric constant [76]. We consider lowering at one of the ends of the
barrier (Figure 19-18).

Figure 19-18: Schottky Lowering.

The entry and exit points xin and xout are found for Ex0.

W eF 
x in out = -------  1  1 – ----------------- 19-143
2F 
4W 
2

For very low fields or very high barriers or both, the second term within the square root can
be much smaller than one. In such cases, a much more accurate expression for xinout is
obtained by a series expansion of the square root:
e
x in = ------------------ 19-144
16W

W
x out = ----- 19-145
F

897 ATLAS User’s Manual

Physical Models MC Device

Finally, the WKB integral is performed numerically on a very fine mesh. Since a typical
barrier is at most of the order of 1 Ry and the effective mass in the oxide is about 0.5m0, the
–1
attenuation factor is the order of a fraction of a 0 , where a0 is the Bohr radius. A mesh step
of a02 has been chosen.

19.7.13 Strain
You can use the strain model in MC DEVICE to account for the effect of strains on the
transport of electrons and holes in bulk and 2D simulations. This model can be applied to
different materials and is also known as the multi-material model [311]. In its present form,
the model can treat semiconductors that have 16-fold degeneracy in their band structures and
where the irreducible wedge (IW) falls in the domain where kx  0 , ky  0 , kz  0 , and
k y  k x . In this version of MC DEVICE, fittings are provided for electron transport in silicon
under biaxial tensile strain. When used in 2D, you may define any number of rectangular
strain regions. In each strain region, the strain varies linearly from the value of the effective
strain (or the effective mole fraction) you provide at the corners of the rectangle.
Band Structure
When you perform a bulk simulation, you need to consider just one value of the strain that
pertains to the bulk. But when you perform device simulations, you often need to consider
strain that varies over the device. For either type of simulation, it is convenient if your strain
model uses one set of strain fittings that pertain to a wide range of strain levels so that these
cases can be modeled with little change in your input. This is the case with the strain model in
MC DEVICE. To test different bulk cases for a given strain type, you change the effective
strain for each case. To test different strain levels within a device, you define the strain over
your device structure. For supported strain types, you can simply set the level of strain to use
the model. In such cases, you do not need to characterize the effect of strain on the band
structure and on carrier scattering since it has already been set.
In Monte Carlo electronic transport modeling, two operations related to the semiconductor
band structure are performed repeatedly. The first operation is called the forward (or direct)
evaluation. In the forward evaluation, the modeling tool determines the carrier energy, E, and
carrier velocity, v, of each MC carrier based on its band index, n, and its wave vector, k. The
second operation is call the reverse (or indirect) evaluation. In the reverse evaluation, the
modeling tool determines the band index, n, and the wave vector, k, after each MC carrier has
scattered. Figure 19-4 shows a flowchart that depicts the forward and reverse evaluations
done in MC DEVICE when the strain model is active. (The evaluations done in MC DEVICE
when the strain model is not active amount to the top two-thirds of Figure 19-4.) Although
not shown in Figure 19-4, the strain model also provides strain corrections for each
component of the carrier velocity. Like the strain correction for the energy, the strain
correction for each velocity component depends quadratically on the effective strain, m. For
example, the strain correction for the x-component of the velocity involves a term that
depends linearly with the effective strain (and uses the coefficients ax) and a term that
depends quadratically on the effective strain (and uses the coefficient bx).
For the forward evaluation, the strain model uses the unstrained band structure along with
fitting coefficients of the quadratic correction terms, a(n,k) and b(n,k). A quadratic correction
was found to be reasonably accurate for modeling the band structure of strained silicon on
relaxed SiGe over the entire Brillouin zone for both valence and conduction bands and for
both the carrier energy and for the carrier velocity components [311]. Since two coefficients

898 ATLAS User’s Manual

Physical Models MC Device

are needed for the correction for every point in the irreducible wedge and for every band, the
amount of data needed to characterize a strain-dependent band structure is three times that of
the unstrained band structure.
For the reverse evaluation, the strain model performs a search to obtain the strained band
index, nm, and strained wave vector, km, of each MC carrier. In this case, the reverse lookup
table that is used to obtain the final states based on the energy is extended to include the two
quadratic fit parameters, a(n,k) and b(n,k). First, the final energy of the carrier is searched in
the energy-ordered lookup table based on the energy of that state in unstrained silicon. This
search yields an unstrained state (n0,k0). The location of the strain-dependent final state (nm,
km) can be reduced to a small local region in the table around the unstrained state (n0,k0). The
local region is small because the modification of the energy of the strained state is small
relative to the energy range of the table. The size of the local region is chosen so that it will
always contain the strained state.
Second, the strained state (nm,km) is determined by randomly choosing a state within the local
region and checking if the strained energy, Em, at the randomly chosen state matches the
known final energy of the carrier, Ef.

Note: This search will correctly sample the states within the local region according to the density of states (DOS).

Forward Evaluation Reverse Evaluation

Initial band index = n Final energy=Ef
Initial wave vector = k

Interpolate to get E0=Em=0(n,k) Search full table for n0,k0=nm=0,km=0(E0=Ef)

Calculate Em as Search local region near n0,k0 for nm,km where

Em=E0+a(n,k)*m+b(n,k)*m2 Em=Em=0(nm,km)+a(nm,km)*m+b(nm,km)*m2=Ef

Figure 19-19: Forward and reverse evaluations for the strain model. m pertains to the effective strain (or to
the effective mole fraction). When the strain model is inactive, the last step of the evaluation is skipped.

899 ATLAS User’s Manual

Physical Models MC Device

Phonon Scattering
MC DEVICE models the transport of charge carriers but not the transport of phonons. In this
context, the term phonon scattering refers to carrier-phonon scattering or the scattering of
charge carriers with phonons. The phonon scattering rates in MC DEVICE are determined at
the start of the simulation by one of two methods. The first method is to calculate the rates
using analytic formulas and is selected by setting TYPE=ANALYTIC on the ESCAT statement.
The second method is to load the rates from data files and is selected by setting
TYPE=FULLBAND on the ESCAT statement.
The first method (TYPE=ANALYTIC) is more flexible and supports a more complete set of
mechanisms in MC DEVICE. For this method, the phonon scattering mechanisms include:
intravalley acoustic, intervalley X-X f-type (at 3 phonon temperatures), intervalley X-X g-
type (at 3 phonon temperatures), and intervalley X-L (at 4 phonon temperatures). Both
absorption and emission of phonons are considered for all carrier phonon scattering
mechanisms. Therefore, there are a total of (1+3+3+4)*2=22 carrier-phonon scattering
mechanisms. Anisotropy in the final state selection is included by having different
mechanisms (e.g., f and g-types) and through the density of states (DOS). The formulas for
these rates are scaled by the full-band DOS at high energies to obtain accurate values for the
rates at high energies. The rates for these mechanisms are described in Section 19.7.5
“Phonon Scattering”. The format of the scat.out file produced when TYPE=ANALYTIC is
given in Table 19-2.
The second method (TYPE=FULLBAND) enables you to calculate the scattering rates using full
band structure although it must be done outside of MC DEVICE. For this method, the phonon
scattering mechanisms include: longitudinal and transverse acoustic scattering and one type
of optical scattering. Both absorption and emission of phonons are considered for all carrier-
phonon scattering mechanisms. Therefore, there are a total of 3*2=6 carrier-phonon
scattering mechanisms. In this case, carrier phonon scattering is anisotropic only through the
density of states. In this case, the rates do not need to be scale by the full-band DOS at high
energies, since they were calculated with full band structure. Table 19-3 shows the format of
the scat.out file produced when TYPE=FULLBAND.

900 ATLAS User’s Manual

Physical Models MC Device

Bulk Simulations
Bulk simulations were performed with MC DEVICE and compared with experimental data to
verify the model. The model was originally developed to treat transport in silicon under
biaxial tensile strain, so much of the testing has been performed for this type of strain. For this
case, the low-field electron mobility versus effective mole fraction is shown in Figure 19-20.
Here, the TYPE=FULLBAND is used on the ESCAT statement and FIELD=(1e3 1e3) on the
BULK statement. The plot of the field-based electron mobility estimate shows that the strain
model in MC DEVICE obtains good agreement with the experimental results obtained by
Welser [302] and with the experimental results obtained by Ismail at x=0.3 [124].

Figure 19-20: Electron mobility versus Substrate Ge fraction, x.

The mobility enhancement saturates at approximately x=0.2 (20% Ge in the substrate). This
indicates that the splitting of the X valleys of the conduction band is sufficiently large at this
mole fraction (strain level) that the entire population of electrons occupies the lower-lying X
valleys where the effective mass is lower. Further splitting of the X valleys provides little
increase in the electron mobility, since all electrons already reside in the lower-lying X
valleys at x=0.2.

Note: The strain model in MC DEVICE predicts the impact of strain on band structure and carrier scattering. The
strain model in MC DEVICE does not predict the impact of defects that can occur at high strain levels and
the corresponding impact on carrier transport.

To use the strain model in MC DEVICE for bulk modeling, you must use some or all of the
following input statements: MMAT, ESCAT, and BULK.
The MMAT statement is used to turn on the strain model.
MMAT TYPE=1

901 ATLAS User’s Manual

Physical Models MC Device

If the TYPE parameter is set to 1|TRUE|YES, the strain mode is active. By default, the TYPE
parameter is set to 0|FALSE|NO, so the strain model is off.
The ESCAT statement is used to control carrier-phonon scattering (see “Phonon Scattering” on
page 900).
ESCAT TYPE=ANALYTIC FBPROC=7 FBBAND=4 ANPROC=24 ANBAND=1 \
FPNCPLFILE="phcpl_c.in" PHNFACFILE="phon_fac_c_uni_300k.in"\
PLOTFILE="scatfl_c_uni.in"
If the TYPE=ANALYTIC (as it is by default), then the scattering rates are calculated inside MC
DEVICE using analytic formulas. If the TYPE=FULLBAND, then scattering rates are loaded from
files in the $MCDEVICEDIR directory as specified by the PHNCPLFILE, PHNFACFILE, and
PLOTFILE parameters.
FBPROC sets the number of scattering processes used for TYPE=FULLBAND. FBBAND sets the
number of bands used for TYPE=FULLBAND. anproc sets the number of scattering processes
used for TYPE=ANALYTIC. anband sets the number of bands used for TYPE=analtyic.
FHNCPLFILE and PHNFACFILE set two filenames related to the phonon scattering rates when
TYPE=FULLBAND. PHNCPLFILE sets the file which contains the phonon coupling parameters.
PHNFACFILE sets the file which contains the sum prefactor in the phonon scattering formulas.
PLOTFILE sets the file from which the phonon scattering rates will be loaded when
TYPE=FULLBAND. Column 1 in PLOTFILE is the carrier energy in eV. Columns 2-7 are the 6
phonon scattering rates. The 6 phonon scattering rates pertain to columns 3-8 of scat.out
produced for this case (see Table 19-3). The rates must be scaled by the squared phonon
coupling parameters (in “phcpl_c.in”) and scaled by the sum prefactors (in
“phon_c_uni_300k.in”) in order to get units of 1/s. This calculation is done inside MC
DEVICE after the data are loaded and the results can be written to “scat.out” (see Table 19-
3).

Note: Fittings for TYPE=FULLBAND are only provided for simulating MC electrons in strained silicon
(CARRIER=E on the ALGO statement). These are not provided for simulating MC holes (CARRIER=H on
the ALGO statement).

The BULK statement is used to set parameters associated with a bulk simulation (see Section
19.4 “Bulk Simulation”).
BULK MMFRAC=0.05
The MMFRAC parameter sets an effective strain (or effective mole fraction) that pertains to the
bulk. In the case of biaxial tensile strained silicon, this pertains to the mole fraction, x, of an
assumed underlying substrate of relaxed Si1-xGex. For other types of strain, MMFRAC pertains
to an effective strain.

902 ATLAS User’s Manual

Physical Models MC Device

Device Simulations
The strain model was tested in 2D by simulating a 50 nm well-tempered n-MOSFET in
saturation with and without strain. The strained n-MOSFET was based on a relaxed Si1-
xGex=Si0.7Ge0.3 substrate, while the unstrained (relaxed) n-MOSFET was based on a pure

silicon substrate. Although no attempt was made to fit the surface scattering or the threshold
voltage shift, relative comparisons of the results can be used to analyze the qualitative effects
of strain. The strained channel device showed a 60% increase in the drive current relative to
the unstrained case. This result agrees with the low-field mobility enhancement expected at
low drain bias. A large velocity enhancement is seen in the strained n-MOSFET, which
reflects the increased low-field mobility [311].
To use the strain model in MC DEVICE in device simulations, you must use some or all of the
following input statements: MMAT, ESCAT, and MMREGION.
The MMAT and the ESCAT statements are described in Section 19.4 “Bulk Simulation”. The
MMAT and ESCAT statements are used the same in 2D device simulations as they are in bulk
simulations.
The MMREGION statements are used to define the effective strain (or effective mole fraction) in
rectangular regions.
MMREGION BOUNDP=(0.025,0.175,-0.062,0.12) FRAC=(0.3,0.3,0.3,0.3)
The BOUNDP parameter gives the boundary of the region within which the strain model will be
applied (see “Regions” on page 834). A BOUND parameter is provided for backward
compatibility. The FRAC parameter gives a list of four values for the effective strain to be used
at the corners of each rectangular region. These four values pertain to the values in the lower-
left, the lower-right, the upper-right, and the upper-left corners, respectively. (These
directions are based on the directions of the axes in TONYPLOT where x is to the right and
where y is up.) The effective strain inside the rectangle is linearly interpolated based on the
values at the corners.

903 ATLAS User’s Manual

Tips MC Device

19.8 Tips
1. Use the RESTART parameter when performing a series of simulations:
There is a switch for using RESTART option on the algo statement. If you set
RESTART = 1|TRUE|YES, the simulator reads in the particle states from the
status.in file. A corresponding status.out file can be written at the end of a
simulation and you may copy status.out to status.in. The way it helps when
performing a series of simulations is as follows. MC DEVICE initially places particles
with the aim of achieving charge neutrality everywhere. Therefore, for any new
simulation the bulk of the particles are places in the source and the drain region
initially. These particles are then redistributed in the channel region. Therefore, when
the restart option is used, the particles are already in the channel satisfying the
boundary conditions of the previous bias conditions and the time that is needed for
achieving this steady state is reduced or eliminated
2. Try changing the error bound for the inner loop for Poisson solver (see TOL=(2e-4,1e-5)
parameter on the POISSON statement):
The TOL parameter takes two values which correspond to the two tolerances used by
the Poisson solver in MC DEVICE. Since it is an iterative solver, it uses a
preconditioner (inner loop) and then a linear search routine (outer loop) to get the
solution. The two values for the TOL parameter are for the maximum error in the
linear search routine and the tolerance of the preconditioner. Making the tolerance of
preconditioner (the second value) more smaller (more strict) may increase the number
on inner loops but reduce the number of outer loops. Thus, reducing the overall time.
There is a trade off here. Depending on the device, you may want to adjust the solver
for the fastest execution time while you keep the tolerance of the result (tolerance of
linear search routine) the same.
3. Real time between Poisson solutions:
It has been observed that solutions are not as accurate when the real time step between
successive Poisson solutions is more than 0.2 fs. Therefore, it is advised that the real
time step between any two Poisson solutions should be 0.2 fs or less. For example, if
your transport time step (on the ALGO statement) is 0.02 fs, then the Poisson TSTEP
(on the POISSON statement) should not be greater than 10.

904 ATLAS User’s Manual

20.5.4 Combining The Iteration Methods

You can start with the GUMMEL scheme. Then switch to BLOCK or NEWTON if convergence has
not occurred within a certain number of iterations. One circumstance where this can be very
helpful is that Gummel iteration can refine initial guess to a point from which Newton
iteration can converge. The number of initial GUMMEL iterations is limited by GUM.INIT.
You may want to use BLOCK iteration and switch to NEWTON if there is no convergence. This is
the recommended strategy for calculations that include lattice heating or energy balance. The
number of initial BLOCK iterations is limited by NBLOCKIT.
Any combination of the parameters: GUMMEL, BLOCK, and NEWTON can be specified on the
METHOD statement. ATLAS will start with GUMMEL if it is specified. If convergence is not
achieved within the specified number of iterations, it will then switch to BLOCK if BLOCK is
specified. If convergence is still not achieved the program will then switch to NEWTON.

913 ATLAS User’s Manual

Non-Linear Iteration Numerical Techniques

20.5.5 Solving Linear Subproblems

The linear subproblems generated by non-linear iteration can be solved by direct or iterative
methods. Direct methods produce solutions in a predictable number of arithmetic operations.
Solutions are affected by roundoff error but are otherwise exact. Iterative methods obtain
solutions by making a series of corrections to an initial guess. Iteration proceeds until the
calculated corrections satisfy specified convergence criteria. The results are not exact to
within roundoff error, and convergence is not guaranteed for general problems. Iterative
methods, however, can be more efficient than direct methods for large linear systems, and
generally require less memory.
There are mathematical proofs regarding the types of linear system for which iterative
techniques converge. All common iterative schemes converge for linear systems that are
symmetric positive definite (SPD). The linearized form of Poisson’s equation is of this type.
The continuity equations are not SPD. But they can be reliably solved by modern, advanced
iterative methods.
The size and structure of the coefficient matrix of the linear system plays an important role in
the choice of direct or iterative methods. The overall problem size is determined by the
number of variables per node (m) and the number of nodes (n). The number of unknowns is
mn. The linear sub problems associated with Newton iteration have a coefficient matrix with
(mn)2 elements. Each linearized subproblem, which is used in Gummel iteration has a
coefficient matrix with n2 elements.
For practical 2D device simulation problems, the number of elements in the coefficient matrix
is typically between 105 and 108. Fortunately, the matrices are sparse (i.e., most of the entries
are zero, and you do not to store them). The sparsity arises because the variables at each node
are coupled to only a few neighboring nodes. Special direct techniques are available for
solving sparse matrices.
Direct techniques are preferred for relatively small problems. They provide reliable solutions
that are exact to within roundoff error in a time that is predictable. Iterative techniques are
preferred for very large problems because they are faster and require less memory. Direct
techniques for solving sparse matrices have a competitive performance for the problem sizes
typically encountered in 2D device simulation, and are used by ATLAS to solve most of the
linear subproblems that arise.

20.5.6 Convergence Criteria for Non-linear Iterations

After a few non-linear iterations, the errors will generally decrease at a characteristic rate as
the iteration proceeds. Non-linear iteration techniques typically converge at a rate that is
either linear or quadratic. The error decreases linearly when Gummel iteration is used (i.e., it
is reduced by about the same factor at each iteration). For Newton iteration, the convergence
is quadratic. In other words, small errors less than one are approximately squared at each
iteration. The non-linear iteration is terminated when the errors are acceptably small. The
conditions required for termination are called convergence criteria. Much effort has gone into
developing reliable default convergence criteria for ATLAS. The default parameters work
well for nearly all situations, and most will never need to change them.
The main technique in ATLAS is Gaussian Elimination (or LU Decomposition) with a
minimum degree of reordering applied to the nodes [70].

914 ATLAS User’s Manual

Non-Linear Iteration Numerical Techniques

20.5.7 Error Measures

A single positive number that characterizes error is obtained by taking a norm of the errors
associated with each unknown. The quantity that ATLAS tries to reduce to zero is the
difference between the left and right hand sides of the equation. It is natural to use this
quantity as the measure of the error. The associated error norm is called the right hand side
(RHS) norm. The units of the RHS norm are C/m for the Poisson equation, and A/m for the
continuity equations.
Carrier Concentrations and CLIM.DD (CLIMIT)
Another measure of error is provided by the size of the calculated corrections for each
unknown. Since the updates are the unknown “xs” at each step, this is called the X norm.
Potential updates are measured in units of kT/q. Updates to carrier concentrations are
measured relative to the previous value at the point. This relative rror (C) is defined as:
K+1 K
Cm – Cm
 C = max --------------------------------- 20-1
K
max  C 0 C m 
m

where C = n or p, for electrons and holes respectively. m is the node identifier. C0 is a

characteristic concentration. K is the iteration number. C0 is specified as or as CLIM.DD or as
CLIM·c* and

CLIM · DD = CLIMT  c 20-2

where:

c = 4 N c N v 20-3

CLIM.DD (or CLIMIT) is specified in the METHOD statement.

It is difficult to specify a reasonable default value for the CLIM.DD parameter in all situations.
In many difficult cases, the round-off numerical errors do not allow for the resolution of very
low concentrations. The default value of CLIMIT is set at 104 (the corresponding default value
for CLIM.DD in Silicon is 4.5·1013cm-3). In simulation of breakdown, a lower value of
CLIM.DD (~108cm-3 for Silicon diodes) should be specified. Otherwise, a “false” solution may
be obtained.
Discussion of CLIM.EB
To estimate errors in the lattice temperature equation and the energy balance equations,
corresponding RHS norms and X norms are calculated in ATLAS. Updates to temperature are
measured relative to some characteristic value of temperature. The CLIM.EB parameter can be
viewed as a regularization parameter for the case of very small electron or hole densities in
the energy balance equations. The CLIM.EB parameter specifies the minimum value of
concentration for which the relaxation term in the energy balance equation will be properly
resolved. The temperatures for points where the concentration is much less than CLIM.EB are
equal to the lattice temperature. The units of CLIM.EB are cm-3 and the default is 0.0.

915 ATLAS User’s Manual

Non-Linear Iteration Numerical Techniques

20.5.8 Terminal Current Criteria

Another qualification for convergence is derived from the relative changes in terminal
currents and the satisfaction of total current conservation. This qualification can be expressed
as the simultaneous satisfaction of the following conditions:
K+1 K K+1
Ii – Ii  E1 Ii + E2 20-4

and

K+1 K+1
 Ii < 0.01 maxi  I i  E2 

I = 1 nc 20-5
where
• Ii is the current through contact i
• K is the iteration number
• E1 and E2 are specified tolerances
• nc is the number of contacts

20.5.9 Convergence Criteria

A summary of the termination criteria that is enough for most purposes will now be given.
Detailed reference information is provided in the next section.
The non-linear iteration is terminated when one of the following four criteria is satisfied.
1. The X norm for every equation falls below a specified tolerance. The specified tolerances
for X norms are:
x
• P tol for potential equation.
x
• C tol for concentration equations.
x
• TL tol for lattice temperature equation.
x
• TC tol for carrier temperature equation.
2. The RHS for all equations and X norms for energy balance falls below a specified
tolerance. The specific tolerances are:
r
• P tol for potential equations.
r
• C tol for concentration equations.
r
• TL tol for lattice temperature equations.
r
• TC tol for carrier temperature equations.
3. For every equation either the X norm or the RHS norm falls below a specified tolerance.
In this case, both the XNORM and RHSNORM parameters must be specified true.

916 ATLAS User’s Manual

Non-Linear Iteration Numerical Techniques

4. If either 1 or 2 or 3 criterion is fulfilled for weaker values of tolerances. In other words,

for specified tolerances multiplied by the W parameter and current criteria 20-4 and 20-5
are satisfied.
To exclude the X-norm criterion, specify ^XNORM in the METHOD statement. To exclude RHS-
norm criterion, specify ^RHSNORM. To exclude the current criterion, make E1 and E2 very
small. You can change all the above mentioned tolerances simultaneously by specifying the
relaxation factor TOL.RELAX in the METHOD statement.

Table 20-1 User-Specifiable Parameters for Convergence Criteria

Symbol Statement Parameter Default

P tol
x METHOD PX.TOL 10-5

C tol
x METHOD CX.TOL 10-5

TL tol
x METHOD TLX.TOL 10-5 (a)

TL tol
x METHOD TOL.TEMP 10-3 (b)

TC tol
x METHOD TCX.TOL 10-5

P tol
r METHOD PR.TOL 5.010-26

C tol
r METHOD CR.TOL 5.010-18

TL tol
r METHOD TLR.TOL 100

TC tol
r METHOD TCR.TOL 100

E1 METHOD IX.TOL 2.010-5

E2 METHOD IR.TOL 5.010-15
W METHOD WEAK 200
TOL.RELAX METHOD TOL.RELAX 1
XNORM METHOD XNORM TRUE

RHSNORM METHOD RHSNORM TRUE

CLIMIT METHOD CLIMIT 104

CLIM.DD METHOD C0 4.51013
CLIM.EB METHOD CLIM.EB 0

917 ATLAS User’s Manual

Non-Linear Iteration Numerical Techniques

Notes a) This refers to the NEWTON and GUMMEL method.


b) This refers to the BLOCK method.

20.5.10 Detailed Convergence Criteria

Only in very difficult situations is more detailed information concerning error estimation and
the specification of convergence criteria needed. The material is organized by algorithm.
Convergence Criteria For Gummel’s Algorithms
Relative update errors are defined as follows.
For potential:
max  K + 1 –  K
m m
 v = ---------------------------------------------
m  20-6
K+1
max  1  mmax 

K+1 K
where mmax is the node where  m –  m has its maximum value.

For electrons:
max n K + 1 – n K
 n = -------------------------------------------
m m m
- 20-7
K
max  C 0 n m 

For holes:
max p K + 1 – p K
 p = -------------------------------------------
m m m
- 20-8
K
max  C 0 P m 

For lattice temperature:

K+1 K
 T L  nmax –  T L  nmax
 T = --------------------------------------------------
- 20-9
K+1
 T L  nmax
L

K+1
where nmax is the node where  T L  i has its maximum value.

For carrier temperature:

 T = max   T   T  20-10
c n p

where
max K+1 K
m  Tn m –  Tn m
 T = -----------------------------------------------------------
- 20-11
n max  T  K + 1
n m
m

918 ATLAS User’s Manual

Non-Linear Iteration Numerical Techniques

max K+1 K
m  Tp m –  Tp m
 T = -----------------------------------------------------------
- 20-12
p max  T  K + 1
p m
m

For drift diffusion, iterations are terminated if the following criteria are satisfied:
x
 v = P tol 20-13

x
 n =  c tol 20-14

x
 v = C tol 20-15

In non-isothermal drift diffusion, iterations are terminated if Equations 20-10 to 20-12 are
satisfied, the current convergence criteria in Equations 20-4 and 20-5 are met and:
x
 T  TL tol 20-16
L

In Gummel’s method with energy balance equations, NITGUMM iterations, which only the non-
linear Poisson equation is solved, will always be done. Gummel’s method with energy
balance equations is terminated if Equations 20-16 is fulfilled, the carrier temperature
convergence criteria
x
 T  TC tol 20-17
c

is achieved, and one of the following conditions is valid:

• NITGUMM < K+1  NITGUMM+NT1 and Equations 20-13 is fulfilled,
• NT1 + NITGUMM < K+1  NITGUMM + NT3 and
x
 n  P tol  w m 20-18

where w = 10
• NT1 + NITGUMM < K + 1  NITGUMM + NT3, the current convergence criteria (Equations
20-4 and 20-5) are satisfied, and inequality (Equations 20-18) is valid for w=100.
• NT1 + NITGUMM < K+1 and (Equations 20-18) is valid for w=100.
• NT1 + NITGUMM < K+1 the current convergence criteria (Equations 20-4 and 20-5) are
satisfied, and (Equations 20-18) is valid for w = 500.
The default values of the iteration parameters are NITGUMM=5, NT0=4, NT1=10, and NT3=
100.

919 ATLAS User’s Manual

Non-Linear Iteration Numerical Techniques

Convergence Criteria For Newton’s Algorithm

Relative update errors are defined as follows.
For potential
K+1 K
 v = max  m – m 20-19
m

For electron concentration:

max n K + 1 – n K
m m m
 n = ---------------------------------------- 20-20
K
max  C 0 n m 

For hole concentration:

max p K + 1 – p K
m m m
 p = ---------------------------------------- 20-21
K
max  C 0 p m 

For lattice temperature and carrier temperature:

max  T  K + 1 –  T  K
m L m L m
 T = ------------------------------------------------------------- 20-22
L T scale

max  T  K + 1 –  T  K
m n m n m
 T = ------------------------------------------------------------ 20-23
n T scale

max  T  K + 1 –  T  K
m p m p m
 T = ------------------------------------------------------------ 20-24
p T scale

where the scaling temperature, Tscale, is by default equal to 300K.

To define the RHS norms used in ATLAS, first represent the non-linear equations obtained
after discretization at every node as

F    = 0 20-25

where  can be , n, p, TL, Tn, and Tp for the potential equation, electron continuity equation,
hole continuity equation, lattice temperature equation, electron temperature equation and hole
temperature equation, respectively.  represents the vector of unknowns.

920 ATLAS User’s Manual

Non-Linear Iteration Numerical Techniques

The RHS norms in ATLAS is then defined as follows:

For the potential equation:
g
 v = max
m F 20-26

For the electron and continuity equations:

g
 n = max F n  C T 20-27
m

g
 p = max F p  C T 20-28
m

where CT = 10-4 4 N c N v kT

For the lattice temperature and carrier temperature equations:

g max
T = m F TL 20-29
L

g max
T = m F Tn 20-30
n

g max
T = m F Tp 20-31
p

Newton iterations are terminated if one of the following criteria is satisfied.

The XNORM parameter is true and:
x
 v  P tol  w 20-32

x
 n  c tol  w 20-33

x
 p  P tol  w 20-34

x
 T  TL tol  w 20-35
L

x
 T  TC tol  w 20-36
n

x
 T  TL tol  w 20-37
p

921 ATLAS User’s Manual

Non-Linear Iteration Numerical Techniques

where w = 1.
The RHSNORM parameter is true, conditions for Equations 20-36 and 20-37 are satisfied for w
= 1 and:
g r
 v  P tol  w 1 20-38

g r
 n  C tol  w 1 20-39

g r
 p  P tol  w 1 20-40

g r
 T  TL tol  w 1 20-41
L

g r
 T  TC tol  w 1 20-42
n

g r
 T  TL tol  w 1 20-43
p

Both the XNORM and RHSNORM parameters are true. The convergence criteria for Equations 20-
36 and 20-37 for carrier temperature are fulfilled for the inequalities (Equations 20-32 and
20-38), (Equations 20-33 and 20-39), (Equations 20-34 and 20-40), (Equations 20-35 and 20-
41), and one of the conditions for every pair.
If the current convergence criteria are satisfied, then condition a) or condition b) or condition
c) is fulfilled for w = w1 = WEAK .
Convergence Criteria For Block Iteration
For the potential equation and the continuity equations, Equations 20-19 to 20-21 and
Equations 20-26 to 20-28 define the X norms and RHS norms for the Newton method. With s
as an index that denotes the number of block iterations, update errors between successive
pairs of block iterations for lattice and carrier temperature are defined by the following
expressions.
For lattice temperature:
s+1 s
B  T L  mmax –  T L  mmax
 T = -------------------------------------------------------
- 20-44
s+1
 T L  mmax
L

where mmax is the number of the node where (TL)m has its maximum value.
For electron temperature:
s+1 s
B  T n  mmax –  T n  mmax
T = --------------------------------------------------------------- 20-45
s+1
 T n  mmax
n

922 ATLAS User’s Manual

Non-Linear Iteration Numerical Techniques

For hole temperature:

s+1 s
B  T p  mmax –  T p  mmax
 T = --------------------------------------------------------------- 20-46
s+1
 T p  mmax
p

Block iterations are terminated if:

B x
 T  TL tol 20-47
L

B x
 T  TC tol 20-48
n

B x
 T  TL tol 20-49
p

and one of the following criteria is fulfilled:

• the XNORM parameter is true and Equations 20-32, 20-33, and 20-34 are valid for w = 1.
• the RHSNORM parameter is true and Equations 20-38, 20-39, and 20-40 are valid for w1 =
1.
• both XNORM and RHSNORM are true and for the pairs of inequalities (Equations 20-32 and
20-38), (Equations 20-33 and 20-39), (Equations 20-34 and 20-40) one of the conditions
is satisfied for each pair.
• the current convergence criteria are satisfied and condition 1, condition 2, or condition 3
is fulfilled for w = w1 = WEAK.

923 ATLAS User’s Manual

Initial Guess Strategies Numerical Techniques

20.6 Initial Guess Strategies

Non-linear iteration starts from an initial guess. The quality of the initial guess (i.e., how
close it is to the final solution) affects how quickly the solution is obtained and whether
convergence is achieved. ATLAS users aren’t required to specify an initial guess strategy. If
no strategy is defined, ATLAS follows certain rules that implement a sensible, although not
necessarily optimum strategy.
There is some interaction between the choice of non-linear iteration scheme and the initial
guess strategy. Decoupled iteration usually converges linearly, although perhaps slowly, even
from a relatively poor initial guess. Newton iteration converges much faster for a good initial
guess, but fails to converge if started from a poor initial guess.
One very simple initial guess strategy is to use the most recent solution as the initial guess. Of
course, there is no previous solution for the first calculation in a series of bias points. In this
case, an initial solution is obtained for equilibrium conditions. There is no need to solve the
current continuity equations at equilibrium, and a solution of Poisson’s equation is quickly
obtained.
You can also modify the initial guess in a way that makes some allowance for the new bias
conditions. Typical strategies include:
• Using two previous solutions and interpolation to project a new solution at each mesh
point.
• Solving a form of current continuity equation with carrier concentrations held constant.
This strategy yields an improved estimate of new potential distribution.
• Modifying the majority carrier quasi-Fermi levels by the same amount as the bias
changes.
You can use the parameters on the SOLVE statement to specify an initial guess strategy. Six
initial guess strategies are available.
• INITIAL starts from space charge neutrality throughout the device. This choice is
normally used to calculate a solution with zero applied bias.
• PREVIOUS uses the currently loaded solution as the initial guess at the next bias point. The
solution is modified by setting a different applied bias at the contacts.
• PROJECTION takes two previous solutions whose bias conditions differ at one contact and
extrapolates a solution for a new applied bias at that contact. This method is often used
when performing a voltage ramp.
• LOCAL sets the applied bias to the specified values, and changes the majority carrier quasi-
Fermi levels in heavily doped regions to be equal to the bias applied to that region. This
choice is effective with Gummel iteration, particularly in reverse bias. It is less effective
with Newton iteration.
• MLOCAL starts from the currently loaded solution and solves a form of the total current
continuity equation that provides an improved estimate of the new potential distribution.
All other quantities remain unchanged. MLOCAL is more effective than LOCAL because it
provides a smooth potential distribution in the vicinity of p-n junctions. It is usually more
effective than PREVIOUS because MLOCAL provides a better estimate of potential. This is
especially true for highly doped contact regions and resistor-like structures.
• NOCURRENT assumes there is negligible current flow in the device and solves the non-
linear Poisson equation to arrive at the initial guess. If the assumption of negligible

924 ATLAS User’s Manual

Initial Guess Strategies Numerical Techniques

current density is indeed correct, then the initial guess should be the solution to the
current continuity equations too. You can use NOCURRENT to avoid ramping a bias to
attain the solution at high bias.
When a regrid is performed, the solution is interpolated from the original grid onto a finer
grid. This provides an initial guess that can be used to start the solution of the same bias point
on the new grid. Although the initial guess is an interpolation of an exact solution, this type of
guess does not provide particularly fast convergence.

20.6.1 Recommendations And Defaults

It is not normally required to specify any initial guess parameter.
The INITIAL parameter is normally used only to obtain a thermal equilibrium solution. The
PREVIOUS, PROJECT, LOCAL, and MLOCAL parameters are used for other bias points.
PREVIOUS and PROJECTION are the most frequently used. PROJECTION is normally preferred
to PREVIOUS when it is available (i.e., when there are two previous solutions differing in the
bias applied to the appropriate terminal). PREVIOUS is required for transient simulations, and
for simulations that use current boundary conditions. LOCAL and MLOCAL tend to work well
for reverse-biased devices, and are especially efficient when trying to increase very large
voltage increments. By default, PROJECTION is used whenever two appropriate solutions are
available. Otherwise, the PREVIOUS guess is used, unless there is no previous solution, in
which case INITIAL is used.

925 ATLAS User’s Manual

Globally Convergent Schemes Numerical Techniques

20.7 Globally Convergent Schemes

ATLAS has a vector of independent variables, x, and a vector of functions of these
independent variables, F(x). A solution xs is sought such that F(xs)=0. A “residual” function,
G(F(x)), is defined. This residual is a single number that gives some measure of how close
F(x) is to the solution. The residual is zero at the solution, and positive everywhere else. The
smaller the residual, the “closer” x is to the solution xs.
The iterative techniques start with the Taylor expansion of F(x) around some vector xi (which
is an initial guess to xs or some previous attempt to calculate xs).
2 2
dF  x i  d F  x i  x
F  x i + x i  = F  x i  + ----------------  x i + -------------------  --------i +  20-50
dx 2 2
dx

2
The x i and above terms are ignored. The new x is made the solution, xi+xi = xs. Therefore,
F(xi + xi) = 0 so
dF  x i  dF  x i 
0 = F  x i  + ----------------  x i  ----------------  x i = – F  x i   J  x i   x i = – F  x i  20-51
dx dx

where J(xi) is the Jacobian. This is the equation solved during the iterative schemes to find
the next value of xi.
Because the Taylor expansion was truncated after the first two terms, the function F(xi + xi)
is modeled as linear in x. This is a good approximation close to xi but will probably become
increasingly bad the further away from xi we are. Specifically, at a distance xi away from xi
the linear approximation may be invalid and therefore
dF  x i 
F  x i  + ----------------  x i  F  x i + x i  20-52
dx

This is why ATLAS often has to do more than one iteration to find the vector xs.
Sometimes, the residual G(F(xi + xi)) is bigger than the residual G(F(xi)), which means we
are jumping to a point that is further away from the solution than the current point. The
globally convergent schemes try to ensure that every step taken moves us closer to the
solution.

926 ATLAS User’s Manual

Globally Convergent Schemes Numerical Techniques

20.7.1 Overview
When
J  x i   x i = – F  x i  20-53

is solved, we get a step xi. A variable k and a function g(k) are introduced so that
g  k  = G  F  x i + k  x i   20-54

thus g(0) is the residual at the current point xi, and g(1) is the residual at the full Newton step
xi+xi. The function g(k) is expected to be a continuous function over the range [0,1]. The
aim of the globally convergent schemes is to find a ki so that g(ki) is sufficiently smaller than
g(0) and then set the next step as
x i + 1 = x i + k i  x i 20-55

Globally convergent scheme usually have two parts. A way to decide if a particular g(k) is
sufficiently small and an algorithm to calculate a new value for k if the current g(k) is not
sufficiently small.
A particular k is accepted if
gk  g0 – k  D 20-56

and the purpose of the second term is often to reject steps where the reduction in the residual
is insignificantly small, as such D is often not very big. However, the residual is expected to
decrease therefore we need D to be positive.
We cannot guarantee there is an acceptable minimum along the Newton step. If there is not,
then the algorithm to calculate k would continue for ever. In order to avoid such situations, a
maximum number of inner iterations needs to be specified.
There is often some numerical overhead when using a damping scheme (the function F(x)
may need to be calculated several times). It is not always necessary to start damping on the
first Newton iteration. The number of Newton iterations to skip before we start damping can
also be specified.

927 ATLAS User’s Manual

Globally Convergent Schemes Numerical Techniques

20.7.2 Line Search Damping

In line search damping, we want
g  k i   g  0  + k i    g  0  20-57

In this case
D = –   g  0  20-58

The gradient of g(k) at k=0 is negative: if an infinitesimal step along the direction x does not
reduce the residual, then the Newton iteration solution has failed. Therefore,  should be
positive, but  can be small, 10-4 by default.
In this algorithm, g(0), g'(0), and g(1) are initially calculated. If g(1) is sufficiently small,
then the full Newton step is accepted. If g(1) is not sufficiently small, then a quadratic is
fitted through the two points and the initial gradient
2
f2  y  = a + b  y + c  y  f2   y  = b + 2  c  y 20-59

with
f 2  0  = g  0  f 2   0  = g  0  f 2  1  = g  1  20-60

The minimum in f2(y) is found and the corresponding y is chosen as the first guess, k1. If g(k1)
is sufficiently small, then this value of k is accepted. If g(k1) is not sufficiently small, there
are now three points and an initial gradient. The two smallest non-zero k are denoted kA and
kB. Initially, kA=k1 and kB=1. Thus, we fit a cubic through the three points and one gradient
2 3 2
f3  y  = a + b  y + c  y + d  y  f3   y  = b + 2  c  y + 3  d  y 20-61

with
f 3  0  = g  0  f 3   0  = g  0  f 3  k A  = g  k A  f 3  k B  = g  k B  20-62

The minimum in f3(y) is found and the corresponding y is chosen as the next guess, kk. If g(kk)
is sufficiently small, then this value of k is accepted. If g(kk) is not sufficiently small, then kA
and kB are redefined to the two smallest non-zero k that have been calculated and the cubic is
recalculated.
Line search damping is activated with the LS.DAMPING parameter on the METHOD statement
and deactivated with the LSD.OFF parameter. The coefficients are specified with the
following in Table 20-2.

928 ATLAS User’s Manual

Globally Convergent Schemes Numerical Techniques

Table 20-2 User-Specifiable Parameters for Line Search Damping

Symbol Statement Parameter Default

METHOD LSD.SKIP.ITER 5
METHOD LSD.MAX.ITER 5
 METHOD LSD.ALPHA 10-4

LSD.SKIP.ITER is the number of Newton iterations that occur before line search damping is
used. LSD.MAX.ITER is the maximum number of line search inner iterations, during a single
Newton iteration, to find a suitable value for k.

20.7.3 Bank-Rose Damping

In Bank-Rose damping, we want g(ki) to be smaller than
g  ki   g  0  – k i    g  0  20-63

in this case
D =   g0 20-64

The residual has to be positive; therefore,  should be positive as well. If  were bigger than
unity then (for some k at least) we would only accept negative residuals (which are
impossible). The default value for  is 10-4.
In this algorithm, g(0) has been initially calculated. The trial k is
1
k = ----------------------------- 20-65
1 + K  g0

If g(k) is sufficiently small, then this value of k is accepted. If g(k) is not sufficiently small,
then K is increased which will reduce k. The initial value of K is zero, which is increased to
unity after the first failed iteration. If K is greater than zero, it is increased by a factor of ten
(KK × 10).
If a value of k is accepted, the corresponding value of K is decreased by a factor of ten
(KK /10) and this value is used as the initial K to damp the subsequent Newton iteration.
Bank-Rose damping is activated with the BR.DAMPING parameter on the METHOD statement
and deactivated with the BRD.OFF parameter. The coefficients are specified with the
following in Table 20-3.

929 ATLAS User’s Manual

Globally Convergent Schemes Numerical Techniques

Table 20-3 User-Specifiable Parameters for Bank-Rose Damping

Symbol Statement Parameter Default

METHOD BRD.SKIP.ITER 5
METHOD BRD.MAX.ITER 5
K METHOD BRD.KAPPA 0.0
 METHOD BRD.DELTA 10-4

BRD.SKIP.ITER is the number of Newton iterations that occur before Bank-Rose damping is
used. BRD.MAX.ITER is the maximum number of Bank-Rose inner iterations, during a single
Newton iteration, to find a suitable value for k. BRD.KAPPA sets the initial value of K for use
in the first Bank-Rose damping iteration.

20.7.4 Explicit Minimum Damping

In explicit minimum damping, we look for the minimum in the g(k) curve. This is
computationally inefficient and should be used only as a last resort.
In this algorithm, g(0), g(0.5), and g(1) are initially calculated. The minimum is “bracketed”
by calculating the residual at the mid-points and keeping the three adjacent values that
contain the minimum residual (if possible the minimum is the middle point of the three).
This algorithm has no check on how good the residual is, so we bracket the minimum a fixed
number of times and return the best k we have found.
Explicit minimum damping is activated with the EM.DAMPING parameter on the METHOD
statement and deactivated with the EMD.OFF parameter. The coefficients are specified with
the following in Table 20-4.

Table 20-4 User-Specifiable Parameters for Explicit Minimum Damping

Symbol Statement Parameter Default

METHOD EMD.SKIP.ITER 5
METHOD EMD.MAX.ITER 5

EMD.SKIP.ITER is the number of Newton iterations that occur before explicit minimum
damping is used. EMD.MAX.ITER is the number of explicit minimum inner iterations, during
a single Newton iteration, to find a suitable value for k. Because explicit minimum is looking
for the actual minimum residual, it always does this for many inner iterations.

930 ATLAS User’s Manual

The DC Curve-Tracer Algorithm Numerical Techniques

20.8 The DC Curve-Tracer Algorithm

Tracing I-V curves for complicated device phenomena, such as breakdown or latchup, can be
very difficult using conventional methods. ATLAS includes a special purpose DC curve-
tracing algorithm that overcomes these problems. This algorithm is based on a dynamic load-
line technique that adapts the boundary conditions for each step. The approach implemented
in ATLAS for curve tracing is based on the work describe in [88].
The key idea is that bias conditions can evolve smoothly between the limits of pure voltage
control and pure current control. This is achieved using external resistors that adapt
dynamically to the shape of the I-V curves to ensure that at each point the load line is
perpendicular to the local tangent of the trace. With this value for the external resistor, the
solution is projected to the next operating point by stepping the external voltage. Once the
solution has converged, a new external resistance is calculated based on the new tangent
information and the process repeats itself.

Figure 20-1: Load the algorithm used in the Curve Tracer

The curve tracing capability is activated by specifying the CURVETRACE parameter in the
SOLVE statement. Prior to this, the CURVETRACE statement is used to set the parameters for the
tracing algorithm. These parameters are the name CONTR.NAME of the ramped electrode
(which will be referred to as a control electrode), the initial voltage increment STEP.INIT, the
upper bound of the tracing curve, and additional parameters if they differ from the default
values. The Upper bound parameter, END.VAL, is used to stop tracing. If the VOLT.CONT
parameter is specified, END.VAL is a voltage. If the CURR.CONT parameter is specified,
END.VAL is a current.

931 ATLAS User’s Manual

The DC Curve-Tracer Algorithm Numerical Techniques

The applied voltage at each step is altered in accordance with the slope of the I-V curve. The
resistor between the applied voltage and the semiconductor is also changed dynamically to
ensure the voltage at the semiconductor (VINT) is smoothly varied along the I-V curve. If
STEP.CONT is specified, the number of operational points on a trace will not exceed specified
parameter STEPS.

932 ATLAS User’s Manual

Transient Simulation Numerical Techniques

20.9 Transient Simulation

When transient simulation is performed, the carrier continuity equations are integrated in the
time domain. Time integration schemes differ in their accuracy, in the number of previous
time levels they employ, and in their stability properties.
Accuracy is usually referred to as being “nth order”, where n is usually an integer between 1
and 4. In the limit of a small timestep, the magnitude of the local truncation error (LTE)
introduced by the time integration scheme is proportional to the nth power of the timestep.
Schemes that require the storage of solutions at timesteps previous to the most recent one are
unattractive due to storage requirements. Single step integration schemes that use the solution
at only one previous time level have a maximum order of 2.
The continuity equations are “stiff” (i.e., they are impacted by phenomena with very short
timescales). Stiffness imposes stringent requirements on the stability of time integration
schemes. Two forms of stability, A-stability and L-stability, are important. A-stability
guarantees that errors introduced at one time step will not increase at the next timestep. L-
stability guarantees that errors will decay even for large time step values. A-stability is a
requirement for any practical scheme. L-stability is extremely desirable to avoid non-physical
“ringing”.
Most device simulation codes use a simple first-order (implicit) backward difference formula
for time integration [83,63,153]. This scheme, which is known as BDF1, is both A-stable and
L-stable. Unfortunately, the scheme is inaccurate for typical timesteps of interest. Second
order accuracy is obtained using the trapezoidal rule (TR) for time integration. This scheme is
A-stable, but it is not L-stable. This means that solutions exhibit non-physical “ringing”,
unless very small timesteps (much smaller than those dictated by LTE considerations) are
used. The BDF2 scheme is second order, and is both A-stable and L-stable. The scheme,
however, uses solutions from two previous time levels and is less accurate than TR.
For drift-diffusion calculations, ATLAS uses a composite TR-BDF2 scheme that was
developed by Bank et. al.[19]. This method is one-step, second order, and both A-stable and
L-stable. An estimate of the LTE is obtained at each timestep. This estimate is also used to
automatically adapt the timestep.
Different schemes are available for transient solutions that include solving for lattice
temperature or carrier temperatures or both. In ATLAS2D, the full TR-BDF2 is available as
either full Newton or a block iteration scheme. To enable the block iteration scheme, use the
BLOCK.TRAN keyword on the METHOD statement. In ATLAS3D, the TR-BDF2 is available as a
full Newton scheme but the block iteration is unavailable. In ATLAS3D, the absolutely stable
half-implicit scheme [219] is available (specify HALFIMPLICIT on the METHOD statement). It
is only available if a lattice temperature solution is requested. It is not for carrier temperature
solutions. Automatic timestep selection with local error control is implemented in this case.
You can specify tolerance using the TOL.TIME parameter in the METHOD statement.

Note: You normally specify only the initial timestep with the TSTEP parameter of the SOLVE statement. After this,
time steps are derived from the LTE and will typically increase.

933 ATLAS User’s Manual

Small Signal and Large Signal Analysis Numerical Techniques

20.10 Small Signal and Large Signal Analysis

There are several ways to predict the small-signal and large-signal high-frequency properties
of semiconductor devices (review of these different techniques was presented by Laux [156]).
Frequency domain perturbation analysis can be used to determine the small-signal
characteristics. Fourier analysis of transient responses, however, can be used for both small-
signal and large signal response.

20.10.1 Frequency Domain Perturbation Analysis

Frequency-domain perturbation analysis of a DC solution can be used to calculate small-
signal characteristics at any user-specified frequency [156]. This scheme works for both 2D
and 3D. The calculation proceeds in the following manner:
1. Variables are represented as the sum of the known DC component and a small unknown
sinusoidal AC component.
2. All equations are expanded.
3. Differentiation in time becomes multiplication by the value of = 2 frequency).
4. Products of AC quantities are neglected since they are small with respect to other
quantities
5. The DC solution is subtracted.
What remains is a complex linear system whose unknowns are the AC components of the
solution. Solving this linear system with appropriate boundary conditions yields small-signal
characteristics.
The coefficient matrix of the complex linear system is simply the Jacobian associated with the
DC operating point with some terms on the leading diagonal supplemented by j. The
Jacobian is available ‘for free’ if the DC solution was calculated using Newton iteration. This
is a very attractive feature of the Newton method. The resulting linear complex system can be
solved by an iterative or a direct method. A form of successive over relaxation (SOR) works
well for frequencies significantly below the cutoff frequency but fails at higher frequencies.
More sophisticated iterative techniques can be used at higher frequencies. It is simpler and
more reliable, however, to switch to direct sparse matrix methods if SOR fails to converge.
Frequency domain perturbation analysis is extremely attractive when the full Newton method
is used to calculate a DC solution. The method works for all frequencies. It requires a
predictable amount of computation, which is quite low with respect to other calculations,
since only a single linear system is solved.
Frequency domain perturbation analysis is invoked in ATLAS by specifying the appropriate
parameters in the SOLVE statement. The AC.ANALYSIS parameter specifies that this analysis
is to be performed. TERMINAL specifies the contact whose terminal voltage is to be perturbed.
A full characterization requires that all but one of the device terminals is perturbed.
FREQUENCY, FSTEP, and NFSTEPS determine the frequencies for which solutions are obtained.
FREQUENCY specifies the initial frequency. FSTEP and NFSTEPS define a loop on frequencies.
If MULT.FREQ is specified, the frequency is multiplied by FSTEP at each increment. This is
useful for characterizing the small-signal response over several decades of frequency. The
solution method can be specified using the SOR, DIRECT, and AUTO parameters. AUTO starts
out using SOR but switches to DIRECT if convergence problems are encountered.

934 ATLAS User’s Manual

Small Signal and Large Signal Analysis Numerical Techniques

You can specify the relaxation parameter by using the S.OMEGA parameter on the SOLVE
statement. Its default value is unity, which usually suffices. If any of the contacts in the
device has a lumped element, however, then the SOR method will probably fail to converge
unless S.OMEGA is set to a value of less than 1. A value of 0.75 is recommended on a first
attempt. For higher frequencies, you need to use the DIRECT method.
To work properly small signal analysis requires a complete Jacobian. Some models may be
non-local, which means they may produce questionable small signal results.
The INCOMPLETE model although “local” lacks rate expressions for carriers. This can also
lead to questionable results, such as currents that do not balance.

20.10.2 Fourier Analysis Of Transient Responses

Frequency-domain perturbation analysis is a post-processing step that must be performed on
a LOG file, which contains transient data. The FOURIER statement performs a Fast Fourier
Transform (FFT) on the time domain data transforming it into the frequency domain. Table
20-5 shows the syntax of the FOURIER statement. For more information about the FOURIER
statement, see Section 21.20 “FOURIER”.

Table 20-5 User-Specifiable Parameters for Fourier Statement

Statement Parameter Default

FOURIER INFILE

FOURIER OUTFILE

FOURIER T.START

FOURIER T.STOP

FOURIER FUNDAMENTAL

FOURIER MAX.HARMONIC

FOURIER NUM.SAMPLES 64
FOURIER INTERPOLATE FALSE
FOURIER COMPLEX.VALUES FALSE

The explanation for the FOURIER parameters are as follows:

• INFILE – input log file. This should contain data from a transient simulation.
• OUTFILE – file output file for the Fourier transform.
The following parameters are optional:
• T.START – start of time data to be used for the FFT. The default value is the first time
point in the input log file.
• T.STOP – end of time data to be used for the FFT. The default value is the last time point
in the input log file.
• FUNDAMENTAL – fundamental frequency. If this is not specified, then the fundamental
frequency is set to 1/(T.STOP - T.START). If the fundamental frequency is specified then
T.STOP is set to T.START + 1/FUNDAMENTAL.

935 ATLAS User’s Manual

Small Signal and Large Signal Analysis Numerical Techniques

• MAX.HARMONIC – maximum harmonic frequency that the FFT should calculate. This will
automatically calculate the correct number of samples (NUM.SAMPLES) required to
generate this frequency. FUNDAMENTAL must be specified when MAX.HARMONIC is used.
• NUM.SAMPLES – number of samples. This should be an integer power of 2 (i.e., 2^n)
where n is a positive integer. The default value is 64 unless the MAX.HARMONIC parameter
is specified. In this case, the number of samples is set to the nearest integer power of 2,
which will generate this frequency.
• INTERPOLATE – performs linear interpolation on input data with non-uniform timesteps.
This interpolates the data on to uniform timesteps. Interpolation of data can introduce
addition (inaccurate) harmonic values into the FFT, which would not occur if uniform
time is taken. INTERPOLATE must be used if the log file contains non-uniform time steps.
• COMPLEX.VALUES – prints the real and imaginary components to file as well as the
magnitude and phase.

Note: FFT works best with uniform time steps. Therefore, set DT.MIN and DT.MAX on the METHOD statement
to the same value. The time step should be set to: time step = 1/(number of samples * fundamental
frequency).

The FFT can then calculate harmonic frequencies up to:

(number of samples/2 - 1)*fundamental frequency.
You can obtain small-signal data at high frequencies by calculating the terminal current
responses to terminal voltage perturbations. Then, Fourier analysis will be performed on
currents and voltages. Their ratio at each frequency provides admittance data for that
frequency. The voltage perturbations are normally selected to have an analytic form with a
known Fourier transform. Care must be taken to self-consistently account for geometric
capacitances when step function voltage perturbations are used. The advantages of this
technique are that it can be used whenever transient calculations are possible. Each transient
solution gives information over a broad range of frequencies. The main disadvantage is that
transients become very long when low frequency effects are investigated.

20.10.3 Overall Recommendations

Use frequency-domain perturbation analysis when it is available. The method works for
arbitrary frequencies and does not require transient calculations. Use Fourier analysis of
transient responses for high frequencies when it is unavailable but transient calculations are
possible.

936 ATLAS User’s Manual

Numerical Precision Numerical Techniques

20.11 Numerical Precision

The convergence and accuracy of a device simulation requires the numerical resolution of the
quantities  n – p  and  n + p  . The first of these quantities appears in the space charge
density, which is used in Equations 3-1 to calculate the electrostatic potential, while the
second is used implicitly in the calculation of current conservation.
In order to resolve these terms, the calculations must maintain at least
ln  n  – ln  p 
P  ----------------------------------- 20-66
ln  2 

significant bits of arithmetic precision. From the theory of carrier statistics we can then
estimate that for a convergent and accurate simulation, we should have
EG
P  -------------  --------- .
1
20-67
ln  2   kT L

Therefore, the required precision P depends on both the lattice temperature and on the
bandgap. In practice, Inequality 20-67 constitutes a loose upper bound, and we can frequently
carry out a successful simulation with somewhat less precision than the relation indicates.
This inequality also tells us that precision is more likely to be an issue with high-bandgap
materials, and at low temperatures.
By default, ATLAS uses a nominal arithmetic precision of 64 bits (i.e., double precision). In
order to support the simulation of high-bandgap materials in particular, ATLAS can now run
at several levels of extended precision as well. The level of precision is selected by means of
a command-line flag (which in DECKBUILD would be passed as one of the arguments to the
SIMFLAGS parameter of the GO statement). The arithmetic precision levels currently supported
by ATLAS are summarized in Table 20-6.

Table 20-6 Arithmetic Precision Levels in ATLAS

Command-line flag (None) -80 -128 -160 -256

Nominal precision Pnom (bits) 64 80 128 160 256

Significant bits 53 64 104 128 209

Type size (bytes) 8 16 16 32 32

Floating-point numbers are composed of three parts: the sign, the exponent, and the
significand [188]. As its name suggests, the size of the significand corresponds to the number
of significant bits in the number. When selecting a precision level for a simulation, you
should check Table 20-6 to verify that the number of significant bits is at least as large as the
required precision P given by Inequality 20-67.
The nominal precision of a floating point number is larger than the number of significant bits,
because it also counts the bits used for the sign and the exponent. In ATLAS, the various flags
controlling the precision levels of the simulator and of the solver always refer to the nominal
precision.
For well-converged solutions, the run-time increases with the precision. The increase is
especially significant at the highest precision levels. Memory requirements also increase with

937 ATLAS User’s Manual

Numerical Precision Numerical Techniques

the precision; relative requirements are suggested by the type sizes listed in Table 20-6. On
the other hand, certain simulations (e.g., of SiC devices) that have difficulty converging at the
lower precision levels are likely to run faster if the precision level is increased. Accordingly,
the optimum precision level for a particular problem is likely to be the minimum level that
yields good convergence during the solution. Based on Table 20-6 and Inequality 20-67,
Table 20-7 provides a guide showing the nominal precision that should be required to
simulate common semiconducting materials at T L = 300 K .

Table 20-7 Required Precision for Common Semiconductors

Material EG (eV) Pnom (bits) Material EG (eV) Pnom (bits)

InSb 0.17 64 ScN 2.15 160

SnTe 0.18 64 AlAs 2.16 160
PbSe 0.26 64 3c-SiC 2.2 160
PbTe 0.29 64 GaP 2.27 160
InAs 0.36 64 ZnTe 2.28 160
PbS 0.37 64 (160-bit Limit) 2.294 160
Ge 0.66 64 AlP 2.43 256
GaSb 0.75 64 CdS 2.48 256
(64-bit Limit) 0.950 64 HgS 2.5 256
Si 1.08 80 BeTe 2.57 256
(80-bit Limit) 1.147 80 ZnSe 2.71 256
InP 1.35 128 6h-SiC 2.9 256
GaAs 1.424 128 4h-SiC 3.26 256
CdTe 1.43 128 ZnO 3.3 256
AlSb 1.63 128 GaN 3.4 256
CdSe 1.75 128 (256-bit Limit) 3.745 256
(128-bit Limit) 1.864 128 C (Diamond) 5.45 (Excessive)
AlN 6.2 (Excessive)

Table 20-7 lists 80 bits as the nominal precision required to simulate silicon devices at 300 K.
From long experience, we know that in most cases a precision of 64 bits is sufficient.
However, the full 80 bits may be required for devices — such as bipolar junction transistors
and TFT’s — where the minority carrier current plays an important functional role.

938 ATLAS User’s Manual

Differences Between 2D and 3D Numerics Numerical Techniques

20.12 Differences Between 2D and 3D Numerics

With respect to numerical techniques, there are several differences between 2D and 3D
simulations.
First, with respect to the nonlinear iteration strategies, all three strategies, NEWTON, GUMMEL
and BLOCK are supported in 2D simulation, whereas, only NEWTON and GUMMEL are supported
for 3D simulations. Implementation of the BLOCK strategy is expected in a future release.
Second, solution of the linear subproblem is handled differently for 2D and 3D simulations.
As previously noted, the computational burden of solving the linear subproblem increases
with the size of the solution domain. For smaller problems, direct methods are quicker while
for larger problems iterative methods are preferred. It turns out that the point at which the
iterative methods become less burdensome roughly coincides with the transition between 2D
and 3D domains. As such, the default method for 2D simulations is a direct solver. For 3D
simulations, the default method is an iterative solver. By default, ILUCGS is applied to 3D
simulations. ILUCGS is an acronym for incomplete lower upper decomposition conjugate
gradient squared. Two alternative iterative solvers are also available for 3D simulations.
BICGST (BICGST on the METHOD statement) is an acronym for biconjugate gradient squared
stabilized. GMRES (GMRES on the METHOD statement) is an acronym for generalized minimum
residual. Direct methods can be used for 3D simulation by specifying DIRECT on the METHOD
statement. Practical experience shows that for 3D simulations either of the iterative methods
are faster than the direct method. But in some cases, the accuracy produced by the iterative
methods can prevent convergence in the nonlinear outer loop. For 2D simulations, only direct
methods are supported.

939 ATLAS User’s Manual

Chapter 21
Statements
Input Language Statements

21.1 Input Language

This chapter contains a complete description (in alphabetical order) of every statement and
parameter used by any of the ATLAS products, except for MIXEDMODE. MIXEDMODE
specific parameters are described in Chapter 12 “MixedMode: Mixed Circuit and Device
Simulator”. MC DEVICE specific parameters are described in Chapter 19 “MC Device”.
Descriptions for the statements in this chapter are as follows:
• The statement name.
• The product for which the statement is applicable.
• The syntax of the statement.
• A list of all statement parameters, their type, default value, and units.
• A description of each parameter.
• An example of the correct usage of each statement.

Note: An error message will be generated if you attempt to specify a statement for a simulator that you haven’t
purchased. For example, the BEAM statement can only be used if you have purchased LUMINOUS or
LUMINOUS 3D.

21.1.1 Syntax Rules

An input deck line is referred to as a statement (or statement line). Since statements and
parameters are not case sensitive, they can be entered using either uppercase or lowercase
letters.
A statement is specified in the general format:
<STATEMENT> <PARAMETER>=<VALUE>
where STATEMENT is the statement name, PARAMETER is the parameter name, and VALUE is
the parameter value. The space character is used to separate the parameters in a statement.
The words and numbers, which follow a statement, are parameters of that statement. A word
is an alphanumeric string, which is terminated either by a space or by a carriage return. A
number is a numeric or alphanumeric string which is terminated either by a space or by a
carriage return. Numerical values may range from 10-38 to 1038. A number may contain the
symbols + (positive), - (negative), and/or E (decimal notation). For example:
+10 -1.234E5 .003 -.12E+10
Four types of parameters are used by the ATLAS products. These are: real, integer, logical,
and character. Table 21-1 explains these parameter types.

941 ATLAS User’s Manual

Input Language Statements

Table 21-1 Types of Parameters

Value
Parameter Description Example
Required

Character Any character string Yes INFILE=NMOS.DOP

Integer Any whole number Yes REGION=2

Logical A true or false condition No GAUSSIAN

Real Any real number Yes X.MIN=0.52

Any parameter that doesn’t have a logical value must be specified in the form: PARAM=VAL,
where PARAM is the name of the parameter, and VAL is the value of the parameter. Logical
parameters must be separated from other parameters or commands by a space.
For example, in the statement:
DOPING UNIFORM CONCENTRATION=1E16 P.TYPE
the UNIFORM and P.TYPE parameters have logical values and the CONCENTRATION parameter
has a value of 1x1016 (real).
Logical parameters can be turned off (switched from true to false) by placing a caret (^) in
front of the logical parameter. For example, in the statement:
DOPING UNIFORM CONCENTRATION=1E16 ^P.TYPE
the P.TYPE parameter has been set to false.
Mnemonics
It is not always necessary to input the entire statement or parameter name. ATLAS only
requires that you input enough letters to distinguish that command or parameter from other
commands or parameters. For example, DOP may be used to abbreviate the DOPING command.
Excessive truncation is not recommended, since future ATLAS syntax might make short
abbreviations become ambiguous.
Continuation Lines
Since it may be necessary for a statement line to contain more than 256 characters, ATLAS
allows you to specify continuation lines. To continue a line, put a backslash (\) character at
the end of the line that is to be continued. When ATLAS encounters the backslash, it will
interpret the next line to be a continuation of the current line. The PISCES-II continuation of
using a (+) at the start of the subsequent line is not supported in ATLAS.
Comments
Comments are indicated either by COMMENT command, or by a pound sign (#). All characters
on a line, which follow a comment indicator (COMMENT or #) will not be analyzed by ATLAS.
Synonyms
Some parameters have synonyms. These are parameters that have a different name but the
same functionality. A parameter’s synonym is listed in the parameter descriptions of the
statements.

942 ATLAS User’s Manual

Input Language Statements

Pseudonyms
Throughout the statement descriptions, pseudonyms are used either to indicate a group of
parameters or to indicate the value of a particular parameter. A < symbol indicates the start of
a pseudonym(s). A > symbol indicates the end of a pseudonym(s). Pseudonyms will be
separated from one another by a space character ( ). For example, LOC might indicate a
group of location parameters, and FILENAME might indicate the name of a file that must be
specified.
Symbols
The following symbols are used in the statement descriptions:
• < Indicates the start of a list of pseudonyms.
• > Indicates the end of a list of pseudonyms.
• | Separates parameters or pseudonyms which are mutually exclusive. Only one of these
parameters may be used in a statement.
• [ Indicates the start of an optional command, parameter, or pseudonym.
• ] Indicates the end of an optional command, parameter, or pseudonym.
Expressions
ATLAS does not support arithmetic expressions in the syntax. You can, however, evaluate
and use expressions by using the SET or EXTRACT statements.

943 ATLAS User’s Manual

A.MESH, R.MESH, X.MESH, Y.MESH, Z.MESH Statements

21.2 A.MESH, R.MESH, X.MESH, Y.MESH, Z.MESH

<n>.MESH specifies the location of grid lines along the <n>-axis in a rectangular mesh for 2D
or 3D simulation. It also specifies radial and angular components for 3D cylindrical meshing
(see Section 2.6.9 “Specifying 3D Cylindrical Structures”).

Note: The commands are equivalent in the X, Y, or Z directions.

Syntax
X.MESH LOCATION=<l> (NODE=<n> [RATIO=<r>])| SPACING=<v>

Parameter Type Default Units

DEPTH Real m
H1 Real m
H2 Real m
H3 Real m
LOCATION Real m
N.SPACES Integer
NODE Integer
RATIO Real 1
SPACING Real m (degrees)
WIDTH Real m
X.MAX Real m
X.MIN Real 0.0 m
Y.MAX Real m
Y.MIN Real 0.0 m

Description

DEPTH Specifies the extent of a mesh section in the Y direction.

H1 Specifies the spacing between consecutive mesh lines at the beginning of a
mesh section.
H2 Specifies the spacing between consecutive mesh lines at the end of a mesh
section.

944 ATLAS User’s Manual

A.MESH, R.MESH, X.MESH, Y.MESH, Z.MESH Statements

H3 Specifies the spacing between consecutive mesh lines at the middle of a mesh
section.
LOCATION Specifies the location of the grid line.
N.SPACES Specifies the number of mesh spacings within a given mesh section.
NODE Specifies mesh line index. There is a limit of 120 mesh lines. These mesh lines
must be assigned in increasing order.
RATIO Specifies the ratio to use when interpolating grid lines between given locations.
Spacing between adjacent grid lines will increase or decrease by the factor
assigned to the RATIO parameter. A RATIO value of between 0.667 and 1.5 is
recommended. RATIO should not be used if SPACING is specified.
SPACING Specifies the mesh spacing at the mesh locations specified by the LOCATION
parameter. If the SPACING parameter is specified, the NODE and RATIO
parameters should not be specified. If the SPACING parameter is used to specify
mesh spacings, the NX, NY, and NZ parameters of the MESH statement should not
be specified. When the mesh spacings are specified using the SPACING
parameter, the mesh size will be calculated.
WIDTH Specifies the extent of a mesh section in the X direction.
X.MAX/Y.MAX Specify the location of the end of a given mesh section (should not be
specified if DEPTH/WIDTH is specified).
X.MIN/Y.MIN Specify the location of the beginning of a given mesh section (should only be
specified for the first section).

Setting Fine Grid at A Junction Example

This example shows how to space grid lines closely around a junction at y=0.85 microns.
Y.MESH LOC=0.0 SPAC=0.2
Y.MESH LOC=0.85 SPAC=0.01
Y.MESH LOC=2 SPAC=0.35

945 ATLAS User’s Manual

BEAM Statements

XRADIUS Real 0.0 m

XSIGMA Real 0.0 m
Y.MAX.CROP Real m
Y.MIN.CROP Real m
Y0.PHOTOGEN Real m
Y.ORIGIN Real 0.0 m

952 ATLAS User’s Manual

BEAM Statements

Parameter Type Default Units

Y.POINT Real 0.0 m

Y.SEMIAXIS Real 0.0 m
Z1 Real 0.0 µm

Z2 Real 0.0 µm

Z.CENTER Real 0.0 m

Z.GAUSSIAN Logical False

Z.MAX.CROP Real m
Z.MEAN Real 0.0 m
Z.MIN.CROP Real m
Z.ORIGIN Real 0.0 m
Z.PERIODIC Logical False
Z.POINT Real m
Z.RADIUS Real 0.0 m
Z.SEMIAXIS Real 0.0 m
Z.SIGMA Real 0.0 m
ZCENTER Real 0.0 m
Z.FACET Real m
ZGAUSSIAN Logical False

ZMEAN Real 0.0 m

ZMAX Real 1.01020 m

ZMIN Real -1.01020 m

Z.MIRROR Logical False

ZRADIUS Real 0.0 m
ZSIGMA Real 0.0 m

953 ATLAS User’s Manual

BEAM Statements

Description

Figure 21-1: LUMINOUS Optical Source Coordinate System

ABS.EDGE Ensures that for energies (wavelengths) less than the bandgap of the material in
question the imaginary part of the complext index is zero. By default, this
parameter is FALSE.
AM0 Enables the built-in AM0 solar spectrum.
AM1.5 Enables the built-in AM1.5 solar spectrum.
AMBIENT.INDEX Specifies the index of refraction of domain outside of all meshed regions.
ANGLE This is the angle of propogation of the optical beam (see Figure 21-1). ANGLE=90
is vertical illumination from the top of the device. The synonym for this
parameter is PHI.
ANISO Signifies that the anisotropic solver should be applied to the problem. This incurs
a roughly 2 increase in computation time.
BACK.REFL Specifies that back side reflections are to be taken into account. When
BACK.REFL is specified, the area outside the device domain is assumed to be a
vacuum (i.e., n = 1.0, k = 0.0).
BPM Use the Beam Propagation Method instead of Ray-Tracing for analysis of light
(LUMINOUS2D) propagation in the device. Use BPM when diffraction of light or coherent effects
are important (see Section 10.4 “Beam Propagation Method in 2D”). BEAM
parameters related specifically to ray tracing (such as RAYTRACE, RAYS, THINEST,
MAX.WINDOW, MIN.WINDOW) are ignored when BPM is specified.

CIRCULAR This is the synonym for ELLIPTICAL.

CON.REFLECT Specifies that all conductors are to be treated as perfect reflectors.

954 ATLAS User’s Manual

BEAM Statements

DEVICE Specifies the name of a device in MIXEDMODE to identify to which device the
beam is directed. The synonym for this parameter is STRUCTURE.
DIEL.FUNC Enables the plotting of the dielectric function for the Tauc-Lorentz Model (see
Section 3.10.3 “Tauc-Lorentz Dielectric Function with Optional Urbach Tail
Model for Complex Index of Refraction”). To plot the model, you should also
specify the energy range using the LAM1 and LAM2 parameters of the
MATERIAL statement with the number of samples, NLAM and the output file name
OUT.INDEX.

DIFFUSIVE Enables the diffusive transfer matrix method. This method permits modeling of
both coherent (specular) and incoherent (diffusive) interactions at material
interfaces (see Section 10.3.4 “Transfer Matrix with Diffusive Interfaces”).
DX.PHOTOGEN Describe the spacings in the x and y directions for tabular photogeneration
DY.PHOTOGEN described in Section 10.6.4 “Tabular Photogeneration (Luminous2D only)”.
E.END Specifies the maximum energy for uniform sampling of the spectrum in energy.
E.NUMBER This is an alias for WAVEL.NUM.
E.START Specifies the minimum energy for uniform sampling of the spectrum in energy.
ELLIPTICAL Specifies that an elliptical source is to be used in LUMINOUS 3D. When the
ELLIPTICAL parameter is specified, the X.CENTER, Z.CENTER, X.RADIUS, and
Z.RADIUS should also be specified. The synonym for this parameter is
CIRCULAR..

ERRCTR Specifies the number of time steps between error estimate updates. This only
applies if NO.ENV is specified.
F.RADIATE Specifies the name of a file containing a C-Interpreter function for specifying
generation rate as a function of position and optionally time. This function can be
used to simulate single event upset (LUMINOUS2D only).
F.REFLECT Specifies the name of a file containing a C-Interpreter function for specifying
reflection coefficient models as a function of wavelength, position, and angle
incidence (LUMINOUS only).
F3.RADIATE This is the same as the F.RADIATE parameter but is applied in 3D. This is
typically used for single event or photogeneration simulations with
LUMINOUS 3D.
FD.AUTO Specifies that the structure will be meshed automatically so that locally the
condition that the mesh spacing be less than the wavelength divided by the value
of the parameter TD.STATE.
FD.AUTO.MIN Specifies the minimum number of mesh lines used to resolve any region in any of
the principal directions when FD.AUTO is also specified.
FILE.PHOTOGEN Specifies the file name for tabular photogeneration described in Section 10.6.4
“Tabular Photogeneration (Luminous2D only)”.

955 ATLAS User’s Manual

BEAM Statements

FRONT.REFL Specifies that front side reflections are to be taken into account. When
FRONT.REFL is specified, the area outside the device domain is assumed to be a
vacuum (i.e.,. n=1.0, k= 0.0).
GAUSSIAN This is the synonym for X.GAUSSIAN.
INPUTRAYS Specifies the filename of a file containing user-defined beam information. The
file format is described in Section 10.6.1 “User-Defined Beams”.
INTEGRATE Specifies whether the user-specified spectrum (contained in the file identified by
the POWER.FILE parameter) should be numerically integrated and averaged over
the wavelength sampling or if the samples should be interpolated directly from
the table.
ITERATION Sets or resets the iteration number passed to the C-Interpreter function
F.CHARGING.

LENS.HEIGHT Specifies an attribute of the composite lenslet (see Figure 10-16).

LENS.INDEX Specifies the index of refraction of a lenslet (LUMINOUS 3D only).
LENS.PLANE Specifies the minimum y-cordinate of the lenslet sphere (LUMINOUS 3D only).
LENS.RADIUS Specifies the radius of the sphere defining a lenslet (LUMINOUS 3D only).
LENS.WIDTH Specifies an attribute of the composite lenslet (see Figure 10-16).
LENS.X Specifies the X coordinate of the center of the sphere defining a lenslet
(LUMINOUS 3D only).
LENS.Y Specifies the Y coordinate of the center of the sphere defining a lenslet
(LUMINOUS 3D only).
LENS.Z Specifies the Z coordinate of the center of the sphere defining a lenslet
(LUMINOUS 3D only).
LENS.XMIN, Specify attributes of the composite lenslet (see Figure 10-16).
LENS.XMAX,
LENS.ZMIN,
LENS.ZMAX

956 ATLAS User’s Manual

BEAM Statements

LENS.XSAGS, Specify the file names of aspheric lenslet Sag-h data files. These files must be in
LENS.ZSAGS the following format.
number_of_pairs
Sag_1 h_1
Sag_2 h_2
. .
. .
. .
Sag_n h_n
Here, n is the number of pairs specified in the first line of the file. The pairs must
be arranged in order of increasing h. The data modeling algorithm assumes that
there is a sample with Sag=0 at h=0. Therefore, such a sample must not be
included in the file.
LID Specifies that the photogeneration rate of the beam will be applied only to the
light induced defect generation model. It will not be included in the carrier
continuity and temperature equations.
LONGIT.STEP Sets the mesh size in the longitudinal direction when using BPM. The default
value is WAVELENGTH/16.0.
MAX.WINDOW Specifies the maximum x-value of the illumination window relative to the
coordinate system of the optical beam (see Figure 21-1). The illumination
window is always clipped to the device domain. The synonym for this parameter
is XMAX.
MC.ARES Specifies the angular resolution for angular distribution functions for Monte
Carlo ray tracing in radians.
MC.SEED Specifies the seed value for random number generation for Monte Carlo ray
tracing.
MEAN This is the synonym for X.MEAN.
METAL.REFLECT Specifies that all metals are to be treated as perfect reflectors.
MIN.POWER Specifies the minimum intensity relative to the source that a given ray will be
traced. This is useful for limiting the numbers of rays traced.
MIN.WINDOW Specifies the minimum x-value of the illumination window relative to the
coordinate system of the optical beam. The synonym for this parameter is XMIN.
MONTE.CARLO Specifies that Monte Carlo ray tracing will be used.
NO.ENV Specifies that charge in intensity is used rather than field envelopes for error
estimation. This saves about 50% in memory.
NORMALIZE Indicates that the POWER.FILE spectral intensity needs to be normalized (integral
over spectrum equals 1). In this case, beam intensity is set by the B<n> parameter
on the SOLVE statement.

957 ATLAS User’s Manual

BEAM Statements

NSAMP Specifies the number of samples used to represent the aspheric lenslet profile
output using OUT.SAG.
NUMBER Specifies the beam number (from 1 to 10). This number is used by the SOLVE
statement to specify the relative intensity of different beams. You may specify
beam numbers in any order that you desire.
NX Specifies the number of rays traced along the source beam’s X axis for
LUMINOUS 3D.
NZ Specifies the number of rays traced along the source beam’s Z axis for LUMINOUS
3D.
N.VS.X Indicate that index variation may occur in the associated direction relative to the
N.VS.Y source beam coordinate system.
N.VS.Z

OUT.POWER Specifies the filename for output of multispectral samples, input using the
POWER.FILE parameter, in a format suitable for display in TONYPLOT.

OUT.SAG Specifies the name of a file used to output an aspheric lenslet profile for
subsequent display in TONYPLOT. OUT.SAG allows a cutline along an arbitrary
direction specified by two points in the XZ plane using the X1, Z1, X2, and Z2
parameters.
OUT.XSAG, Specify the file names for output log files that can be displayed using TONYPLOT.
OUT.ZSAG These files contain the data input from the LENS.XSAGS and LENS.ZSAGS files
and the results of the data modeling.
OUTFILE Specifies the name of a structure file for visualizing lenslet information and ray
tracing. This can be distinguished from the RAYTRACE parameter, which will save
ray trace information only without lenslets. The alias for this parameter is
TD.FILE.

PARALLEL Specifies that multiple processors can be used in parallel during ray trace
calculation.
PERIODIC Specifies that for ray tracing, the structure is to be treated as periodic in the X and
Z directions. Rays exiting the sides of the device are wrapped around to the other
side of the device.
PHI This is the synonym for ANGLE.
PML.CROP Specifies that PMLs will be omitted from any FDTD structure file.

958 ATLAS User’s Manual

BEAM Statements

POLARIZE Specifies the polarization of the optical beam at the origin. The polarization angle
is the angle between the E vector and the incidence plane.
POWER.SCAL Specifies a scale factor. This factor is multiplied by each of the relative powers in
the spectrum file when multi-spectral simulations are performed. The
POWER.SCAL parameter can be used to perform unit conversions.

POWER.FILE Specifies the filename of a spectrum file. The spectrum file must be in the
following format.
number of pairs
wavelength_1 power_1
wavelength_2 power_2
. . . . . .
. . . . . .
wavelength_n power_n

Note: The power file must contain at least two power/wavelengths pairs.

QUANTUM.EFF This is a quantum efficiency factor which specifies the number of carrier pairs
generated per photon absorbed.
RAYTRACE Specifies the name of a file where the results of a ray trace are saved. The ray
trace may be viewed using TONYPLOT3D (LUMINOUS 3D only). This deos not
include lens information. For lens information, use the OUTFILE parameter.
RAY.CHECK Enables diagnostic printing during ray tracing.
RAYAREA This is a parameter that specifies area in µm2 (thickness in 2D in um) of each ray
in the BEAM statement. This parameter is used when the specified INPUTRAYS
file does not contain ray area information.
RAYS Specifies the number of rays you want to split the optical beam. LUMINOUS 2D
will automatically split the beam into enough rays to resolve the geometry. Use
of the RAYS parameter will cause further splitting of the optical beam
(LUMINOUS only).
REFLECTS Specifies the number of reflections that will be traced. When the value of the
REFLECTS parameter is increased, the total number of rays traced increases non-
linearly. We recommend that this parameter be used wisely. For example, a
single ray incident on three material layers will produce 4 rays if REFLECTS=0 is
specified, 10 rays if REFLECTS=1 is specified, and 24 rays if REFLECTS=2 is
specified.
REL.POWER Specifies the relative power in the beam when mono-spectral simulations are
performed. This factor is multiplied by the power parameters specified in the
SOLVE statement to give the total optical power in the beam.

SAG.ITS Specifies the number of interations used to model the aspherical lenslet Sag-h
data using non-linear least squares fitting.

959 ATLAS User’s Manual

BEAM Statements

SEQ.LOG Specifies the number of name sequenced log plot files to be saved.
SEQ.STR Specifies the number of name sequenced structure plot files to be saved.
SEQ.NFP Specifies the number of name sequenced near-field pattern plot files to be saved.
SEQ.FFP Specifies the number of name sequenced far-field pattern plot files to be saved.
SIGMA This is the synonym for X.SIGMA.
STRUCTURE This is a synonym for DEVICE.
SUB.EXTINCT Specifies the extinction coefficient used at the exit of a layer stack for
analysis using the matrix method (see Section 10.3 “Matrix Method”).
SUB.INDEX Specifies the index of refraction at the exit of a layer stack for analysis using the
matrix method (see Section 10.3 “Matrix Method”).
SUBSTRATE Indicates that the transfer matrix method in LUMINOUS2D will treat the final
layer as the substrate (i.e., it is treated as being infinitely thick).
TAUC.DOS Specifies that midgap density of states will be calculated from the absorption
coefficient and output with absorption as specified by the OUT.INDEX parameter
of the MATERIAL statement.
TBLACK Specifies the black body temperature of a 1 steradian target for multi-spectral
analysis.
THETA Specifies the angle of rotation for the source beam direction of propogation
relative to the XY plane.
THINEST Specifies the width of the thinnest ray to be traced (LUMINOUS only).
TR.MATRIX Specifies that optical absorption and photogeneration analysis in LUMINOUS 2D
will be done using the Matrix Method as described in Section 10.3 “Matrix
Method”.
TRANSV.STEP Sets the mesh size in the transverse direction when using BPM. The default
value is WAVELENGTH/16.0.
USER.SPECTRUM Specifies that sampling in wavelength used for simulation is the same as that
contained in the POWER.FILE.
VERBOSE Enables a higher level of diagnostic run-time printing.
WAVELENGTH Specifies the optical wavelength of the source beam (in the vacuum) for
mono-spectral simulations.
WAVEL.END Specifies the maximum wavelength of the source beam (in the vacuum) when
multi-spectral simulations are performed.
WAVEL.NUM Specifies the number of wavelengths which will be used when multi-spectral
simulations are performed. Spectral illumination is selected when the
WAVEL.NUM parameter is greater than 1. Once multi-spectral simulations are
selected, you must specify the POWER.FILE, WAVEL.START, and WAVEL.END
parameters.

960 ATLAS User’s Manual

BEAM Statements

WAVEL.SCAL Specifies the scale factor for the wavelengths which are used in multispectral
simulations. Each of the wavelengths in the spectrum file is multiplied by this
scale factor.
WAVEL.START Specifies the minimum wavelength of the source beam (in the vacuum) when
multi-spectral simulations are performed.
X0.PHOTOGEN Describe the origin of the tabular photogeneration described in Section 10.6.4
Y0.PHOTOGEN “Tabular Photogeneration (Luminous2D only)”.
X1, X2, Z1, and Z2 Specify the coordinates of two points in the XZ plane for extraction of an
aspheric lenslet profile along the cutline defined by the two points. The profile
is output to the file specified by the OUT.SAG parameter for subsequent display
in TONYPLOT.
X.CENTER Specifies the X coordinate of the center of an elliptical source related to the
beam coordinate system (which is centered at the beam origin) for LUMINOUS
3D. Note that to enable elliptical sources, the ELLIPTICAL parameter should
also be specified. The synonym for this parameter is XCENTER.
X.FACET Specify the location of a plane intersecting perpendicular the associated axis
Y.FACET used for extraction of near and far field data.
Z.FACET

X.GAUSSIAN Specifies that a LUMINOUS 3D source is to have a Gaussian intensity profile in

the X direction related to the beam coordinate system (which is centered at the
beam origin). When X.GAUSSIAN is specified, X.MEAN and X.SIGMA should
also be specified. The synonyms for this parameter are XGAUSSIAN and
GAUSSIAN.

X.MEAN Specifies the location of the mean value of a Gaussian source in the X
direction related to the beam coordinate system (i.e., from the X.CENTER
position). When X.MEAN is specified, X.GAUSSIAN and X.SIGMA should also be
specified. The synonyms for this parameter are XMEAN and MEAN.
X.MIRROR Indicates that the extreme boundaries of the simulation domain in the beam
referenced coordinate system X coordinate are to be treated as mirrors for ray
tracing.
X.ORIGIN Specifies the X coordinate of the optical beam origin (see Figure 21-1). The
beam must originate outside all device regions.
X.RADIUS Specifies the X axis radius of an elliptical source related to the beam
coordinate system (which is centered at the beam origin) for LUMINOUS 3D.
Note that to enable elliptical sources, the ELLIPTICAL should also be
specified. The synonym for this parameter is XRADIUS.
X.SEMIAXIS, Specify attributes of the elliptical lense (see Equation 10-79).
Y.SEMIAXIS,
Z.SEMIAXIS

961 ATLAS User’s Manual

BEAM Statements

X.SIGMA Specifies the standard deviation in the X direction of the beam coordinate
system for a Gaussian source in LUMINOUS 3D. When X.SIGMA is specified,
X.GAUSSIAN and X.MEAN should also be specified. The synonym for this
parameter is XSIGMA.
XMAX Specifies the maximum X coordinate in the source beam coordinate system for
ray tracing in LUMINOUS 3D.
XMIN Specifies the minimum X coordinate in the source beam coordinate system for
ray tracing in LUMINOUS 3D.
Y.ORIGIN Specifies the Y coordinate of the optical beam origin (see Figure 21-1). The
beam must originate outside all device regions.
Z.CENTER Specifies the Z coordinate of the center of an elliptical source related to the
beam coordinate system (which is centered at the beam origin). Note that to
enable elliptical sources, the ELLIPTICAL parameter should also be specified.
The synonym for this parameter is ZCENTER.
Z.GAUSSIAN Specifies that a LUMINOUS 3D source is to have a Gaussian intensity profile in
the Z direction related to the beam coordinae system (which is centered at the
beam origin). When Z.GAUSSIAN is specified, Z.MEAN and Z.SIGMA should also
be specified. The synonym for this parameter is ZGAUSSIAN.
Z.MEAN Specifies the location of the mean value of a Gaussian source in the Z
direction related to the beam coordinate system. When Z.MEAN is specified,
Z.GAUSSIAN and Z.SIGMA should also be specified. The synonym for this
parameter is ZMEAN.
Z.MIRROR Indicates that the extreme boundaries of the simulation domain in the beam
referenced coordinate system Z coordinate are to be treated as mirrors for ray
tracing.
Z.RADIUS Specifies the X axis radius of an elliptical source related to the beam
coordinate system. When Z.MEAN is specified, Z.GAUSSIAN and Z.SIGMA
should also be specified. The synonym for this parameter is ZRADIUS.
Z.SIGMA Specifies the standard deviation in the Z direction of the beam coordinate system
for a Gaussian source in LUMINOUS 3D. When Z.SIGMA is specified,
Z.GAUSSIAN and Z.MEAN should also be specified. The synonym for this
parameter is ZSIGMA.
ZMAX Specifies the maximum Z coordinate in the source beam coordinate system for
ray tracing in LUMINOUS 3D.
ZMIN Specifies the minimum Z coordinate in the source beam coordinate system for
ray tracing in LUMINOUS 3D.
Z.ORIGIN Specifies the Z coordinate of the optical beam origin. The beam must
originate outside all device regions (LUMINOUS 3D only).

FDTD Parameters

962 ATLAS User’s Manual

BEAM Statements

BIG.INDEX Specifies that the parameters TD.SRATE and PROP.LENS will be

multipled by the maximum real part of the index of refraction.
COSINE Specifies that the FDTD source is modeled as a cosine wave.
BACK.FACET These are logical parameters for extracting near and far field patterns at the
BOTTOM.FACET surfaces of the FDTD domain.
FRONT.FACET
LEFT.FACET
RIGHT.FACET
TOP.FACET

DT Specifies the uniform time step for FDTD analysis.

FACET Specifies the root name of files for output of near and far field patterns.
FDTD Enables FDTD modeling of the beam propogation.
INSTATE Specifies the root file name of the FDTD “state” previously saved to disk using
the OUTSTATE parameter. INSTATE will load the previous state if available.
MEMORIZE Specifies a user-defined word to be used as a key for retrieval of FDTD state
from memory using the REMEMBER parameter.
MESHONLY Specifies that only mesh structure is loaded when INSTATE is specified.
NOZLENS Specifies that lens variation in the z direction is ignored. The x, y
characteristics are taken at z=0.
NX Specifies the number of uniformly spaced grid lines to be used in the FDTD
mesh in the X direction.
NY Specifies the number of uniformly spaced grid lines to be used in the FDTD
mesh in the Y direction.
NZ Specifies the number of uniformly spaced grid lines to be used in the FDTD
mesh in the Z direction.
OUT.BACK Indicate which PMLs are included in calculation of output coupling for LED
OUT.BOTTOM analysis.
OUT.FRONT
OUT.LEFT
OUT.RIGHT
OUT.TOP

OUTSTATE Specifies the root name of files written to disk saving the current FDTD “state”
for later retrieval using the INSTATE parameter.
PHASE Specifies the initial phase of the FDTD source.
PLANE Specifies that the FDTD source is a plane wave located at y'=0.0.
POINT Specifies that the FDTD source is a point source. The location is specified by
X.POINT, Y.POINT, and Z.POINT.

963 ATLAS User’s Manual

BEAM Statements

PROP.LENG Specifies the distance limit for which the light will be propagated before FDTD
simulation is concluded.
REMEMBER Specifies a user-defined word that is associated with the “state” of the FDTD
saved in memory by the MEMORIZE parameter. The REMEMBER parameter will
restore the previous state is available.
S.BOTTOM Specifies that the plane source is located at the bottom (maximum Y
coordinate) of the device.
S.BACK Specifies that the plane source is located at the back (minimum Z
coordinate) of the device.
S.FRONT Specifies that the plane source is located at the front (maximum Z
coordinate) of the device.
S.LEFT Specifies that the plane source is located at the left (minimum X
coordinate) of the device.
S.RIGHT Specifies that the plane source is located at the right (maximum X
coordinate) of the device.
S.TOP Specifies that the plane source is located at the top (minimum Y
coordinate) of the device.
SINE Specifies that the FDTD source is modeled as a sine wave.
STR.DECI Specifies the number of decimal digits of precision to maintain when
writing structure files.
STR.PRECIS This is alias for STR.DECI.
SUBDX Specify the spacing in microns that the FDTD mesh will be sampled at for
SUBDY subsampled output of FDTD meshes.
SUBDZ

SUBNX Specify the number of uniform samples that the FDTD mesh will be
SUBNY sampled at for subsampled output of FDTD meshes.
SUBNZ

SX Specifies the spacing of uniformly spaced grid lines to be used in the FDTD
mesh in the X direction.
SY Specifies the spacing of uniformly spaced grid lines to be used in the FDTD
mesh in the Y direction.
SZ Specifies the spacing of uniformly spaced grid lines to be used in the FDTD
mesh in the Z direction.
TE Specifies that the FDTD source is a transverse electric wave.
TD.END Specifies that a time-domain structure is written at the end of the number of
specified iterations (TD.LIMIT or TD.WAVES) or when the estimated error is
reduced below the specified maximum (TD.ERRMAX). This file will have a root
name given by the parameter TD.FILE with a ".str" suffix.

964 ATLAS User’s Manual

BEAM Statements

TD.ERRMAX Specifies the maximum residual (estimated) error for termination of FDTD
time stepping.
TD.EVERY Specifies how often FDTD time periodic saves are performed. A value of 1
indicates that files are saved during every time step.
TD.FILE Specifies the root name of an output file saved periodically during the time
domain simulation of FDTD. The alias for this parameter is OUTFILE.
TD.LIMIT Specifies the maximum number of timesteps to simulate in FDTD analysis. If
TD.LIMIT is unspecified, the time domain simulation will continue
indefinitely.
TD.LOG Specifies a log file for capture of time domain data for FDTD.
TD.MANY Specifies how many FDTD files are saved during periodic time domain saves.
TD.ONE Specifies that only one file name will be written for periodic FDTD saving.
Each time a save is performed it is written over the previous save.
TD.PER Specifies the number of structure files to output during a cycle at the source in
FDTD time stepping.
TD.SRATE Defines the spatial sampling rate in terms on the number of samples per
wavelength (in vacuum).
TD.WAVES Specifies the maximum number of cycles (wavelengths) at the source for
FDTD time stepping.
TD.ZCUT Specifies that structure files output by specification of TD.FILE in
LUMINOUS3D will be output at 2D cutplanes. The z location of the 2D
cutplane is specified by the value of TD.ZCUT.
TM Specifies that the FDTD source is a transverse magnetic wave.
TIMING Enables the output of timing summaries for the solution process of FDTD.
WAVE.SAMP This is alias for TD.SRATE.
X.MAX.CROP Provide cropping (truncation) of structure files output by TD.FILE along the
X.MIN.CROP principal axes.
Y.MAX.CROP
Y.MIN.CROP
Z.MAX.CROP
Z.MIN.CROP

X.POINT Specify the location of a point source for FDTD.

Y.POINT

X.PERIODIC Specifies that for ray tracing, the structure is to be treated as periodic in the X
direction. Rays exiting the sides of the device are wrapped around to the other
side of the device.
Z.PERIODIC Specifies that for ray tracing, the structure is to be treated as periodic in the Z
direction. Rays exiting the sides of the device are wrapped around to the other
side of the device.

965 ATLAS User’s Manual

BEAM Statements

Monochromatic Beam Example

This beam has a monochromatic spectrum with a wavelength of 0.6 m. The beam originates
at x=0.5 and y=-2.0. It has a 90 degree propogation angle and a beam width of 0.2 m which
is centered at the beam origin.During the ray trace calculation the rays will be terminated
when the power level along the ray falls to 5% of the original power.
BEAM NUM=1 WAVELENGTH=0.6 X=0.5 Y=-2.0 ANG=90.0 MIN=-0.1 MAX=0.1 \
MIN.POWER=0.05
Multi-spectral Beam Example
A multi-spectral beam (at a 45 degree angle) which originates at x=0.0 and y=-1.0. The multi-
spectral source is imported from the spectrum file, source.spc. The spectrum is discretized
into our wavelengths between 0.4 m and 0.6 m.
BEAM NUM=2 X=0.0 Y=-1.0 ANG=45.0 \
POWER.FILE=SOURCE.SPC WAVEL.START=0.4 \
WAVEL.END=0.6 WAVEL.NUM=4

LUMINOUS3D Lens Example

BEAM NUM=1 X.ORIGIN=2.5 Y.ORIGIN=-1.0 Z.ORIGIN = 2.5 ANG=90.0 WAVEL=0.6 \
NX=10 NZ=10 LENS.X=2.5 LENS.Y=-0.5 LENS.Z=2.5 \
LENS.INDEX=2.03 LENS.RADIUS=0.25 LENS.PLANE=-0.5

Gaussian Intensity Profile Example

The following beam statement will define a beam window 2 m wide that is centred at
(X.ORIGIN, Y.ORIGIN) with a Gaussian peak of 0.01 W/cm2 with a standard deviation of
0.05 m.
beam num=1 x.origin=5.0 y.origin=-1.0 angle=90.0 wavelength=0.6 \
xmin=-1 xmax=1 GAUSSIAN MEAN=0 SIGMA=0.05 RAYS=200
SOLVE B1=1E-2

Note: It is recommended that you use the RAYS parameter to define a large number of rays across the beam to
ensure that the Gaussian profile is adequately reproduced. The rays are evenly spaced across the beam so
it is necessary to use a large number of them.

966 ATLAS User’s Manual

CHARACTERIZE Statements

21.4 CHARACTERIZE
The CHARACTERIZE statement initiates optical characterization. Characterization involves
the capture of absorption versus depth data at each subsequent solution.
Syntax
CHARACTERIZE ABSORB=<string> [<parameters>]

Parameter Type Default Units

ABSORPTION Character
BEAM Integer
NDEP Integer
DEP1 Real microns
DEP2 Real microns
S.DEP1 Real microns
S.DEP2 Real microns
SEQ.ABS Integer 0

Description

ABSORPTION Specifies the file name root string for capture of absorption versus depth. A
characterization file will be written at each subsequent solution.
BEAM Specifies the beam index that associated with the beam to be characterized.
NDEP Specifies the number of uniform samples to be taken with depth.
DEP1 Specifies the minimum y coordinate in microns for data capture.
DEP2 Specifies the maximum y coordinate in microns for data capture.
S.DEP1 Specifies the sample spacing at the minimum y coordinate for data capture in
microns.
S.DEP2 Specifies the sample spacing at the minimum y coordinate for data capture in
microns.
SEQ.ABS Specifies the maximum number of seqeuenced files to capture.

967 ATLAS User’s Manual

COMMENT, # Statements

21.5 COMMENT, #
COMMENT allows comments to be placed in an ATLAS input file. ATLAS will print and
display comment lines.
Syntax
COMMENT [<string>]
# [<string>]
string is any alphabetic, numeric, or alphanumeric sequence of characters. The synonym for
this parameter for #.
Example
COMMENT ATLAS is a copyright of Silvaco International
# ATLAS is a copyright of Silvaco International

Note: The $ was allowed as a comment character in previous versions of ATLAS. This should be avoided and
replaced by the # or COMMENT statement.

968 ATLAS User’s Manual

CONTACT Statements

21.6 CONTACT
CONTACT specifies the physical attributes of an electrode.

Note: If the CONTACT statement is not used for a given electrode, the electrode is assumed to be charge-neutral
(Ohmic).

Syntax
CONTACT NUMBER=<n>|NAME=<ename>|ALL [<wfp>] [<bc>] [<lcr>] [<link>]

Parameter Type Default Units

ALL Logical False

ALPHA Real 0 cm
ALUMINUM Logical False
BARRIER Logical False
BETA Real 1.0
CAPACITANCE Real 0 F/m
CGTUNN Logical True
COMMON Character
CON.RESIST Real 0 ·cm2
CURRENT Logical False
DDMS.TUNNEL Logical False
DE.TUNNEL Real 10-4
DEVICE Character
DOS.NEGF Real 0.1 1/eV
E.TUNNEL Logical False
E.HIGHK Real 0.0 eV
ELE.CAP Integer
EXCLUDE_NEAR Logical False
EXT.ALPHA Real 0 W/(cm2K)
EXT.TEMP Real 300 K
F.ETUNNEL Character
F.WORKF Character

969 ATLAS User’s Manual

CONTACT Statements

Parameter Type Default Units

FACTOR Real 0
FARADS Logical False
FLOATING Logical False
FG.CAP Real 0.0 F/m
GAMMA Real 1.0
H.TUNNEL Logical False
H1SRV Real  cm/s
H2SRV Real  cm/s
HENRYS Logical False
HIGHK.AL Logical False
HIGHK.AU Logical False
HIGHK.HF Logical False
HIGHK.MG Logical False
HIGHK.MO Logical False
HIGHK.NI Logical False
HIGHK.PT Logical False
HIGHK.TA Logical False
HIGHK.TIN Logical False
HIGHK.W Logical False
IFL.AL203 Logical False
IFL.HF02 Logical False
IFL.SI3N4 Logical False
IFL.SI02 Logical False
IFL.ZR02 Logical False
INDUCTANCE Real 0 H·m
ME.TUNNEL Real 1.0
MH.TUNNEL Real 1.0
MO.DISILICIDE Logical False
MOLYBDENUM Logical False

970 ATLAS User’s Manual

CONTACT Statements

Parameter Type Default Units

MULT Logical False

NAME Character
NEGF.TUNNEL Logical False
NEUTRAL Logical True
N.POLYSILICON Logical False
NSURF.REC Logical False
NUMBER Integer
OHMS Logical False
ON.AL2O3 Logical False
ON.HFO2 Logical False
ON.SIO2 Logical False
ON.SI3N4 Logical False
ON.ZRO2 Logical False
P.POLYSILICON Logical False
PARABOLIC Logical False
PIPINYS Logical False
PSURF.REC Logical False
REF.ELEC Character
REF.ENUM Integer
REFLECT Logical False
REFLECT Real
RESISTANCE Real 0 m
RESONANT Logical False
S.HIGHK Real 0.0
SCOTT.MALLIARAS Logical False
SHORT Logical False
STRUCTURE Character
SURF.REC Logical False
QTUNN.CMASS Real 1.0

971 ATLAS User’s Manual

CONTACT Statements

Parameter Type Default Units

QTUNN.VMASS Real 1.0

THERM Logical True
TU.DISILICIDE Logical False
TUNGSTEN Logical False
VSURFN Real see description cm/s
VSURFP Real see description cm/s
WORKFUN Real 0 V

Description

ALL Defines the same properties for all electrodes.

Contact Parasitics

CAPACITANCE Specifies a lumped capacitance value to be attached to the contact.

CON.RESIST Specifies a distributed contact resistance. You cannot specify both CON.RESIST
and RESIST.
FARADS Specifies that the CAPACITANCE value will be in the units of Farads.
HENRYS Specifies that the INDUCTANCE value will be in the units of Henrys.
INDUCTANCE Specifies an external inductance which is related to the specified electrode. A
synonym is L.
OHMS Specifies that the RESISTANCE value will be in the units of ohms.
RESISTANCE Specifies a lumped resistance value. You may not specify both RESISTANCE
and CON.RESIST.

Note: There are restrictions on the allowed numerical methods and types of analysis possible when any form of
parasitic element is attached to a contact. See Chapter 2 “Getting Started with ATLAS” for details.

RESONANT Only applies to small signal AC analysis. If set, it models the lumped elements as
forming a parallel resonant circuit. The resistor and inductor are joined in series,
while the capacitor is in parallel with them. The driving voltage used in the
perturbation analysis is then applied to the circuit. If you do not set RESONANT, then
the capacitor will be connected to AC ground (see Figure 3-5).

976 ATLAS User’s Manual

CONTACT Statements

High-K Parameters
E.HIGHK Specifies the charge neutrality workfunction used in
Equation 3-199.
HIGHK.AL, HIGHK.AU, HIGHK.HF, Specify material dependent default metal workfunctions as
HIGHK.MG, HIGHK.MO, HIGHK.NI, given in Table 3-30.
HIGHK.PT, HIGHK.TA, HIGHK.W,
HIGHK.TIN

IFL.AL203, IFL.HF02, IFL.SI02, Specify material dependent charge neutrality workfunction and
IFL.SI3N4, IFL.ZR02 slope parameters for an interfacial layer as described in Table 3-
32.
ON.AL2O3, ON.HFO2, ON.SIO2, Specify material dependent default charge neutrality
ON.SI3N4, ON.ZRO2 workfunction and slope paramaters as given in Table 3-31.
S.HIGHK Specifies the slope parameter used in Equation 3-199.

Electrode Linking Parameters

These parameters allow one electrode to be biased as a function of another electrode. This
allows separate regions of the same physical contact to be linked together. For example, in the
statement:
CONTACT NAME=MYDRAIN COMMON=DRAIN FACTOR=4.6
The electrode, MYDRAIN, is linked to the electrode, DRAIN. The bias on MYDRAIN will always
be equal to the bias on the drain plus 4.6. If the optional MULT parameter had been specified,
the bias on MYDRAIN would be equal to the bias on the drain multiplied by 4.6.

COMMON Specifies the electrode name to which the contact refered to by NAME is linked.
Although the electrodes are linked, separate currents will be saved for both
electrodes unless SHORT is also specified. The electrode refered to in NAME should not
appear on any SOLVE statments since its bias is now determined as a function of the
electrode refered to by COMMON.
SHORT Specifies that the electrode referred to by NAME is shorted to the electrode
specified by the COMMON parameter. This implies that the two electrodes will be
treated as one and only one value will be written to log files and in the run time
output.
FACTOR Specifies the constant offset voltage (or current) between the electrodes refered by
NAME and COMMON. By default, FACTOR is added the defined voltage.

MULT Specifies that FACTOR is a multiplier.

977 ATLAS User’s Manual

CONTACT Statements

Floating Gate Capacitance Parameters

In some cases, you may want to simulate floating gate structures, in 2D, which have control
gates that are longer in the unsimulated dimension than the floating gate. In these cases,
specify the following parameters to account for addition capacitance between the floating
gate and the control gate and other electrodes. Up to four extra capacitances are allowed so in
the following <n> is an integer number between 1 and 4.

EL<n>.CAP Specifies the name of the electrode to which the extra capacitance is linked.
FG<n>.CAP Specifies the additional capacitance per unit length to be added between the
floating gate and electrode specified in EL<n>.CAP.

Schottky Barrier and Surface Recombination Example

This example defines all electrodes except number 2 (aluminum) to be neutral. Electrode
number 2 also includes finite surface recombination velocities and barrier lowering. A
definition in the second statement overrides that definition in the first statement.
CONTACT ALL NEUTRAL
CONTACT NUMBER=2 ALUMINUM SURF.REC BARRIER
Parasitic Resistance Example
This example attaches a lumped resistor with a value of 105m to the substrate. A
distributed contact resistance of 10-6·cm2 is included on the drain.
CONTACT NAME=substrate RESISTANCE=1E5
CONTACT NAME=drain CON.RESIST=1E-6
Floating Gate Example
This syntax defines a floating contact with a workfunction equal to 4.17eV. An extra 1fF/um
capacitance is added between this electrode and the electrode named cgate.
CONTACT NAME=fgate FLOATING N.POLY EL1.CAP=cgate
FG1.CAP=1e-15

Note: The MODELS PRINT command can be used to echo the back contact workfunction and parasitic
elements settings to the run-time output.

978 ATLAS User’s Manual

CURVETRACE Statements

21.7 CURVETRACE
CURVETRACE sets the parameters for the automatic curve tracing routine.
Syntax
CURVETRACE <params>

Parameter Type Default Units

ANGLE1 Real 5 Degree

ANGLE2 Real 10 Degree
ANGLE3 Real 15 Degree
BEG.VAL Real 0.0 V
CONTROLRES Real 1.01.010 

CONTR.ELEC Integer
CONTR.NAME Character
CURR.CONT Logical
END.VAL Real
MAXDV1 Real 0.1 V
MAXDV2 Real 1.0 V
MINCUR Real
MINDL Real 0.1
NEXTST.RATIO Real 2.0
NO.BACKTRACE Logical False
STEP.CONT Logical
STEP.INIT Real 0.1 V
STEPS Real 0.1
THETA.BACKTRACE Real 0.01 Degree
TURNINGPOINT Logical False
V.BACKTRACE Real 1 V
VOLT.CONT Logical False

979 ATLAS User’s Manual

CURVETRACE Statements

Description

ANGLE1, ANGLE2, and ANGLE3 These are the critical angles (in degrees) affecting the
smoothness and step size of the trace. If the difference in slopes of
the last two solution points is less than ANGLE1, the step size will
be increased for the next projected solution. If the difference lies
between ANGLE1 and ANGLE2, the step size remains the same. If
the difference is greater than ANGLE2, the step size is reduced.
ANGLE3 is the maximum difference allowed, unless overridden by
the MINDL parameter. ANGLE2 should always be greater than
ANGLE1 and less than ANGLE3.

BEG.VAL This is the value of the voltage at the starting point of the curve
trace for the controlling electrode.
CONTR.ELEC This is the number of the electrode that is designated as a
control electrode.
CONTR.NAME This is the name of the control electrode.
CURR.CONT Denotes that a maximum current on the control electrode,
specified by END.VAL is used as the upper bound on the trace.
CONTROLRES Specifies the control resistance value to be set at the control
electrode after the current at the control electrode becomes greater
than the value specified by the MINCUR parameter. Using a large
value of the control resistance allows you to set the value of the
current (instead of the internal voltage) on the control electrode. A
large value of the control resistance is suitable for electron devices
where the shape of the characteristic curve is vertical or exhibits
the presence of turning points (where it is required to switch the
electrode control from voltage to current).
END.VAL This is used to stop tracing if the voltage or current of control
electrode equals or exceeds END.VAL.
MAXDV1 Specifies maximum value of the voltage bias step at the
beginning of the CURVETRACER algorithm. This limitation is used
until the current at the control electrode is greater than the value
specified by the MINCUR parameter.
MAXDV2 Specifies maximum value of the voltage bias step to be used after
the current at the control electrode becomes greater than the value
specified by the MINCUR parameter.
MAXDV1 and MAXDV2 Control (together with NEXTST.RATIO) the maximum voltage step
at each point of the IV curve. Using large values for them means
that few points are used to trace the curve. This results in a faster
simulation but can affect the stability of the tracing algorithm and
decrease the accuracy of the IV curve.

980 ATLAS User’s Manual

CURVETRACE Statements

MINCUR This may be used to set a small current value in order to switch
from internal control electrode bias ramping to external
ramping with load resistor. This parameter is recommended for
small current breakdown simulation.
MINDL This is the minimum normalized step size allowed in the trace.
Usually, you don’t need to adjust this parameter. Increasing MINDL
will reduce the smoothness of the trace by overriding the angle
criteria, resulting in more aggressive projection and fewer
simulation points. Reducing MINDL will enhance the smoothness
and increase the number of points in the trace.
NEXTST.RATIO Specifies which factor to use to increase the voltage step on the
smooth parts of the I-V curve.
NO.BACKTRACE Stops curve tracing if it detects back tracing.

Note: If you define the curvetrace using the CONTINUE parameter on the SOLVE statement and you specify
_TMA on the command line, then the default value of NO.BACKTRACE is true.

STEPS This is the number of operational points on a trace if STEP.CONT was specified.
STEP.CONT Specifies that the trace will proceed for a certain number of simulation points.
STEP.INIT Specifies initial voltage step size.

Note: To set a sweep of increasingly negative voltage in CURVETRACE, you only need to set STEP.INIT to be
negative. Since all parameters are multiplier of STEP.INIT, the whole voltage sweep will be negative.

THETA.BACKTRACE Specifies the angular difference at which the curve is considered to be

backtracing.
TURNINGPOINT Specifies that binary output solution files will be saved whenever the
slope of the IV curve changes sign (i.e., there is a turning point). The
name of the output file is soln.num, where num is the number of the
current solution.
V.BACKTRACE Specifies the (internal) voltage difference at which the curve is
considered to be backtracing.
VOLT.CONT Denotes that a maximum voltage on the control electrode, specified by
END.VAL is used as the upper bound on the trace.

Diode Breakdown Example

To trace a diode breakdown curve using current value as a termination criteria, the following
statement may be used:
CURVETRACE CURR.CONT END.VAL=0.01 CONTR.NAME=anode \
MINCUR=5E-12 NEXTST.RATIO=1.1 STEP.INIT=0.1
SOLVE CURVETRACE

981 ATLAS User’s Manual

DATASET Statements

21.8 DATASET
The DATASET statement provides a common interface to import distributed data. The interface
includes a common interface to allow distortions such as rescale, windowing, stretch, and
rotation.
Syntax
DATASET INFILE=<string> NAME=<string> <filetype> [<parameters>]

Parameter Type Default Units

ABS.CHAR Logical False

AXIS.XYZ Logical True
AXIS.XZY Logical False
AXIS.YXZ Logical False
AXIS.YZX Logical False
AXIS.ZXY Logical False
AXIS.ZYX Logical False
DATA.XYZ Logical True
DATA.XZY Logical False
DATA.YXZ Logical False
DATA.YZX Logical False
DATA.ZXY Logical False
DATA.ZYX Logical False
FX1 Real 0.0 um
FY1 Real 0.0 um
FZ1 Real 0.0 um
FX2 Real 0.0 um
FY2 Real 0.0 um
FZ2 Real 0.0 um
INFILE Character
NAME Character
TX1 Real 0.0 um
TY1 Real 0.0 um

982 ATLAS User’s Manual

DATASET Statements

Parameter Type Default Units

TZ1 Real 0.0 um

TX2 Real 0.0 um
TY2 Real 0.0 um
TZ2 Real 0.0 um
XMULT Real 1.0
YMULT Real 1.0
ZMULT Real 1.0

Description
ABS.CHAR Specifies that the file contains photogeneration rate data
extracted using the CHARACTERIZE statement.
AXIS.XYZ Specify how the data in the file is to be ordered. In this case X,
AXIS.XZY then Y, then Z. Since the file is sequential, the first character
AXIS.YXZ (X) represents the fastest varying ordinate, the second
AXIS.YZX represents Y, and the third represents Z. Specifying AXIS.ZYX
AXIS.ZXY would reverse this order.
AXIS.ZYX

DATA.XYZ These are similar to the AXIS.XYZ except now we are referring
DATA.XZY how the data sets are to be interpreted. In this case, we are
DATA.YXZ specifying that the first set is the direction cosine with respect
DATA.YZX to X, the second one with respect to Z, and the last data set with
DATA.ZXY respect to Y.
DATA.ZYX

INFILE This is the name of the source data file.

FX1, FX2, FY1, FY2 (FZ1, FZ2) Specify the coordinates corresponding to rows and columns
(and pages) of data in the source file. In this case, there are
7121 data points in the data. We use FX and FY since we
specified AXIS.XY(Z) first. The letter F is meant to indicate
"FROM". The numbers 1 and 2 refer to the specification of two
points in 2 (3) space describing the limits of the what subset of
the data contained in the file are mapped.
NAME This is a convenient tag that we can refer to later in the MODELS
statement
TX1, TX2, TY1, TY2 (TZ1, TZ2) Specify the corners in the device coordinate where the data is to
be mapped. Scaling is automatic.
XMULT, YMULT, and ZMULT These are multipliers times each of the data sets.

983 ATLAS User’s Manual

DBR Statements

21.9 DBR
The DBR statement is used to define Distributed Bragg Reflectors (DBR). DBRs are periodic
structures composed of alternating layers of two different materials. The alias for this
parameter is SUPERLATTICE.
Syntax
DBR <parameters>

Parameter Type Default Units

BOTTOM Logical False

CALC1.STRAIN Logical False

CALC2.STRAIN Logical False
HALF.CYCLES Real 2

LAYERS Real 2
LED1 Logical False
LED2 Logical False
MAT1 Character

MAT2 Character

N1 Real 0

N2 Real 0

NA1 Real 0.0 cm-3

NA2 Real 0.0 cm-3

NAME1 Character
NAME2 Character
ND1 Real 0.0 cm-3

ND2 Real 0.0 cm-3

NU12 Integer 5
NU21 Integer 5
POLARIZ1 Logical False
POLARIZ2 Logical False
POLAR1.CHARG Real 0 cm-2

POLAR2.CHARG Real 0 cm-2

984 ATLAS User’s Manual

DBR Statements

POLAR1.SCALE Real 1.0

POLAR2.SCALE Real 1.0
QWELL1 Logical False
QWELL2 Logical False
SPA1 Real 0.0 m
SPA2 Real 0.0 m
STRAIN1 Real 0.0

STRAIN2 Real 0.0

TH12 Real
TH21 Real
THICK1 Real 0.0 m
THICK2 Real 0.0 m
TOP Logical False

WELL1.CNBS Integer 1
WELL2.CNBS Integer 1
WELL1.FIELD Logical True
WELL2.FIELD Logical True
WELL1.GAIN Real 1.0
WELL2.GAIN Real 1.0
WELL1.NX Integer 10
WELL2.NX Integer 10
WELL1.NY Integer 10
WELL2.NY Integer 10
WELL1.VNBS Integer 1
WELL2.VNBS Integer 1
Y.START Real 0.0 m
Y.FINISH Real 0.0 m
Y.FINAL Real 0.0 m
Y.END Real 0.0 m
Y1.COMP Real 0.0

985 ATLAS User’s Manual

DBR Statements

Y2.COMP Real 0.0

X1.COMP Real 0.0

X2.COMP Real 0.0

Description
The DBR statement is a short cut for specifying a set of REGION statements, which specifies a
stack of alternating layers of two materials. The material compositions of the layers are
specified by the MAT1, MAT2, NA1, NA2, ND1, ND2, STRAIN1, STRAIN2, X1.COMP, X2.COMP,
Y1.COMP, and Y2.COMP parameters. The number of layers are specified by the HALF.CYCLES
parameter. The locations of the layers are specified by the Y.START, Y.FINISH or the TOP/
BOTTOM parameters. The meshing of the layers is specified by the N1, N2 or the SPA1, SPA2
parameters.
For more information about DBR, see “Specifying Distributed Bragg Reflectors” on page
506.

BOTTOM Specifies that the DBR is to be added at the bottom (starting at

the maximum previously specified Y coordinate and extending
in the positive Y direction) of the structure.
CALC1.STRAIN Specifies that the strain in the region/cycle is calculated from
CALC2.STRAIN the lattice mismatch with adjacent regions.
HALF.CYCLES Specifies the total number of layers. Note that if HALF.CYCLES
is odd, there will be more layers of one material than the other.
Its alias is LAYERS.

Note: Don’t confuse HALD.CYCLES with fractions of the optical emisson wavelength. HALF.CYCLES relates
to layers.

LAYERS This is an alias for HALF.CYCLES.

LED1 Specifies that the region/cycle is to be treated as a light emitting region and
LED2 included in postprocessing for LED analysis. See Chapter 11 “LED: Light
Emitting Diode Simulator”.
MAT1 Specifies the material name for layers of material 1.
MAT2 Specifies the material name for layers of material 2.
N1 Specifies the integer number of mesh lines per layer of material 1. Note that if
N1 is specified, then SPA1 shouldn’t be specified.

N2 Specifies the integer number of mesh lines per layer of material 2. Note that if
N2 is specified, then SPA2 shouldn’t be specified.

NAME1 Specify the names of the regions composed of material 1 and material 2. You
NAME2 can conveniently use these names to modify material parameters and models on
the MATERIAL, MODEL, IMPACT, and MOBILITY statements.

986 ATLAS User’s Manual

DBR Statements

NA1 Specifies the ionized acceptor concentration in layers of material 1.

NA2 Specifies the ionized acceptor concentration in layers of material 2.
ND1 Specifies the ionized donor concentration in layers of material 1.
ND2 Specifies the ionized donor concentration in layers of material 2.
NU12 This is the number of grid lines in the graded region between the 1st and 2nd
half cycle in the direction specified by the BOTTOM or TOP parameters.
NU21 This is the number of grid lines in the graded region between the 2nd and 1st
half cycle in the direction specified by the BOTTOM or TOP parameters.
POLARIZ1 Enables the automatic calculation of added interface charge due to
POLARIZ2 spontaneous and piezoelectric polarization. See Section 3.6.12 “Polarization in
Wurtzite Materials”.
POLAR1.CHARG Specify polarization charge densities to replace those calculated using
POLAR2.CHARG Equations 3-601 and 3-602.
POLAR1.SCALE Specifies a constant scale factor multiplied by the calculated spontaneous and
POLAR2.SCALE piezoelectric polarization charges when you enable polarization by setting the
POLARIZATION parameter of the DBR statement. See Section 3.6.12
“Polarization in Wurtzite Materials”.
QWELL1 Specifies that the region/cycle is treated as a quantum well for calculation of
QWELL2 radiative recombination or gain or both for certain optoelectronic models.
SPA1 Specifies the spacing between mesh lines in layers of material 1. Note that this
should be smaller than or equal to THICK1.
SPA2 Specifies the spacing between mesh lines in layers of material 2. Note that this
should be smaller than or equal to THICK2.
STRAIN1 Specifies the strain in the layer of material 1. (Negative strain is
compressive).
STRAIN2 Specifies the strain in the layer of material 2. (Negative strain is
compressive).
TOP Specifies that the DBR is to be added at the top (starting at the minimum
previously specified Y coordinate and extending in the negative Y direction) of
the structure.
TH12 This is the thickness of grading region between 1st half cycle and 2nd half cycle
going in the direction specified by the BOTTOM or TOP parameters.
TH21 This is the thickness of grading region between 2nd half cycle and 1st half cycle
going in the direction specified by the BOTTOM or TOP parameters.
THICK1 Specifies the thickness of each layer of material 1.
THICK2 Specifies the thickness of each layer of material 2.

987 ATLAS User’s Manual

DBR Statements

WELL1.CNBS Specify the number of bound states retained for calculation of radiative
WELL2.CNBS recombination or gain if the region/cycle is treated as a quantum well as
WELL1.VNBS specified by the QWELL1, QWELL2 parameters.
WELL2.VNBS

WELL1.FIELD When enabled, specify that the calculations of bound state energies should
WELL2.FIELD include the effects of the local field.
WELL1.GAIN Specify a constant scale factor multiplied by the calculated gain to give the net
WELL2.GAIN gain used for certain optoelectronic calculations.
WELL1.NX Specify the number of slices (WELL#.NX) and the number of samples per slice
WELL2.NX (WELL#.NY) are used in the solution of Schrodinger's equation to obtain the local
WELL1.NY bound state energies for calculation of radiative recombination or gain or both
WELL2.NY for certain optoelectronic models.
Y.START Specifies the starting location of a DBR that begins with a layer of material 1
and alternates materials in an increasing Y direction.
Y.FINISH Specifies the starting location of a DBR that begins with a layer of material 1
and alternates materials in an decreasing Y direction.
Y.FINAL These are aliases for Y.FINISH.
Y.END

Y1.COMP Specifies the composition fraction y in layers of material 1.

Y2.COMP Specifies the composition fraction y in layers of material 2.
X1.COMP Specifies the composition fraction x in layers of material 1.
X2.COMP Specifies the composition fraction x in layers of material 2.

Note: DBR statements can be intermixed with X.MESH, Y.MESH, and REGION statements.

988 ATLAS User’s Manual

DEFECTS Statements

SIGTDH Real 1.010-16 cm2

STRUCTURE Character
TAT.TRAP Logical False
T0.AMPHOTERIC Real 300 K
TFILE Character
WGA Real 0.1 eV
WGD Real 0.1 eV
WTA Real 0.025 eV
WTD Real 0.05 ev
X.MIN Real m

X.MAX Real m

Y.MIN Real m

Y.MAX Real m

Z.MIN Real m

Z.MAX Real m

Description
The DEFECTS statement is used to describe the density of defect states in the band gap. You
can specify up to four distributions, two for donor-like states and two for acceptor-like states.
Each type of state may contain one exponential (tail) distribution and one Gaussian
distribution.

AFILE Specifies the file name where the acceptor state density
distribution, as a function of energy, will be stored. You can examine
this file by using TONYPLOT.
AMPHOTERIC Specifies the amphoteric defect model will be used.
CONTINUOUS Specifies that the continuous defect integral model will be used.
DEVICE Specifies which device the statement applies in mixed mode
simulation. The synonym for this parameter is STRUCTURE.
DFILE Specifies the file name where the donor state density distribution, as a
function of energy, will be stored. You can examine this file by using
TONYPLOT.

991 ATLAS User’s Manual

DEFECTS Statements

EGA Specifies the energy that corresponds to the Gaussian distribution peak
for acceptor-like states. This energy is measured from the
conduction band edge.
EGD Specifies the energy that corresponds to the Gaussian distribution peak
for donor-like states. This energy is measured from the valence band
edge.
EP.AMPHOTERIC Specifies the most probable potential defect energy.
EPT0.AMPHOTERIC Specifies the time constant used in time amphoteric reflect
generation model.
EPBETA.AMPHOTERIC Specifies the exponential constant used in the amphoteric defect
generation model.
EU.AMPHOTERIC Specifies the defect electron correlation energy.
EV0.AMPHOTERIC Specifies the characteristic energy.
F.TFTACC Specifies the name of a file containing a C-Interpreter function,
describing the distribution of acceptor state densities as a function of
energy.
F.TFTDON Specifies the name of a file containing a C-Interpreter function,
describing the distribution of donor state densities as a function of
energy.
FILE.AMPHOTERIC Specifies the file name where the dangling bond density of states
distribution, as a function of energy, will be stored. If you specify
FILEX.AMPHOTERIC, FILEY.AMPHOTERIC, and
FILEZ.AMPHOTERIC, then the file will be saved at the coordinate
closest to the one specified by FILEX.AMPHOTERIC,
FILEY.AMPHOTERIC, and FILEZ.AMPHOTERIC.

FILE.INT Specifies the integrated density of states with respect to energy will be
stored in the AFILE, DFILE, and TFILE files for discrete defects.
FILEX.AMPHOTERIC Specifies the X coordinate used in FILE.AMPHOTERIC.
FILEY.AMPHOTERIC Specifies the Y coordinate used in FILE.AMPHOTERIC.
FILEZ.AMPHOTERIC Specifies the Z coordinate used in FILE.AMPHOTERIC.
HCONC.AMPHOTERIC Specifies the density of hydrogen.
INT_LIM1 Specifies the lower limit for the numerical integration of the
CONTINUOUS method.

INT_LIM2 Specifies the upper limit for the numerical integration of the
CONTINUOUS method.

LIMIT1 Lower energy limit in TAT.TRAP integration range .

LIMIT2 Upper energy limit in TAT.TRAP integration range.

992 ATLAS User’s Manual

DEFECTS Statements

MATERIAL Specifies which material from the table in Appendix B “Material

Systems” will apply to the DEFECT statements. If a material is
specified, then the regions defined as being composed of that
material will be affected.
NA.MIN Specifies the minimum value of the acceptor trap DOS that will be
used in the discrete DEFECTS model.
ND.MIN Specifies the minimum value of the donor trap DOS that will be used in
the discrete DEFECTS model.
NGA Specifies the total density of acceptor-like states in a Gaussian
distribution.
NGD Specifies the total density of donor-like states in a Gaussian
distribution.
NSISI.AMPHOTERIC Specifies the density of Si-Si bonds.
NTA Specifies the density of acceptor-like states in the tail distribution at the
conduction band edge.
NTD Specifies the density of donor-like states in the tail distribution at the
valence band edge.
NUM.AMPHOTERIC Specifies the number of energy levels used in the DOS integration for
amphoteric defects.
NUMBER or REGION Specifies the region index to which the DEFECT statement applies.
NUMA Specifies the number of discrete levels that will be used to simulate the
continuous distribution of acceptor states.
NUMD Specifies the number of discrete levels that will be used to simulate the
continuous distribution of donor states.
NV0.AMPHOTERIC Specifies the density of states of the valance-band tail exponential
region extrapolated to the valence-band edge. If this parameter is not
specified, then the valence band density of states (NV300 on the
MATERIAL statement) will be used instead.

SIGGAE Specifies the capture cross-section for electrons in a Gaussian

distribution of acceptor-like states.
SIGGAH Specifies the capture cross-section for holes in a Gaussian
distribution of acceptor-like states.
SIGGDE Specifies the capture cross-section for electrons in a Gaussian
distribution of donor-like states.
SIGGDH Specifies the capture cross-section for holes in a Gaussian
distribution of donor-like states.
SIGMA.AMPHOTERIC Specifies the defect pool width.
SIGN0.AMPHOTERIC Specifies the electron capture cross-section for neutral defects.

993 ATLAS User’s Manual

DEFECTS Statements

SIGNP.AMPHOTERIC Specifies the electron capture cross-section for positive defects.

SIGP0.AMPHOTERIC Specifies the hole capture cross-section for neutral defects.
SIGPN.AMPHOTERIC Specifies the hole capture cross-section for negative defects.
SIGTAE Specifies the capture cross-section for electrons in a tail distribution of
acceptor-like states.
SIGTAH Specifies the capture cross-section for holes in a tail distribution of
acceptor-like states.
SIGTDE Specifies the capture cross-section for electrons in a tail
distribution of donor-like states.
SIGTDH Specifies the capture cross-section for holes in a tail distribution of
donor-like states.
STRUCTURE This is a synonym for DEVICE.
TAT.TRAP Specifies that the DEFECTS will only be used in the Ielmini trap assisted
tunneling models. Applies to insulators and semiconductors. With this
flag, you must specify either NGA, EGA,WGA for acceptors or NGD, EGD,
WGD for donors.

TFILE Specifies the file name where the acceptor and donor state density
distributions, as a function of energy referenced from Ev, will be stored.
You can examine this file by using TONYPLOT.
T0.AMPHOTERIC Specifies the freeze-in temperature.
WGA Specifies the characteristic decay energy for a Gaussian
distribution of acceptor-like states.
WGD Specifies the characteristic decay energy for a Gaussian
distribution of donor-like states.
WTA Specifies the characteristic decay energy for the tail distribution of
acceptor-like states.
WTD Specifies the characteristic decay energy for the tail distribution of
donor-like states.
X.MIN, X.MAX, Y.MIN, Y.MAX, Specify the bounding box for the DEFECT statement.
Z.MIN, and Z.MAX

994 ATLAS User’s Manual

DEFECTS Statements

TFT Example
The following statement lines specify distributed defect states which would typically be used
for polysilicon.
DEFECTS NTA=1.E21 NTD=1.E21 WTA=0.033 WTD=0.049 \
NGA=1.5E15 NGD=1.5E15 EGA=0.62 EGD=0.78 \
WGA=0.15 WGD=0.15 SIGTAE=1.E-17 \
SIGTAH=1.E-15 SIGTDE=1.E-15 SIGTDH=1.E-17 \
SIGGAE=2.E-16 SIGGAH=2.E-15 SIGGDE=2.E-15 \
SIGGDH=2.E-16

995 ATLAS User’s Manual

DEGRADATION Statements

21.11 DEGRADATION
DEGRADATION specifies parameters for MOS device degradation modeling.
Syntax
DEGRADATION <params>

Parameter Type Default Units

F.NTA Character
F.NTD Character
F.SIGMAE Character
F.SIGMAH Character
KC.AE0 Real 1.5 eV
KC.AMBIENTH Real 1.0 cm-3
KC.BETA Real 0.0
KC.BETAPERP Real 0.0 cm/V
KC.BETAPARL Real 0.0 cm/V
KC.COUPLED Logical False
KC.DELTPARL Real 0.0 [Q=1] cm
KC.DELTPERP Real 0.0 [Q=1] cm
KC.E.HCCOEF Real 0.0 cm2/A
KC.E.FNCOEF Real 0.0 cm2/A
KC.E.RHOHC Real 1.0
KC.E.RHOFN Real 1.0
KC.H.HCCOEF Real 0.0 cm2/A
KC.H.FNCOEF Real 0.0 cm2/A
KC.H.RHOHC Real 1.0
KC.H.RHOFN Real 1.0
KC.KF0 Real 10-5 s
KC.KR0 Real 3.010-9 s
KC.NIT0 Real 0.0 cm2
KC.RHOPARL Real 1.0

996 ATLAS User’s Manual

DEGRADATION Statements

Parameter Type Default Units

KC.RHOPERP Real 1.0

KC.SIHTOT Real 1012 cm2
KC.T0 Real 300.0 Kelvin
KC.VOLUME Real 0.0 cm3
NTA Real 1.01011 cm-2
NTD Real 1.01010 cm-2
PL.E.ALPHA Real 0.5
PL.E.EA Real 0.25 eV
PL.E.NU0 Real 1.0 s-1
PL.H.ALPHA Real 0.5
PL.H.EA Real 0.25 eV
PL.H.NU0 Real 1.0 s-1
RD.AE Real 1.5 eV
RD.AESLOPE Real -5.610-8 [Q=1] cm
RD.AEVAR Real 0.1 eV
RD.COUPLED Logical True
RD.E.HCCOEF Real 0.0 cm2/A
RD.E.FNCOEF Real 0.0 cm2/A
RD.E.OFFSET Real 0.0 eV
RD.H.HCCOEF Real 0.0 cm2/A
RD.H.FNCOEF Real 0.0 cm2/A
RD.H.OFFSET Real 0.0 eV
RD.INVHCOEF Real 0.0 cm3
RD.KF0 Real 10-5 s-1
RD.KR0 Real 310-9 cm3s-1
RD.NIT0 Real 0.0 cm2

997 ATLAS User’s Manual

DEGRADATION Statements

Parameter Type Default Units

RD.SIHTOT Real 1012 cm2

SIGMAE Real 1.01017 cm-2
SIGMAH Real 1.01017 cm-2

Description

F.NTA Specifies the file name for a C-Interpreter function that specifies arbitrary
density distribution of the acceptor-like traps on the interface.
F.NTD Specifies the file name for a C-Interpreter function that specifies arbitrary
density distribution of the donor-like traps on the interface.
F.SIGMAE Specifies the name of a file containing a C-Interpreter function specifying the
distribution of acceptor trap cross-sections.
F.SIGMAH Specifies the name of a file containing a C-Interpreter function specifying the
distribution of donor trap cross-sections.
NTA Specifies the uniform acceptor-like trap density on the interface.
NTD Specifies the uniform donor-like trap density on the interface.
SIGMAE Specifies the acceptor-like trap capture cross section.
SIGMAH Specifies the donor-like trap capture cross section.

Kinetic Model

KC.AE0 Activation energy.

KC.AMBIENTH Scaling hydrogen density in repassivation model.
KC.BETA Energy parameter for kinetic model.
KC.BETAPERP Energy scaling parameter.
KC.BETAPARL Energy scaling parameter.
KC.COUPLED Enables strong implicit coupling scheme between equations.
KC.DELTPARL Energy scaling parameter.
KC.DELTPERP Energy scaling parameter.
KC.E.HCCOEF Hot electron current scaling factor.
KC.E.FNCOEF Tunnel electron current scaling factor.
KC.E.RHOHC Hot electron current scaling exponent.

998 ATLAS User’s Manual

DEGRADATION Statements

KC.E.RHOFN Tunnel electron current scaling exponent.

KC.H.HCCOEF Hot hole current scaling factor.
KC.H.FNCOEF Tunnel hole current scaling factor.
KC.H.RHOHC Hot hole current scaling exponent.
KC.H.RHOFN Tunnel hole current scaling exponent.
KC.COUPLED Enable strong coupling in kinetic model.
KC.KF0 Depassivation rate constant.
KC.KR0 Repassivation rate constant.
KC.NIT0 Initial density of depassivated Si-H bonds.
KC.RHOPARL Energy scaling exponent.
KC.RHOPERP Energy scaling exponent.
KC.SIHTOT Total passivated and depassivated dangling bond density.
KC.T0 Effective temperature.
KC.VOLUME Repassivation volume.

Reaction-Diffusion Model
RD.AE Median activation energy for depassivation process.
RD.AESLOPE Coefficient for field dependent contribution to activation energy.
RD.AEVAR Energy width of depassivation activation energy distribution.
RD.COUPLED Enables strong implicit coupling scheme between equations.
RD.E.HCCOEF Coefficient for electron hot channel current.
RD.E.FNCOEF Coefficient for electron Fowler-Nordheim current.
RD.E.OFFSET Energy offset for electron hot channel current.
RD.H.HCCOEF Coefficient for hole hot channel current.
RD.H.FNCOEF Coefficient for hole Fowler-Nordheim current.
RD.H.OFFSET Energy offset for hole hot channel current. \\
RD.INVHCOEF Coefficient of inversion hole contribution to depassivation rate.
RD.KF0 Depassivation rate constant.
RD.KR0 Repassivation rate constant.
RD.NIT0 Initial density of depassivated Si-H bonds.
RD.SIHTOT Total passivated and depassivated dangling bond density.

999 ATLAS User’s Manual

DEGRADATION Statements

Power-Law Model
PL.E.ALPHA Exponent for power law dependence of degradation lifetime of
electron traps.
PL.E.EA Activation energy for degradation rate.
PL.E.NU0 Rate constant for degradation rate.
PL.H.ALPHA Exponent for power law dependence of degradation lifetime of hole
traps.
PL.H.EA Activation energy for degradation rate.
PL.H.NU0 Rate constant for degradation rate.

MOS Interface State Example

DEGRADATION NTA=1.E-12 SIGMAE=5.E-18
This syntax defines a density of acceptor states uniformly distributed along the silicon-oxide
interface. The trapping cross section is also defined. Traps will be filled by gate current in
transient mode simulations leading to a shift in device parameters.

1000 ATLAS User’s Manual

DEVDEGBULKTRAP Statements

21.12 DEVDEGBULKTRAP
DEVDEGBULKTRAP specifies parameters for traps created by Stress Induced Leakage Current.
Syntax
DEVDEGBULKTRAP <params>

Parameter Type Default Units

ELEC.DEPTH Real -999.0 eV

HOLE.DEPTH Real -999.0 eV
NT.N Real 0.0 cm-3
NT.P Real 0.0 cm-3
SIGMAN.P Real 1.0e-15 cm2
SIGMAP.N Real 1.0e-14 cm2
SIGMAT.N Real 1.0e-16 cm2
SIGMAT.P Real 1.0e-14 cm2
TAU.N Real 1.0e300 s
TAU.P Real 1.0e300 s

Description
ELEC.DEPTH Energy below conduction band of acceptor traps.
HOLE.DEPTH Energy above valence band of donor traps.
NT.N Sets the uniform density of electron traps (acceptor-like) in the gate insulator.
NT.P Sets the uniform density of hole traps (donor-like) in the gate insulator.
SIGMAN.P Coefficient of term giving capture rate of valence band holes into acceptor-like
trap.
SIGMAP.N Coefficient of term giving capture rate of conduction band electrons into
donor-like trap.
SIGMAT.N Coefficient of term giving capture rate of conduction band electrons into
acceptor-like trap.
SIGMAT.P Coefficient of term giving capture rate of valence band electrons into donor-
like trap.
TAU.N Lifetime for electron de-trapping from acceptor-like traps to conduction band.
TAU.P Lifetime for hole de-trapping from donor-like traps to valence band.

1001 ATLAS User’s Manual

DEVDEGBULKTRAP Statements

See Section 3.6.8 “Device Level Reliability Modeling” for further description of using this
statement.
Example
DEVDEGBULKTRAP NT.P=1.0E16 NT.N=1.0E17 TAU.N=10000.0 TAU.P=10.0 \
SIGMAT.N=1.0E-15 \
SIGMAT.P=1.0E-15 ELEC.DEPTH=1.5 HOLE.DEPTH=1.5

1002 ATLAS User’s Manual

OUTFILE Specifies the name of an output file for use with REGRID. The first DOPING
statement should use this parameter to specify a filename. All doping information from
the first DOPING statement and all subsequent DOPING statements in the input file are
saved to this file. The REGRID statement can read this file and interpolate doping on the
new grid.

Note: The file from OUTFILE cannot be used in TONYPLOT or in the MESH statement.The SAVE command
should be used after all of the DOPING commands required to save a file for plotting the doping profile.

Statement Applicability Parameters

DEVICE Specifies which device the statement applies to in the MIXEDMODE simulation. The
synonym for this parameter is STRUCTURE.
MATERIAL Restricts the applicability of the statement to regions of the specified material.
NAME Restricts the applicability of the statement to regions with the specified name.
REGION Restricts the applicability of the statement to regions with the specified region number.
STRUCTURE This is a synonym for DEVICE.

Note: If you don’t specify the DEVICE, MATERIAL, NAME and REGION parameters, the DOPING statement
will apply to all regions.

1008 ATLAS User’s Manual

DOPING Statements

Interface Doping Profile Location Parameters

You can use these parameters with the parameters NAME, MATER, and REGION to describe
interfaces between regions to be doped with an analytic profile. The NAME, MATER, and
REGION parameters are used to describe the regions that receive the doping while the INAME,
IMATER, and IREGION parameters describe regions incident on the regions receiving the
doping where the doping is to be placed (i.e., along the interface).
The locations of of these interface profiles may also be restricted by the location parameters
X.MAX, X.MIN, Y.MAX, and Y.MIN.

IMATER Specifies the material of all neighboring regions where an interface profile is
placed.
INAME Specifies the name of a neighboring region where an interface profile is placed.
IREGION Specifies the region number of the a neighboring region where the interface
profile is placed.
OUTSIDE Specifies that interface regions will placed at all interfaces with the outside of the
simulation domain.

Analytical Profile Types

These parameters specify how ATLAS will generate a doping profile from analytical
functions.

DOP.SIGMA Specifies the variance for random gaussian dopant distribution.

DOP.SEED Specifies a seed value for random gaussian dopant distribution.
ERFC Specifies the use of a ERFC analytical function to generate the doping
profile. If ERFC is specified, the following parameters must also be specified:
• Polarity parameters N.TYPE or P.TYPE
• One of the following groups of profile specifications:
• Group 1:CONCENTRATION and JUNCTION
• Group 2:DOSE and CHARACTERISTIC
• Group 3:CONCENTRATION and CHARACTERISTIC
GAUSSIAN Specifies the use of a gaussian analytical function to generate the doping profile.
If GAUSSIAN is specified, the following parameters must also be specified:
• Polarity parameters N.TYPE or P.TYPE
• One of the following groups of profile specifications:
• Group 1:CONCENTRATION and JUNCTION
• Group 2:DOSE and CHARACTERISTIC
• Group 3:CONCENTRATION and CHARACTERISTIC
UNIFORM Specifies the use of uniform (constant) analytical functions to generate the
doping profile. If UNIFORM is specified, the N.TYPE, P.TYPE, and
CONCENTRATION parameters must be specified. Doping is introduced into a box
defined by the boundary parameters (see the “Boundary Conditions” on page
974). The box by default includes the entire region.

1009 ATLAS User’s Manual

DOPING Statements

F.COMPOSIT Specifies the name of a file containing a C-Interpreter function specifying the
spatial distribution of composition fractions.
F.DOPING Specifies the name of a file containing a C-Interpreter function specifying the
spatial distribution of dopants.
F3.DOPING Specifies the name of a file containing a C-Interpreter function specifying the
spatial distribution of dopants for a 3D device.

1010 ATLAS User’s Manual

DOPING Statements

File Import Profile Types

These parameters specify how ATLAS will generate a doping profile from a file. Files can be
user-defined or from process simulation.

1D.PROC Specifies is an alias for TMA.SUPREM3.

2D.ASCII Specifies that a 2D doping profile, which is defined on a rectangular Cartessian
grid, should be loaded from a file specified by INFILE. 2D.ASCII must be specified
along with either the N.TYPE, P.TYPE or NET parameters. This first column of the
file should contain the X coordinates. The second column should contain the Y
coordinates. The third column should contain the doping data.
ASCII There has two separate meanings. The first meaning is that it specifies the file type
as ASCII when it’s combined with other format parameters. The second meaning is
when this parameter is used alone, it specifies ASCII data files containing
concentration versus depth information. The alias for this parameter is 1D.PROC.

In the second meaning, this parameter must be written in the form:

ASCII INFILE=<filename>
where filename is the name of the ASCII input file. The data file must be in the
following format:
depth concentration
depth concentration
depth concentration
...

where depth is specified in m and concentration is specified in cm-3. An input

file name, a dopant type, and boundary parameters must be specified. Positive
concentrations are assumed to be n-type and negative concentrations are assumed to
be p-type unless the N.TYPE or P.TYPE parameters are used.
ATHENA.1D Specifies that the doping file is a ATHENA 1D export file. This parameter acts in a
similar way to the SSUPREM3 parameter.
ATHENA Reads 2D doping information from ATHENA standard structure file (SSF) or
PISCES-II format files. The PISCES-II format is an obsolete file format.
Doping information obtained from this file will be added to each point of the
current ATLAS mesh. If points in the ATLAS mesh do not coincide with points in
the ATHENA mesh, doping for ATLAS mesh points will be interpolated from
ATHENA doping information. If this profile type is used, the INFILE parameter
must also be specified.

Note: The X.STRETCH function available in previous versions of ATLAS has been replaced by similar more
powerful functions in DEVEDIT. This feature should no longer be used in ATLAS.

1011 ATLAS User’s Manual

DOPING Statements

FILE This is an alias for INFILE.

IN.FILE This is a synonym for INFILE. The alias for this parameter is FILE.
INFILE Specifies the name of the appropriate input file. The synonym for this
parameter is IN.FILE.
MASTER Specifies that the INFILE is written in the Silvaco standard structure file (SSF)
format. This file format is the default output format of ATHENA and
SSUPREM3. This parameter is typically combined with the SSUPREM3,
ATHENA 1D, or ATHENA parameters. If neither of these are used the default is
SUPREM3.
N.COLUMN Specifies which column of an 2D ASCII table corresponds to the net donor
concentration when 2D.ASCII is specified.
P.COLUMN Specifies which column of an 2D ASCII table corresponds to the net acceptor
concentration when 2D.ASCII is specified.
SUPREM3 Specifies the INFILE was produced by SSUPREM3 in standard structure file
(SSF) format or binary or an ASCII export format. If this profile type is used, an
input file name, a dopant, and boundary parameters must be specified. When
SSUPREM3 produces an output file, the doping profiles are stored by dopant.
Therefore, a dopant parameter should be specified in order to import the correct
doping profile into ATLAS. If a specific dopant is not specified the total donors
and acceptor concentrations are loaded.
TMA.SUPREM3 Specifies that the file specified by INFILE is in TMA SUPREM3 binary
format. The alias for this parameter is 1D.PROC.

Note: Files containing 1D doping profiles can be loaded into BLAZE, BLAZE3D, DEVICE3D, or S-PISCES. Files
containing 2D doping profiles can only be loaded into S-PISCES.

X.COLUMN Specifies which column of an 2D ASCII table corresponds to the X coordinate value
when 2D.ASCII is specified.
Y.COLUMN Specifies which column of an 2D ASCII table corresponds to the Y coordinate value
when 2D.ASCII is specified.

2
N  Y  = PEAK · exp –  ---------------------
Y
21-3
 Y CHAR
·

Location Parameters
DIRECTION Specifies the axis along which a one-dimensional profile is directed in a two-
dimensional device (x or y). DIR=y will typically be used for implanted
profiles.
REGION Specifies the region number where doping is to be added.
START Specifies the depth in the Y direction where the profile should start.

Lateral Extent Parameters

These parameters must be specified when a 1D doping profile type is used (MASTER,
GAUSSIAN, ASCII, ERFC, or UNIFORM). These boundary parameters set the doping boundaries
before applying lateral spreading. This is equivalent to setting implant mask edges.

WIDTH Specifies the extent of the profile in the X direction. Specifying WIDTH is
equivalent to specifying X.MAX such that X.MAX=X.MIN+WIDTH.
R.MIN, R.MAX, A.MIN, For an ATLAS3D device created using the MESH CYLINDRICAL option,
and A.MAX these parameters restrict the radial and angular positions respectively of
the analytical doping profile. The principal direction of the analytical
doping profile is the Z direction in this case.
X.MIN, X.MAX, Y.MIN, Specify the x, y and z bounds of a rectangular shaped region or box in
Y.MAX, Z.MAX, and Z.MIN the device. The dopant profile within this box will be constant with a
density equal to the value specified by the CONC parameter. Outside this
box the profile decreases from the peak, CONC, with distance, from the
box along the principal axes. The relationship between the
concentration, outside the box, to distance will depend upon the
profile type as specified by the GAUSSIAN, MASTER, ATHENA, ATLAS, and
UNIFORM parameters.

X.LEFT, X.MIN Specifies the left boundary of a vertical 1D profile.

1015 ATLAS User’s Manual

1018 ATLAS User’s Manual

DOPING Statements

DOP REGION=1 MASTER ATHENA.1D BORON RATIO.LAT=0.7 INF=bipolar.exp

# EMITTER
DOP REGION=1 MASTER ATHENA.1D ARSENIC RATIO.LAT=0.6 \
INF=bipolar.exp X.LEFT=12.0 X.RIGHT=13.0 DOPOFF=1e16
Athena Doping Interface Example
This example demonstrates how to use an SSF format ATHENA file to interpolate doping
onto a ATLAS grid and save the doping information for subsequent regrid operations.This is
an alternative to the ATHENA/ATLAS interface, which is described in Section 2.6.1
“Interface From ATHENA”.
DOPING ATHENA MASTER INFILE=NMOS.DOP OUTFILE=NMOS.DOP
REGRID DOPING ABS LOG RATIO=4 OUTFILE=NMESH1.STR
DOPFILE=NMOS.DOP
3D Doping Definition Example
The following example illustrates the formation of a Gaussian highly doped n-type area in a
three-dimensional structure.
DOPING GAUSS N.TYPE CONC=1e20 PEAK=0.0 CHAR=0.2 X.LEFT=0.5 \
X.RIGHT=1.0 Z.LEFT=0.5 Z.RIGHT=1.0
For a cylindrical structure you can use the following syntax. In this case, it is for
complementary error function doping.
doping erfc start=0.0 char=0.005 ratio.lat=0.1 conc=1.0e20 n.type
r.min=0.01 a.min=45.0 a.max=225.0

1019 ATLAS User’s Manual

DOPING Statements

3D Doping From ASCII 1D File

You can read a 1D doping profile from an ASCII file and apply it to the entire or part of a 3D
device. The X.DIR, Y.DIR, or Z.DIR parameters specify the direction of the profile as
applied in the 3D device. The starting position of the doping profile will be set by the
relevant X.MIN, Y.MIN, or Z.MIN parameter. The ending position of the doping profile will
be the minimum of the relevant X.MIN, Y.MIN, or Z.MIN plus the spatial extent of the doping
profile in the ASCII file, or the X.MAX, Y.MAX, Z.MAX parameter or a physical boundary of the
device. You can specify a general parallelogram in the plane perpendicular to the direction of
doping direction. To specify it, use the appropriate combination of X1, X2, Y1, Y2, Z1, Z2,
X.MIN, X.MAX, Y.MIN, Y.MAX, Z.MIN, Z.MAX. For example, we consider the Y.DIR parameter
specified. We can then specify a parallelogram in the XZ plane as shown in Figure 21-2.

x
Z.MIN Z.MAX

X.MIN
X1=X2

X.MAX

Z1 Z2 z

Figure 21-2: Parallelogram Geometry Example

In the figure, two sides are parallel to the Z axis. Z1 and Z2 specify the Z coordinate of the
midpoints of these two sides. Z.MIN and Z.MAX together specify the lengths of the sides, and
Z.MIN will be negative and equal to -Z.MAX. (In fact, the positions are evaluated as
Z1+Z.MIN and Z1+Z.MAX for one side and Z2+Z.MIN and Z2+Z.MAX for the other. Therefore,
having Z1 and Z2 and so on refer to the side midpoint is just a convenient convention.)
You must specify X1 and X2 as having the same value, and specify X.MIN and X.MAX the side
lengths the same way as for Z.MIN and Z.MAX.
Setting Z1 and Z2 to the same value will result in a rectangle. By then changing X1 and X2 so
they are not equal will result in a parallelogram with sides parallel to the X axis. If a different
direction of doping variation is defined, then a parallelogram can be set up in the plane
perpendicular to it in analogously to the above example.
Outside the specified parallelogram the doping level will be either the value of the last point
in the ASCII file, or the value specified by the BACKDOPE parameter. A smooth transition of
the doping from the value inside the parallelogram to the value outside can be applied
elsewhere using the LAT.CHAR or RATIO.LAT parameters.

1020 ATLAS User’s Manual

DOPING Statements

X1, X2, Y1, Y2, Z1, Z2 By convention, specify the midpoints of the sides making up the
parallelogram.
X.MIN, X.MAX Determine the X coordinate of the start and end points of the
placement of the doping profile read in from an ASCII file (if you
specify X.DIR). Otherwise, they specify the lateral extent of the
parallelogram sides defining the location of the doping in a plane
perpendicular to the specified direction of the doping variation. In
this case, they are added to X1 and to X2 to determine the
coordinates of the ends of sides.
Y.MIN, Y.MAX, Z.MIN, Z.MAX These are the same as X.MIN and X.MAX except they apply to the y
and Z directions respectively.

Example:
doping ascii inf=ydop.dat y.dir n.type x1=6.0 x2=4.0 x.min=-2
x.max=2
z.min=3.0 z.max=7.0 y.min=2.0 y.max=10.0 lat.char=0.005
backdope=0.0
This will apply the doping profile specified in ydop.dat as n-type doping in the Y direction
between the positions y=2 microns and y=10 microns (unless the length range in ydop.dat is
less than this distance). In the XZ plane, it will be applied to a parallelogram with sides
parallel to the X direction of length 4 microns. The Z coordinates of these sides are 3.0
microns and 7.0 microns. The doping will transition from the value inside the parallelogram
to 0.0 outside on a length scale of 0.005 microns.
3D Doping From 2D ATHENA Master Files
In a 2D ATHENA master file, the doping profile is defined over a box in two dimensions.
This doping is put over 2D-sections normal to the plane of a parallelogram to be defined in
the doping statement. This parallelogram must lie in one of the XY, YZ, and XZ planes and
have two edges parallel to one of the coordinate axes.
As the mesh in the ATHENA master file doesn’t necessarily coincide with the three-
dimensional mesh in ATLAS, an interpolation routine is required to import the doping in
ATLAS. You can then choose between a linear and a logarithmic interpolation algorithm.
Here’s a list of the parameters that are used for this kind of doping.

ASPECT.RATIO Specifies the aspect ratio of the doping will be preserved when loaded into the
doping parallelogram.
DOP.XMAX Specifies the maximum X coordinate of the doping in the ATHENA file to be
loaded into the doping parallelogram.
DOP.XMIN Specifies the minimum X coordinate of the doping in the ATHENA file to be
loaded into the doping parallelogram.
DOP.YMAX Specifies the maximum Y coordinate of the doping in the ATHENA file to be
loaded into the doping parallelogram.

1021 ATLAS User’s Manual

DOPING Statements

DOP.YMIN Specifies the minimum Y coordinate of the doping in the ATHENA file to be
loaded into the doping parallelogram.
INT.LIN Specifies that a linear interpolation is to be used to import doping from
ATHENA into ATLAS.
INT.LOG Specifies that the ATHENA doping is imported in ATLAS by using a
logarithmic interpolation algorithm.
INT.OPTM Enables an optimized interpolation routine, which attempts to reduce CPU
time due to the doping interpolation.
LAT.CHAR Defines the characteristic length, CL, of the lateral spreading (in the y>y2 and
y<y1 planes). If RATIO.LAT is used instead, then the characteristic length is
assumed to come from this parameter by the height of the parallelogram (i.e.,
CL=RL X OCL, OCL being the height of the parallelogram (y2-y1 in this case). If
both LAT.CHAR and RATIO.LAT aren’t specified, then no lateral spreading is
done.
XY Specifies that the parallelogram containing doping in ATLAS3D is in the xy
plane.
YX This is a synonym for XY.
YZ Specifies that a parallelogram containing doping in ATLAS3D is in the YZ
plane.
ZY This is a synonym for YZ.
XZ Specifies that a parallelogram containing doping in ATLAS3D is in the XZ
plane.
ZX This is a synonym for ZX.
X1, X2, Y1, Y2, Z1, Z2 Specify an average segment in one of the three coordinate planes defining the
orientation of the parallelogram.
X.DIR, Y.DIR, Z.DIR Specify that a parallelogram containing doping in ATLAS3D has two edges in
the X direction, Y direction, and Z direction.
X.FLIP Specifies that the orientation of the ATHENA file X axis will be flipped.
X.MIN, X.MAX Define the minimum and maximum lateral extent of a parallelogram along the
X direction lying in the xy or zx plane, starting from its average segment (they
must be specified with X1 and X2). If a parallelogram is defined in the YZ
plane, X.MIN can then be used to specify the initial coordinate to start the
doping along the X direction.
X.SCALE Specifies that the ATHENA file doping will be scaled in the X direction.
Y.FLIP Specifies that the orientation of the ATHENA file Y axis will be flipped.

1022 ATLAS User’s Manual

DOPING Statements

Y.MIN, Y.MAX Define the minimum and maximum lateral extent of a parallelogram along the
Y direction lying in the XY or YZ plane, starting from its average segment
(they must be specified with Y1 and Y2). If a parallelogram is defined in the zx
plane, Y.MIN can then be used to specify the initial coordinate to start the
doping along the Y direction.
Y.SCALE Specifies that the ATHENA file doping will be scaled in the Y direction.
Z.MIN, Z.MAX Define the minimum and maximum lateral extent of a parallelogram along the
Z direction lying in the yz or zx plane, starting from its average segment (they
must be specified with Z1 and Z2). If a parallelogram is defined in the xy
plane, Z.MIN can then be used to specify the initial coordinate to start the
doping along the Z direction.
Z.SCALE Specifies that the ATHENA file doping will be scaled in the Z direction.

The following three cases, which correspond to the parallelograms in the xy, yz and zx planes,
describe how the doping from the ATHENA master file is put into the ATLAS structure .
First Case: Parallelogram In The XY Plane
SUB-CASE A: Parallelogram along the X direction.
Sections of the ATLAS structure in zx planes are considered, which intersect for y1<y<y2,
the segment (xa, xb) inside the parallelogram. Each of these sections defines a box where the
2D doping is imported from ATHENA. Particularly, all the coordinates in ATHENA are scaled
and translated so that the edge (xc, xd) of the ATHENA box containing the doping coincide
with the segment (xa, xb) of the ATLAS box. This is done for all the sections inside the
parallelogram. See Figure 21-3.

Figure 21-3: Parallelogram in the XY plane in the X direction and doping from the ATHENA2D master file
Examples:
doping athena master inf=athena.str boron xy \
x1=2.0 x2=3.0 y1=1.5 y2=4.0 x.dir \
x.min=-2.0 x.max=1.0 z.min=0.6 lat.char=0.05 int.log int.optm
For the Z coordinate, z.min is used to specify the minimum value where to start putting the
doping into the zx-sections (which properly defines the segment (xa, xb) in this plane).

1023 ATLAS User’s Manual

DOPING Statements

In addition, a lateral spreading can be partially accomplished by extending the parallelogram

into the planes, y>y2 and y<y1, where the doping is spread out according to a Gaussian law.
SUB-CASE B: Parallelogram along the Y direction
Sections of the ATLAS structure are considered in the YZ planes. A transformation of
coordinates in ATHENA is accomplished in order to place the ATHENA segment (yc, yd) into
the ATLAS one (ya, yb). See Figure 21-4.

Figure 21-4: Parallelogram in the XY plane in the Y direction and doping from the ATHENA2D master file
Examples:
doping athena master inf=athena.str boron xy \
x1=2.0 x2=3.0 y1=1.5 y2=4.0 y.dir \
y.min=-2.0 y.max=1.0 z.min=0.6 lat.char=0.05 int.lin int.optm

Second Case: YZ Plane

The YZ plane containing the parallelogram is defined by the yz parameter (or zy) in the
DOPING statement.
If parallelograms along the Z direction are defined (specify Z.DIR in doping statement), the
doping from ATHENA is put over sections in the zx planes. If the parallelograms along the Y
direction are defined (using Y.DIR in DOPING statement), the doping from ATHENA is put
over sections in the xy planes.
The minimum X coordinate in ATLAS to start adding doping can be specified by X.MIN in
DOPING statement.

Examples:
doping athena master inf=athena.str boron zy z.dir \
z1=0.3 z2=0.6 y1=0.35 y2=0.25 z.min=-0.2 z.max=0.6 \
x.min=0.2 ratio.lat=0.05

doping athena master inf=athena.str boron zy y.dir \

z1=0.3 z2=0.6 y1=0.35 y2=0.25 y.min=-0.2 y.max=0.6 \
x.min=0.2 ratio.lat=0.05

1024 ATLAS User’s Manual

DOPING Statements

Third Case: ZX Plane

The ZX plane containing the parallelogram is defined by the zx parameter (or xz) in DOPING
statement. If parallelograms along the Z direction are defined (using Z.DIR in the DOPING
statement), the doping from ATHENA is placed over sections in the YZ planes.
If the parallelograms along the X direction are defined (using X.DIR in DOPING statement),
the doping from ATHENA is placed over the sections in the XY planes.
The minimum Y coordinate in ATLAS to start adding doping can be specified by Y.MIN in
doping statement.
Examples:
doping athena master inf=athena.str boron int.lin \
xz z.dir z1=0.45 z2=0.55 x1=0.2 x2=0.7 x.min=-0.4 \
x.max=0.2 y.min=0.0 lat.char=0.02
doping athena master inf=athena.str boron int.lin \
xz x.dir z1=0.4 z2=0.8 x1=0.2 x2=0.7 z.min=-0.4 \
z.max=0.2 y.min=0.5 ratio.lat=0.5

1025 ATLAS User’s Manual

DOSEXTRACT Statements

21.14 DOSEXTRACT
DOSEXTRACT specifies the grain boundary and interface trap density of states as a function of
energy [137] are to be extracted from IV and CV files and saved to log files.
Syntax
DOSEXTRACT [<parameters>]

Parameter Type Default Units

ACCEPTOR Logical False

CVFILE Character
DONOR Logical False
EG300 Real eV
EI Real eV
EMAX Real eV
GRAIN.SIZE Real m

GRAINFILE Character
INTFILE Character
IVFILE Character
LENGTH Real m

MIN.DENSITY Real 1 cm-2

MU Real cm2/Vs
NUM.DIVISIONS Real 25
NUM.GRAINS Real
OUTFILE Character
TEMPERATURE Real 300 K
TOX Real m

TSI Real m

VDRAIN Real V
WIDTH Real m

1026 ATLAS User’s Manual

DOSEXTRACT Statements

Description

ACCEPTOR Specifies that acceptor traps are to be extracted.

CVFILE Specifies the file name which contains the capacitance/gate voltage data.
DONOR Specifies that donor traps are to be extracted.
EG300 Specifies the value of bandgap of the material. EMAX will be calculated from this
value if it is specified along with EI.
EI Specifies the intrinsic energy level of the material.
EMAX Specifies the maximum energy seperation from the intrinsic energy level (EI) that
should be calculated.
GRAIN.SIZE Specifies the size of the grains.
GRAINFILE Specifies the file name where the grain boundary DOS distribution, as a function of
energy, will be stored.
INTFILE Specifies the file name where the interface DOS distribution, as a function of energy,
will be stored.
IVFILE Specifies the file name that contains the drain current/gate voltage data.
LENGTH Specifies the x-direction length under the gate.
MIN.DENSITY Specifies the minimum trap density to be used for the grain boundary traps.
MU Specifies the majority carrier mobility.
NUM.DIVISIONS Specifies the number of divisions used from the oxide-silicon interface to the
substrate.
NUM.GRAINS Specifies the number of grains in the device.
OUTFILE Specifies the file name where the grain boundary and interface DOS distribution, as a
function of energy, will be stored.
TEMPERATURE Specifies the temperature used in the DOS extraction.
TOX Specifies the oxide thickness.
TSI Specifies the silicon thickness.
VDRAIN Specifies the drain voltage used when the data in CVFILE and IVFILE were created.
WIDTH Specifies the width of the device.

Example
dosextract tox=0.1 tsi=0.05 width=25 length=5 vd=0.1 mu=50
num.grains=4 ei=0.54
eg=1.08 cvfile=cv.dat ivfile=iv.dat donor grainfile=grain.dat
intfile=int.dat

1027 ATLAS User’s Manual

ELECTRODE Statements

21.15 ELECTRODE
ELECTRODE specifies the locations and names of electrodes in a previously defined mesh.
Syntax
ELECTRODE NAME=<en> [NUMBER=<n>] [SUBSTRATE] <pos> <reg>

Parameter Type Default Units

A.MAX Real degrees

A.MIN Real degrees
BOTTOM Logical False
FLOATING Logical False
IX.HIGH Integer right side of structure
IX.LOW Integer left side of structure
IX.MAX Integer right side of structure
IX.MIN Integer left side of structure
IY.MAX Integer bottom side of structure
IY.MIN Integer top side of structure
IY.HIGH Integer bottom of structure
IY.LOW Integer top of structure
IZ.MAX Integer
IZ.MIN Integer
IZ.HIGH Integer
IZ.LOW Integer
LEFT Logical False
LENGTH Real length of structure m
MATERIAL Character Contact
MODIFY Logical False
NAME Character
NUMBER Integer defined #(electrodes)+ 1
R.MAX Real m
R.MIN Real m
RIGHT Logical False
SUBSTRATE Logical False

1028 ATLAS User’s Manual

ELECTRODE Statements

Parameter Type Default Units

THERMAL Logical False

TOP Logical False
X.MAX Real right side of structure m
X.MIN Real left side of structure m
Y.MAX Real Y.MIN m
Y.MIN Real top of the structure m
Z.MIN Real m
Z.MAX Real m

1029 ATLAS User’s Manual

ELECTRODE Statements

Description

FLOATING Marks an enclosed electrode as a floating semiconductor region. See Section 3.5.10
“Floating Semiconductor Regions”.
MATERIAL Specifies a material for the electrode (see Table B-1). This material will be displayed
in TONYPLOT. The electrode material can also be used to define the electrode thermal
characteristics (thermal conductivity) and optical characteristics (complex index of
refraction). Setting the material here does not apply any electrical property such as
workfunction to the terminal. All electrical properties of electrodes are set on the
CONTACT statement.

MODIFY To mark an electrode read from a Standard Structure File as a floating semiconductor
region specify both MODIFY and FLOATING flags. See Section 3.5.10 “Floating
Semiconductor Regions”.
NAME Specifies an electrode name. The electrode name can be referenced by other ATLAS
statements to modify characteristics of the specified electrode. For reference by the
CONTACT or THERMALCONTACT statements any valid character string can be used and
properly cross-referenced. But when setting voltages, currents and charge from the
SOLVE statement certain electrode names are recognized in a simplified syntax. By
prepending the electrode name with "V" for voltage, "I" for current and "Q" for
charge, you can directly and conveniently set the electrode bias, current or charge
respectively. For example:
SOLVE VGATE=1.0
can be used to assign 1 volt bias to the electrode named "GATE". In such a
manner, the following list of names can be used to set voltage, current or charge.
• GATE
• FGATE
• CGATE
• NGATE
• PGATE
• VGG

The following list of names can be used to assign only voltage or current.
• DRAIN
• SOURCE
• BULK
• SUBSTRATE
• EMITTER
• COLLECTOR
• BASE
• ANODE
• CATHODE
• WELL

1030 ATLAS User’s Manual

ELECTRODE Statements

• NWELL
• PWELL
• CHANNEL
• GROUND
• NSOURCE
• PSOURCE
• NDRAIN
• PDRAIN
• VDD
• VSS
• VEE
• VBB
• VCC

NUMBER Specifies an electrode number from 1 to 50. Electrode numbers may be specified in
any order. If NUMBER is not specified, electrodes will be automatically numbered in
sequential order. This parameter cannot re-number electrodes already defined in
ATLAS or other programs.
pos This is one of the position parameters described below.
reg This is a set of the region parameters described on the next page.
SUBSTRATE Places the specified electrode at the bottom of the device and names the
electrode, substrate.
THERMAL Specifies that the electrode is treated as a boundary condition for heatflow
simulation using GIGA.

Position Parameters

BOTTOM or SUBSTRATE Specifies that the electrode is positioned along the bottom of the device.
LEFT Specifies that the electrode starts at the left-hand edge of the device. The
electrode will be positioned from left to right along the top of the device.
RIGHT Specifies that the electrode starts at the right-hand edge of the device. The
electrode will be positioned from right to left along the top of the device.
TOP Specifies that the electrode is positioned along the top of the device.

1031 ATLAS User’s Manual

ELECTRODE Statements

Region Parameters
Device coordinates may be used to add regions to both rectangular and irregular meshes. In
either case, boundaries must be specified with the A.MAX, A.MIN, R.MAX, R.MIN, X.MAX,
X.MIN, Y.MAX, Y.MIN, Z.MAX, and Z.MIN parameters.

LENGTH Specifies the length of the electrode in the X direction. It is not necessary to specify
X.MIN, X.MAX, and LENGTH. If two of these parameters are specified, the value of the
third parameter will be calculated.
A.MAX Specifies the maximum angle of a 3D cylindrical electrode.
A.MIN Specifies the minimum angle of a 3D cylindrical electrode.
R.MAX Specifies the maximum radius of a 3D cylindrical electrode.
R.MIN Specifies the minimum radius of a 3D cylindrical electrode.
X.MAX Specifies the maximum x-boundary of the electrode.
X.MIN Specifies the minimum x-boundary of the electrode.
Y.MAX Specifies the maximum y-boundary of the electrode.
Y.MIN Specifies the minimum y-boundary of the electrode.
Z.MIN Specifies the minimum z-boundary of the electrode.
Z.MAX Specifies the maximum z-boundary of the electrode.

Note: If an electrode has been shortened to fit the current mesh, a warning message will be generated by
ATLAS. Electrode placement can only occur at previously defined mesh nodes.

Grid Indices
As an alternative to the region parameters, you can use grid indices to define a region only
when the mesh is rectangular although these parameters are not recommended. To define a
region with a rectangular mesh:
1. Use the X.MESH and Y.MESH statements to specify grid indices.
2. Use the IX.HIGH, IX.LOW, IY.HIGH, and IY.LOW parameters to specify x and y values.

IX.HIGH Specifies the maximum x-value of the grid index. The alias for this parameter is
IX.MAX.

IX.LOW Specifies the minimum x-value of the grid index. The alias for this parameter is
IX.MIN

IY.HIGH Specifies the maximum y-value of the grid index. The alias for this parameter is
IY.MAX.

1032 ATLAS User’s Manual

ELECTRODE Statements

IY.LOW Specifies the minimum y-value of the grid index. The alias for this parameter is
IX.MIN.
Nodes, which have x and y grid indices, between IX.LOW and IX.HIGH and
between IY.LOW and IY.HIGH are designated electrode nodes. Normally,
horizontal planar electrodes will be used. In this case, IY.LOW equals IY.HIGH.
IZ.HIGH Specifies the maximum z-value of the grid index. The alias for this parameter is
IZ.MAX.

IZ.LOW Specifies the minimum z-value of the grid index. The alias for this parameter is
IX.MIN.

IX.MAX, IX.MIN, These are aliases for IX.HIGH, IX.LOW, IY.HIGH, IY.LOW, IZ.HIGH, and IZ.LOW.
IY.MIN, IZ.MAX,
IZ.MIN

MOS Electrode Definition Example

This example defines electrodes for a typical MOS structure.
ELEC X.MIN=0.5 LENGTH=0.25 NAME=gate
ELEC LENGTH=0.25 Y.MIN=0 LEFT NAME=source
ELEC LENGTH=0.25 Y.MIN=0 RIGHT NAME=drain
ELEC SUBSTRATE
3D Electrode Definition Example
The following example illustrates electrode definition for a 3D structure.
ELECTRODE NAME=ANODE X.MIN=0.5 X.MAX=1.0 \
Z.MIN=0.5 Z.MAX=1.0

Note: In ATLAS, it is preferred to refer to ELECTRODES by name rather than number. Some functions,
however, may require the electrode number. The syntax, MODELS PRINT, can be used to echo electrode
numbers to the run-time output.

1033 ATLAS User’s Manual

ELIMINATE Statements

21.16 ELIMINATE
ELIMINATE terminates mesh points along lines in a rectangular grid defined within ATLAS in
order to reduce the local mesh density.
Syntax
ELIMINATE X.DIRECTION|Y.DIRECTION [<boundary>]

Parameter Type Default Units

COLUMNS Logical False

IX.LOW Integer
IX.HIGH Integer
IY.LOW Integer
IY.HIGH Integer
ROWS Logical False
X.MIN Real m
X.MAX Real m
Y.MIN Real m
Y.MAX Real m

Description
The ELIMINATE statement is used to remove points along every other line within the chosen
range. Successive eliminations of the same range remove points along every fourth line. For
horizontal elimination, the vertical bounds should be decreased by one at each re-elimination
of the same region. For vertical elimination, the horizontal bounds should be decreased by
one at each re-elimination of the same region.

ROWS or X.DIR Eliminates points along horizontal lines.

COLUMNS or Y.DIR Eliminates points along vertical lines.

Boundary Parameters

X.MIN, X.MAX, Y.MIN, and Y.MAX Specify the location of the boundaries of an area in
coordinates, where the elimination is applied.

1034 ATLAS User’s Manual

ELIMINATE Statements

The following are provided for backward compatibility only. Their use is not recommended.

IX.HIGH Specifies the mesh line number high boundary in the X direction.
IX.LOW Specifies the mesh line number low boundary in the X direction.
IY.HIGH Specifies the mesh line number high boundary in the Y direction.
IY.LOW Specifies the mesh line number low boundary in the Y direction.

Substrate Mesh Reduction Example

This example removes vertical points between the depth of 10m and 20m.
ELIM Y.DIR Y.MIN=10 Y.MAX=20 X.MIN=1 X.MAX=8
ELIM Y.DIR Y.MIN=10 Y.MAX=20 X.MIN=1 X.MAX=7

Note: In some cases, applications of the ELIMINATE statement can cause internal inconsistencies in the
mesh. When this occurs, an error message will appear, warning you that there are triangles that are not
associated with any region.

Note: The ELIMINATE statement only works on meshes defined using ATLAS syntax. You can eliminate
mesh points on arbitrary meshes in DEVEDIT

1035 ATLAS User’s Manual

EXTRACT Statements

21.17 EXTRACT
EXTRACT statements are used to measure parameters from both log and solution files.

Note: These commands are executed by DECKBUILD. This statement is documented in the DECKBUILD
USER ’S MANUAL.

Terminal Current Extraction Example

By default, EXTRACT works on the currently open log file. For example, to extract peak drain
current from a run immediately after solution, type:
LOG OUTF=myfile.log
SOLVE .........
EXTRACT NAME=”peak Id” max(i.”drain”)
Extraction Example from Previously Generated Results
To extract the same data from a previously run simulation, use the INIT parameter.
EXTRACT INIT INFILE=”myfile.log”
XTRACT NAME=”peak Id” max(i.”drain”)
Solution Quantities Extraction Example
To use EXTRACT with solution files, use the INIT parameter. To find the integrated number
of electrons in a 1D slice at X=1.0, use:
SAVE OUTF=mysolve.str or SOLVE ...... MASTER OUTF=mysolve.str
EXTRACT INIT INFILE=”mysolve.str”
EXTRACT NAME=”integrated e-” area from curve(depth,n.conc \
material="Silicon" mat.occno=1 x.val=1.0)

Note: EXTRACT commands are generally case sensitive.

1036 ATLAS User’s Manual

EYE.DIAGRAM Statements

21.18 EYE.DIAGRAM
EYE.DIAGRAM specifies that an eye diagram should be generated from the specified log file.
An eye diagram is created by dividing up transient data in periods of fixed size and then
overlaying it.
Syntax
EYE.DIAGRAM INFILE OUTFILE PERIOD + [OPTIONAL PARAMETERS]

Parameter Type Default Units

INFILE Character
OUTFILE Character
PERIOD Real s
T.START Real s
T.STOP Real s

Description

INFILE Specifies the input log file. This should contain data from a transient simulation.
OUTFILE Specifies the file output file for the eye diagram.
PERIOD Specifies the window period.
T.START Specifies the inital time value to be used. The default valueis the first time point in the
input log file.
T.STOP Secifies the final time value to be used. The default value is the last time point in the
input log file.

Examples
EYE.DIAGRAM INFILE=laser.log OUTFILE=eye.log PERIOD=2e-10
T.START=1.5e-9
In this example, laser.log is a log file from the transient simulation of a buried
heterostructure laser when it is biased using a pseudo-random bit sequence. The data values
from time=1.5ns onwards are used to create the eye diagram with a data period of 0.2ns. The
results are then saved to the log file, eye.log. Figure 21-5 shows the eye diagram in
TONYPLOT.

1037 ATLAS User’s Manual

EYE.DIAGRAM Statements

Figure 21-5: Eye diagram for BH Laser diode with a psuedo-random bit sequence input

1038 ATLAS User’s Manual

FDX.MESH, FDY.MESH Statements

21.19 FDX.MESH, FDY.MESH

FD<n>.MESH specifies the location of grid lines along the <n>-axis in a rectangular mesh used
in FDTD simulation. The syntax is equivalent for X and Y directions.
Syntax
FDX.MESH NODE=<n> LOCATION=<n>
FDX.MESH SPACING=<n> LOCATION=<n>
FDY.MESH NODE=<n> LOCATION=<n>
FDY.MESH SPACING=<n> LOCATION=<n>

Parameter Type Default Units

LOCATION Real m
NODE Integer
SPACING Real m

Description
NODE Specifies the mesh line index. These mesh lines are assigned consecutively.
LOCATION Specifies the location of the grid line.
SPACING Specifies the mesh spacing at the mesh locations specified by the LOCATION
parameter. If the SPACING parameter is specified, then the NODE parameter should
not be specified.

FDTD Mesh Example

This syntax defines a mesh of 3333 covering the area bounded by (0.3,0.0) to (2.4,1.0).
FDX.M n=1 l=0.3
FDX.M n=33 l=2.4
FDY.M n=1 l=0.0
FDY.M n=33 l=1.0

Note: The mesh defined in these statements for the FDTD Solution is entirely separate from the electrical device
simulation mesh defined on the MESH statement. The FDTD Mesh is defined in the coordinate system
relative to the direction of beam propagation (see Figure 10-1).

Setting Locally Fine Grids Example

This example shows how to space grid lines closely around a topological feature such as a
junction at y=0.85 microns.
LY.MESH LOC=0.0 SPAC=0.2
LY.MESH LOC=0.85 SPAC=0.01
LY.MESH LOC=2 SPAC=0.35

1039 ATLAS User’s Manual

FOURIER Statements

21.20 FOURIER
FOURIER enables you to do Fourier transformations.
Syntax
FOURIER INFILE OUTFILE + [ OPTIONAL PARAMETERS ]

Parameter Type Default Units

COMPLEX.VALUES Logical False

FUNDAMENTAL Real Hz
INFILE Character
INTERPOLATE Logical False
MAX.HARMONIC Real Hz
NO.DC Logical False
NUM.SAMPLES Integer 64
OUTFILE Character
T.START Real s
T.STOP Real s

Description
The FOURIER statement performs a Fast Fourier Transform on log file data.

COMPLEX.VALUES Specifies that the real and imaginary components are saved to file as well as the
magnitude and phase values The synonym for this parameter is REAL.VALUES.
FUNDAMENTAL Specifies the fundamental frequency. If the fundamental frequency is
specified, T.STOP is set to T.START+1/FUNDAMENTAL. If this is not specified,
the fundamental frequency is set to (T.STOP-T.START)/NUM.SAMPLES.
INFILE Specifies the input log file. This should contain data from a transient simulation.
INTERPOLATE Specifies that the input data should be linearly interpolated such that data at
uniform time steps are created. Interpolation of data can introduce addition
(inaccurate) harmonic values into the Fourier transform. INTERPOLATE must be
used if the log file contains non-uniform time steps.
MAX.HARMONIC Specifies the maximum harmonic frequency that the Fourier transform should
calculate. This will automatically calculate the correct number of samples
(NUM.SAMPLES) required to generate this frequency. FUNDAMENTAL must be
specified when MAX.HARMONIC is used.
NO.DC Specifies that the DC component will not be saved to the log file.

1040 ATLAS User’s Manual

FOURIER Statements

NUM.SAMPLES Specifies the number of discrete samples. This should be an integer power of 2
(i.e., 2n where n is a positive integer). The default value is 64 unless the
MAX.HARMONIC parameter is specified. In this case the number of samples is set
to the nearest integer power of 2 which will generate this frequency.
OUTFILE Specifies the output file for the Fourier transform data.
T.START Specifies the start of time data to be used for the Fourier transform. The default
value is the first time point in the input log file.
T.STOP Specifies the end of time data to be used for the Fourier transform. The default
value is the last time point in the input log file.

Example 1
In this example, the transient data previously written to log file, hemt1.log, is transformed
from the time domain to the frequency domain. The fundamental frequency is set to 0.5 GHz,
and harmonic frequencies up to 16 GHz are calculated. Since the data in hemt1.log has non-
uniform time steps, the INTERPOLATE flag must be enabled. The complex values as well as
the magnitude and phase values are stored in fftout1.log.
FOURIER INFILE=hemt1.log FUNDAMENTAL=5e8 MAX.HARMONIC=1.6E10 \
OUTFILE=fftout1.log INTERPOLATE COMPLEX.VALUES
Example 2
In this example, the log file values between 31.25 ps and 2ns are transformed into the
frequency domain. The fundamental frequency is automatically determined from the time
period set by T.START and T.STOP. The data values from this time period are interpolated
into 64 samples, giving a maximum harmonic frequency of 15.5 GHz. The magnitude and
phase values are then stored in fftout2.log.
FOURIER INFILE=hemt1.log T.START=3.125e-11 T.STOP=2e-9
NUM.SAMPLES=64 \
OUTFILE=fftout2.log INTERPOLATE

1041 ATLAS User’s Manual

GO Statements

21.21 GO
GO quits and restarts ATLAS and also defines certain global parameters for ATLAS execution

Note: This command is executed by DECKBUILD. This statement is documented in the DECKBUILD USER ’S
MANUAL.

Example starting a given ATLAS Version

To start a given version of ATLAS the syntax is set by the simflags argument. To start
version 5.12.0.R, type:
go atlas simflags=”-V 5.12.0.R”
Parallel ATLAS Example
To define the number of processors to be used in parallel ATLAS, use the P flag. For example,
to start parallel ATLAS use four processors. For example:
go atlas simflags=”-V 5.12.0.R -P 4”

1042 ATLAS User’s Manual

IMPACT Statements

21.22 IMPACT
IMPACT specifies and set parameters for impact ionization models.
Syntax
IMPACT <model>

Parameter Type Default Units

A.NT Real 0.588

A.PT Real 0.588
AE0001 Real 1.76108 cm-1
AE1120 Real 3.30107 cm-1
AH0001 Real 3.41108 cm-1
AH1120 Real 2.50107 cm-1
AN1 Real 7.03105 cm-1
AN2 Real 7.03105 cm-1
AN0.VALD Real 4.3383
AN1.VALD Real -2.4210-12
AN2.VALD Real 4.1233
ANGLE Real 0.0 Degrees
ANISO Logical False
ANOKUTO Real 0.426 V
ANLACKNER Real 1.316106 cm

AP0.VALD Real 2.376

AP1 Real 6.71105 cm-1
AP2 Real 1.682106 cm-1
AP1.VALD Real 0.01033
AP2.VALD Real 1.0
APOKUTO Real 0.243 V
APLACKNER Real 1.818106 cm

B.NT Real 0.248

1043 ATLAS User’s Manual

IMPACT Statements

Parameter Type Default Units

B.PT Real 0.248

BE0001 Real 2.10107 V/cm

BE1120 Real 1.70107 V/cm

BETAN Real 1.0

BETAP Real 1.0
BH0001 Real 2.96107 V/cm

BH1120 Real 1.690107 V/cm

BN1 Real 1.231106 V/cm

BN2 Real 1.231106 V/cm

BN1.VALD Real 0.0

BN0.VALD Real 0.235
BNLACKNER Real 1.474106 V/cm

BNOKUTO Real 4.81105 V/cm

BP1 Real 1.231106 V/cm

BP2 Real 2.036106 V/cm

BP0.VALD Real 0.17714

BP1.VALD Real -0.002178
BPOKUTO Real 6.53105 V/cm

BPLACKNER Real 2.036106 V/cm

C0 Real 2.510-10
CHIA Real 3.0105
CHIB Real 5.0104
CHI.HOLES Real 4.6104
CN2 Real 0.0 cm-1K-1
CN0.VALD Real 1.6831104
CN1.VALD Real 4.3796
CN2.VALD Real 1.0

1044 ATLAS User’s Manual

IMPACT Statements

Parameter Type Default Units

CN3.VALD Real 0.13005

CNOKUTO Real 3.0510-4 K

CP2 Real 0.0 cm-1K-1

CP0.VALD Real 0.0
CP1.VALD Real 0.00947
CP2.VALD Real 2.4929
CP3.VALD Real 0.0
CPOKUTO Real 5.3510-4 K

CROWELL Logical False

CSUB.N Real 2.01014
CSUB.P Real 4.01012
DN0.VALD Real 1.233735106
DN1.VALD Real 1.2039103
DN2.VALD Real 0.56703
DNOKUTO Real 6.8610-4 K

DP0.VALD Real 1.4043106

DP1.VALD Real 2.9744103
DP2.VALD Real 1.4829
DPOKUTO Real 5.6710-4 K

DEVICE Character
DIRECTED Logical False
E.DIR Logical True
E.SIDE Logical False
E.VECTOR Logical False
ECN.II Real 1.231106
ECP.II Real 2.036106
EGRAN Real 4.0105 V/cm

1045 ATLAS User’s Manual

IMPACT Statements

Parameter Type Default Units

ENERGY.STEP Real 0.0025

ETH.N Real 1.8 eV
ETH.P Real 3.5 eV
EXN.II Real 1.0
EXP.II Real 1.0
F.EDIIN Character
F.EDIIP Character
GRADQFL Logical False
GRANT Logical False
IINOFF Character
IIPOFF Character
JNX.MIN Real 0.0 A/cm2
JNY.MIN Real 0.0 A/cm2
JNZ.MIN Real 0.0 A/cm2
JPX.MIN Real 0.0 A/cm2
JPY.MIN Real 0.0 A/cm2
JPZ.MIN Real 0.0 A/cm2
ICRIT Real 4.010-3 A/cm2
INFINITY Real 0.001
LACKNER Logical False
LAMDAE Real 6.210-7 cm

LAMDAH Real 3.810-7 cm

LAN300 Real 6.210-7 cm

LAP300 Real 3.810-7 cm

LENGTH.REL Logical False

LREL.EL Real 3.3510-2 µm

VAL.BN1 Real 0.0

VAL.CN0 Real 1.6831104

VAL.CN1 Real 4.3796

VAL.CN2 Real 1.0

VAL.CN3 Real 0.13005

VAL.DN0 Real 1.233735106

VAL.DN1 Real 1.2039103

VAL.DN2 Real 0.56703

VAL.BP0 Real 0.17714

VAL.BP1 Real -0.002178

VAL.CP0 Real 0.0

VAL.CP1 Real 0.00947

VAL.CP2 Real 2.4924

VAL.CP3 Real 0.0

VAL.DP0 Real 1.4043106

VAL.DP1 Real 2.9744103

VAL.DP2 Real 1.4829

VALDINOCI Logical False

VANOVERS Logical False
ZAPPA Logical False

ZAP.AN Real 1.9 eV

ZAP.BN Real 41.7 angstroms

ZAP.CN Real 46.0 meV

ZAP.DN Real 46.0 meV

ZAP.EN Real 46.0 meV

ZAP.AP Real 1.4 eV

ZAP.BP Real 41.3 angstroms

1049 ATLAS User’s Manual

IMPACT Statements

Parameter Type Default Units

ZAP.CP Real 36.0 meV

ZAP.DP Real 36.0 meV

ZAP.EP Real 36.0 meV

Description
The impact ionization model for continuity equations allows the accurate prediction of
avalanche breakdown for many devices. Since impact ionization is a two-carrier process, the
following statement must be specified after setting impact ionization models.
METHOD CARRIERS=2
Model Selection Flags

ANISO Specifies the 4hSiC Model Equations 5-132 through 5-135.

CROWELL Specifies the Crowell and Sze formulae [59].
GRADQFL Enables Gradient of Quasi-Fermi levels as drivers.
GRANT Selects Grant’s impact ionization model [89] (see Equations 3-407 through 3-423).
LACKNER Enables the Lackner model [150].
SELB Selects the impact ionization model described by Selberherr [250].
OKUTO Enables the Okuto-Crowell model [198].
N.CHEN Enable the tabular ionization rates [44].
P.CHEN

N.CONCANNON Set the Concannon substrate current model.

P.CONCANNON

N.MCCLINTOCK Enables the tabular ionization rates [181].

P.MCCLINTOCK

N.OGUZMAN Enables the tabular ionization rates [200].

P.OGUZMAN

N.TUT Enables the tabular ionization rates [275].

P.TUT

N.UPSILON Enables non-linear dependence of effective field on electron temperature if you use
the -ISE flag.
P.UPSILON Enables non-linear dependence of effective field on hole temperature if you use the
-ISE flag.

SIC4H0001 Specifies that the y axis concides with the 0001 crystal orientation for the
anisotropic impact ionization model.

1050 ATLAS User’s Manual

IMPACT Statements

SIC4H1120 Specifies that the y axis concides with the 1120 crystal orientation for the
anisotropic impact ionization model.
TOYABE This is a synonym for SELB.
VALDINOCI Enables the Valdinoci impact ionization model. See Equations 3-394, 3-407, and
3-415. The alias for this parmeter is II.VALDI.
VANOVERS Enables the VanOverstraeten de Man model [285].
ZAPPA Enables Zappa’s model for ionization rates in InP.

Note: If no model selection flag is set, the model parameters from Grant [89] are used. See “Grant’s Impact
Ionization Model” on page 233.

Okuto-Crowell Model Parameters

ANOKUTO See Section 3.6.4 “Impact Ionization Models”, Equation 3-435, Table 3-92.
APOKUTO See Section 3.6.4 “Impact Ionization Models”, Equation 3-436, Table 3-92.
BNOKUTO See Section 3.6.4 “Impact Ionization Models”, Equation 3-435, Table 3-92.
BPOKUTO See Section 3.6.4 “Impact Ionization Models”, Equation 3-436, Table 3-92.
CNOKUTO See Section 3.6.4 “Impact Ionization Models”, Equation 3-435, Table 3-92.
CPOKUTO See Section 3.6.4 “Impact Ionization Models”, Equation 3-436, Table 3-92.
DNOKUTO See Section 3.6.4 “Impact Ionization Models”, Equation 3-435, Table 3-92.
DPOKUTO See Section 3.6.4 “Impact Ionization Models”, Equation 3-436, Table 3-92.
NGAMOKUTO See Section 3.6.4 “Impact Ionization Models”, Equation 3-435, Table 3-92.
PGAMOKUTO See Section 3.6.4 “Impact Ionization Models”, Equation 3-436, Table 3-92.
NDELOKUTO See Section 3.6.4 “Impact Ionization Models”, Equation 3-435, Table 3-92.
PDELOKUTO See Section 3.6.4 “Impact Ionization Models”, Equation 3-436, Table 3-92.

van Overstraeten de Man Model Parameters

N.VANHW See Section 3.6.4 “Impact Ionization Models”, Equation 3-405, Table 3-88.
P.VANHW See Section 3.6.4 “Impact Ionization Models”, Equation 3-406, Table 3-88.

Note: You must also use the Selberherr model parameters with this model.

1051 ATLAS User’s Manual

IMPACT Statements

Lackner Model Parameters

ANLACKNER See Section 3.6.4 “Impact Ionization Models”, Equation 3-438, Table 3-93.
APLACKNER See Section 3.6.4 “Impact Ionization Models”, Equation 3-439, Table 3-93.
BNLACKNER See Section 3.6.4 “Impact Ionization Models”, Equation 3-438, Table 3-93.
BPLACKNER See Section 3.6.4 “Impact Ionization Models”, Equation 3-439, Table 3-93.
N.LACKHW See Section 3.6.4 “Impact Ionization Models”, Equation 3-440, Table 3-93.
P.LACKHW See Section 3.6.4 “Impact Ionization Models”, Equation 3-440, Table 3-93.

Non-linear Hydrodynamic Model

To enable this model, use the command line flag -ISE.

N.UPSILON See Section 3.6.4 “Impact Ionization Models”, Equation 3-453, Table 3-97.
P.UPSILON See Section 3.6.4 “Impact Ionization Models”, Equation 3-454, Table 3-97.
NISELAMBDA See Section 3.6.4 “Impact Ionization Models”, Equation 3-453, Table 3-97.
PISELAMBDA See Section 3.6.4 “Impact Ionization Models”, Equation 3-454, Table 3-97.
NISEDELTA See Section 3.6.4 “Impact Ionization Models”, Equation 3-453, Table 3-97.
PISEDELTA See Section 3.6.4 “Impact Ionization Models”, Equation 3-454, Table 3-97.

Electric Field Model Selection Flags

The default electric field to use for ionization rate calculations is calculated as the modulus of
the electric field over a triangle. In ATLAS3D, this is combined with the z-component of the
field at each corner of the prism to give an overall field modulus at each node. This
corresponds to Equation 3-387 for a triangle.

ANGLE Specifies the direction along which field magnitude will be calculated for use in
calculating ionization rate when you specify DIRECTED.
DIRECTED Specifies that the field for ionization rate calculations will be calculated along a
user-specified direction. That direction is specified by the ANGLE parameter.
E.DIR Specifies that the impact ionization rate will be calculated as a function of the
electric field in the direction of the current (see Equation 3-389). Its alias is NEW.
E.SIDE Specifies that the impact ionization rate will be calculated as a function of the
electric field along the side of the triangle (see Equation 3-388). This model is
only supported in ATLAS2D. Its alias is OLD.

1052 ATLAS User’s Manual

IMPACT Statements

E.VECTOR Specifies that the vector electric field will be used in the calculated of the impact
ionization rate (see Equation 3-387).
OLD This is the model that corresponds to Equation 3-388 and is only supported in
ATLAS2D.
NEW This is the model that corresponds to Equation 3-389, where the component of the
field in the direction of the current is used in the ionisation rate calculation. It is
implemented in both ATLAS2D and ATLAS3D. If the dot product of E and J is
negative, then the field component is taken as 0. Consequently, impact ionisation
may only occur when a current is dominated by the drift term.

Model Localization Parameters

DEVICE Specifies the device in MIXEDMODE simulation to which the statement should
apply. The synonym for this parameter is STRUCTURE.
MATERIAL Specifies what material from Table B-1 the statement should apply. If a material is
specified then all regions defined as being composed of that material will be
affected.
NAME Specifies what region that the IMPACT statement should apply. Note that the name
must match the name specified in the NAME parameter of the REGION statement or
the region number.
REGION Specifies that index of the region to which the impact parameters apply.
STRUCTURE This is a synonym for DEVICE.

Valdinoci Model Parameters

AN0.VALD This is an alias for VAL.AN0.

AN1.VALD This is an alias for VAL.AN1.
AN2.VALD This is an alias for VAL.AN2.
AP0.VALD This is an alias for VAL.AP0.
AP1.VALD This is an alias for VAL.AP1.
AP2.VALD This is an alias for VAL.AP2.
BN0.VALD This is an alias for VAL.BN0.
BN1.VALD This is an alias for VAL.BN1.
BP0.VALD This is an alias for VAL.BP0.
BP1.VALD This is an alias for VAL.BP1.
CN0.VALD This is an alias for VAL.CN0.

1053 ATLAS User’s Manual

IMPACT Statements

CN1.VALD This is an alias for VAL.CN1.

CN2.VALD This is an alias for VAL.CN2.
CN3.VALD This is an alias for VAL.CN3.
CP0.VALD This is an alias for VAL.CP0.
CP1.VALD This is an alias for VAL.CP1.
CP2.VALD This is an alias for VAL.CP2.
CP3.VALD This is an alias for VAL.CP3.
DN0.VALD This is an alias for VAL.DN0.
DN1.VALD This is an alias for VAL.DN1.
DN2.VALD This is an alias for VAL.DN2.
DP0.VALD This is an alias for VAL.DP0.
DP1.VALD This is an alias for VAL.DP1.
DP2.VALD This is an alias for VAL.DP2.
VAL.AN0 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-408. The alias for this parameter is AN0.VALD.
VAL.AN1 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-408. The alias for this parameter is AN1.VALD.
VAL.AN2 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-408. The alias for this parameter is AN2.VALD.
VAL.BN0 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-409. The alias for this parameter is BN0.VALD.
VAL.BN1 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-409. The alias for this parameter is BN1.VALD.
VAL.CN0 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-410. The alias for this parameter is CN0.VALD.
VAL.CN1 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-410. The alias for this parameter is CN1.VALD.
VAL.CN2 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-410. The alias for this parameter is CN2.VALD.
VAL.CN3 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-410. The alias for this parameter is CN3.VALD.
VAL.DN0 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-411. The alias for this parameter is DN0.VALD.
VAL.DN1 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-411. The alias for this parameter is DN1.VALD.

1054 ATLAS User’s Manual

IMPACT Statements

VAL.DN2 Specifies the value of a temperature dependent impact ionization parameter in

Equation 3-411. The alias for this parameter is DN2.VALD.
VAL.AP0 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-412.
VAL.AP1 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-412. The alias for this parameter is AP1.VALD.
VAL.AP2 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-412. The alias for this parameter is AP2.VALD.
VAL.BP0 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-413. The alias for this parameter is BP0.VALD.
VAL.BP1 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-413. The alias for this parameter is BP1.VALD.
VAL.CP0 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-414. The alias for this parameter is CP0.VALD.
VAL.CP1 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-414. The alias for this parameter is CP1.VALD.
VAL.CP2 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-414. The alias for this parameter is CP2.VALD.
VAL.CP3 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-414. The alias for this parameter is CP3.VALD.
VAL.DP0 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-415. The alias for this parameter is DP0.VALD.
VAL.DP1 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-415. The alias for this parameter is DP1.VALD.
VAL.DP2 Specifies the value of a temperature dependent impact ionization parameter in
Equation 3-415. The alias for this parameter is DP2.VALD.

Zappa Model Parameters

ZAP.AN Specifies the value of a temperature dependent ionization parameter in Equation 3-

416.
ZAP.BN Specifies the value of a temperature dependent ionization parameter in Equation 3-
417.
ZAP.CN Specifies the value of a temperature dependent ionization parameter in Equation 3-
417.
ZAP.DN Specifies the value of a temperature dependent ionization parameter in Equation 3-
418.

1055 ATLAS User’s Manual

IMPACT Statements

ZAP.EN Specifies the value of a temperature dependent ionization parameter in Equation 3-

418.
ZAP.AP Specifies the value of a temperature dependent ionization parameter in Equation 3-
419.
ZAP.BP Specifies the value of a temperature dependent ionization parameter in Equation 3-
420.
ZAP.CP Specifies the value of a temperature dependent ionization parameter in Equation 3-
420.
ZAP.DP Specifies the value of a temperature dependent ionization parameter in Equation 3-
421.
ZAP.EP Specifies the value of a temperature dependent ionization parameter in Equation 3-
421.

Crowell Model Parameters

LAMBDAE Specifies the mean free path for electrons. The alias for this parameter is LAN300.
LAMDAH Specifies the mean free path for holes. The alias for this parameter is LAP300.
LAN300 This is an alias for LAMBDAE.
LAP300 This is an alias for LAMBDAH.
OP.PH.EN This is an alias for OPPHE.
OPPHE Specifies the optical phonon energy. The alias for this parameter is OP.PH.EN.

Selberrherr Model Parameters

AN1, AN2, BN1, BN2, Specify the basic set of parameters for Selberherr’s impact ionization model.
EGRAN Index 1 (AP1, BP1, AN1, and BN1) corresponds to field values greater than
EGRAN, and index 2 (AP2, BP2, AN2, and BN2) corresponds to field values less
than EGRAN. The aliases for these parameters are N.IONIZA, P.IONIZA,
ECN.II, and ECP.II.

BETAN This is for electrons and BETAP is for holes correspond to coefficients for the
power of ECRIT/E. The aliases for these parameters are EXN.II and EXP.II.
EXN.II This is an alias for BETAN.
EXP.II This is an alias for BETAP.

1056 ATLAS User’s Manual

IMPACT Statements

Temperature Dependence Parameters

A.NT Specifies the value of the temperature-dependent parameter in Equation 3-393.

A.PT Specifies the value of the temperature-dependent parameter in Equation 3-394.
B.NT Specifies the value of the temperature-dependent parameter in Equation 3-395.
B.PT Specifies the value of the temperature-dependent parameter in Equation 3-396.
M.ANT Specifies the value of the temperature-dependent parameter in Equation 3-393.
M.APT Specifies the value of the temperature-dependent parameter in Equation 3-394.
M.BNT Specifies the value of the temperature-dependent parameter in Equation 3-395.
M.BPT Specifies the value of the temperature-dependent parameter in Equation 3-396.
CN2, CP2, DN2, and These are specifiable coefficients in the temperature dependent models
DP2 described in Equations 3-394 and 3-395. The aliases for these parameters are
N.ION.1, P.ION.1, N.ION.2, and P.ION.2.

N.ION.1, P.ION.1, These are the aliases for CN2, CP2, DN2, and DP2.
N.ION.2, and
P.ION.2

Parameters for use with Energy Balance

F.EDIIN Specifies the name of the file containing a C-Interpreter function describing the
values of the parameters in Equation 3-391 as a function of electron
temperature.
F.EDIIP Specifies the name of the file containing a C-Interpreter function describing the
values of the parameters in Equation 3-392 as a function of hole temperature.
LENGTH.REL Specifies the use of energy relaxation length for the impact ionization model
with the energy balance model. If LENGTH.REL is specified, TAUSN and TAUSP
cannot be specified and have any affect.
LREL.EL Specifies an energy relaxation length for electrons if LENGTH.REL is
specified.
LREL.HO Specifies an energy relaxation length for holes if LENGTH.REL is specified.
TAUSN Specifies the relaxation time for electrons in the temperature dependent impact
model.
TAUSP Specifies the relaxation time for holes in the temperature dependent impact
model.

1057 ATLAS User’s Manual

IMPACT Statements

Note: When energy balance simulations are run, the Toyabe impact ionization model is used. This model is used
regardless of the SELB or CROWELL settings. See “Toyabe Impact Ionization Model” on page 237 for more
information about this model.

Concannon Model Parameters

CSUB.N This is an empirical tuning factor used in Concannon’s Substrate Current Model
(see Equation 3-447) for electrons.
CSUB.P This is an empirical tuning factor used in Concannon’s Substrate Current Model
(see Equation 3-448) for holes.
ETH.N Specifies the ionization threshold energy for electrons used in Concannon’s
Substrate Current Model (see Equation 3-447).
ETH.P Specifies the ionization threshold energy for holes used in Concannon’s Substrate
Current Model (see Equation 3-448).
C0 Specifies the electron distribution weight factor used in Concannon’s Substrate
Current Model (see Equation 3-451).
CHIA Specifies the electron distribution function constant used in Concannon’s
Substrate Current Model (see Equation 3-451).
CHIB Specifies the electron distribution function constant used in Concannon’s
Substrate Current Model (see Equation 3-451).
CHI.HOLES Specifies the hole distribution function constant used in Concannon’s Substrate
Current Model (see Equation 3-452).
ENERGY.STEP Specifies the energy step for numeric integration used in Concannon’s
Substrate Current Model
INFINITY Specifies the limit for the highest energy in numeric integration used in
Concannon’s Substrate Current Model.

Anisotropic Impact Ionization Parameters

AE0001 Anisotropic impact ionization parameter for Equations 5-132 through 5-135.
AE1120 Anisotropic impact ionization parameter for Equations 5-132 through 5-135.
AH0001 Anisotropic impact ionization parameter for Equations 5-132 through 5-135.
AH1120 Anisotropic impact ionization parameter for Equations 5-132 through 5-135.
BE0001 Anisotropic impact ionization parameter for Equations 5-132 through 5-135.

1058 ATLAS User’s Manual

IMPACT Statements

BE1120 Anisotropic impact ionization parameter for Equations 5-132 through 5-135.
BH0001 Anisotropic impact ionization parameter for Equations 5-132 through 5-135.
BH1120 Anisotropic impact ionization parameter for Equations 5-132 through 5-135.

Wide Bandgap Semiconductor Parameters

JNX.MIN Specifies minimum electron current in x-direction to be used in calculating impact

generation rate for E.SIDE model.
JNY.MIN Specifies minimum electron current in y-direction to be used in calculating impact
generation rate for E.SIDE model.
JNZ.MIN Specifies minimum electron current in z-direction to be used in calculating impact
generation rate for E.SIDE model.
JPX.MIN Specifies minimum hole current in x-direction to be used in calculating impact
generation rate for E.SIDE model.
JPY.MIN Specifies minimum hole current in y-direction to be used in calculating impact
generation rate for E.SIDE model.
JPZ.MIN Specifies minimum hole current in z-direction to be used in calculating impact
generation rate for E.SIDE model.

Tabular Selberherr Parameters

Parameter Type Default Units

SIGP0.AMPHOTERIC Real 1.010-16 cm2

SIGPN.AMPHOTERIC Real 1.010-16 cm2
SIGTAE Real 1.010-16 cm2
SIGTAH Real 1.010-14 cm2
SIGTDE Real 1.010-14 cm2
SIGTDH Real 1.010-16 cm2
STRUCTURE Character
TFILE Character
T0.AMPHOTERIC Real 300 K
WGA Real 0.1 eV
WGD Real 0.1 eV
WTA Real 0.025 eV
WTD Real 0.05 ev
X.MIN Real left of structure m
X.MAX Real right of structure m
Y.MIN Real top of structure m
Y.MAX Real bottom of structure m
Z.MIN Real back of structure m
Z.MAX Real front of structure m

Description
The INTDEFECTS statement is used to describe the density of defect states in the band gap at
semiconductor interfaces. You can specify up to four distributions, two for donor-like states
and two for acceptor-like states. Each type of state may contain one exponential (tail)
distribution and one Gaussian distribution.

1062 ATLAS User’s Manual

INTDEFECTS Statements

DEVICE Specifies which device the statement applies in mixed mode simulation. The
synonym for this parameter is STRUCTURE.
DFILE Specifies the file name where the donor state density distribution, as a function
of energy, will be stored. You can examine this file by using TONYPLOT.
EGA Specifies the energy that corresponds to the Gaussian distribution peak for
acceptor-like states. This energy is measured from the conduction band edge.
EGD Specifies the energy that corresponds to the Gaussian distribution peak for
donor-like states. This energy is measured from the valence band edge.
EP.AMPHOTERIC Specifies the most probable potential defect energy.
EU.AMPHOTERIC Specifies the defect electron correlation energy.
EV0.AMPHOTERIC Specifies the characteristic energy.
F.TFTACC Specifies the name of a file containing a C-Interpreter function, discribing the
distribution of acceptor state densities as a function of energy.
F.TFTDON Specifies the name of a file containing a C-Interpreter function, describing the
distribution of donor state densities as a function of energy.
FILE.AMPHOTERIC Specifies the file name where the dangling bond density of states distribution, as
a function of energy, will be stored. If you specify FILEX.AMPHOTERIC,
FILEY.AMPHOTERIC, and FILEZ.AMPHOTERIC, then the file will be saved at the
coordinate closest to the one specified by FILEX.AMPHOTERIC,
FILEY.AMPHOTERIC, and FILEZ.AMPHOTERIC.

FILEX.AMPHOTERIC Specifies the X coordinate used in FILE.AMPHOTERIC.

FILEY.AMPHOTERIC Specifies the Y coordinate used in FILE.AMPHOTERIC.
FILEZ.AMPHOTERIC Specifies the Z coordinate used in FILE.AMPHOTERIC.
HCONC.AMPHOTERIC Specifies the density of hydrogen.
INTMATERIAL Specifies that the interface models defined on this command should be applied
to every interface between the two given materials.
For example, INTMATERIAL="oxide/silicon".
INTNAME Specifies that the interface models defined on this command should be applied
to every interface between the two given regions (where the regions are defined
by their region name). For example, INTNAME="channel/substrate".
INTNUMBER Specifies that the interface models defined on this command should be applied
to every interface between the two given regions (where the regions are defined
by their region number). For example, INTNUMBER="2/3".
INTREGION Specifies that the interface models defined on this command should be applied
to every interface between the two given regions (where the regions are defined
by either their region name or their region number).
For example, INTREGION="channel/substrate".

1063 ATLAS User’s Manual

INTDEFECTS Statements

NGA Specifies the total density of acceptor-like states in a Gaussian distribution.

NGD Specifies the total density of donor-like states in a Gaussian distribution.
NSISI.AMPHOTERIC Specifies the density of Si-Si bonds.
NTA Specifies the density of acceptor-like states in the tail distribution at the
conduction band edge.
NTD Specifies the density of donor-like states in the tail distribution at the valence
band edge.
NUM.AMPHOTERIC Specifies the number of energy levels used in the DOS integration for
amphoteric defects.
NV0.AMPHOTERIC Specifies the density of states of the valance-band tail exponential region
extrapolated to the valence-band edge. If this parameter is not specified, then the
valence band density of states (NV300 on the MATERIAL statement) will be used
instead.
NUMBER or REGION Specifies the region index to which the DEFECT statement applies.
NUMA Specifies the number of discrete levels that will be used to simulate the
continuous distribution of acceptor states.
NUMD Specifies the number of discrete levels that will be used to simulate the
continuous distribution of donor states.
R1MATERIAL Specifies that the interface models defined on this command should be applied
R2MATERIAL to every interface between the two given materials.
For example, R1MATERIAL=oxide R2MATERIAL=silicon.
R1NAME Specifies that the interface models defined on this command should be applied
R2NAME to every interface between the two given regions (where the regions are defined
by their region name). For example, R1NAME=channel R2NAME=substrate.
R1NUMBER Specifies that the interface models defined on this command should be applied
R2NUMBER to every interface between the two given regions (where the regions are defined
by their region number). For example, R1NUMBER=2 R2NUMBER=3.
R1REGION Specifies that the interface models defined on this command should be applied
R2REGION to every interface between the two given regions (where the regions are defined
by either their region name or their region number).
For example, R1REGION=channel R2REGION=substrate.
S.I Specifies that the INTDEFECTS statement should apply to semiconductor-
insulator interfaces.
S.M Specifies that the INTDEFECTS statement should apply to semiconductor-metal
interfaces.
S.S Specifies that the INTDEFECTS statement should apply to semiconductor-
semiconductor interfaces.

1064 ATLAS User’s Manual

INTDEFECTS Statements

S.X Specifies that the INTDEFECTS statement should apply to all semiconductor-
insulator interfaces, including those to the outside domain.
SIGGAE Specifies the capture cross-section for electrons in a Gaussian distribution of
acceptor-like states.
SIGGAH Specifies the capture cross-section for holes in a Gaussian distribution of
acceptor-like states.
SIGGDE Specifies the capture cross-section for electrons in a Gaussian distribution of
donor-like states.
SIGGDH Specifies the capture cross-section for holes in a Gaussian distribution of donor-
like states.
SIGMA.AMPHOTERIC Specifies the defect pool width.
SIGN0.AMPHOTERIC Specifies the electron capture cross-section for neutral defects.
SIGNP.AMPHOTERIC Specifies the electron capture cross-section for positive defects.
SIGP0.AMPHOTERIC Specifies the hole capture cross-section for neutral defects.
SIGPN.AMPHOTERIC Specifies the hole capture cross-section for negative defects.
SIGTAE Specifies the capture cross-section for electrons in a tail distribution of acceptor-
like states.
SIGTAH Specifies the capture cross-section for holes in a tail distribution of acceptor-like
states.
SIGTDE Specifies the capture cross-section for electrons in a tail distribution of donor-
like states.
SIGTDH Specifies the capture cross-section for holes in a tail distribution of donor-like
states.
STRUCTURE This is a synonym for DEVICE.
TFILE Specifies the file name where the acceptor and donor state density distributions,
as a function of energy referenced from Ev, will be stored. You can examine this
file by using TONYPLOT.
T0.AMPHOTERIC Specifies the freeze-in temperature.
WGA Specifies the characteristic decay energy for a Gaussian distribution of
acceptor-like states.
WGD Specifies the characteristic decay energy for a Gaussian distribution of donor-
like states.
WTA Specifies the characteristic decay energy for the tail distribution of acceptor-like
states.
WTD Specifies the characteristic decay energy for the tail distribution of donor-like
states.

1065 ATLAS User’s Manual

INTDEFECTS Statements

X.MIN Specifies the left boundary of a box, where an interface must exist, where
defects are to be applied.
X.MAX Specifies the right boundary of a box, where an interface must exist, where
defects are to be applied.
Y.MIN Specifies the top boundary of a box, where an interface must exist, where
defects are to be applied.
Y.MAX Specifies the bottom boundary of a box, where an interface must exist, where
defects are to be applied.
Z.MIN Specifies the back boundary of a box, where an interface must exist, where
defects are to be applied.
Z.MAX Specifies the front boundary of a box, where an interface must exist, where
defects are to be applied.

1066 ATLAS User’s Manual

INTERFACE Statements

21.24 INTERFACE
INTERFACE specifies interface parameters at semiconductor/insulator boundaries. All
parameters apply only at the boundary nodes except where stated.
Syntax
INTERFACE [<params>]

Parameter Type Default Units

A.MAX Real 360.0 Degrees

A.MIN Real 0.0 Degrees
ABSORPTION Real 0.0
ADF.TABLE Character
AR.ABSORB Real 0.0 cm-1
AR.MATERIAL Character
AR.INDEX Real 1
AR.THICK Real 0 m
BESONOS Logical False
CHARGE Real 0.0 cm-2
COATING Integer 1
CONSTANT Logical False
CR Real 1.0
CT Real 1.0
DEVICE Character
DD.TUNNEL Real 0.1
DIFFUSIVE Logical False
DISPERSION Real 0.0
DY.TUNNEL Real 0.1
DYNASONOS Logical True
ELLIPSE Logical False
F.QF Character
GAUSS Logical False
I.I Logical False

1067 ATLAS User’s Manual

INTERFACE Statements

Parameter Type Default Units

IITHERM Logical False

INT.RESIST Real 0.0 cm2

INTMATERIAL Character
INTNAME Character
INTNUMBER Character
INTREGION Character
LAM1 Real 0.0 m

LAM2 Real 0.0 m

LAMBERT Logical False

LAYER Integer 1
LORENTZ Logical False
MATERIAL Character
N.I Logical False
NLAM Integer 0
NOM.ANGLE Real 0.0 Degrees
NOS.ANGLE Real 0.0 Degrees
NOM.DIST Real 1.0 Microns

NOS.DIST Real 1.0 Microns

NR Real 2.0
NT Real 3.0
OPTICAL Logical False
OUTADF Character
OUT.HR Character
OUT.HT Character
QF Real 0.0 cm-2
P1.X Real m
P2.X Real m
P1.Y Real m
P2.Y Real m

1068 ATLAS User’s Manual

INTERFACE Statements

Parameter Type Default Units

R.MAX Real Device radius m

R.MIN Real 0.0 m
R1MATERIAL Character
R1NAME Character
R1NUMBER Integer
R1REGION Character
R2MATERIAL Character
R2NAME Character
R2NUMBER Integer
R2REGION Character
REFLECT Real
REGION Integer
S.C Logical False
S.I Logical True
S.M Logical False
S.N Real 0.0 cm/s
S.P Real 0.0 cm/s
S.S Logical False
S.X Logical False
SCATTERED Integer 11
SIGMA Real 20.0 nm
SITHERM Logical False
SEMIMINOR Real 0.31
SPECULAR Real 0.0
STRUCTURE Character
TCR.TABLE Character
TCT.TABLE Character
THERMIONIC Logical False
TRIANGLE Logical False

1069 ATLAS User’s Manual

INTERFACE Statements

Parameter Type Default Units

TUNNEL Logical False

X.MAX Real right hand side of structure m
X.MIN Real left hand side of structure m
Y.MAX Real bottom of structure m
Y.MIN Real top of structure m
Z.MIN Real m
Z.MAX Real m

Description
The INTERFACE statement consists of a set of boundary condition parameters for the interface
and a set of parameter to localize the effect of these parameters.
Boundary Condition Parameters

ABSORPTION Specifies the fraction of the incident power lost to absorption in diffusive
reflection.
ADF.TABLE The name of the file containing a table describing the angular distribution
function. The first entry is the number of samples. The following rows contain
two numbers each. The first number is the angle in degrees, and the second
number is the value of the ADF.
AR.ABSORB Specifies the absorption coefficient of a anti-refractive coating layer. Default
value is 0.0 cm-1.
AR.INDEX Specifies the real component refractive index for the anti-reflective coating model
in LUMINOUS. See Section 10.9 “Anti-Reflective (AR) Coatings for Ray Tracing
and Matrix Method” for more information about the anti-reflective coating
model.
AR.MATERIAL Specifies the material type for an anti-reflective layer. The alias for this parameter
is MATERIAL.
AR.THICK Specifies the thickness of an anti-reflective coating layer for the reflection model
in LUMINOUS. This layer should generally not exist in the device mesh structure.
See Section 10.9 “Anti-Reflective (AR) Coatings for Ray Tracing and Matrix
Method” for more information about the anti-reflective coating model.
BESONOS If DYNASONOS is enabled, this additionally enables the BESONOS model. This
allows the use of a Band-Engineered tunnel insulator stack.

1070 ATLAS User’s Manual

INTERFACE Statements

CHARGE Specifies interface charge density (cm-2) applied between two materials. The
additional parameters (S.I, S.S, and S.X) allow this value to be applied at
semiconductor-insulator, semiconductor-semiconductor, and semiconductor-
domain edges respectively. A value of 1e10 cm-2 represents 1e10 electronic
charges per cm-2 at the interface. A positive value will introduce a positive charge
value and a negative value will introduce a negative charge value.
COATING Specifies the number of a multilayer anti-reflective coating (ARC) referred to in
the INTERFACE statement. If the COATING parameter is not set, the first coating is
assumed. You must specify coatings in order (i.e., you must define COATING=2
before you define COATING=3). Different coatings should not overlap. If this
occurs, the later coating is assumed in the overlapping part. For more information
about ARC, see Section 10.9 “Anti-Reflective (AR) Coatings for Ray Tracing and
Matrix Method”.
CONSTANT Specifies a constant angular distribution function as described in Equation 10-42.
CR Specifies a user tuning parameter of the reflection haze function (see Equation
10-41).
CT Specifies a user tuning parameter of the transmission haze function (see Equation
10-40).
DD.TUNNEL Controls the numerical integration in Equation 5-54. The integral in considered
converged when the argument of the outer integral is less than the value of
DD.TUNNEL.

Note: The numerical integral starts at the peak of the barrier where the argument should be equal to 1.0.
Reducing this value increases simulation time but may improve accuracy.

DIFFUSIVE Enables diffusive reflection and transmission at an interface.

DISPERSION Specifies the spread of diffusive reflection.
DY.TUNNEL Controls the step size in the numerical integration in Equation 5-54. The value of
DY.TUNNEL specifies the size of the energy step, dEx. The energy step is chosen
as the product of the value for DY.TUNNEL and the energy change over the first
triangle adjacent to the edge at the location in question. Reducing this value
increases simulation time but may improve accuracy.
DYNASONOS Enables the DYNASONOS model unless exlicitly cleared by using
^DYNASONOS. In this case, the SONOS model will be enabled.

ELLIPSE Specifies a elliptical angular distribution function as described in Equation 10-

47.
F.QF Specifies the name of a file containing a C-INTERPRETER function describing the
density of the interface fixed charge as a function of position.
GAUSS Enables Gaussian distribution of diffusive reflection (see Equation 10-9).

1071 ATLAS User’s Manual

INTERFACE Statements

I.I Specifies that the application of interface models specified in this INTERFACE
statement should include insulator-insulator interfaces.
IITHERM This sets up the interface between two insulators as being one that can model
thermionic emission if they are both subsequently changed to wide bandgap
semiconductors using the SEMICONDUCTOR parameter of the MATERIAL
statement. To do this, use the THERMIONIC and I.I parameters on the
INTERFACE statement.

INT.RESIST Specifies the value of a non-zero interface resistance between a conductor region
and a semiconductor region.
LAYER Describes the order of layers in a coating. If LAYER is not specified, the first (top)
layer is assumed. You must specify the layers in order (i.e., you must define
LAYER=2 before you define LAYER=3). Note that you only need to specify the
coordinates of the interface for the first layer of each coating.
LAMBERT Specifies a Lambertian angular distribution function as described in Equation 10-
46.
LAM1 Specify the beginning and ending wavelengths in microns for outputting haze
LAM2 functions. See OUT.HT and OUT.HR.
LORENZ Enables Lorenz distribution of diffusive reflection (see Equation 10-10).

Note: When specifying INT.RESIST, you should also specify S.C on the INTERFACE statement.

MATERIAL Specifies material name for the layer of a coating defined by the statement.
Setting the material of a coating in this way enables you to apply any refractive index
model supported by MATERIAL statement in ATLAS. See Section 10.9 “Anti-
Reflective (AR) Coatings for Ray Tracing and Matrix Method” for more information
about anti-reflective coatings. The alias for this parameter is AR.MATERIAL.
N.I Enables either the SONOS or DYNASONOS model.
NLAM Specifies the number of samples to output haze functions. See OUT.HT and OUT.HR.
NR Specifies a user tuning parameter of the reflection haze function (see Equation 10-
41).
NT Specifies a user tuning parameter of the transmission haze function (see Equation 10-
40).
OPTICAL Specifies the optical properties of the interface defined in the statement are
modeled using Transfer Matrix Method. See Section 10.3 “Matrix Method” for more
information.
OUTADF Specifies the file name for outputting the angular distribution function.
OUT.HR Specify file names for outputting reflective and transmissive haze functions for
OUT.HT display in TONYPLOT. See NLAM, LAM1, and LAM2.

1072 ATLAS User’s Manual

INTERFACE Statements

RELECT Specifies the reflection coefficient for ray tracing in LUMINOUS. The value should be
between 0.0 abd 1.0 inclusive. You should also specify OPTICAL and values for P1.X,
P1.Y, and P2.Y.

QF Specifies fixed oxide charge density (cm-2) applied at a semiconductor to insulator

interface. A value of 1e10 cm-2 represents 1e10 electronic charges per cm-2 at the
interface. A positive value will introduce a positive charge value and a negative
value will introduce a negative charge value.
S.C Specifies that the application of interface models specified in this INTERFACE
statement should include semiconductor-conductor interfaces.
S.I Specifies that the application of interface models specified in this INTERFACE
statement should include semiconductor-insulator interfaces.
S.M Specifies that the application of interface models in this INTERFACE statement should
include semiconductor-metlal interfaces.
S.N Specifies the electron surface recombination velocity.
S.P Specifies the hole surface recombination velocity.
S.S Specifies that the application of interface models specified in this INTERFACE
statement should include semiconductor-semiconductor interfaces.
S.X Specifies that the application of interface models specified in this INTERFACE
statement should include any interfaces including interfaces with the outside domain.
SCATTERED Specifies the number of rays scattering event for diffusive reflections.
SIGMA Characterizes the mean feature size for rough interfaces as used in Equations 10-40
and 10-41.
SITHERM This sets up the interface between a semiconductor and an insulator as being one
which can model thermionic emission if the insulator is subsequently changed to
wide bandgap semiconductors using the SEMICONDUCTOR parameter of the MATERIAL
statement. Additionally, any hot carrier current generated at the interface is injected
into the carrier continuity equations in the wide bandgap semiconductor. The usual
algorithm for propagating the hot carrier current by drift only is inactive.
SEMIMINOR Specifies the semiminor axis of the elliptical angular distribution function described
in Equation 10-47.
SPECULAR Specifies the fraction of the incident power that experiences specular reflection.
TCR.TABLE Specify the names of files containing tables describing the wavelength dependent
TCT.TABLE haze function parameters CR and CT. The first entry of the tables is the number of
samples. The following rows contain pairs, where the first number is the
wavelength in microns and the second number is the value for CR or CT.

1073 ATLAS User’s Manual

INTERFACE Statements

THERMIONIC Specifies that carrier transport across an interface is modeled by thermionic emission.
This model will only be applied at semiconductor/semiconductor boundaries. See
Section 5.2.2 “The Thermionic Emission and Field Emission Transport Model” for
more information on thermionic emission.
TRIANGLE Specifies a triangular angular distribution function as described in
Equation 10-43.
TUNNEL Specifies that the carrier transport across the interface will account for
thermionic field emission. When TUNNEL is specified, THERMIONIC should also be
specified. See Section 5.2.2 “The Thermionic Emission and Field Emission Transport
Model” for more information on thermionic emission.

Note: To use the INTERFACE statement to specify thermionic or tunneling interfaces, place the INTERFACE
statement immediately after all MESH, REGION, ELECTRODE and DOPING statements and before any
CONTACT, MODEL, MATERIAL, MOBILITY or IMPACT statements.

Position Parameters

A.MAX Specifies the maximum angle for the application of the statement. This
is used with 2D circular or 3D cylindrical structures.
A.MIN Specifies the minimum angle for the application of the statement. This
is used with 2D circular or 3D cylindrical structures.
INTMATERIAL Specifies that the interface models defined on this command should be
applied to every interface between the two given materials.
For example, INTMATERIAL="oxide/silicon".
INTNAME Specifies that the interface models defined on this command should be
applied to every interface between the two given regions (where the
regions are defined by their region name). For example,
INTNAME="channel/substrate".

INTNUMBER Specifies that the interface models defined on this command should be
applied to every interface between the two given regions (where the
regions are defined by their region number). For example,
INTNUMBER="2/3".

1074 ATLAS User’s Manual

INTERFACE Statements

R1MATERIAL Specifies that the interface models defined on this command should be
R2MATERIAL applied to every interface between the two given materials. For
example, R1MATERIAL=oxide R2MATERIAL=silicon.
R1NAME and R2NAME Specifies that the interface models defined on this command should be
applied to every interface between the two given regions (where the
regions are defined by their region name). For example,
R1NAME=channel R2NAME=substrate.

R1NUMBER and R2NUMBER Specifies that the interface models defined on this command should be
applied to every interface between the two given regions (where the
regions are defined by their region number). For example,
R1NUMBER=2 R2NUMBER=3.

R1REGION and R2REGION Specifies that the interface models defined on this command should be
applied to every interface between the two given regions (where the
regions are defined by either their region name or their region
number). For example, R1REGION=channel R2REGION=substrate.
R.MAX Specifies the maximum radius for the application of the statement.
This is used with 2D circular or 3D cylindrical structures.
R.MIN Specifies the minimum radius for the application of the statement. This
is used with 2D circular or 3D cylindrical structures.
X.MIN, X.MAX, Y.MIN, and Define a bounding box. Any semiconductor/insulator interfaces found
Y.MAX within this region are charged. If there is only one interface in a device,
a non-planar surface may be defined using a box which contains the
whole device.
X.MIN Specifies the left X coordinate of the bounding box.
X.MAX Specifies the right X coordinate of the bounding box.
Y.MIN Specifies the bottom Y coordinate of the bounding box.
Y.MAX Specifies the top Y coordinate of the bounding box.
Z.MIN Specifies the front Z coordinate of the bounding box. It is used in
3D modules only.
Z.MAX Specifies the back Z coordinate of the bounding box. It is used in
3D modules only.
P1.X, P1.Y, P2.X, and Define a bounding box. Within this box must lie the interface that is to
P2.Y be represented as the anti-reflective coating. See Section 10.9 “Anti-
Reflective (AR) Coatings for Ray Tracing and Matrix Method” for
more information about the anti-reflective coating model.

1075 ATLAS User’s Manual

INTERFACE Statements

Note: For anti-reflective coatings in Luminous3D, use X.MIN, X.MAX, Y.MIN, Y.MAX, Z.MIN, and Z.MAX to
define the bounding box.

P1.X Specifies the left X coordinate of the bounding box.

P2.X Specifies the right X coordinate of the bounding box.
P1.Y Specifies the bottom Y coordinate of the bounding box.
P2.Y Specifies the top Y coordinate of the bounding box.
DEVICE Specifies which device in a MIXEDMODE simulation applies to the statement. The
synonym for this parameter is STRUCTURE.
REGION Specifies which region number applies to the statement.
STRUCTURE This is a synonym for DEVICE.

MOS Example
This example defines an interface with both fixed charge and recombination velocities.
INTERFACE X.MIN=-4 X.MAX=4 Y.MIN=-0.5 Y.MAX=4 \
QF=1E10 S.N=1E4 S.P=1E4
SOI Example
To define different fixed charge on the front and back interfaces of an SOI transistor, you
need two INTERFACE statements.
In the syntax below, the first statement will apply 5.1010 cm-2 charge to any silicon/oxide
interface above Y=0.01m. The second statement applied a higher charge to any interface
below Y=0.01m. Note that charges are only applied at the material interfaces so the Y
coordinate needs only to be somewhere within the silicon film.
INTERFACE Y.MAX-0.01 QF=5e10
INTERFACE Y.MIN=0.01 QF=2e11
Interface Charge for III-V Devices
By default, the INTERFACE statement is applied to semiconductor-insulator interfaces.
Interface charge can, however, be added at the interfaces between two semiconductor regions
or at the edges of semiconductor regions.
The CHARGE parameter defines the interface charge value in cm-2. The S.I, S.S, and S.X
parameters control whether the charge is placed between semiconductor-insulator regions,
semiconductor-semiconductor regions, or at the semiconductor domain edges. You can
control the location of the added charge by using the position parameters.
2D Circular and 3D Cylindrical Limits Example
INTERFACE S.S A.MIN=0.0 A.MAX=180.0 R.MIN=0.1 R.MAX=0.2
CHARGE=1.0e12
This restricts the application of interface charge to being between 0 and 180° and between a
radius of 0.1 um and 0.2 um. The angular and radial limits can also be used with the QF, S.N,
S.P, and THERMIONIC parameters.

1076 ATLAS User’s Manual

INTTRAP Statements

21.25 INTTRAP
INTTRAP activates interface defect traps at discrete energy levels within the bandgap of the
semiconductor and sets their parameter values.
Syntax
INTTRAP <type> E.LEVEL=<r> DENSITY=<r> <capture parameters>

Parameter Type Default Units

ACCEPTOR Logical False

DEGEN.FAC Real 1
DENSITY Real cm-2
DEPTH Real 5.0e-3 m

DEVICE Character
DONOR Logical False
E.LEVEL Real eV
F.DENSITY Character
F.MSCRATES Character
GRA.3.DET Logical False
GRA.3.STO Logical False
GRA.4.DET Logical False
GRA.4.STO Logical False
GRA.5.DET Logical False
GRA.5.STO Logical False
GRA.EA Real 1.0 eV
GRA.EA.SD Real 0.025 eV
GRA.EB.SD Real 0.025 eV
GRA.EB.ELEC Real 0.054 eV
GRA.EB.HOLE Real 0.054 eV
GRA.EC.ELEC Real 0.02 eV
GRA.EC.HOLE Real 0.02 eV
GRA.EC.SD Real 0.025 eV
GRA.ED Real 0.1 eV

1077 ATLAS User’s Manual

INTTRAP Statements

INTTRAP Statements

Parameter Type Default Units

Y.MAX Real bottom of structure m

Z.MIN Real back of structure m
Z.MAX Real front of structure m

Description

DEVICE Specifies which device the statement applies to in MIXEDMODE simulation. The
synonym for this parameter is STRUCTURE.
DONOR Specifies a donor-type trap level.
ACCEPTOR Specifies an acceptor-type trap level.
DEGEN.FAC Specifies the degeneracy factor of the trap level used to calculate the density.
DENSITY Sets the maximum density of states of the trap level.
DEPTH The depth that the uniform interface trap density penetrates into the insulator. This
requires the HEIMAN model.
E.LEVEL Sets the energy of the discrete trap level. It is equal to the energy distance between
conductance band and trap level for acceptor trap, and to energy distance between
trap level and valence band for donor trap.
F.DENSITY Specifies the name of a file containing a C-Interpreter function describing the
density of donor/acceptor interface traps as a function of position.
HEIMAN Enables the hysteresis interface traps model.
HPOINTS Number of uniformly spaced points inside the insulator for which the interface trap
density is calculated. This requires the HEIMAN model.
I.C Specifies that the INTTRAP statement should apply to Insulator-Conductor
interfaces.
I.I Specifies that the INTTRAP statement should apply to Insulator-Insulator interfaces.
I.M Specifies that the INTTRAP statement should apply to Insulator-Electrode interfaces.
INTMATERIAL Specifies that the interface models defined on this command should be applied to
every interface between the two given materials.
For example, INTMATERIAL="oxide/silicon".
INTNAME Specifies that the interface models defined on this command should be applied to
every interface between the two given regions (where the regions are defined by
their region name). For example, INTNAME="channel/substrate".
INTNUMBER Specifies that the interface models defined on this command should be applied to
every interface between the two given regions (where the regions are defined by
their region number). For example, INTNUMBER="2/3".

1082 ATLAS User’s Manual

INTTRAP Statements

INTREGION Specifies that the interface models defined on this command should be applied to
every interface between the two given regions (where the regions are defined by
either their region name or their region number).
For example, INTREGION="channel/substrate".
MIDGAP Specifies that the energy of the trap be set to the middle of the bandgap.
R1MATERIAL Specifies that the interface models defined on this command should be applied to
R2MATERIAL every interface between the two given materials.
For example, R1MATERIAL=oxide R2MATERIAL=silicon.
R1NAME Specifies that the interface models defined on this command should be applied to
R2NAME every interface between the two given regions (where the regions are defined by
their region name). For example, R1NAME=channel R2NAME=substrate.
R1NUMBER Specifies that the interface models defined on this command should be applied to
R2NUMBER every interface between the two given regions (where the regions are defined by
their region number). For example, R1NUMBER=2 R2NUMBER=3.
R1REGION Specifies that the interface models defined on this command should be applied to
R2REGION every interface between the two given regions (where the regions are defined
by either their region name or their region number). For example,
R1REGION=channel R2REGION=substrate.

S.C Specifies that the INTTRAP statement should apply to semiconductor-insulator

interfaces.
S.I Specifies that the INTTRAP statement should apply to semiconductor-insulator
interfaces.
S.M Specifies that the INTTRAP statement should apply to semiconductor-metal
interfaces.
S.S Specifies that the INTTRAP statement should apply to semiconductor-
semiconductor interfaces.
S.X Specifies that the INTTRAP statement should apply to all semiconductor-
insulator interfaces, including those to the outside domain.
SR.EMIT Activation energy for temperature dependent emission coefficients in SR.HEIMAN
model.
SR.HEIMAN Enables the structural relaxation enhancement of the Heiman model.
SR.TRAP Activation energy for temperature dependent capture coefficents in SR.HEIMAN
model.
STRUCTURE This is a synonym for DEVICE.
X.MIN Specifies the left boundary of a box, where an interface must exist and where traps
are to be applied.
X.MAX Specifies the right boundary of a box, where an interface must exist and where traps
are to be applied.

1083 ATLAS User’s Manual

INTTRAP Statements

Y.MIN Specifies the top boundary of a box, where an interface must exist and where traps
are to be applied.
Y.MAX Specifies the bottom boundary of a box, where an interface must exist and where
traps are to be applied.
Z.MIN Specifies the back boundary of a box, where an interface must exist and where traps
are to be applied.
Z.MAX Specifies the front boundary of a box, where an interface must exist and where traps
are to be applied.

2-State NBTI Degradation Models Parameters

GRA.3.DET Enables the deterministic 3-state model.

GRA.3.STO Enables the stochastic 3-state model.
GRA.4.DET Enables the deterministic 4-state model.
GRA.4.STO Enables the stochastic 4-state model.
GRA.5.DET Enables the deterministic 5-state model.
GRA.5.STO Enables the stochastic 5-state model.
GRA.EA Activation energy for process of relaxation back to precursor state from
neutral switching state.
GRA.EA.SD Variation parameter for GRA.EA.
GRA.EB.SD Variation parameter for GRA.EB.ELEC and GRA.EB.HOLE.
GRA.EB.ELEC Activation energy for MPFAT process of electron emission by the precursor.
GAR.EB.HOLE Activation energy for MPFAT process of electron emission by the precursor.
GRA.EC.ELEC Activation energy for the switching trap electron rates.
GRA.EC.HOLE Activation energy for the switching trap hole rates.
GRA.EC.SD Variation parameter for GRA.EC.ELEC and GRA.EC.HOLE.
GRA.ED Activation energy for the dangling bond depassivation rate.
GRA.ET1 Energy level of precursor state relative to valence band.
GRA.ET2 Energy level of switching trap relative to valence band.
GRA.ET4 Energy level of Pb dangling bond state relative to valence band.
GRA.ET1.SD Variation parameter for GRA.ET1.
GRA.ET2.SD Variation parameter for GRA.ET2.

1084 ATLAS User’s Manual

INTTRAP Statements

GRA.ET4.SD Variation parameter for GRA.ET4.

GRA.FC Field scaling factor for MPFAT tunneling process.
GRA.FC.SD Variation parameter for GRA.FC.
GRA.GAMMA Coefficient for field dependent thermal activation energy in dangling bond
depassivation rates.
GRA.GAMMA.SD Variation parameter for GRA.GAMMA.
GRA.ED.SD Standard deviation parameter for GRA.ED.
GRA.NU Attempt frequency for some processes.
GRA.NU.SD Variation parameter for GRA.NU.
GRA.SAMPLES Number of sample traps per interface point for stochastic models.
GRA.SIGN Electron capture cross-section of trap (alias for SIGN).
GRA.SIGN.SD Variation parameter for GRA.SIGN.
GRA.SIGP Hole capture cross-section of trap (alias for SIGP).
GRA.SIGP.SD Variation parameter for GRA.SIGP.

Capture Parameters
Either the cross section or lifetime parameters should be used to define the capture
parameters.

SIGN Specifies the capture cross section of the trap for electrons.
SIGP Specifies the capture cross section of the trap for holes.
TAUN Specifies the lifetime of electrons in the trap level.
TAUP Specifies the lifetime of holes in the trap level.

EC
E.level for acceptor trap
Eta

Etd
E.level for donor trap
EV

Figure 21-6: Acceptor and Donor Interface Trap Energy Levels

Multiple Interface Trap States Example

The following example sets three discrete interface trap levels within the silicon bandgap.
These trap levels will capture carriers, which slows down the switching speed of any device.
In this example, the capture cross sections are used to define the properties of each trap.
inttrap e.level=0.49 acceptor density=2.e10 degen=12 \
sign=2.84e-15 sigp=2.84e-14
inttrap e.level=0.41 acceptor density=1.e10 degen=12 \
sign=7.24e-16 sigp=7.24e-15
inttrap e.level=0.32 donor density=1.e10 degen=1 \
sign=1.00e-16 sigp=1.00e-17

Note: For semiconductor bulk trap levels, see Section 21.65 “TRAP”.

1088 ATLAS User’s Manual

LASER Statements

21.26 LASER
LASER defines physical models and model parameters for laser and Vertical Cavity Surface-
Emitting Lasers (VCSEL) simulation. For more information about VCSEL, see Chapter 9
“VCSEL Simulator”.
Syntax
LASER<parameters>

Parameter Type Default Units

ABSORPTION Logical False

ATRAP Real 0.5
CAVITY.LENGTH Real 100.0 cm
COUPLED Logical False
DBR1.START Real 0.0 m

DBR1.FINAL Real 0.0 m

DBR2.START Real 0.0 m

YMAX Real max. Y m

Description
Localization Parameters

DEVICE Specifies which device the LASER statement should apply to in MIXEDMODE
simulation. The synonym for this parameter is STRUCTURE.
LX.MAX Sets the upper x boundary of solution domain for vector Helmholtz solver.
LX.MIN Sets the lower x boundary of solution domain for vector Helmholtz solver.
LY.MAX Sets the upper y boundary of solution domain for vector Helmholtz solver.
LY.MIN Sets the lower y boundary of solution domain for vector Helmholtz solver.
MATERIAL Specifies which material from the table in Appendix B “Material Systems” that the
LASER statement should apply. If a material is specified, then all regions defined as
being composed of that material will be affected.
NAME Specifies which region the LASER statement should apply. Note that the name must
match the name specified in the NAME parameter of the REGION statement or the
region number.
REGION Specifies the region number to which these parameters apply. If there is more than
one semiconductor region, specification of different parameters for each region is
allowed. If LASER is not specified, all regions in the structure are changed.
STRUCTURE This is a synonym for DEVICE.

1092 ATLAS User’s Manual

LASER Statements

Laser Model Parameters

ABSORPTION Enables the absorption loss in photon rate equations.

ATRAP Specifies the photon rate equation cutback ratio (see Section 8.4.3
“Numerical Parameters” ).
DBR1.START These parameters specify the locations of the front and rear DBR
DBR1.FINAL mirrors for a VCSEL device (See Figure 21-7). The
DBR2.START reflectivity of front and rear DBR mirrors will be calculated from
DBR2.FINAL the optical intensity profile.
CAVITY.LENGTH Specifies the cavity length in the longitudinal direction (in m).
COUPLED Specifies that the solution process for the photon rate equation is
fully coupled to the Jacobian of the drift-diffusion equations.
EFFMODE In case of effective index (1.5D) solver, this sets the number of
modes in vertical direction.
EINIT and EFINAL Specify the lower and upper photon energies. LASER will
calculate multiple longitudinal photon rates within this range.
Using wide ranges can slow down simulation.
ESEP (DELTAE) Specifies the photon energy separation. If this isn’t specified,
LASER will automatically calculate the number of longitudinal
modes based on the cavity length and the energy range.
ETRANS Enables the selection of a specific transverse mode. The selected
transverse mode will be the mode with the closest energy to
TRANS_ENERGY.

F.MIRROR Specifies the name of a file containing a C-Interpreter function that

defines the front and rear mirror reflectivities as a function of
wavelength.
FAR.NX and FAR.NY Describe the number of samples to output for the far-field
pattern in the X and Y directions. For more information about the
far-field pattern, see Section 8.4.5 “Generation of Near-Field and
Far-Field Patterns”.
FAR.SPANX and FAR.SPANY Specifies the range of angles, where the farfield pattern is to be
calculated:
[-FAR.SPANX/2; FAR.SPANX/2] and
[-FAR.SPANY/2; FAR.SPANY/2]

ABS.FCARRIER Enables the free carrier absorption model

FCARRIER Enables the free carrier loss model in LASER.
GAINMOD Sets the number of the gain model.
HELM.GEOM Sets dimenionality and direction of optical eigenmode solver.
Possible choices are 2DXY, 1DX,1DY,15DX, and 15DY.

1093 ATLAS User’s Manual

LASER Statements

HELM.WHICH Determines which part of the eigenvalue spectrum has to be found.

21.27 LED
LED saves LED output characteristics after each subsequent solution step.
Syntax
LED <parameters>

Parameter Type Default Units

AMBIEN.IMAG Real 0.0

AMBIEN.REAL Real 1.0
ANGLE.OUTPUT Real 30.0 degrees
ANGPOWER Character
DIPOLE Logical False
EDGE_ONLY Logical False
EDGE_REFLECT Logical False
EMIN Real eV
EMAX Real eV
ESAMPLE Logical False
HORIZONTAL Real 2/3
INCOHERENT Logical False
INCOH.TOP Logical False
L.WAVE Real 0.8 m

LMAX Real m

LMIN Real m

MIN.POWER Real 110-4

MIR.BOTTOM Logical False
MIR.TOP Logical False
NSAMP Integer 100
NUMRAYS Integer 180
ONEFILE Logical False
POLAR Real 0.0
REFLECTS Integer 0
OUT.USPEC Character

1098 ATLAS User’s Manual

LED Statements

Parameter Type Default Units

SIDE Logical False

SMOOTHNESS Real 8.0
SOURCE.TERM Logical False
SOURCE_TERM Logical False
SPECT.ANGLE Character
SPECTRUM Character
SURFACE_ONLY Logical False
TEMPER Real 300.0
TR.MATRIX Logical False
USER.SPECT Character
X Real 0.0 m

XMAX Real 0.0 m

XMIN Real 0.0 m

XNUM Integer 0
Y Real 0.0 m

YMAX Real 0.0 m

Y.MAX Real 0.0 m

YMIN Real 0.0 m

Y.MIN Real 0.0 m

YNUM Integer 0
ZMAX Real 0.0 m

Z.MAX Real 0.0 m

ZMIN Real 0.0 m

Z.MIN Real 0.0 m

1099 ATLAS User’s Manual

LED Statements

Description

AMBIEN.IMAG Specifies the imaginary index of refraction for the ambient outside the device
domain for LED reverse ray tracing.
AMBIEN.REAL Specifies the real index of refraction for the ambient outside the device domain for
LED reverse ray tracing.
ANGLE.OUTPUT Specifies the angular interval in degrees where spectrum versus angle are written
to the file specified by the SPECT.ANGLE parameter.
ANGPOWER Enables the reverse ray-tracing algorithm (see Section 11.5 “Reverse Ray-
Tracing”) for analysis of output coupling of light from the structure of a Light
Emitting Diode. ANGPOWER specifies the name of the output file for the angular
power density vs. output angle dependence.
DIPOLE Specifies a particular angular distribution of the internal radiating field that
corresponds to a preferred in-plane orientation of dipoles often relevant to OLED
devices (see Section 11.5 “Reverse Ray-Tracing”).
EDGE_ONLY Specifies that for LED reverse ray tracing only the rays exiting from the edges are
considered. Rays exiting at the top and bottom surfaces are not shown.
EDGE_REFLECT Specifies the rays in reverse ray tracing reaching the edges of the device are totally
reflected.
EMAX and EMIN Specify the energy range for saving a spectrum file.
ESAMPLE Specifies that samples are taken evenly in energy unlike even sampling in
wavelength.
HORIZONTAL Specifies the proportion of horizontally oriented dipoles for the dipole emission
model.
INCOHERENT Specifies that an incoherent material in LED/OLED structure is used for the
purpose of combined transfer matrix method/reverse ray tracing (TTM+RRT)
optical analysis.
INCOH.TOP When specified with the SOURCE.TERM parameter, it allows a combination of
transfer matrix method in thin coherent layers and ray tracing in thick
incoherent layers for LED reverse ray tracing.
LMAX and LMIN Specify the range of wavelength for saving spectrum files.
L.WAVE Specifies the wavelength for reverse ray-tracing (see Section 11.5 “Reverse Ray-
Tracing”).
MIN.POWER Specifies the minimum relative power of a ray (see Section 11.5 “Reverse Ray-
Tracing”). The ray is not traced after its power falls below MIN.POWER value. This
is useful to limit the number of rays traced. The default value is MIN.POWER=
1e-4.

MIR.TOP Specifies that the top surface of the device be treated as an ideal mirror in reverse
ray-tracing (see Section 11.5 “Reverse Ray-Tracing”).

1100 ATLAS User’s Manual

LED Statements

MIR.BOTTOM Specifies that the bottom surface of the device be treated as an ideal mirror in
reverse ray-tracing (see Section 11.5 “Reverse Ray-Tracing”).
NSAMP Specifies the number of samples to use for a spectrum plot.
NUMRAYS Specifies the number of rays starting from the origin in reverse ray-tracing (see
Section 11.5 “Reverse Ray-Tracing”). The default is 180. Acceptable range is 36-
3600.
ONEFILE Specifies that successive files (output at each solution) will overwrite previous
specified by SPECTRUM, ANGPOWER or SPECT.ANGLE or all three. By default,
unique output files are written by incrementing the specified name.
POLAR Specifies polarization of the emitted photons in degrees while linearly
polarized light is assumed (see Section 11.5 “Reverse Ray-Tracing”). Parallel
(TM-mode, POLAR=90.0) and perpendicular (TE-mode, POLAR=0.0)
polarizations result in significantly different output coupling values. Use
POLAR=45.0 if there is no preferred direction of polarization of emitted photons
(unpolarized light emission).
REFLECTS Specifies a number of reflections to be traced for each ray in reverse ray-
tracing (see Section 11.5 “Reverse Ray-Tracing”). The default value is
REFLECTS=0. The maximum allowed value is REFLECTS=10. Setting the number
of reflections to 3 or 4 is often a good choice.
SIDE Specifies that the rays reaching the sides of the device are terminated there and do
not contribute to the total light output (see Section 11.5 “Reverse Ray-Tracing”).
SOURCE.TERM Indicates that dipole source terms are combined with transfer matrix
calculation of light output for LEDs. The alias is SOURCE_TERM.
SPECT.ANGLE Specifies the output file name where LED spectrum is written as a function of
emission angle.
SPECTRUM Specifies the name of a file for saving spectrum plots.
SURFACE_ONLY Specifies that for LED reverse ray tracing only the rays exiting from the top and
bottom surfaces are considered. Rays exiting at the edges are not shown.
TEMPER This is the temperature (needed for using appropriate refractive indexes of the
materials in reverse ray-tracing). The default setting of 300 K will be used if
TEMPER is unspecified.

TR.MATRIX Specifies that the transfer matrix method should be used to handle thin film LEDs.
OUT.USPEC Specifies the TONYPLOT compatible file name associated with the USER.SPECT
parameter.
X This is the X coordinate of light origin for a single point reverse ray-tracing (see
Section 11.5 “Reverse Ray-Tracing”).
XMAX This is the maximum X coordinate of a rectangular area containing multiple origin
points in reverse ray-tracing (see Section 11.5 “Reverse Ray-Tracing”).

1101 ATLAS User’s Manual

LED Statements

XMIN This is the minimum X coordinate of a rectangular area containing multiple origin
points in reverse ray-tracing (see Section 11.5 “Reverse Ray-Tracing”).
XNUM Specifies the number of points along X axis within a rectuangular area in multiple
origin reverse ray-tracing (see Section 11.5 “Reverse Ray-Tracing”).
Y This is the Y coordinate of light origin for a single point reverse ray-tracing (see
Section 11.5 “Reverse Ray-Tracing”).
YMAX This is the maximum Y coordinate of a rectangular area containing multiple origin
points in reverse ray-tracing (see Section 11.5 “Reverse Ray-Tracing”).
YMIN This is the minimum Y coordinate of a rectangular area containing multiple origin
points in reverse ray-tracing (see Section 11.5 “Reverse Ray-Tracing”).
YNUM Specifies the number of points along Y axis within a rectuangular area in multiple
origin reverse ray-tracing (see Section 11.5 “Reverse Ray-Tracing”).

1102 ATLAS User’s Manual

LENS Statements

21.28 LENS
The LENS statement is used to specify lenslet characteristics used for light propagation
analysis in LUMINOUS.
Syntax
LENS <parameters>

Parameter Type Default Units

ANGLE Real 54.74 degrees

APEX Real m

AR.INDEX Real 1.0

AR.THICKNESS Real 0.0
ASPHERIC Logical False
COMPOSITE Logical False
DENSITY Real 0.0 per sq cm
ELLIPSE Logical False
EXTINCT Real 0.0
F.LENS Character
GIZA Logical False
HEIGHT Real m

INDEX Real 1.0

K Real 0.0
MATERIAL Character
MEAN Real 0.1 m

N Real 1.0
N.FEATURES Integer 0
NEGATIVE Logical False
NSAMP Integer
PC.DX Real 0.0 m

PC.DZ Real 0.0 m

PLANE Real
PYRAMID Logical False

1103 ATLAS User’s Manual

LENS Statements

Parameter Type Default Units

RADIUS Real m

RANDOM Logical False

SAG.OUT Character
SAG.ITS Integer
SEED Integer -10
SIGMA Real 0.1 m

SPHERIC Logical False

TWODEE Logical False
USER Logical False
WIDTH Real m

X.LOC Real m

X.MAX Real m

X.MEAN Real 0.1 m

X.MIN Real m

X.SAGS Character
X.SEMI Real m

X.SIGMA Real 0.1 m

X.SOUT Character
Y.LOC Real m

Y.SEMI Real m

Y.SIGMA Real 0.1 m

Z.LOC Character
Z.MAX Real m

Z.MEAN Real 0.1 m

Z.MIN Real m

Z.SAGS Character
Z.SEMI Real m

Z.SIGMA Real 0.1 m

Z.SOUT Character

1104 ATLAS User’s Manual

LENS Statements

Description

ANGLE Specifies the elevation angke for all 4 sides of the pyramid(s) enabled by the GIZA flag.
APEX Specifies the height of a pyramid lenslet.
AR.INDEX Specifies the anti-reflective coating layer index of refraction for ray tracing.
AR.THICKNESS Specifies the anti-reflective coating layer thickness in microns for ray tracing.
ASPHERIC Specifies that the lenslet specification describes an aspheric lens.
COMPOSITE Specifies that the lenslet specification describes a composite lens.
DENSITY Specifies the density of primitives to be randomly placed when the parameter
RANDOM is specified along with one of the primitive flags (e.g., SPHERE, GIZA, or
CYLINDER).

ELLIPSE Specifies that the lenslet specification describes an elliptical lens.

EXTINCT Specifies the extinction coefficient of the lenslet.
F.LENS Specifies the file name of a C-Interpreter function for user-definable lenslets.
HEIGHT Specifies the height of a composite lens.
INDEX Specifies the real part of the index of refraction of the lenslet.
GIZA Specifies an equillateral pyramid with an elevation angle specified by the ANGLE
parameter.
K This is an alias for EXTINCT.
MATERIAL Specifies the material name used to determine the complex index of refraction
versus wavelength for the lens.
MEAN Specifies the mean value of the random step size in the X and Z directions in microns
for RANDOM lenslets.
N This is an alias for INDEX.
N.FEATURES Specifies the total number of primitive lens features to add when the RANDOM parameter
is specified along with one of the flags associated with one of the primitives (e.g.,
SPHERE, GIZA, or CYLINDER).

NEGATIVE Specifies that for spherical and elliptical lenses the concave surface is to be used rather
than the convex side.
NSAMP Specifies the number of samples for outputting aspheric lens data.
PC.DX Specify the periodicity in x and z directions for lenslet representation of a photonic
PC.DZ crystal.
PLANE Specifies the Y location of the planar background of the lenslet.
PYRAMID Specifies that the lenslet specification describes a Pyramid lens.

1105 ATLAS User’s Manual

LENS Statements

RADIUS Specifies the radius of a spherical lenslet.

RANDOM Specifies that the lenslet is a random textured surface lenslet.
SAG.OUT Specifies the file name for aspheric lens output.
SAG.ITS Specifies the number of iterations to be used for the Levenberg-Marquardt non-linear
least squares algorithm for aspheric lenslet specification.
SEED Specifies the random number generator seed value for random textured surface lenslets.
SIGMA Specifies the standard deviation of the Gaussian distributed random step size in the X,
Y, and Z directions for RANDOM lenslets in microns.
SPHERIC Specifies that the lenslet specification describes a spheric lens.
TWODEE Specifies that there is no variation in the lens in the z direction (i.e., the z
coordinate is set to 0 in the analytic formula describing the lens surface).
USER Specifies that the lenslet will use the C-Interpreter function specified by the F.LENS
parameter for a user-definable lenslet.
WIDTH Specifies the periphery width of a composite lenslet.
X.LOC Specify the center coordinates of a spheric or pyramid lenslet.
Y.LOC
Z.LOC

X.MEAN Specifies the mean value of the random step size in the X direction in microns for
RANDOM lenslets.

X.MIN Specify the perimeter of a composite or pyramid lenslet.

X.MAX
Z.MIN
Z.MAX

X.SAGS Specify the file names of the input sag data for an aspheric lenslet.
Z.SAGS

X.SEMI Specify the semimajor axes of a elliptical lenslet.

Y.SEMI
Z.SEMI

X.SIGMA Specifies the standard deviation of the Gaussian distributed random step size in the X
direction for RANDOM lenslets in microns.
X.SOUT Specify the output file names for the X and Z aspheric lenslet.
Z.SOUT

Y.SIGMA Specifies the standard deviation of the Gaussian distributed random step size in the Y
direction for RANDOM lenslets in microns.
Z.MEAN Specifies the mean value of the random step size in the Z direction in microns for
RANDOM lenslets.

Z.SIGMA Specifies the standard deviation of the Gaussian distributed random step size in the Z
direction for RANDOM lenslets in microns.

1106 ATLAS User’s Manual

LOAD Statements

21.29 LOAD
LOAD loads previous solutions from files as initial guesses to other bias points.
Syntax
LOAD [ASCII|MASTER] [NO.CHECK] <files>

Parameter Type Default Units

ASCII Logical False

IN.FILE Character
INFILE Character
IN1FILE Character
IN2FILE Character
LAS.SPECTRUM Logical False
MASTER Logical False
NO.CHECK Logical False
TWOD Logical False

Description
ASCII Specifies that any original PISCES format files read or written by this
statement will be in an ASCII rather than in a binary format.
LAS.SPECTRUM Loads the spectrum file (specified by the IN.FILE parameter) into LASER.

Note: The number of longitudinal and transverse modes in the spectrum file must be the same as created by the
LASER statement.

MASTER Specifies that any files read by this statement will be in a standard structure file rather
than the original PISCES format. Specify this parameter if you are using TONYPLOT to
plot simulation results.
NO.CHECK Prevents checking material parameter differences between loaded binary files and the
values set in the current input file.
TWOD Allows loading of a 2D solution into a 3D structure. Note that the values from the 2D
solution are loaded uniformly in the Z direction.

File Parameters
The LOAD statement requires that one of the following file parameter syntax be used.
LOAD INFILE=<filename>
or
LOAD IN1FILE=<filename> IN2FILE=<filename>

1107 ATLAS User’s Manual

LOAD Statements

IN.FILE This is a synonym for INFILE.

INFILE Specifies a single input filename for solution data. Use this parameter when you want
to load only one solution, which is the most common case. The synonym for this
parameter is IN.FILE.
IN1FILE Specifies a filename for present solution data. Use this parameter if two input files
are needed to perform an extrapolation for an initial approximation (i.e., the
PROJECT parameter of the SOLVE statement).

IN2FILE Specifies an input filename for previous solution data. Use this parameter if two
input files are needed to perform an extrapolation for an initial approximation (i.e.,
the PROJECT parameter of the SOLVE statement). The solution specified by this
parameter is the first to be overwritten when new solutions are obtained.

Simple Save and Load Examples

This example saves and loads the master format solution file, SOL.STR.
SAVE OUTF=SOL.STR.
....
LOAD INFILE=SOL.STR MASTER
As before but using the SOLVE syntax.
SOLVE OUTF=SOL.STR MASTER
..
LOAD INF=SOL.STR MASTER
When the save and load operations are not done within the same ATLAS run see the note
below.
Binary Format Example
Saving and loading using the binary format. This is quicker but these files cannot be plotted in
TONYPLOT.
SOLVE OUTF=SOLVE_TMP
..
LOAD INF=SOLVE_TMP

Note: The function to calculate the difference between two files is now inside TONY P LOT . It has been
discontinued from the LOAD statement

Note: The LOAD statement loads only the saved solution quantities into ATLAS. The mesh, electrodes, doping,
regions, contact settings, material parameters, models, and numerical methods must all be specified in
advance of any LOAD statement. See “Re-initializing ATLAS at a Given Bias Point” on page 95.

1108 ATLAS User’s Manual

LOG Statements

21.30 LOG
LOG allows all terminal characteristics of a run to be saved to a file. Any DC, transient, or AC
data generated by SOLVE statements after the LOG statement is saved. Any parameters
specified by the PROBE statement are also stored in the logfile. If a log file is already open,
the open log file is closed and a new log file is opened.
Syntax
LOG [OUTFILE=<filename>] [MASTER] [acparams]

Parameter Type Default Units

ABCD.PARAM Logical False

APPEND Logical False
CLOSE Logical False
CSVFILE Character
DEVDEG Logical False
FILE Character
GAINS Logical False
H.PARAM Logical False
IMPEDANCE Real 50 

INPORT Character
IN2PORT Character
J.DISP Logical False
J.ELECTRON Logical False
J.HEI Logical False
J.HHI Logical False
J.HOLE Logical False
J.TUN Logical False
LCOMMON Real 0 H
LGROUND Real 0 H
LIN Real 0 H
LOUT Real 0 H
MASTER Logical True
NO.TRAP Logical False

1109 ATLAS User’s Manual

LOG Statements

Parameter Type Default Units

NOISE Logical False

NOISE.ALL Logical False
NOISE.I Logical False
NOISE.IV Logical False
NOISE.I.ALL Logical False
NOISE.V Logical False
NOISE.VI Logical False
NOISE.V.ALL Logical False
OFF Logical False
OLD Logical False
OUTPORT Character
OUT2PORT Character
OUT.FILE Character
OUTFILE Character
RCOMMON Real 0 

RGROUND Real 0 

RIN Real 0 

ROUT Real 0 

S.PARAM Logical False

SIM.TIME Logical False s
SONOS.CURR Logical False
WIDTH Real 1 m

Y.PARAM Logical False

Z.PARAM Logical False

1110 ATLAS User’s Manual

LOG Statements

File Output Parameters

APPEND Specifies that the output I-V information should be appended to an existing
log file. Make sure that the existing log files contain the same type of data
(e.g., DC, AC, transient) as the subsequent SOLVE statements.
CLOSE This is an alias for OFF.
CSVFILE Specifies the file name that will be used to store DC, AC, or transient IV
information in Comma Seperated Value (CSV) format.
DEVDEG When used with Device degradation models, it causes the total interface
charges to be written to a LOG file.
FILE This is an alias for OUTFILE.
J.DISP Specifies that displacement currents are written to the log file.
J.ELECTRON Specifies that electron currents are to be written into the log file.
J.HEI Specifies that hot electron currents are written to the logfile.
J.HHI Specifies that hot hole currents are written to the logfile.
J.HOLE Specifies that hole currents are to be written into the log file.
J.TUN Specifies that tunneling currents are to be written into the log file.
MASTER Specifies that AC data and I-V information will be saved in a standard
structure file format. This is the default format.
NO.TRAP Specifies that no data will be written to the file for DC bias cut back steps.
OFF Specifies that any currently open log file will be closed and log file
output is discontinued. The alias for this parameter is CLOSE.
OLD Specifies that AC data and IV information will be saved in the original
PISCES-II file format. A synonym for this parameter is PISCES.
OUT.FILE This is a synonym for OUTFILE.
OUTFILE Specifies the log file that will be used to store DC, AC, or transient I-V
information. The synonym for this parameter is OUT.FILE.
SONOS.CURR Causes the output of SONOS Tunneling Insulator Current and SONOS
Blocking Insulator Current from the DYNASONOS model.

Note: The older ACFILE syntax is not supported and should not be used. AC results are stored in the file
specified by OUTFILE as long as the first SOLVE statement after the LOG statement contains AC
analysis.

1111 ATLAS User’s Manual

LOG Statements

RF Analysis Parameters
If S.PARAM, H.PARAM, Z.PARAM, GAINS, or ABCD.PARAM is specified, the capacitance and
conductance data will be converted into the requested set of AC parameters. See Appendix C
“RF and Small Signal AC Parameter Extraction” for more information.

S.PARAM Elects s parameter analysis. For S-parameter analysis, you can also choose to set
any of the parasitic element parameters.
H.PARAM Selects H parameter analysis.
Y.PARAM Selects Y parameter analysis
Z.PARAM Selects Z parameter analysis.
ABCD.PARAM selects ABCD parameter analysis.
GAINS Selects the calculation of several types of gains used in RF analysis [222]. These
are the stability factor, unilateral power gain (GUmax), maximum unilateral
transducer power gain (GTmax), maximum available power gain (Gma), and the
maximum stable power gain (Gms). The magnitude of H21 is also
calculated.
IMPEDANCE Specifies the matching impedance for s-parameter calculation.
INPORT Specifies the electrode name for the primary input port used when performing any
AC parameter calculations.

IN2PORT Specifies the electrode name of the secondary input port.

OUTPORT Specifies the electrode name for the output ports used when performing any AC
parameter calculations.
OUT2PORT Specifies the electrode n of the secondary output port.

NOISE Parameters
To perform noise analysis on a one-port device, define the INPORT. To perform noise analysis
on a two-port device, define the INPORT and the OUTPORT.

NOISE Selects Fmin, ZO, and gn (the two-port noise figures of merit, 2pNFOM) for output.
NOISE.ALL Selects all noise results for output. These are 2pNFOM, total and individual noise
voltage sources, and total and individual noise current sources.
NOISE.I Selects the 2pNFOM and the correlation of the total noise current sources.
NOISE.I.ALL Selects the 2pNFOM, the correlation of the total noise current sources, and the
correlation of the noise current sources from the individual noise mechanisms (GR,
II, electron and hole diffusion; electron and hole flicker).

1112 ATLAS User’s Manual

LOG Statements

NOISE.V Selects the 2pNFOM and the correlation of the total noise voltage sources.
NOISE.VI Selects the 2pNFOM and the correlations of both the total noise voltage sources and
NOISE.IV the total noise current sources.
NOISE.V.ALL Selects the 2pNFOM, the correlation of the total noise voltage sources, and the
correlation of the noise voltage sources from the individual noise mechanisms (GR,
II, electron and hole diffusion; electron and hole flicker).

Note: The 2pNFOM have no meaning for a one-port device and therefore don’t output. If NOISE is the only
parameter defined for a one-port device, then the correlation of the total noise voltage sources outputs
(rather than outputting nothing).

Parasitic Element Parameters

For RF parameter extraction, you can also set any of the parasitic element parameters. By
setting the parasitic element parameters, you can apply lumped parasitic resistances or
inductances to the terminal of the two-port device during the RF parameter extraction. These
parameters will not affect the capacitance or conductance matrices calculated by ATLAS.

RIN Specifies the lumped parasitic resistance on the input to the two port device for s-
parameter extraction. The value of RIN is in Ohms.
ROUT Specifies the lumped parasitic resistance on the output to the two port device for s-
parameter extraction. The value of ROUT is in Ohms.
RCOMMON This is an alias for RGROUND.
RGROUND Specifies the lumped parasitic resistance on the ground or common side of the two
port device for s-parameter extraction. The value of RGROUND is in Ohms.
LIN Specifies the lumped parasitic inductance on the input to the two port device for s-
parameter extraction. The value of LIN is in Henrys.
LOUT Specifies the lumped parasitic inductance on the output to the two port device for s-
parameter extraction. The value of LOUT is in Henrys.
LCOMMON This is an alias for LGROUND.
LGROUND Specifies the lumped parasitic inductance on the ground or common side of the two
port device for s-parameter extraction. The value of LGROUND is in Henrys.
SIM.TIME Saves the time taken for a bias point into the log file and measured in seconds. Note
that if multiple jobs are using the same CPU that this method may not be a true
reflection of processor speed.
WIDTH Specified an output width (in Z direction) to apply during the s-parameter
calculation. Note that this parameter affects only the derived RF parameters and not
currents, capacitances or conductances. The WIDTH parameter of the MESH statement
can be used to scale these. Using both these WIDTH parameters will lead to a
multiplication of the two widths for the RF parameters.

1113 ATLAS User’s Manual

LOG Statements

Simple Logfile Definition Example

This example saves all I-V data in file, myfile.log.
LOG OUTF=myfile.log
Results should be plotted using TONYPLOT.
RF Analysis Example
To generate s-parameters, assume the input is gate/source and the output is drain/source. A
width of 100 microns is also defined along with 100ohm resistance on the input. For example:
LOG OUTF=mysparams.log S.PARAM INPORT=gate OUTPORT=drain \
IN2PORT=source OUT2PORT=source WIDTH=100 RIN=100
Transient or AC Logfile Example
The contents of LOG files varies for different types of simulations (e.g., DC, transient, AC).
The content is set by the first SOLVE statement after the LOG statement. Therefore, the
following syntax is required.
SOLVE VDRAIN=0.5
LOG OUTF=myfile.log
SOLVE VDRAIN=3.0 RAMPTIME=1e-9 DT=1e-11 TSTOP=1e-7
Correct transient parameters would have not been stored, if the LOG statement had been
placed before the first SOLVE statement, which is DC.

1114 ATLAS User’s Manual

LX.MESH, LY.MESH Statements

21.31 LX.MESH, LY.MESH

L<n>.MESH specifies the location of grid lines along the <n>-axis in a rectangular mesh used
in LASER simulation. The syntax is equivalent for X and Y directions.
Syntax
LX.MESH NODE=<n> LOCATION=<n>
LX.MESH SPACING=<n> LOCATION=<n>
LY.MESH NODE=<n> LOCATION=<n>
LY.MESH SPACING=<n> LOCATION=<n>

Parameter Type Default Units

LOCATION Real m
NODE Integer
SPACING Real m

LASER Mesh Example

This syntax defines a mesh of 3333 covering the area bounded by (0.3,0.0) to (2.4,1.0).
LX.M n=1 l=0.3
LX.M n=33 l=2.4
LY.M n=1 l=0.0
LY.M n=33 l=1.0

Note: The mesh defined in these statements for the Laser Helmholtz Solver is entirely separate from the electrical
device simulation mesh defined on the MESH statement.

Setting Locally Fine Grids Example

This example shows how to space grid lines closely around a topological feature such as a
junction at y=0.85 microns.
LY.MESH LOC=0.0 SPAC=0.2
LY.MESH LOC=0.85 SPAC=0.01
LY.MESH LOC=2 SPAC=0.35

1115 ATLAS User’s Manual

MATERIAL Statements

21.32 MATERIAL
MATERIAL associates physical parameters with materials in the mesh. The parameter default
values for standard semiconductors are shown in Appendix B “Material Systems”.
Syntax
MATERIAL <localization> <material_definition>

BN1.VALD Real 0.0

BN0.VALD Real 0.235
BP Real 1.0
BP2 Real 2.036106 V/cm

BP0.VALD Real 0.17714

BP1.VALD Real -0.002178
BQP.NGAMMA Real 1.2
BQP.NALPHA Real 0.5
BQP.PGAMMA Real 1.0
BQP.PALPHA Real 0.5
BSTR Real 0.0
C.DEFPOT Real -4.75 eV
C.DIRECT Real 0.0 cm3/s

1119 ATLAS User’s Manual

MATERIAL Statements

Parameter Type Default Units

C1.RAJ Real 5.5

C2.RAJ Real 4.0
C11 Real 0.0 GPa
C12 Real 0.0 GPa
C13 Real 0.0 GPa
C33 Real 0.0 GPa
CDL.COUPLING Real 1.0 s-1cm-3
CDL.ETR1 Real 0.0 eV
CDL.ETR2 Real 0.0 eV
CDL.TN1 Real 1.0 s
CDL.TN2 Real 1.0 s
CDL.TP1 Real 1.0 s
CDL.TP2 Real 1.0 s
CHI.EG.TDEP Real 0.5
CON.BGN Real 0.5
CN Real 0.0
CN2 Real 0.0 cm-1K-1
CN0.VALD Real 1.6831104
CN1.VALD Real 4.3796
CN2.VALD Real 1.0
CN3.VALD Real 0.13005
COPT Real 0.0 cm3/s
CP Real 0.0
CP2 Real 0.0 cm-1K-1
CP0.VALD Real 0.0
CP1.VALD Real 0.00947
CP2.VALD Real 2.4929
CP3.VALD Real 0.0
CRELMAX.N Real 0.5

1120 ATLAS User’s Manual

MATERIAL Statements

Parameter Type Default Units

CRELMAX.P Real 0.5

CRELMIN.N Real 1e-10
CRELMIN.P Real 1e-10
CSTR Real 0.0
D.DEFPOT Real 1.1 eV
D.TUNNEL Real 10-6 cm

D0.H1 Real 1.7510-4 cm2s-1

D0.H2 Real 1.7510-4 cm2s-1
D1 Real 0.0 eV
D2 Real 0.0 eV
D3 Real 0.0 eV
D4 Real 0.0 eV
DADACHI Real see Appendix B eV
DEC.DEG Real see Table 5-13
DEC.C1 Real 0.0 eV
DEC.C2 Real 0.0 eV
DEC.C3 Real 0.0 eV
DEC.ISO Real 0.0 eV
DEC.D2 Real 0.0 eV
DEC.D4 Real 0.0 eV
DEV.HH Real 0.0 eV
DEV.LH Real 0.0 eV
DEV.ISO Real 0.0 eV
DEV.SO Real 0.0 eV
DEGENERACY Integer 2
DELTA1 Real 0.0 eV
DELTA2 Real 0.0 eV
DELTA3 Real 0.0 eV
DEVICE Character

1121 ATLAS User’s Manual

MATERIAL Statements

Parameter Type Default Units

DGN.GAMMA Real 3.6

DGP.GAMMA Real 3.6
DINDEXDT Real 0.0 1/K
DIST.SBT Real 10-6 cm

DKCQ.EXCITON Real 1.110-8 s-1

DKNRS.EXCITON Real 0 s-1
DKSS.EXCITON Real 310-8 cm3s-1
DN2 Real 0.0 cm-1K-1
DN0.VALD Real 1.233735106
DN1.VALD Real 1.2039103
DN2.VALD Real 0.56703
DOPE.SPECT Character
DP2 Real 0.0 cm-1K-1
DP0.VALD Real 1.4043106
DP1.VALD Real 2.9744103
DP2.VALD Real 1.4829
DPHEFF.EXCITON Real 1
DRST.EXCITON Real 0.25
DTAUS.EXCITON Real 510-9 s

DRHODT Real 0.0 µcm/k

DSTR Real 0.0
E.CONDUC Real 0.0 /cm
E.FULL.ANISO Logical False E.FULL.ANISO
E31 Real 0.0 cm-2
E33 Real 0.0 cm-2
EA.CUBIC Real See Table 5-24. eV
EA.H1 Real 0.1685 cm2s-1

1122 ATLAS User’s Manual

MATERIAL Statements

1125 ATLAS User’s Manual

MATERIAL Statements

Parameter Type Default Units

FERRO.PS Real 0.0 C/cm2

FERRO.PR Real 0.0 C/cm2
FSTR Real 0.0
G.2D.LANGEVIN Real 1
GAIN.SAT Real 1.0107
GAIN0 Real 2000.0 cm-1
GAIN00 Real -200.0 cm-1
GAIN1N Real 0 cm2
GAIN1P Real 0 cm2
GAIN1MIN Real 3.010-16 cm2
GAIN2NP Real 0 cm5
GAMMA Real
GAMMA.RAJ Real 1108.0 K
GCB Real 2.0
GN1 Real 3.010-16 cm-2
GN2 Real 4.010-15 cm-2
G3MAX.N Real 100.0
G3MAX.P Real 100.0
G.SURF Real 1.0 cm2
GVB Real 4.0
H1TOH2RATE Real 1.0E-2 cm-3s-1
H2TOH1RATE Real 100.0 s-1
H1SRV Real 0.0 cm/s
H2SRV Real 0.0 cm/s
HC.A Real See Appendix B J/Kcm3
HC.B Real See Appendix B J/K2cm3
HC.BETA Real

1126 ATLAS User’s Manual

MATERIAL Statements

Parameter Type Default Units

HC.C Real See Appendix B J/K3cm3

HC.C1 Real J/K/kg
HC.C300 Real J/K/kg
HC.D Real See Appendix B JK/cm3
HC.RHO Real g/cm3
HOOGEN Real 0.0
HNS.AE Real 6.710-32 cm6/s

HNS.AH Real 7.210-32 cm6/s

HNS.BE Real 2.4510-31 cm6/s

HNS.BH Real 4.510-33 cm6/s

HNS.CE Real -2.210-32 cm6/s

HNS.CH Real 2.6310-32 cm6/s

HNS.HE Real 3.4667

HNS.HH Real 8.25688
HNS.N0E Real 1.01018 cm-3

HNS.N0H Real 1.01018 cm-3

HNU1 Real 1.0 eV

HNU2 Real 1.0 eV
HOOGEP Real 0.0
HOPN.BETA Real 1.5
HOPN.GAMMA Real 510-7 cm-1
HOPN.V0 Real 11011 Hz

HOPP.BETA Real 1.5

HOPP.GAMMA Real 11011 cm-1
HOPP.V0 Real 11011 Hz

HUANG.RHYS Real 3.5

IF.CHAR Real 0.0 microns

1127 ATLAS User’s Manual

MATERIAL Statements

Parameter Type Default Units

IG.ELINR Real 6.1610-6 cm

IG.HLINR Real 6.1610-6 cm

IG.ELINF Real 9.210-7 cm

IG.HLINF Real 9.210-7 cm

IINOFF Character
IIPOFF Character
IMAG.INDEX Real See Appendix B
INDEX.FILE Character
INDX.IMAG Character
INDX.REAL Character
INSULATOR Logical False
J.ELECT Real 0.0 A/cm2
J.MAGNET Real 0.0 V/cm2
JTAT.M2 Real 10-20 V2cm3

K.SHIFT Real 0.0 eV

K.SINGLET Real
KAUGCN Real 1.8310-31 cm6/s
KAUGCP Real 2.7810-31 cm6/s
KAUGDN Real 1.18
KAUGDP Real 0.72
KD.LID Real 510-15 cm3s-1
KH.LID Real 510-18 cm3s-1
KISC.EXCITON Real 0 s-1
KNRS.EXCITON Real 0 s-1
KNRT.EXCITON Real 0 s-1
KSP.EXCITON Real 0 cm-3s-1
KSRHCN Real 3.010-13 cm3

1128 ATLAS User’s Manual

MATERIAL Statements

Parameter Type Default Units

KSRHCP Real 11.7610-13 cm3

KSRHGN Real 1.77
KSRHGP Real 0.57
KSRHTN Real 2.510-3 s

KSRHTP Real 2.510-3 s

KSS.EXCITON Real 0 cm-3s-1

KST.EXCITON Real 0 cm-3s-1
KTP.EXCITON Real 0 cm-3s-1
KTT.EXCITON Real 0 cm-3s-1
L.2D.LANGEVIN Real 1.6 nm
L1SELL Real Appendix B µm
L2SELL Real Appendix B µm
LAM1 Real 0 µm
LAM2 Real 0 µm
LAMDAE Real 6.210-7 cm

LAMDAH Real 3.810-7 cm

LAMHN Real 9.210-7 cm

LAMHP Real 9.210-7 cm

LAMRN Real 6.1610-6 cm

LAMRP Real 6.1610-6 cm

LAN300 Real 6.210-7 cm

LAP300 Real 3.810-7 cm

LDS.EXCITON Real 0.01 µm

LDT.EXCITON Real 0.632 µm
LT.TAUN Real 0.0
LT.TAUP Real 0.0
LUTT1 Real 0.0

1129 ATLAS User’s Manual

MATERIAL Statements

Parameter Type Default Units

LUTT2 Real 0.0

LUTT3 Real 0.0
M.DSN Real See Appendix B
M.DSP Real See Appendix B
M.VTHN Real
M.VTHP Real
MATERIAL Character
MAG.LOSS Real 0.0 mho/cm
MBULKSQ Real eV
MC Real 0.0
ME.TUNNEL Real
MH.TUNNEL Real
ME.SBT Real
MH.SBT Real
MHH Real 0.49
MINIMA Real 6
ML Real 0.916
MLH Real 0.16
MSO Real 0.23
MSTAR Real 0.0
MTT Real 0.0
MT1 Real 0.191
MT2 Real 0.191
MUN Real See Appendix B
MUNOFF Character
MUPOFF Character
MUP Real See Appendix B
MV Real 0.0
MZZ Real 0.0

1130 ATLAS User’s Manual

MATERIAL Statements

Parameter Type Default Units

N.ANISO Logical False

N.ALPHA Real 0.0 degrees
N.BETA Real 0.0 degrees
N.GAMMA Real 0.0 degrees
N.ION.1 Real 0.0 cm-1K-1
N.ION.2 Real 0.0 cm-1K-2
N.IONIZA Real 7.03105 cm-1
N.SHIFT Real 0.0 eV
N.XX Real 1.0
N.YY Real 1.0
N.ZZ Real 1.0
NK.SHIFT Real 0.0 eV
N0.BGN Real 1.01017 cm-3
NAME Character
NC.F Real 1.5
NC300 Real See Appendix B cm-3
NDX.ADACHI Logical False
NDX.SELLMEIER Logical False
NHNU Integer 0
NI.MIN Real 0.0 cm-3
NK.EV Logical False
NK.NM Logical False
NLAM Integer 0
N.SCH.GAMMA Real 1.0 s
N.SCH.MAX Real 3.010-6 s

N.SCH.MIN Real 0.0 s

N.SCH.NREF Real 1.01016 cm-3
NSRHN Real See Appendix B cm-3

1131 ATLAS User’s Manual

MATERIAL Statements

Parameter Type Default Units

NSRHP Real See Appendix B cm-3

NTC.GAUSS Real NTV.GAUSS/1.0e21 cm-3
NTV.GAUSS Real NTC.GAUSS/1.0e21 cm-3
NTRANSPARENT Real 2.01018 cm-3
NUE.EXTR Real
NUH.EXTR Real
NV.F Real 1.5
NV300 Real See Appendix B cm-3
OP.PH.EN Real 0.063 eV
OPPHE Real 0.063 eV
OUT.DSPEC Character
OUT.INDEX Character
OUT.USPEC Character
OXCH.ONLY Logical False
P.ION.1 Real 0.0 cm-1K-1
P.ION.2 Real 0.0 cm-1K-1
P.IONIZA Real 1.682106 cm-1
P.PASSLER Real 2.33
P.SCH.GAMMA Real 1.0 s
P.SCH.MAX Real 3.010-6 s

P.SCH.MIN Real 0.0 s

P.SCH.NREF Real 1.01016 cm-3
PCM.AEA Real 3.2 eV
PCM.AP Real 1.0
PCM.ARHO Real cm

PCM.ATAU Real S
PCM.ATC Real K
PCM.CEA Real 3.7 eV

1132 ATLAS User’s Manual

MATERIAL Statements

Parameter Type Default Units

PCM.CP Real 1.0

PCM.CRHO Real cm

PCM.CTAU Real S
PCM.CTC Real K
PCM.LATHEAT Real 0.0 J
PDA.N Real 0.2
PDA.P Real 0.2
PDEXP.N Real -2.5
PDEXP.P Real -2.5
PERM.ANISO Real -999.0
PERMEABILITY Real 1.0
PERMITTIVITY Real See Appendix B
PHEFF.EXCITON Real 1
PHONON.ENERGY Real 0.068 eV
PIP.ACC Real 2.0
PIP.ET Real 1.0 eV
PIP.NT Real 0.0 cm2
PIP.OMEGA Real 0.07 eV
POWER Real
PSP Real 0.0 cm-2
QE.EXCITON Real 0.0
QR.EXCITON Real 0.0
REAL.INDEX Real See Appendix B
REGION All regions
RESISTIVITY Real µcm
RST.EXCITON Real 0.25
RST.TT.EXCITON Real RST.EXCITON

S.BINDING Real 0.1 eV

S0SELL Real See Appendix B

1133 ATLAS User’s Manual

MATERIAL Statements

Parameter Type Default Units

S1SELL Real See Appendix B

S2SELL Real See Appendix B
SIG.LID Real 1.010-8 cm-3s-1
SIGC.GAUSS Real SIGV.GAUSS/0.1 eV
SIGV.GAUSS Real SIGC.GAUSS/0.1 eV
SEMICONDUC Logical False
SO.DELTA Real see Table 3-135 eV
SOPRA Character
SPECIES1.AF Real 1.0e15 Hz
SPECIES1.EA Real 0.25 eV
SPECIES1.HOP Real 1.0e-6 cm
SPECIES2.AF Real 1.0e15 Hz
SPECIES2.EA Real 0.25 eV
SPECIES2.HOP Real 1.0e-6 cm
SPECIES3.AF Real 1.0e15 Hz
SPECIES3.EA Real 0.25 eV
SPECIES3.HOP Real 1.0e-6 cm
STABLE Integer
STRUCTURE Character
T.PASSLER Real 406.0
TAA.CN Real 1.010-12 cm3/s
TAA.CP Real 1.010-12 cm3/s
TANI.CONST Logical False
TANI.POWER Logical False
TANI.POLYNOM Logical False
TANI.RECIP Logical false
TAUMOB.EL Real 0.410-12 s

TAUMOB.HO Real 0.410-12 s

TAUN0 Real See Appendix B s

1134 ATLAS User’s Manual

MATERIAL Statements

Parameter Type Default Units

TAUP0 Real See Appendix B s

TAUREL.EL Real 0.410-12 s

TAUREL.HO Real 0.410-12 s

TAUS.EXCITON Real 110-9 s

TAUT.EXCITON Real 110-9 s

TC.A Real See Appendix B cm --

------ K
W

TC.B Real See Appendix B cm

---
W

TC.C Real See Appendix B cm

--
----
WK

TC.CONST Real W/cmK

TC.NPOW Real
TC.D Real K
TC.E Real W/cm
TC.FULL.ANISO Logical False
TCOEFF.N Real 2.55
TCOEFF.P Real 2.55
TCON.CONST Logical
TCON.POWER Logical
TCON.POLYNOM Logical
TCON.RECIPRO Logical
TE.MODES Logical False
TLU.A Real 0 eV
TLU.C Real 0 eV
TLU.E0 Real 0 eV
TLU.EC Real 0 eV
TLU.EG Real 0 eV
TLU.EPS Real 1.0
TMUN Real

1135 ATLAS User’s Manual

MATERIAL Statements

Parameter Type Default Units

TMUP Real
TRE.T1 Real s
TRE.T2 Real s
TRE.T3 Real s
TRE.W1 Real eV
TRE.W2 Real eV
TRE.W3 Real eV
TRH.T1 Real s
TRH.T2 Real s
TRH.T3 Real s
TRH.W1 Real eV
TRH.W2 Real eV
TRH.W3 Real eV
U.DEFPOT Real 10.5 eV
UBGN.B Real 3.11012
UBGN.C Real -3.910-5
USER.DEFAULT Character
USER.GROUP Character SEMICONDUCTOR
USER.SPECT Character
V0.BGN Real 9.010-3 eV

VAL.AN0 Real 4.3383

VAL.AN1 Real -2.4210-12

VAL.AN2 Real 4.1233

VAL.AP0 Real 2.376

VAL.AP1 Real 0.01033

VAL.AP2 Real 1.0

VAL.BN0 Real 0.235

VAL.BN1 Real 0.0

1136 ATLAS User’s Manual

MATERIAL Statements

Parameter Type Default Units

VAL.CN0 Real 1.6831104

VAL.CN1 Real 4.3796

VAL.CN2 Real 1.0

VAL.CN3 Real 0.13005

VAL.DN0 Real 1.233735106

VAL.DN1 Real 1.2039e3

VAL.DN2 Real 0.56703

VAL.BP0 Real 0.17714

VAL.BP1 Real -0.002178

VAL.CP0 Real 0.0

VAL.CP1 Real 0.00947

VAL.CP2 Real 2.4924

VAL.CP3 Real 0.0

VAL.DP0 Real 1.4043106

VAL.DP1 Real 2.9744103

VAL.DP2 Real 1.4829

VSAT Real cm/s

VSATN Real cm/s
VSATP Real cm/s
WELL.DELTA Real see Table 3-135 eV
WELL.EPS Real 0 eV
WELL.GAMMA0 Real 210-3 eV

WELL.TAUIN Real 3.310-13 s

WELL.TAUN Real 1e-12 sec

WELL.TAUP Real 1e-12 sec
X.X Real 1.0
X.Y Real 0.0
X.Z Real 0.0

1137 ATLAS User’s Manual

MATERIAL Statements

Parameter Type Default Units

XDIR.ANISO Logical False

Y.X Real 0.0
Y.Y Real 1.0
Y.Z Real 0.0
YDIR.ANISO Logical False
ZDIR.ANISO Logical True
Z.SCHENK Real 1.0
Z.X Real 0.0
Z.Y Real 0.0
Z.Z Real 1.0
ZAMDMER.Z0 Real 7.010-7 cm

Description
The MATERIAL statement is used set basic material parameters related to band structure and
parameters for certain mobility, recombination or carrier statistics models. Parameters for
temperature dependence are noted in a separate section below.
Localization of Material Parameters

DEVICE Specifies which device the MATERIAL statement should apply to in

MIXEDMODE simulation. The synonym for this parameter is STRUCTURE.
MATERIAL Specifies which material from the table in Appendix B “Material Systems” that
the MATERIAL statement should apply. If a material is specified, then all regions
defined as being composed of that material will be affected.

Note: You can specify the following logical parameters to indicate the material instead of assigning the
MATERIAL parameter: SILICON, GAAS, POLYSILI, GERMANIU, SIC, SEMICOND, SIGE,
ALGAAS, A-SILICO, DIAMOND, HGCDTE, INAS, INGAAS, INP, S.OXIDE, ZNSE, ZNTE,
ALINAS, GAASP, INGAP and MINASP.

1138 ATLAS User’s Manual

MATERIAL Statements

NAME Specifies which region the MATERIAL statement should apply. Note that
the name must match the name specified in the NAME parameter of the
REGION statement or the region number.

REGION Specifies the region number to which these parameters apply. If there is
more than one semiconductor region, specification of different
parameters for each region is allowed. If REGION is not specified, all
regions in the structure are changed.
STRUCTURE This is a synonym for DEVICE.

Band Structure Parameters

A.DEFPOT, B.DEFPOT, C.DEFPOT, Specify the deformation potentials used in calculating the
D.DEFPOT, and U.DEFPOT effects of strains on the bandgap of silicon (see Section 3.6.13
“Stress Effects on Bandgap in Si”).
AFFINITY Specifies the electron affinity.
ALIGN Specifies the fraction of the bandgap difference that is applied
to the conduction band edge, relative to the minimum bandgap
material in the device. Note that specifying this parameter
overrides any electron affinity specification. See Section 5.1.2
“Alignment” for information on setting the band alignment.
ARICHN Specifies the effective Richardson constant for electrons.
ARICHP Specifies the effective Richardson constant for holes.
CHI.EG.TDEP This is a ratio that specifies what fraction of the change in
bandgap due to temperature change ascribed to the electron
affinity. CHI.EG.TDEP is not active when ALIGN is specified.
CRELMAX.N and CRELMAX.P Specifies the maximum normalized electron and hole densities
used to calculate g3 for Gaussian band structures.

CRELMIN.N and CRELMIN.P Specifies the minimum normalized electron and hole densities
used to calculate g3 for Gaussian band structures.

D.TUNNEL Specifies the maximum tunneling distance for the universal

Schottky tunneling model (see Section 3.5.2 “Schottky
Contacts”). The alias for this parameter is DIST.SBT.
DEC.DEG Overrides the default values of dEc/dEv given in Table 5-13.
DEC.C1 This is a conduction band shift for 1st pair of electron
valleys.
DEC.C2 This is a conduction band shift for 2nd pair of electron
valleys.

1139 ATLAS User’s Manual

MATERIAL Statements

DEC.C3 This is a conduction band shift for 3rd pair of electron

valleys.
DEC.D2 This is a conduction band shift for 2 electron valleys if
DEC.C1 is not set.

DEC.D4 This is a conduction band shift for 4 electron valleys if

DEC.C2 is not set.

DEC.ISO This is a conduction band shift for isotropic electron band,

when NUM.DIRECT=1.
DEV.HH This is a valence band shift for heavy holes.
DEV.LH This is a valence band shift for light holes.
DEV.ISO This is a valence band shift for isotropic hole band, when
NUM.BAND=1.

DEV.SO This is a valence band shift for split-off holes.

DIST.SBT This is an alias for D.TUNNEL.
EG300 Specifies energy gap at 300K (see Equation 3-38). All
semiconductor materials in ATLAS must have a defined EG300.
EG1300, EG2300, EG12BOW, Specify parameters of the General Ternary bandgap model
EG1ALPH, EG2ALPH, EG1BETA, and described in Equations 3-42 through 3-44.
EG2BETA

EPS11, EPS12, EPS13, EPS22, Specify the strain tensor used in the calculation of the strain
EPS23, and EPS33 effects on bandgap of silicon (see Section 3.6.13 “Stress Effects
on Bandgap in Si”).
EPS.XX XX component of dielectric permittivity tensor used by vector
Helmholtz solver.
EPS.YY YY component of dielectric permittivity tensor used by vector
Helmholtz solver.
EPS.ZZ ZZ component of dielectric permittivity tensor, used by vector
Helmholtz solver.
F.BANDCOMP Specifies the name of a file containing a C-Interpreter
function that defines temperature and composition
dependent band parameter models.
F.CBDOSFN Specifies the C-Interpreter file for the Conduction band
effective density of states as a function of Lattice/Electron
temperature. This function is called cbdosfn.
F.TCAP Specifies the name of a file containing a C-Interpreter
function that defines the lattice thermal capacity as a
function of the lattice temperature, position, doping and fraction
composition.

1140 ATLAS User’s Manual

MATERIAL Statements

F.TCOND Specifies the name of a file containing a C-Interpreter

function that defines the lattice thermal conductivity as a
function of the lattice temperature, position, doping and fraction
composition.
F.EPSILON Specifies the name of a file containing a C-Interpreter
function that defines temperature and composition dependent
static dielectric constant models.
F.FERRO Specifies the name of a file containing a C-Interpreter
function that defines dielectric permittivity as a function of
electric field and position (x,y). You need a license to use this
parameter.
F.VBDOSFN Specifies the C-Interpreter file for the Valence band effective
density of states as a function of Lattice/Hole temperature. This
function is called vbdosfn.
G3MAX.N and G3MAX.P Specifies the maximum value of g3 for Gaussian band
structures.
M.DSN Specifies the electron density of states effective mass. When
specified, it is in Equation 3-31 to calculate electron density of
states.
M.DSP Specifies the hole density of states effective mass. When
specified it is in Equation 3-32 to calculate hole density of
states.
M.VTHN Specifies the electron effective mass for calculation of
thermal velocity in the thermionic heterojunction model (see
Equation 5-52).
M.VTHP Specifies the hole effective mass for calculation of thermal
velocity in the thermionic heterojunction model (see Equation
5-53).
ME.SBT and MH.SBT These are aliases for ME.TUNNEL and MH.TUNNEL.
ME.TUNNEL and MH.TUNNEL Specify the electron and hole effective masses for tunneling
used in the universal Schottky tunneling model (see Section
3.5.2 “Schottky Contacts”). The aliases for these parameters are
ME.SBT and MH.SBT.

NC.F Specifies the conduction band density of states temperature (see

Equation 3-31).
NV.F Specifies the valence band density of states temperature (see
Equation 3-32).
NC300 Specifies the conduction band density at 300K. (see Equation 3-
31).

1141 ATLAS User’s Manual

MATERIAL Statements

NI.MIN Specifies the minimum allowable value of the intrinsic carrier

density.
NTC.GAUSS and NTV.GAUSS Specifies the total density of states of a conduction and valence
band Gaussian DOS.
NV300 Specifies valence band density at 300K (see Equation 3-32).
PERMITTIVITY Specifies dielectric permittivity of the material. All
materials in an ATLAS structure must have a defined
permittivity.
SIGC.GAUSS and SIGV.GAUSS Specifies the standard deviation of a conduction and valence
band Gaussian DOS.

L.2D.LANGEVIN Specifies the prefactor for the Juska 2D Langevin recombination
model. See Equation 15-53.
N.SCH.MIN These are used in the Scharfetter doping concentration
P.SCH.MIN dependent lifetime model. They are applied in Equations 3-337
N.SCH.MAX and 3-338.
P.SCH.MAX
N.SCH.GAMMA
P.SCH.GAMMA
N.SCH.NREF
P.SCH.NREF

TCOEFF.N and TCOEFF.P These are used in the alternative lifetime lattice temperature
dependence model.
TAA.CN and TAA.CP These are used in the carrier concentration dependent
recombination lifetime model of Equations 3-347 and 3-348.
ZAMDMER.Z0 The nominal spatial separation of recombination centers in the
ZAMDMER model of electron trap lifetime in LT-GaAs.

1145 ATLAS User’s Manual

MATERIAL Statements

Impact Ionization Parameters

AN0.VALD This is an alias for VAL.AN0.

AN1.VALD This is an alias for VAL.AN1.
AN2.VALD This is an alias for VAL.AN2.
AP0.VALD This is an alias for VAL.AP0.
AP1.VALD This is an alias for VAL.AP1.
AP2.VALD This is an alias for VAL.AP2.
BETAN This is for electrons and BETAP for holes correspond to coefficients for the
power of ECRIT/E. The aliases for these parameters are EXN.II and
EXP.II.

BN0.VALD This is an alias for VAL.BN0.

BN1.VALD This is an alias for VAL.BN1.
BP0.VALD This is an alias for VAL.BP0.
BP1.VALD This is an alias for VAL.BP1.
CN2, CP2, DN2, and DP2 These are specifiable coefficients in the temperature dependent models
described in Equations 3-397 and 3-398. The aliases for these parameter
are N.ION.1, P.ION.1, N.ION.2, and P.ION.2.
CN0.VALD This is an alias for VAL.CN0.
CN1.VALD This is an alias for VAL.CN1.
CN2.VALD This is an alias for VAL.CN2.
CN3.VALD This is an alias for VAL.CN3.
CP0.VALD This is an alias for VAL.CP0.
CP1.VALD This is an alias for VAL.CP1.
CP2.VALD This is an alias for VAL.CP2.
CP3.VALD This is an alias for VAL.CP3.
DN0.VALD This is an alias for VAL.DN0.
DN1.VALD This is an alias for VAL.DN1.
DN2.VALD This is an alias for VAL.DN2.
DP0.VALD This is an alias for VAL.DP0.
DP1.VALD This is an alias for VAL.DP1.
DP2.VALD This is an alias for VAL.DP2.

1146 ATLAS User’s Manual

MATERIAL Statements

ECN.II This is an alias for BN2.

ECP.II This is an alias for BP2.
EXN.II This is an alias for BETAN.
EXP.II This is an alias for BETAP.
LAMBDAE Specifies the mean free path for electrons. The alias for this parameter is
LAN300.

LAMDAH Specifies the mean free path for holes. The alias for this parameter is
LAP300.

LAN300 This is an alias for LAMBDAE.

LAP300 This is an alias for LAMBDAH.
N.ION.1 These are the aliases for CN2, CP2, DN2 and DP2.
P.ION.1
N.ION.2
P.ION.2

N.IONIZA This is an alias for AN2.

OP.PH.EN This is an alias for OPPHE.
OPPHE Specifies the optical phonon energy. The alias for this parameter is
OP.PH.EN.

P.IONIZA This is an alias for AP2.

VAL.AN0 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-408. The alias for this parameter is AN0.VALD.
VAL.AN1 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-408. The alias for this parameter is AN1.VALD.
VAL.AN2 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-408. The alias for this parameter is AN2.VALD.
VAL.BN0 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-409. The alias for this parameter is BN0.VALD.
VAL.BN1 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-409. The alias for this parameter is BN1.VALD.
VAL.CN0 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-410. The alias for this parameter is CN0.VALD.
VAL.CN1 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-410. The alias for this parameter is CN1.VALD.
VAL.CN2 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-410. The alias for this parameter is CN2.VALD.

1147 ATLAS User’s Manual

MATERIAL Statements

VAL.CN3 Specifies the value of a temperature dependent impact ionization

parameter in Equation 3-410. The alias for this parameter is CN3.VALD.
VAL.DN0 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-411. The alias for this parameter is DN0.VALD.
VAL.DN1 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-411. The alias for this parameter is DN1.VALD.
VAL.DN2 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-411. The alias for this parameter is DN2.VALD.
VAL.AP0 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-412.
VAL.AP1 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-412. The alias for this parameter is AP1.VALD.
VAL.AP2 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-412. The alias for this parameter is AP2.VALD.
VAL.BP0 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-413. The alias for this parameter is BP0.VALD.
VAL.BP1 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-413. The alias for this parameter is BP1.VALD.
VAL.CP0 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-414. The alias for this parameter is CP0.VALD.
VAL.CP1 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-414. The alias for this parameter is CP1.VALD.
VAL.CP2 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-414. The alias for this parameter is CP2.VALD.
VAL.CP3 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-414. The alias for this parameter is CP3.VALD.
VAL.DP0 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-415. The alias for this parameter is DP0.VALD.
VAL.DP1 specifies the value of a temperature dependent impact ionization parameter
in Equation 3-415. The alias for this parameter is DP1.VALD.
VAL.DP2 Specifies the value of a temperature dependent impact ionization
parameter in Equation 3-415. The alias for this parameter is DP2.VALD.

1148 ATLAS User’s Manual

MATERIAL Statements

Klaassen Model parameters

KSRHTN Coefficient for Klaassen’s concentration and temperature dependent SRH lifetime
model.
KSRHTP Coefficient for Klaassen’s concentration and temperature dependent SRH lifetime
model.
KSRHCN Coefficient for Klaassen’s concentration and temperature dependent SRH lifetime
model.
KSRHCP Coefficient for Klaassen’s concentration and temperature dependent SRH lifetime
model.
KSRHGN Coefficient for Klaassen’s concentration and temperature dependent SRH lifetime
model.
KSRHGP Coefficient for Klaassen’s concentration and temperature dependent SRH lifetime
model.
KAUGCN Coefficient for Klaassen’s concentration dependent Auger model.
KAUGCP Coefficient for Klaassen’s concentration dependent Auger model.
KAUGDN Coefficient for Klaassen’s concentration dependent Auger model.
KAUGDP Coefficient for Klaassen’s concentration dependent Auger model.
NSRHN Specifies the SRH concentration parameter for electrons (see Equation 3-333).
NSRHP Specifies the SRH concentration parameter for holes (see Equation 3-334).
TAUN0 Specifies SRH lifetime for electrons (see Equation 3-333).
TAUP0 Specifies SRH lifetime for holes (see Equation 3-334).

See also Equations 3-335 and 3-336 and Equations 3-370 and 3-372 for more information
about Klaassen models.

1149 ATLAS User’s Manual

MATERIAL Statements

Carrier Statistics Model Parameters

ASYMMETRY Specifies the relative degree where band gap narrowing applies to the
conduction band versus the valence band. The value of the ASYMMETRY
parameter is multiplied by the total change in band gap due to band gap
narrowing and that product is applied to the conduction band edge. For
example, if the ASYMMETRY parameter has a value of 1.0, then the change in
band gap due to band gap narrowing is applied only to the conduction band
edge and the valence band edge remains unaffected.
BGN.E Specify the parameters of the band gap narrowing model given in
BGN.N Equation 3-46. The aliases for these parameters are V0.BGN, N0.BGN, and
BGN.C C0N.BGN. BGN.E and BGN.N are also used in the Bennett-Wilson and del
Alamo Bandgap narrowing models (see Equations 3-51 and 3-52).
BGN.E.ACC Parameters for the BGN and BGN2 models when used in 4H-SiC or 6H-SiC.
BGN.E.DON
BGN.N.ACC
BGN.N.DON

BGN.LIND.ANC Parameters for the Lindefelt Bandgap narrowing model described in Section
BGN.LIND.ANV 3.2.13 “Lindefelt Bandgap Narrowing Model”.
BGN.LIND.APC
BGN.LIND.APV
BGN.LIND.BNC
BGN.LIND.BNV
BGN.LIND.BPC
BGN.LIND.BPV

BGN.SHNK.EPS Relative dielectric permittivity for Schenk bandgap narrowing model.

BGN.SHNK.GE Electron degeneracy factor for Schenk bandgap narrowing model.

BGN.SHNK.GH Hole degeneracy factor for Schenk bandgap narrowing model.

BGN.SHNK.ME Electron effective mass for Schenk bandgap narrowing model.

BGN.SHNK.MH Hole effective mass for Schenk bandgap narrowing model.

CON.BGN This is an alias for BGN.C.

EA.CUBIC Acceptor ionization energy in SiC for cubic symmetry site.
EA.HEXAGONAL Acceptor ionization energy in SiC for hexagonal symmetry site.
ED.CUBIC Donor ionization energy in SiC for cubic symmetry site.
ED.HEXAGONAL Donor ionization energy in SiC for hexagonal symetry site.
EAB Specifies acceptor energy level (see Equation 3-74).
EDB Specifies donor energy level (see Equation 3-73).
F.BGN Specifies the name of a file containing a C-INTERPRETER function for the
specification of temperature, composition and doping dependent bandgap
narrowing models.

1150 ATLAS User’s Manual

MATERIAL Statements

GCB Specifies the conduction-band degeneracy factor (see Equation 3-73).

GVB Specifies the valence-band degeneracy factor (see Equation 3-74).
N0.BGN This is an alias for BGN.N.
V0.BGN This is an alias for BGN.E.

Energy Balance Parameters

TAUMOB.EL Specifies the relaxation time for electrons in the temperature dependent mobility
model (see Equation 3-157).
TAUMOB.HO Specifies the relaxation time for holes in the temperature dependent
mobility model (see Equation 3-158).
TAUREL.EL Specifies the relaxation time for electrons in the energy balance model (see
Equation 3-151).
TAUREL.HO Specifies the relaxation time for holes in the energy balance model (see
Equation 3-152).
TRE.T1, TRE.T2, These are used in the temperature dependent energy relaxation time model
TRE.T3, TRE.W1, based on table data from Laux-Fischetti Monte-Carlo simulation (see Table 3-
TRE.W2, TRE.W3, 21).
TRH.T1, TRH.T2,
TRH.T3, TRH.W1,
TRH.W2, and TRH.W3

Hot Carrier Injection Parameters

IG.ELINR Specifies the electron mean free path between redirecting collisions. The alias for
this parameter is LAMRN.
IG.HLINR Specifies the hole mean free path between redirecting collisions. The alias for this
parameter is LAMRP.
IG.ELINF Specifies the electron mean free path length for scattering by optical phonons. The
alias for this parameter is LAMHN.
IG.HLINF Specifies the hole mean free path length for scattering by optical phonons. The alias
for this parameter is LAMHP.
LAMRN This is an alias for IG.ELINR.
LAMRP This is an alias for IG.HLINR.
LAMHN This is an alias for IG.ELINF.
LAMHP This is an alias for IG.HLINF.

1151 ATLAS User’s Manual

MATERIAL Statements

Lattice Temperature Dependence Parameters

A.PASSLER Specifies a parameter of the Passler temperature dependent bandgap

model given in Equation 3-39.
EGALPHA Specifies the alpha coefficient for temperature dependence of
bandgap (see Equation 3-38).
EGBETA Specifies the beta coefficient for temperature dependence of
bandgap (see Equation 3-38).
F.PDN Uses C-Interpreter function for phonon drag contribution to the
electron thermopower.
F.PDP Uses C-Interpreter function for phonon drag contribution to the hole
thermopower.
HC.A, HC.B, HC.C, and HC.D Specify the values of the four coefficient of the heat capacity
equation (see Equation 7-6).
HC.BETA, HC.C1, HC.C300, and Specify user overrides for the parameters of Equation 7-7.
HC.RHO

LT.TAUN Specifies the temperature dependence for electron lifetimes (see

Equation 7-31).
LT.TAUP Specifies the temperature dependence for hole lifetimes
(see Equation 7-32).
P.PASSLER Specifies a parameter of the Passler temperature dependent bandgap
model given in Equation 3-39.
PCM.AEA Specifies the activation energy for amorphization of a phase change
material.
PCM.AP Specifies the time exponent for amorphous change of a phase
change material (see Section 7.2.9 “Phase Change Materials
(PCMs)”).
PCM.ARHO Specifies the amorphous resistivity of a phase change material
(see Section 7.2.9 “Phase Change Materials (PCMs)”).
PCM.ATC Specifies the critical temperature for amorphous change of a phase
change material (see Section 7.2.9 “Phase Change Materials
(PCMs)”).
PCM.CEA Specifies the activation energy for crystallization of a phase change
material.
PCM.CP Specifies the time exponent for crystallization of a phase change
material (see Section 7.2.9 “Phase Change Materials (PCMs)”).
PCM.CRHO Specifies the crystalline resistivity of a phase change material (see
Section 7.2.9 “Phase Change Materials (PCMs)”).

1152 ATLAS User’s Manual

MATERIAL Statements

PCM.CTC Specifies the critical temperature for crystallization of a phase

change material (see Section 7.2.9 “Phase Change Materials
(PCMs)”).
PCM.ATAU Specifies the time constant for amorphous change of a phase change
material (see Section 7.2.9 “Phase Change Materials (PCMs)”).
PCM.CTAU Specifies the time constant for crystallization of a phase change
material (see Section 7.2.9 “Phase Change Materials (PCMs)”).
PCM.LATHEAT Specifies the latent of change from crystalline to amorphous state in
a phase change material (see Section 7.2.9 “Phase Change Materials
(PCMs)”).
PDA.N Specifies the value of a parameter in the model of phonon drag
contribution to the electron thermopower (see Equation 7-13).
PDA.P Specifies the value of a parameter in the model of phonon drag
contribution to the hole thermopower (see Equation 7-14).
PDEXP.N Specifies the value of a parameter in the model of phonon drag
contribution to the electron thermopower (see Equation 7-13).
PDEXP.P Specifies the value of a parameter in the model of phonon drag
contribution to the hole thermopower (see Equation 7-14).
POWER Specifies the value of thermal power generated in a power source
associated with a region in THERMAL3D (see Chapter 17 “Thermal
3D: Thermal Packaging Simulator”).
T.PASSLER Specifies a parameter of the Passler temperature dependent bandgap
model given in Equation 3-39.
TC.A, TC.B, and TC.C Specify the three thermal conductivity coefficients (see Table 7-1).
TC.CONST Specifies the equilibrium value of thermal conductivity, k(T0), in
Table 7-1. The synonym for this parameter is TC.CONST. TC.0 is an
alias for TC.CONST.
TC.D Specifies the value of the parameter D in Table 7-1.
TC.E Specifies the value of the parameter E in Table 7-1.
TCON.CONST Specifies that thermal conductivity should be modeled as
constant with respect to temperature. The value of the thermal
conductivity is given by the value of the TC.C0 parameter.
TCON.POWER Specifies that the temperature dependence of thermal conductivity
should be modeled using Table 7-1.
TCON.POLY Specifies that the temperature dependence of thermal conductivity
should be modeled using Table 7-1.
TCON.RECIP Specifies that the temperature dependence of thermal conductivity
should be modeled using Table 7-1.

1153 ATLAS User’s Manual

PERMEABILITY Specifies the relative permeability.

S0SELL, S1SELL, These are parameters for Sellmeier's refractive index model. See Equation 3-
S2SELL, L1SELL, and 700.
L2SELL

TLU.A, TLU.C, TLC.E0, These are used to specify the parameters of the Tauc-Lorentz dielectric
TLU.EC, TLU.EG, and function model for complex index of refraction described in Section 3.10.3
TLU.EPS “Tauc-Lorentz Dielectric Function with Optional Urbach Tail Model for
Complex Index of Refraction”

Note: The index file must be ordered by increasing wavelength .

REAL.INDEX Specifies the real portion of the refractive index of the semiconductor.
Wavelength dependent defaults exist for certain materials as documented in
Appendix B “Material Systems”.

LASER Parameters

ABSORPTION.SAT Specifies the absorption saturation intensity used in the non-linear absorption
loss model given in Equation 8-39.
ALPHAA Specifies the bulk absorption coefficient in Equation 8-17.
ALPHAR Specifies the line width broadening factor in Equation 8-18.
EMISSION.FACTOR Specifies a scale factor accounting the proportion of light directionally coupled
into the lasing mode.
EPSINF Specifies the high frequency relative dielectric permittivity () (see Equation
8-17). If this parameter is not specified, it will be set equal to the static
dielectric permittivity of the material.
F.ALPHAA Specifies the name of a file containing a C-INTERPRETER function for the bulk
absorption coefficient.
FC.RN This is the free-carrier loss model parameter in Equation 8-19.
FC.RP This is the free-carrier loss model parameter in Equation 8-20.
GAIN.SAT Specifies the non-linear gain saturation factor used in Equation 8-38.
GAIN0 Specifies the parameter in Sections 3.9.2 “The Default Radiative
Recombination Model” and 3.9.3 “The Standard Gain Model”.
GAIN00 Specifies the parameter in Section 3.9.4 “The Empirical Gain Model”.
GAIN1N Specifies the parameter in Section 3.9.4 “The Empirical Gain Model”.
GAIN1P Specifies the parameter in Section 3.9.4 “The Empirical Gain Model”.
GAIN1MIN Specifies the parameter in Section 3.9.4 “The Empirical Gain Model”.

1156 ATLAS User’s Manual

MATERIAL Statements

GAIN2NP Specifies the parameter in Section 3.9.4 “The Empirical Gain Model”.
GAMMA Specifies the parameter in Sections 3.9.2 “The Default Radiative
Recombination Model” and 3.9.3 “The Standard Gain Model”. If this parameter
is not specified, it will be calculated using Equation 3-634.
GN1 Specifies the parameter in Section 3.9.5 “Takayama's Gain Model”.
GN2 Specifies the parameter in Section 3.9.5 “Takayama's Gain Model”.
NTRANSPARENT Specifies the parameter in Section 3.9.5 “Takayama's Gain Model”.

NOISE Parameters
F.MNSNDIFF Specifies the name of a file containing a C-Interpreter function for the
microscopic noise source for electron diffusion noise.
F.MNSPDIFF Specifies the name of a file containing a C-Interpreter function for the
microscopic noise source for hole diffusion noise.
F.MNSNFLICKER Specifies the name of a file containing a C-Interpreter function for the
microscopic noise source for electron flicker noise.
F.MNSPFLICKER Specifies the name of a file containing a C-Interpreter function for the
microscopic noise source for hole flicker noise.
HOOGEN Specifies the Hooge constant for electron flicker noise.
HOOGEP Specifies the Hooge constant for hole flicker noise.

Organic Transport Parameters

ML Specifies the conduction band longitudinal effective mass.

MLH Specifies the light hole effective mass.
MSO Specifies the split off (crystal lattice) effective mass.
MSTAR Specifies the conduction band effective mass dispersion used in
Equation 3-637.
MTT Specifies the transverse effective mass (wurtzite).
MT1 and MT2 Specify the transverse effective masses (zincblende).
MUNOFF and MUPOFF Specify file names for writing electron and hole mobility field
characteristics for plotting in TONYPLOT.
MV Specifies the valence band effective mass.
MZZ Specifies the effective mass along the crystal axis (wurtzite).
PHONON.ENERGY This is a parameter of the Schenk trap assisted tunneling model.
PIP.OMEGA This is the phonon energy in eV for the Pipinys model.
PIP.ACC This is the electron-phonon interaction constant for the Pipinys
model.
PIP.ET This is the trap depth in eV for the Pipinys model.
PIP.NT This is the occupied interface state density for the Pipinys model.
PSP Specifies the spontanteous polarization.
RESISTIVITY Specifies the resistivity of conductor regions.
SO.DELTA Specifies the spin orbital splitting energy in a quantum well. The
alias for this parameter is WELL.DELTA.
SOPRA Identifies the name of a file from the SOPRA database. See
Section B.13.1 “SOPRA Database”
SPECIES1.AF Attempt frequency for ionic species 1 diffusion coefficient (see
Section 3.16 “Generic Ion Transport Model”).
SPECIES1.EA Activation energy for ionic species 1 diffusion coefficient (see
Section 3.16 “Generic Ion Transport Model”).
SPECIES1.HOP Hopping distance for ionic species 1 diffusion coefficient (see
Section 3.16 “Generic Ion Transport Model”).
SPECIES2.AF Attempt frequency for ionic species 2 diffusion coefficient (see
Section 3.16 “Generic Ion Transport Model”).
SPECIES2.EA Activation energy for ionic species 2 diffusion coefficient (see
Section 3.16 “Generic Ion Transport Model”).
SPECIES2.HOP Hopping distance for ionic species 2 diffusion coefficient (see
Section 3.16 “Generic Ion Transport Model”).

1161 ATLAS User’s Manual

MATERIAL Statements

SPECIES3.AF Attempt frequency for ionic species 3 diffusion coefficient (see

Section 3.16 “Generic Ion Transport Model”).
SPECIES3.EA Activation energy for ionic species 3 diffusion coefficient (see
Section 3.16 “Generic Ion Transport Model”).
SPECIES3.HOP Hopping distance for ionic species 3 diffusion coefficient (see
Section 3.16 “Generic Ion Transport Model”).
STABLE Specifies the selected strain table for the Ishikawa strain effects
model. See Section 3.9.12 “Ishikawa's Strain Effects Model”.
TE.MODES Specifies whether TE or TM modes will be used for calculation of
asymmetry factors for the LI model.
USER.DEFAULT Specifies which material the user-defined material should use for its
default parameters.
USER.GROUP Specifies the material group for the user-defined material.
USER.GROUP can be either SEMICONDUCTOR, INSULATOR, or
CONDUCTOR.

WELL.EPS Specifies the high frequency permittivity used in calculating gain

and radiative recombination in a quantum well.
WELL.GAMMA0 Specifies the Lorentzian gain broadening factor from Section 3.9.11
“Lorentzian Gain Broadening”.
WELL.TAUIN Specifies the Lorentzian gain broadening factor from Section 3.9.11
“Lorentzian Gain Broadening”.
WELL.TAUN Specifies the quantum well capture escape model electron capture
time.
WELL.TAUP Specifies the quantum well capture escape model hole capture time.
Z.SCHENK This is a parameter of the Schenk trap assisted tunneling model.

Conductor Parameters

DRHODT Specifies the temperature coefficient of resistivity in µWcm/K.

RESISTIVITY Specifies the material resisitivity in µWcm.

1162 ATLAS User’s Manual

MATERIAL Statements

Defect Generation Material Parameters

KD.LID Specifies the light induced SiHD creation constant. See Equation 14-52.
KH.LID Specifies the disassociation constant of a hyrdogen atom from a SiHD defect.
See Equation 14-52.
SIG.LID Specifies the capture coefficient of free carriers by dangling bonds. See
Equation 14-52.

Material Coefficient Definition Examples

Numbered region
This example specifies SRH lifetimes and concentration independent low-field mobilities for
region number 2. All other parameters use default values and parameters in other regions are
unaffected.
MATERIAL TAUN0=5.0E-6 TAUP0=5.0E-6 MUN=3000 MUP=500 REGION=2

All regions
This example defines carrier lifetimes and the refractive index for all semiconductor regions.
MATERIAL TAUP0=2.E-6 TAUN0=2.E-6 REAL.INDEX=3.7 \
IMAG.INDEX=1.0E-2

Named Material
This shows the definition of bandgap for all InGaAs regions in the structure:
MATERIAL MATERIAL=InGaAs EG300=2.8
All materials are divided into three classes: semiconductors, insulators and conductors. See
Appendix B “Material Systems” for more information about the parameters required for each
material class.

Note: You can use the MODEL PRINT command to echo back default material parameters or MATERIAL
parameter settings to the run-time output.

1163 ATLAS User’s Manual

MEASURE Statements

21.33 MEASURE
MEASURE extracts selected electrical data from the solution.

Note: This statement is almost obsolete. Its functions have been replaced by the EXTRACT , OUTPUT , or
PROBE statement.

Syntax
MEASURE <dt> [<boundary>] [OUTFILE=<filename>]

Parameter Type Default Units

CONTACT Integer
E.CRIT Real1E-8
ELECTRON Logical False
HOLE Logical False
IONIZINT Logical False
LRATIO Real 1.0
METAL.CH Logical False
N.CURRENT Logical False
N.LAYER Real 15
N.LINES Integer 50
N.RESIST Logical False
NET.CARR Logical False
NET.CHAR Logical False
OUTFILE Character
P.CURRENT Logical False
P.RESIST Logical False
REGIONS Integer All regions
SUBSTR Character
U.AUGER Logical False
U.RADIATIVE Logical False
U.SRH Logical False
U.TOTAL Logical False

1164 ATLAS User’s Manual

MEASURE Statements

Parameter Type Default Units

X.MIN Real Left of device m

X.MAX Real Right of device m
Y.MIN Real Top of device m
Y.MAX Real Bottom of device m

Description

dt This is used to specify the type of information to be measured.

boundary Specifies which nodes will be measured.
OUTFILE Specifies a filename where simulation results and bias information will be
written.

Data Type Parameters

Net carrier concentration, charge concentration, electron concentration, or hole concentration
can be integrated over a section of a device. The charge on part of an electrode can be
calculated, just like the current through that part. This is useful for capacitance studies in
conjunction with the difference mode of the LOAD statement. The resistance of a structure
cross-section, such as a diffused line, can also be calculated.

E.CRIT Specifies the critical electric field used to calculate integration integrals.
ELECTRON Extracts integrated electron concentration.
HOLE Extracts integrated hole concentration.
IONIZINT Enables the calculation of ionization integrals. Other integral ionization
parameters will be ignored unless IONIZINT is specified.
LRATIO Specifies the ratio between electric field lines used in ionization integral
calculation. The value of this parameter should be set from 0.5 to 1.5.
METAL.CH Extracts integrated charge on a contact.
N.CURRENT Extracts n-current through an electrode.
N.LINES Specifies the number of ionization integrals.
N.RESIST Extracts n-resistance of a cross-section.
NET.CARR Extracts integrated carrier concentration.
NET.CHAR Extracts integrated net charge.
NLAYERS Controls the distance from the contact where electric field lines start.
P.CURRENT Extracts p-current through an electrode.

1165 ATLAS User’s Manual

MEASURE Statements

P.RESIST Extracts p-resistance of a cross-section.

SUBSTR Selects the substrate electrode for electric field lines. You don’t need to specify
this parameter if a substrate electrode has been defined in the ELECTRODE
statement.
U.AUGER Specifies that the integrated Auger recombination rate is to be extracted.
U.RADIATIVE Specifies that the integrated radiative recombination rate is to be extracted.
U.SRH Specifies that the integrated SRH recombination rate is to be extracted.
U.TOTAL Specifies that the integrated total recombination rate is to be extracted.

Boundary Parameters
Boundary parameters: X.MIN, X.MAX, Y.MIN, and Y.MAX define a bounding box. Only nodes
falling within this bounding box are included in the integration. The default bounds are the
boundaries of the entire device.

CONTACT Specifies the contact number. For electrode quantities (current and metal charge),
a contact must be selected. Only nodes falling within the bounds and belonging to
the contact are included in the integration. When IONIZINT is specified, this is the
electrode used to start electric field lines.
REGIONS Specifies a particular set of regions. If REGIONS is specified, only nodes within
the specified bounds that are part of a particular set of regions will be
integrated.
X.MAX Specifies the X coordinate of the right edge of the bounding box.
X.MIN Specifies the X coordinate of the left edge of the bounding box.
Y.MAX Specifies the Y coordinate of the top of the bounding box.
Y.MIN Specifies the Y coordinate of the bottom of the bounding box.

Resistance Example
This example extracts the resistance of a p-type line diffused into a lightly doped n-substrate.
Since the p-conductivity of the substrate is negligible, the integration bounds can include the
whole device.
MEASURE P.RESIST
Gate Charge Example
In this example, the charge on the lower surface of a gate electrode is integrated. There is 0.05
m of gate oxide on the surface, which is located at y=0.
MEASURE METAL.CH CONT=1 X.MIN=-2.0 X.MAX=2.0 \
Y.MAX=-0.0499 Y.MIN=-0.0501

1166 ATLAS User’s Manual

MEASURE Statements

Ionization Integral Example

This example shows how to extract the maximum ionization integral in the device.
MEASURE IONIZINT CONTACT=3 SUSTR=4 N.LINES=200 \
NLAYERS=15 LRATIO=1.1
This syntax was the original implementation of ionization integrals in early ATLAS versions.
It has been superseded. See Sections 21.57 “SOLVE” and 21.43 “OUTPUT” for the
recommended approach to extract ionization integrals.

1167 ATLAS User’s Manual

MESH Statements

21.34 MESH
MESH generates a mesh or reads a previously generated mesh.
Syntax
MESH <prev>|<new> [<output>]

Parameter Type Default Units

ATHENA Logical False

AUTO Logical False
CONDUCTOR Logical False
CYLINDRICAL Logical False
DATAFILE.ISE Character
DIAG.FLIP Logical True
ELEC.BOT Logical False
GRIDFILE.ISE Character
FLIP.POS Real Midpoint m
FLIP.Y Logical False
IN.FILE Character
INFILE Character
MASTER.IN Logical True
MASTER.OUT Logical True
MAX.ANGLE Real 360 Angles
MINOBTUSE Logical False
NX Integer
NY Integer
NZ Integer
OUT.FILE Character
OUTFILE Character
PERIODIC Logical False
PISCES.IN Logical False
RECTANGULAR Logical True
SCALE Integer 1

1168 ATLAS User’s Manual

MESH Statements

Parameter Type Default Units

Note: See Sections 2.6.1 “Interface From ATHENA” and 2.6.2 “Interface From DevEdit”, or the on-line examples
for details of the interfaces from ATHENA or DEVEDIT to ATLAS.

1172 ATLAS User’s Manual

METHOD Statements

21.35 METHOD
METHOD sets the numerical methods to be used to solve the equations and parameters
associated with the these algorithms.
Syntax
METHOD <gp> <mdp>

Parameter Type Default Units

2NDORDER Logical True

AC.FILL.LEVEL Real 0
AC.FILL.RATIO Real 0.5
AC.GMRES Real False
AC.PRECONDITIONER Real 0
AC.ZIP.BICGST Logical False 0
ACONTINU Real 0.5
ALT.SCHRO Logical False
ATRAP Real 0.5
AUTONR Logical False
BICGST Logical False
BLOCK Logical False
BLOCK.TRAN Logical False
BQP.ALTEB Logical False
BQP.NEWTON Logical False
BQP.NMIX Real 1
BQP.NOFERMI logical false
BQP.PMIX Real 1
BQPX.TOL Real 2.510-7
BQPR.TOL Real 1.010-26 (2D)
1.010-18 (3D)
BR.DAMPING Logical False
BRD.DELTA Real 0.0001

BRD.KAPPA Real 0.0

1173 ATLAS User’s Manual

METHOD Statements

Parameter Type Default Units

BRD.MAX.ITER Integer 5

BRD.OFF Logical False

BRD.SKIP.ITER Logical False

CARRIERS Real 2
C.ITLIMIT Integer 500
C.STABIL Real 1.0x10-10
C.RESID Real 1.0x10-8
CLIM.DD Real 4.51013 cm-3
CLIM.EB Real 0 cm-3
CLIM.LAT Real 105 cm-3
CLIMIT Real 10000
CONT.RHS Logical True
CONTINV Logical True
CR.TOLER Real 5.010-18
CUR.PROJ Logical False
CX.TOLER Real 1.010-5
DIRECT Logical False
DG.INIT logical false
DG.N.INIT logical false
DG.P.INIT logical false
DT.MAX Real 1.01010 s

DT.MIN Real 1.010-25 s

DTUPDATE Real 10-9 K

DVLIMIT Real 0.1

DVMAX Real 2.5 V
EF.TOL Real 10-12
ELECTRONS Logical True
EM.DAMPING Logical False

1174 ATLAS User’s Manual

METHOD Statements

Parameter Type Default Units

EMD.MAX.ITER Integer 5

EMD.OFF Logical False

EMD.SKIP.ITER Integer 5

ETR.TOLE Real 100

ETX.TOLE Real
EXTRAPOLATE Logical False
FAIL.QUIT Logical False
FAIL.SAFE Logical True
FGCNTRL Logical True
FILL.LEVEL Real 0
FILL.RATIO Real 0.5
FIX.QF Logical False
FLUX.JOULE Logical True
GCARR.ITLIMIT Real 1
GCON.ITLIMIT Real 0
GMRES Logical False
GUM.INIT Integer 15
GUMITS Integer 100
GUMMEL Logical False
GUMMEL.NEWTON Logical False
HALFIMPLICIT Logical False
HCIR.TOL Real 5.0x10-11
HCIX.TOL Real 5.0x10-4
HCTE.THERM Logical False
HOLES Logical True
ICCG Logical False
IR.TOL Real 5.010-15
ITLIMIT Integer 25
IX.TOL Real 2.010-5

1175 ATLAS User’s Manual

METHOD Statements

Parameter Type Default Units

LAS.PROJ Logical False

LS.DAMPING Logical False
LSD.ALPHA Real 0.0001

LSD.MAX.ITER Integer 5
LSD.OFF Logical False
LSD.SKIP.ITER Integer 5
LTR.TOLE Real 100
LTX.TOLE Real
LTE2STEP Logical False
LU1CRI Real 3.010-3
LU2CRI Real 3.010-2
L2NORM Logical True
MAG2DLSQ Logical False
MAG3DINT.N Logical False
MAG3DINT.P Logical False
MAXTRAPS Integer 4
MAX.TEMP Real 2000 K
MEINR Logical True
MIN.TEMP Real 120 K
NBLOCKIT Integer 15
NEG.CONC Logical False
NEW.EI Logical False
NEWTON Logical True
NITGUMM Integer 5
NOGRADMSTAR Logical False
NT0 Integer 4
NT1 Integer 10
NT2 Integer 25
NT3 Integer 100

1176 ATLAS User’s Manual

METHOD Statements

Parameter Type Default Units

NRCRITER Real 0.1

NO.POISSON Logical False
PC.FILL_LEVEL Real 0
PC.SCALE Real 1
PCM Logical False
PR.TOLER Real 1.010-26
PRECONDIT Real 0
PX.TOLER Real 1.010-5
RAT.FGCNTRL Real 1.0
RATIO.TIME Real 0.2
REL.STOP Logical False
RHSNORM Logical False
RXNORM Logical True
QMAX.FGCNTRL Real 100.0 Q

QUASI Logical True

Equation Solver Parameters

CARRIERS Specifies the number of carrier continuity equations that will be solved Valid
values are 0, 1 and 2. CARRIERS=0 implies that only Poisson’s Equation will be
solved. CARRIERS=1 implies that only one carrier solution will be obtained. When
this is specified, also specify either HOLES or ELECTRONS. CARRIERS=2 implies
that solutions will be obtained for both electrons and holes.
ELECTRONS Specifies that only electrons will be simulated for single carrier simulation.
HOLES Specifies that only holes will be simulated for single carrier simulations.

Solution Tolerance Parameters

The default convergence criteria used in ATLAS consists of a combination of relative and
absolute values. The program will converge if either criterion is met. This is particularly
useful when low-carrier concentrations would not converge using just relative criteria.
Current convergence criteria are also used. Terminal currents are monitored at each iteration
and overall convergence is allowed if currents converge along with absolute potential error.

BQPX.TOL This is the convergence criterion for the update vector in the BQP model.
BQPR.TOL This is the convergence criterion for the Right hand side (residual) in the BQP
model.
CR.TOLER Specifies an absolute tolerance for the continuity equation.
CX.TOL or C.TOL This is the relative tolerance for the continuity equation. The XNORM parameter
uses the CX.TOL and PX.TOL parameters to calculate convergence criteria.
EF.TOL Specifies the Schrodinger solver tolerance when NEW.EI is specified.
ETR.TOLE This is an alias for TCR.TOL.
ETX.TOLE This is an alias for TCX.TOL.
HCIR.TOL This is the absolute current convergence criteria for energy transport models.
HCIX.TOL This is the relative current convergence criteria for energy transport models.
IR.TOL Specifies absolute current convergence criteria.
IX.TOL Specifies relative current convergence criteria.
LTR.TOLE This is an alias for TLR.TOL.
LTX.TOLE This is an alias for TLX.TOL.
PR.TOLER Specifies an absolute tolerance for the Poisson Equation.
PX.TOL This is the relative tolerance for the potential equation. The XNORM parameter uses
the CX.TOL and PX.TOL parameters to calculate convergence criteria.

1181 ATLAS User’s Manual

METHOD Statements

RHSNORM Specifies that only absolute errors will be used to determine convergence. If
RHSNORM is selected Poisson error are measured in C/m and the continuity error
is measured in Am.
RXNORM Specifies that both relative and absolute convergence criteria will be used in the
solution method. This is the equivalent of specifying both XNORM and
RHSNORM. This is the default and it is not recommended to change this.

TCR.TOL Specifies the absolute (RHSNORM) tolerance for convergence of the carrier
temperature equations. The alias for this parameter is ETR.TOL.
TCX.TOL Specifies the relative (XNORM) tolerance for convergence of the carrier
temperature equations. The alias for this parameter is ETX.TOL.
TOL.TIME Specifies maximum local truncation error for transient simulations.
TOL.LTEMP Specifies the temperature convergence tolerance in block iterations using the
lattice heat equation.
TOL.RELAX Specifies a relaxation factor for all six Poisson, continuity, and current
convergence parameters (PX.TOL, CX.TOL, PR.TOL, CR.TOL, IX.TOL, and
IR.TOL).

TLR.TOL Specifies the relative (XNORM) tolerance for convergence of the lattice
temperature equation. The alias for this parameter is LTR.TOL.
TLX.TOL Specifies the relative (XNORM) tolerance for convergence of the lattice
temperature equation. The alias for this parameter is LTX.TOL.
WEAK Specifies the multiplication factor for weaker convergence tolerances applied
when current convergence is obtained.
XNORM Specifies that only the relative errors will be used to determine convergence for
the drift-diffusion equations. If XNORM is used Poisson updates are measured in kT/
q. and carrier updates are measured relative to the local carrier concentration.

Note: Generally, the solution tolerances should not be changed. Convergence problems should be tackled by
improving the mesh or checking the model and method combinations. Chapter 2: “Getting Started with
ATLAS” has useful hints.

General Parameters

ACONTINU This is an alias for ATRAP.

ATRAP Specifies the multiplication factor which reduces the electrode bias steps when a
solution starts to diverge. This parameter has no effect unless the TRAP parameter
is specified. The alias for this parameter for ACONTINU.
BQP.NMIX Specify a mixing parameter with value between 0 and 1, which causes
BQP.PMIX relaxation of fermi-dirac factor for electron and holes in the BQP model.

1182 ATLAS User’s Manual

METHOD Statements

CLIM.DD This is analogous to CLIMIT except it is expressed in a dimensional value

representing the minimum carrier concentration that can be resolved.
CLIM.EB This can be treated as a regularization parameter for the case of very small carrier
concentrations for energy balance simulation. It specifies the minimum value of
carrier concentration for which the relaxation term in the energy balance equation
will still be properly resolved. Carrier temperatures for points where the
concentration is much less than CLIM.EB, will tend to the lattice temperature.
CLIM.LAT Specifies the minimum resolvable carrier concentration for Joule heating
calculation.
CLIMIT Specifies a concentration normalization factor. See “Carrier Concentrations and
CLIM.DD (CLIMIT)” on page 915 for a complete description.
CONT.RHS This is an alias for TRAP.
CONTINU This is an alias for TRAP.
CUR.PROJ Enables the use of projection method for initial guesses with current
boundary conditions.
DTUPDATE Specifies the change in local temperature during lattice heating simulations,
which local composition and temperature dependent physical models (such as
mobility and bandgap) are recalculated.
FAIL.QUIT Requires that premature due to excessive trapping during continuation will exit
the simulator.
FAIL.SAFE Requires that premature exit from a static ramp due to compliance or
excessive trapping during continuation will exit with the device in the last
converged state.
FIX.QF Fixes the quasi-Fermi potential of each non-solved for carrier to a single value,
instead of picking a value based on local bias.
FGCNTRL Enables the FGCG model stabilisation algorithm for addition of charge to floating
gate.
FLUX.JOULE If TRUE, then it uses a flux-like discretization for including Joule heating into the
lattice temperature equation. If FALSE, then it uses a source-like discretization for
including Joule heating into the lattice temperature equation. This applies when
LAT.TEMP and JOULE.HEAT are enabled. It’s TRUE by default. The source-like
discretization is provided primarily for compatibility with VICTORY DEVICE.
HCTE.THERM Specifies that hot carrier simulation will use a thermodynamically correct Joule
heating term. This term has the form  J n  J n     n N  rather than the default
 Jn  E  .

ITLIMIT Specifies the maximum number of allowed outer loops (Newton loops or
GITLIMIT Gummel continuity iterations).
K.SHIFT Specifies the energy shift applied to the input imaginary index spectrum in
electron volts.

1183 ATLAS User’s Manual

METHOD Statements

LAS.PROJ Extrapolates that photon rates for the initial guess during bias ramp.
MAXTRAPS Specifies the number of times the trap procedure will be repeated in case of
divergence. The value of MAXTRAPS may range from 1 to 10. The alias for this
parameter is STACK.
MIN.TEMP These are specified to control the absolute range of lattice temperatures allowed
MAX.TEMP during Gummel loop iterations with lattice temperature. These parameters help
insure that lattice temperatures converge during the outer loop iterations.
N.SHIFT Specifies the energy shift applied to the input real index spectrum in electron
volts.
NBLOCKIT Specifies the maximum number of BLOCK iterations. If METHOD BLOCK NEWTON is
specified, the solver will switch to Newton’s method after NBLOCKIT Block-
Newton iterations.
NEG.CONC This flag allows negative carrier concentrations.
NK.SHIFT Specifies the energy shift applied to the complex index of refraction in electron
volts.
NO.POISSON This flag allows you to omit the solution of the Poisson’s Equation (see Equation
3-1). This feature doesn’t work with Gummel’s Method (see Section 20.5.2
“Gummel Iteration”).
NOGRADMSTAR Specifies that the gradient of the log of the effective mass will be omitted from
drift-diffusion and hydrodynamic/energy balance calculations. This may help
convergence properties.
PC.FILL_LEVEL Specifies the fill level to be used in the ZIP.BICGST iterative solver.
PC.SCALE Specifies the scaling method to be used in the ZIP.BICGST iterative solver.
PCM Sets certain models and parameter values for optimal simulation of phase change
materials (see Table 7-12).
QMAX.FGCNTRL Threshold charge parameter for FGCNTRL stabilization algorithm.
RAT.FGCNTRL Step size ratio parameter for FGCNTRL stabilization algorithm.
RG.MSCTRAP Clearing this flag decouples the multistate trap recombination rates from the
corresponding carrier or ionic transport continuity equations.
SPECIES.INS Restricts the generic ionic species to insulator regions only.
SPECIES.MAXX Gives the maximum relative update of the species concentrations during Newton
iteration.
STACK This is an alias for MAX.TRAPS.
TMIN.FACT Specifies the minimum electron or hole temperature allowable during non- linear
iteration updates. TMIN.FACT is normalized to 300K.

1184 ATLAS User’s Manual

METHOD Statements

TRAP Specifies that if a solution process starts to diverge, the electrode bias steps taken
from the initial approximation are reduced by the multiplication factor ATRAP.
The aliases for this parameter is CONT.RHS and CONTINU.
TTNMA Specifies the maximum allowable electron carrier temperature.
TTPMA Specifies the maximum allowable hole carrier temperature.
VSATMOD.INC Specifies that the derivatives of the negative differential mobility (MODEL FLDMOB
EVSATMOD=1) will not be included into the Jacobian until the norm of Newton
update for potential is less than the value specified by VSATMOD.INC. This is
useful since the negative differential mobility model is highly nonlinear and
causes numerical stability problems.

Gummel Parameters

DVLIMIT Limits the maximum potential update for a single loop.

GCARR.ITLIMIT Specifies the maximum number of interations that the
carrier continuity equations will solve when using GUMMEL.
GCON.ITLIMIT Specifies the number of extra GUMMEL loop iterations that will
be performed after all equations have converged.
GUM.INIT Specifies the maximum number of Gummel iterations in order
to obtain an initial approximation for successive Newton
iterations. This parameter is used when METHOD GUMMEL
NEWTON is specified

GUMITS Specifies the maximum number of Gummel iterations.

LU1CRI and LU2CRI Specifies amount of work per Poisson loop. The inner norm is
required to either decrease by at least LU1CRI before
returning, or reach a factor of LU2CRI below the projected
Newton error, whichever is smaller. If the inner norm is
exceeds the projected Newton error, the quadratic
convergence is lost.
SINGLEPOISSON Specifies that only a single Poisson iteration is to be
performed per Gummel loop. In the default state, the
continuity equation is only treated after the Poisson
iteration has fully converged. This technique is useful where
the carrier concentration and potential are strongly coupled
but the initial guess is poor precluding the use of NEWTON.
NITGUMM, NT0, NT1, NT2, and NT3 Specify Gummel iteration control parameters described in
Section 20.5.10 “Detailed Convergence Criteria”.

1185 ATLAS User’s Manual

METHOD Statements

Newton Parameters

2NDORDER Specifies that second-order discretization will be used when transient simulations are
performed.
AUTONR Implements an automated Newton-Richardson procedure, which attempts to reduce
the number of LU decompositions per bias point. We strongly recommend that you
use this parameter to increase the speed of NEWTON solutions. Iterations using AUTONR
will appear annotated with an A in the run-time output. Often an extra iteration is
added when using this parameters since the final iteration of any converged solution
cannot be done using AUTONR.
DT.MAX Specifies the maximum time-step for transient simulation.
DT.MIN Specifies the minimum time-step for transient simulations.
DVMAX Sets the maximum allowed potential update per Newton iteration. Large voltage steps
are often required when simulating high voltage devices. If any simulation requires
voltage steps of 10V or more, set DVMAX to 100,000. Reducing DVMAX may serve to
damp oscillations in solutions in some cases leading to more robust behavior.
Excessive reduction, however, in DVMAX is not recommended since the maximum
voltage step allowed will be limited by DVMAX*ITLIMIT. The alias for this parameter
is N.DVLIM.
EXTRAPOLATE Specifies the use of second-order extrapolation to compute initial estimates for
successive time-steps for transient simulations.
L2NORM Specifies the use of L2 error norms rather than infinity norms when calculating time
steps for transient simulations.
LTE2STEP Use an alternative method for evaluating Local Truncation Error during a transient
run. This method uses two adjacent time steps and is less prone to numerical noise
than the default method.
NRCRITER Specifies the ratio by which the norm from the previous Newton loop must decrease
in order to be able to use the same Jacobian (LU decomposition) for the current
Newton loop.
N.DVLIM This is an alias for DVMAX.
RATIO.TIME Specifies the minimum time step ratio allowed in transient simulations. If the
calculated time step divided by the previous time step is less than RATIO.TIME.
ATLAS will cut back the transient solution instead of continuing on to the next time
point.

1186 ATLAS User’s Manual

METHOD Statements

TAUTO Selects automatic, adaptive timesteps for transient simulations from local
truncation error estimates. Automatic time-stepping is the default for second-order
discretization, but is not allowed for first-order.
TSTEP.INCR Specifies the maximum allowable ratio between the time step sizes of successive
(increasing) time steps during transient simulation.
QUASI Specifies a quasistatic approximation for transient simulations. This is useful in
simulating transient simulations with long timescales where the device is in
equilibrium at each timestep.

Quantum Parameters

BQP.NOFERMI Forces the BQP model to only use its formulation derived using Maxwell
Boltzmann statistics. If this is not set, then the BQP formulation used will be the same
as the carrier statistics specified for the electrons or holes or both.
BQP.ALTEB Enables an alternative iteration scheme that can be used when solving the BQP
equation and the energy balance equation. This may produce a more robust
convergence than for the default scheme. The scheme used with BQP.ALTEB set,
however, will be slower.
DG.INIT Enables an algorithm to initialize the Density Gradient method within the SOLVE INIT
statement. This is not self-consistent with the solution of Poisson's equation, which
makes it limited in usefulness. You can specify the carrier type to be initialized
using either DG.N.INIT or DG.P.INIT. DG.INIT will enable both DG.N.INIT and
DG.P.INIT.

SCHUR Specifies that the Schur Compliment method is used for the quantum well rate
equations for the quantum well capture escape model.
SP.NUMITER Sets the maximum number of Schrodinger, NEGF, or DDMS-Poisson iterations, after
which ATLAS proceeds to the next bias point.
SP.NUMOSC The number of previous Schrodinger-Poisson iterations, whose residual is
compared to the residual of the current iteration in order to check whether the residual
is oscillating.
SP.STAOSC The number of Schrodinger-Poisson iteration, after which ATLAS starts checking for
oscillatory behavior of the residual.
SP.MINDIF Minimum difference between logarithm of current and previous residuals of
Schrodinger-Poisson below which the residual is regarded as oscillating and ATLAS
proceeds to the next bias point.
WELL.ERROR Specifies the largest relative error of bound carrier density to regulate convergence for
the quantum well capture escape model.

1187 ATLAS User’s Manual

METHOD Statements

WELL.ITERMAX Specifies the maximum number of Newton iterations for the coupled system for the
quantum well capture escape model.
WELL.PROJ Enables the projection method for the 2D densities for the quantum well capture
escape model.
WELL.SELFCON Enables charge self-consistency for the capture escape model by ensuring band edges
as seen by the bulk carriers are modified to remove wells in the QWELL regions and
bound carrier density is added to the Poisson equation.

Globally Convergent Scheme Parameters

BR.DAMPING Turns on Bank-Rose damping.

BRD.OFF Turns off Bank-Rose damping.
BRD.SKIP.ITER Indicates how many Newton iterations should be performed before Bank-Rose
damping is used.
BRD.MAX.ITER Indicates the maximum number of Bank-Rose inner iterations that are performed
before ATLAS indicates a failure to find a suitable step.
BRD.DELTA The minimum acceptable reduction in the residual, scaled to the full Newton
step, is BRD.DELTA*g(0) (where g(0) is the residual at the current point).
BRD.KAPPA The next step in a Bank-Rose inner iteration is k = 1/(1 + K*g(0)) (where
g(0) is the residual at the current point). This sets the initial value of K.

EM.DAMPING Turns on explicit minimum damping.

EMD.OFF Turns off explicit minimum damping.
EMD.SKIP.ITER Indicates how many Newton iterations should be performed before explicit
minimum damping is used.
EMD.MAX.ITER Indicates the number of explicit minimum inner iterations used to bracket the
minimum.
LS.DAMPING Turns on line search damping.
LSD.OFF Turns off line search damping.
LSD.SKIP.ITER Indicates how many Newton iterations should be performed before line search
damping is used.
LSD.MAX.ITER Indicates the maximum number of line search inner iterations that are performed
before ATLAS indicates a failure to find a suitable step.
LSD.ALPHA The minimum acceptable reduction in the residual, scaled to the full Newton
step, is -LSD.ALPHA*g'(0) (where g'(0) is the gradient, in the direction of the
Newton step, of the residual at the current point).

1188 ATLAS User’s Manual

METHOD Statements

Numerical Method Definition Example

The default numerical method is the equivalent of:
METHOD NEWTON CARRIERS=2
For more complex problems including those involving floating regions the following is
recommended
METHOD GUMMEL NEWTON GUM.INIT=5
When impact ionization is combined with floating regions as in SOI or guard ring breakdown
simualation the above syntax can also be used. Quicker solutions, however, can be obtained
using the following SINGLEPOISSON technique:
METHOD GUMMEL NEWTON GUM.INIT=5 SINGLE
TRAP Parameter Example
This example illustrates the trap feature (often used to capture knees of I-V curves for
junction breakdown).
The first SOLVE statement solves for the initial, zero bias case. In the second SOLVE
statement, we attempt to solve for V2=3 volts and V3=5 volts. If such a large bias change
caused the solution algorithms to diverge for this bias point, the bias steps would be
multiplied by ATRAP (0.5).
An intermediate point (V2=1.5 volts, V3=2.5 volts) would be attempted before trying to
obtain V2=3 volts and V3=5 volts again. If the intermediate point can not be solved for either
case, then the program will continue to reduce the bias step (the next would be V2=0.75 volts
and V3=1.25 volts) up to MAXTRAPS times.
METHOD TRAP ATRAP=0.5
SOLVE INIT
SOLVE V2=3 V3=5 OUTFILE=OUTA
Transient Method Example
In this transient simulation example, second-order discretization is used (by default), but the
required LTE (10-3) is smaller than the default. Since the Jacobian is exact for the second part
(BDF-2) of the composite timestep, there should be very few factorizations for the BDF-2
interval when AUTONR is specified.
METHOD NEWTON TOL.TIME=1E-3 AUTONR

Note: For recommendations on METHOD parameters for different simulations, see Chapter 2 “Getting Started with
ATLAS” or the on-line examples.

1189 ATLAS User’s Manual

MOBILITY Statements

21.36 MOBILITY
MOBILITY allows specification of mobility model parameters.
Syntax
MOBILITY [NUMBER=<n>] [REGION=<n>] [MATERIAL=<name>]
[NAME=<region_name>] <parameters>

Parameter Type Default Units

A.BROOKS Real 1.561021 (cm V s )-1

A.PASV.N Real 0.1 nm
A.PASV.P Real 0.1 nm
ACCN.SF Real 0.87
ACCP.SF Real 0.87
ALBRCT.N Logical False
ALBRCT.P Logical False
AL1N.WATT Real -0.16
AL1P.WATT Real -0.296
AL2N.WATT Real -2.17
AL2P.WATT Real -1.62
AL3N.WATT Real 1.07
AL3P.WATT Real 1.02
ALP.PASV.N Real nm-1
ALP.PASV.P Real nm-1
ALPHAN Real 0.0
ALPHAN.ARORA Real -0.57
ALPHAN.CAUG Real 0.0
ALPHAP Real 0.0
ALPHAP.ARORA Real -0.57
ALPHAP.CAUG Real 0.0
ALPHAN.FLD Real 2.4107 cm/s

ALPHAP.FLD Real 2.4107 cm/s

1190 ATLAS User’s Manual

MOBILITY Statements

Parameter Type Default Units

ALPHA1N.KLA Real 0.68

ALPHA1P.KLA Real 0.719
ALPHAN.RCS Real -0.3
ALPHAN.RPS Real -0.415
ALPHAN.TAS Real 2
ALPHAP.RCS Real -0.3
ALPHAP.RPS Real -0.415
ALPHAP.TAS Real 3.4
ALPHN.CVT Real 0.680
ALPHP.CVT Real 0.71
ALN.CVT Real 6.8510-21
ALP.CVT Real 7.8210-21
ALPN.UM Real 0.68
ALPP.UM Real 0.719
AN.PLBRCT Real
AN.CCS Real 4.611017 cm-3
AN.CVT Real 2.58
AN.IIS Real 4.611017 cm-3
AP.IIS Real 1.01017 cm-3
ANALYTIC.N Logical False
ANALYTIC.P Logical False
AP.CCS Real 1.01017 cm-3
AP.CVT Real 2.18
ARORA.N Logical False
ARORA.P Logical False
ASN.YAMA Real 1.5410-5 cm/V

ASP.YAMA Real 5.3510-5 cm/V

B.BROOKS Real 7.631019 cm-3

1191 ATLAS User’s Manual

MOBILITY Statements

Parameter Type Default Units

B1N.TAS Real 1.75

B1P.TAS Real 1.5
B2N.TAS Real -0.25
B2P.TAS Real -0.3
BETAN Real 2.0
BETAP Real 1.0
BETAN.ARORA Real -2.33
BETAN.CAUG Real -2.3
BETAN.CVT Real 2.00
BETAN.RCS Real 2.1
BETAN.RPS Real -0.505
BETAN.TAS Real 2
BETAP.ARORA Real -2.33
BETAP.CAUG Real -2.2
BETAP.CVT Real 2.00
BETAP.RCS Real 2.1
BETAP.RPS Real -0.505
BETAP.TAS Real 1
BN.CCS Real 1.521015 cm-3
BP.CCS Real 6.251014 cm-3
BN.CVT Real 4.75107 cm/(K·s)

BN.IIS Real 1.521015 cm-3

BN.LSM Real 4.75107 cm2/(V*s)
BP.CVT Real 9.925106 cm/(K·s)

BP.IIS Real 6.251014 cm-3

BP.LSM Real 9.925106 cm2/(V*s)
CA.KLA Real 0.50
CCSMOB.N Logical False

1192 ATLAS User’s Manual

MOBILITY Statements

Parameter Type Default Units

CCSMOB.P Logical False

CCUTOFF.N Real 0.1

CCUTOFF.P Real 0.1

CD.KLA Real 0.21

CHEN.N Logical False
CHEN.P Logical False
CONMOB.N Logical False
CONMOB.P Logical False
CVT.N Logical False
CVT.P Logical False
CN.ARORA Real 1.4321017 cm-3
CN.CVT Real 1.74105
CN.LSM Real 1.74105
CN.UCHIDA Real 0.78125 cm2/Vsµm6
CP.ARORA Real 2.671017 cm-3
CP.CVT Real 8.842105
CP.LSM Real 8.842105
CP.UCHIDA Real 0.78125 cm2/Vsµm6
CRN.CVT Real 9.681016 cm-3
CRN.LSM Real 9.681016 cm-3
CRP.CVT Real 2.231017 cm-3
CRP.LSM Real 2.231017 cm-3
CSN.CVT Real 3.431020 cm-3
CSN.LSM Real 3.431020 cm-3
CSP.CVT Real 6.101020 cm-3
CSP.LSM Real 6.101020 cm-3
D.CONWELL Real 1.041021 (cmVs )-1

1193 ATLAS User’s Manual

MOBILITY Statements

Parameter Type Default Units

DELN.CVT Real 5.821014 V/s

DELP.CVT Real 2.05461014 V/s

DELTAN.CAUG Real 0.73

DELTAN.RCS Real 0.035
DELTAP.CAUG Real 0.70
DELTAP.RCS Real 0.035
DEVICE Character
DP.CVT Real 1/3
DN.CVT Real 1/3
DN.LSM Real 3.581018 cm2/(V*s)
DN.TAS Real 3.210-9
DP.LSM Real 4.11015 cm2/(V*s)
DP.TAS Real 2.3510-9
EON Real 4.0103 V/cm

E1N.SHI Real 6.3103 V/cm

E2N.SHI Real 0.77106 V/cm

EOP Real 4.0103 V/cm

E1P.SHI Real 8.0103 V/cm

E2P.SHI Real 3.9105 V/cm

ECRITN Real 4.0103 V/cm

ECRITP Real 4.0103 V/cm

ECN.MV Real 6.48104 V/cm

N1P.TAS Real 8.41016
NAME Character
NCRITN.ARORA Real 1.4321017 cm-3
NCRITP.ARORA Real 2.671017 cm-3
NCRITN.CAUG Real 1.0721017 cm-3
NCRITP.CAUG Real 1.6061017 cm-3
NEWCVT.N Logical False
NEWCVT.P Logical False
NHANCE Logical False
NREF1N.KLA Real 9.6816 cm-3
NREF1P.KLA Real 2.2317 cm-3
NREFD.KLA Real 4.020 cm-3
NREFA.KLA Real 7.220 cm-3
NREFN Real 1.0721017 cm-3
NREFN.YAMA Real 3.01016 cm-3
NREFP Real 1.6061017 cm-3
NREFP.YAMA Real 4.01016 cm-3
NRFN.UM Real 9.681016 cm-3
NRFP.UM Real 2.231017 cm-3
NUN Real -2.3
NUP Real -2.2
OLDSURF.N Logical False
OLDSURF.P Logical False
OXLEFTN Real m

OXLEFTP Real m

OXRIGHTN Real m

1201 ATLAS User’s Manual

MOBILITY Statements

Parameter Type Default Units

OXRIGHTP Real m

OXBOTTOMN Real m

OXBOTTOMP Real m

OZGUR.N Logical False

OZGUR.P Logical False
P.ANGLE Real 0.0 Degrees
P.BETA0 Real 1.213 (CANALI)
0.6 (MEINR)
P.BETAEXP Real 0.17 (CANALI)
0.01 (MEINR)
P.BROOKS Logical False
P.CANALI Logical False
P.CONWELL Logical False
P.LCRIT Real 1.0 cm
P.MASETTI Logical False
P.MEINR Logical False
P1N.SHI Real 0.28
P1P.SHI Real 0.3
P2N.SHI Real 2.9
P2P.SHI Real 1.0
P1N.TAS Real 0.09
P1P.TAS Real 0.334
P2N.TAS Real 4.5310-8
P2P.TAS Real 3.1410-7
PASVEER.N Logical False
PASVEER.P Logical False
PC.LSM Real 9.231016 cm-3
PCN.CVT Real 0.0 cm-3
PCP.CVT Real 0.231016 cm-3

1202 ATLAS User’s Manual

MOBILITY Statements

Parameter Type Default Units

PHANCE Logical False

PFMOB.N Logical False
PFMOB.P Logical False
PRINT Logical False
R1.KLA Real 0.7643
R2.KLA Real 2.2999
R3.KLA Real 6.5502
R4.KLA Real 2.3670
R5.KLA Real -0.8552
R6.KLA Real 0.6478
RCS.N Logical False
RCS.P Logical False
RPS.N Logical False
RPS.P Logical False
RN.TAS Real 2
RP.TAS Real 3
REGION Integer
S1.KLA Real 0.89233
S2.KLA Real 0.41372
S3.KLA Real 0.19778
S4.KLA Real 0.28227
S5.KLA Real 0.005978
S6.KLA Real 1.80618
S7.KLA Real 0.72169
SCHWARZ.N Logical False
SCHWARZ.P Logical False
SHI.N Logical False
SHI.P Logical False
SIG.PASV.N Logical False

1203 ATLAS User’s Manual

MOBILITY Statements

Parameter Type Default Units

SIG.PASV.P Logical False

SN.YAMA Real 350
SP.YAMA Real 81.0
STRUCTURE Character
SURFMOB.N Logical False
SURFMOB.P Logical False
TASCH.N Logical False
TASCH.P Logical False
TAUN.CVT Real 0.125
TAUP.CVT Real 0.0317
TETN.UM Real -2.285
TETP.UM Real -2.247
THETAN.FLD Real 0.8
THETAP.FLD Real 0.8
THETAN.KLA Real 2.285
THETAN.SHI Real 2.285
THETAP.KLA Real 2.247
THETAP.SHI Real 2.247
TMUBN.TAS Real 2.5
TMUBP.TAS Real 1.4
TMUN Real 1.5
TMUP Real 1.5
TNOMN.FLD Real 600.0 K
TNOMP.FLD Real 600.0 K
TN.UCHIDA Real 4.0 nm
TP.UCHIDA Real 4.0 nm
ULN.YAMA Real 4.9106 cm/s

ULP.YAMA Real 2.928106 cm/s

VSATN Real cm/s

1204 ATLAS User’s Manual

MOBILITY Statements

Parameter Type Default Units

VSATP Real cm/s

VSAT.QFN Logical False
VSAT.QFP Logical False
VSAT.ZAKN Logical False
VSAT.ZAKP Logical False
VSN.YAMA Real 1.036107 cm/s

VSP.YAMA Real 1.200107 cm/s

VTHN.PFMOB Real cm/s

VTHP.PFMOB Real cm/s
UCHIDA.N Logical False
UCHIDA.P Logical False
XIN Real -3.8
XIP Real -3.7
XMINN.WATT Real -1.032 m

XMAXN.WATT Real 1.032 m

YMAXN.WATT Real -1.0x32 m

XMINP.WATT Real -1.032 m

XMAXP.WATT Real 1.032 m

YMAXP.WATT Real -1.032 m

YCHARN.WATT Real 1.032 m

YCHARP.WATT Real 1.032 m

Z11N.TAS Real 0.0388

Z11P.TAS Real 0.039
Z22N.TAS Real 1.7310-5
Z22P.TAS Real 1.5110-5

1205 ATLAS User’s Manual

MOBILITY Statements

Description

MATERIAL Specifies which material from Table B-1 should be applied to the MOBILITY statement.
If a material is specified, then all regions defined as being composed of that material will
be affected.

DEVICE Specifies the device in MIXEDMODE simulation (see Chapter 12 “MixedMode: Mixed
Circuit and Device Simulator”) to be applied to the MOBILITY statement. The
synonym for this parameter is STRUCTURE.
NAME Specifies the name of the region to be applied to the MOBILITY statement. Note that the
name must match the name specified in the NAME parameter of the REGION statement or
the region number.
REGION Specifies the number of the region to be applied to the MOBILITY statement.
PRINT Acts the same as PRINT on the MODELS statement.
STRUCTURE This is a synonym for DEVICE.

Mobility Model Flags

ALBRCT.N Enable the Albrecht mobility model (see “The Albrecht Model” on page 428).
ALBRCT.P

ANALYTIC.N Specifies that the analytic concentration dependent model is to be used for
electrons (see Equation 3-202).
ANALYTIC.P Specifies that the analytic concentration dependent model is to be used for holes
(see Equation 3-203).
ARORA.N Specifies that the Arora analytic concentration dependent model is to be used for
electrons (see Equation 3-204).
ARORA.P Specifies that the Arora analytic concentration dependent model is to be used for
electrons (see Equation 3-205).
CHEN.N When specified, enable a set of velocity overshoot model parameters
CHEN.P corresponding to Chen et. al. [44] for electrons and holes to simulate electron
overshoot in GaN.
CONMOB.N Specifies that doping concentration dependent model to be used for electrons.
CONMOB.P Specifies that a doping concentration dependent model is to be used for holes.

1206 ATLAS User’s Manual

MOBILITY Statements

CCSMOB.N Specifies that carrier-carrier scattering model is to be used for electrons (see
Equations 3-208-3-212).
CCSMOB.P Specifies that carrier-carrier scattering model is to be used for holes (see
Equations 3-208-3-212).
EGLEY.N Enables the low field strained silicon mobility model for electrons (see Section
3.6.14 “Low-Field Mobility in Strained Silicon”).
EGLEY.P Enables the low field strained silicon mobility model for holes (see Section 3.6.14
“Low-Field Mobility in Strained Silicon”).
EVSATMOD Specifies which parallel field dependent mobility model (see Equation 3-306 and
Equation 5-59) should be used for electrons. The value of EVSATMOD should be
assigned as follows:
• 0 - Use the standard saturation model (Equation 3-306). You also
can apply the option parameter, MOBTEM.SIMPL (see “Carrier
Temperature Dependent Mobility” on page 198 for more
information).
• 1 - Use the negative differential velocity saturation model
(Equation 5-59).
• 2 - Use a simple velocity limiting model.
In most cases, the default value of 0 should be used.
HVSATMOD Specifies which parallel field dependent mobility model (see Equation 3-307 and
Equation 5-60) should be used for holes. The value of HVSATMOD should be
assigned as follows:
• 0 - Use the standard saturation model (Equation 3-307).
• 1 - Use the negative differential velocity saturation model
(Equation 5-60).
• 2 - Use a simple velocity limiting model.
In most cases, the default value of 0 should be used.
EXP.WATT.N Turns on exponertial modification to Watt’s mobility model for electrons (see
Equation 3-304).
EXP.WATT.P Turns on exponential modification to Watt’s mobility model for holes
(see Equation 3-305).
FLDMOB Specifies transverse field degradation for electrons as follows:
1. No transverse degradation.
2. Use the Watt or Tasch transverse field models depending on the
settings of FIELDMOB1 and FIELDMOB2.
3. Use the Yamaguchi transverse field dependent model.
4. Use the CVT transverse field dependent model.
FLDMOB.N Specifies a lateral electric field dependent model for electrons (see Equation 3-
306).
FLDMOB.P Specifies a lateral electric field dependent model for holes (see Equation 3-307).

1207 ATLAS User’s Manual

MOBILITY Statements

KLA.N Turns on Klaassen’s mobility model for electrons (see Equations 3-220 through 3-
245).
KLA.P Turns on Klaassen’s mobility model for holes (see Equations 3-220 through
3-245).
MOBMOD.N Specifies transverse field degradation for electrons as follows:
1. No transverse degradation.
2. Use the Watt or Tasch transverse field models depending on the
settings of FIELDMOB1 and FIELDMOB2.
3. Use the Yamaguchi transverse field dependent model.
4. Use the CVT transverse field dependent model.
MOBMOD.P Specifies transverse field degradation for holes as follows:
1. No transverse degradation.
2. Use the Watt or Tasch transverse field models depending on the
settings of FIELDMOB1 and FIELDMOB2.
3. Use the Yamaguchi transverse field dependent model.
4. Use the CVT transverse field dependent model.
MOD.WATT.N Turns on modified Watt mobility model for electrons (see Equation 3-304).
MOD.WATT.P Turns on modified Watt Mobility Model for holes (see Equation 3-305).
MSTI.PHOS Selects the parameter set for Phosphorous doping in Masetti model. For more
information about this parameter, see Tables 3-37, 3-38, and 3-39.
N.ANGLE Specifies angle for application of electron mobility parameters in simulation of
anisotropic mobility.
N.CANALI Specifies that the Canali velocity saturation model is to be used for electrons (see
Equations 3-310 through 3-317).
N.CONWELL Specifies that the Conwell-Weisskopf carrier-carrier scattering model is to be used
for electrons (see Equation 3-214).
N.MASETTI Enables Masetti mobility model for electrons. For more information about this
parameter, see Tables 3-37, 3-38, and 3-39.
N.MEINR Specifies that the Meinerzhagen-Engl hydrodynamic scattering model is to be
used for electrons (see Equation 3-326).
NEWCVT.N Specifies that a new surface mobility degradation calculation is used in the CVT
mobility model for electrons in place of the original (see “Darwish CVT Model”
on page 180).
NEWCVT.P Specifies that a new surface mobility degradation calculation is used in the CVT
mobility model for holes in place of the original (see “Darwish CVT Model” on
page 180).
NHANCE Specifies that low field mobility values will be modified by mobility
enhancement values contained in the mesh structure file.

1208 ATLAS User’s Manual

MOBILITY Statements

OZGUR.N When specified, enable a set of velocity overshoot model parameters

OZGUR.P corresponding to Ozgur et. al. [199] for electrons and holes to simulate electron
overshoot in ZnO.
P.ANGLE Specifies angle for application of hole mobility parameters in simulation of
anisotropic mobility.
P.BROOKS Specifies that the Brooks-Herring carrier-carrier scattering model is to be used for
holes (see Equation 3-217).
P.CANALI Specifies that the Canali velocity saturation model is to be used for holes (see
Equations 3-310 through 3-317).
P.CONWELL Specifies that the Conwell-Weisskopf carrier-carrier scattering model is to be used
for holes (see Equation 3-214).
P.MASETTI Enables Masetti mobility model for holes. For more information about this
parameter, see Tables 3-37, 3-38 and 3-39.
P.MEINR Specifies that the Meinerzhagen-Engl hydrodynamic scattering model is to be
used for holes (see Equation 3-327).
PASVEER.N Enables the generalized Gaussian disorder mobility model for electrons (see
Equations 15-53 through 15-61).
PASVEER.P Enables the generalized Gaussian disorder mobility model for holes (see
Equations 15-53 through 15-61).
PHANCE Specifies that low field hole mobilities will be modified by mobility
enhancement factors contained in the mesh structure file.
RCS.N Enables the remote coulomb scattering mobility model for electrons (see
Equation 3-294).
RCS.P Enables the remote coulomb scattering mobility model for holes (see Equation 3-
295).
RPS.N Enables the remote phonon scattering mobility model for electrons (see
Equation 3-296).
RPS.P Enables the remote phonon scattering mobility model for holes (see Equation 3-
297).
SURFMOB.N Invokes the effective field based surface mobility model for electrons
(see Equation 3-300).
SURFMOB.P Invokes the effective field based surface mobility model for holes (see Equation 3-
301).
SCHWARZ.N Specifies the use of transverse electric field-dependent mobility models for
electrons. See TFLDMB1 or SCHWARZ in the MODELS statement.
SCHWARZ.P Specifies the use of transverse electric field-dependent mobility models for holes.
See TFLDMB1 or SCHWARZ in the MODELS statement.
SHI.N Turns on Shirahata’s Mobility Model for electrons (see Equation 3-292).

1209 ATLAS User’s Manual

MOBILITY Statements

MU1P.CAUG MUP.MIN THETAP.KLA TETP.UM*

MU2N.CAUG MUN.MAX MUMAXN.KLA MMXN.UM

MU2P.CAUG MUP.MAX MUMAXP.KLA MMXP.UM

BETAN.CAUG NUN MUMINN.KLA MMNN.UM

BETAP.CAUG NUP MUMINP.KLA MMNP.UM

DELTAN.CAUG ALPHAN NREF1N.KLA NRFN.UM

DELTAP.CAUG ALPHAP NREF1P.KLA NRFP.UM

GAMMAN.CAUG XIN ALPHA1N.KLA ALPN.UM

GAMMAP.CAUG XIP ALPHA1P.KLA ALPP.UM

NCRITN.CAUG NREFN BN.CVT BN.LSM

NCRITP.CAUG NREFP BP.CVT BP.LSM

MU1N.ARORA MUN1.ARO CN.CVT CN.LSM

MU1P.ARORA MUP1.ARO CP.CVT CP.LSM

MU2N.ARORA MUN2.ARO TAUN.CVT EXN4.LSM

MU2P.ARORA MUP2.ARO TAUP.CVT EXP4.LSM

ALPHAN.ARORA EXN1.ARO MU0N.CVT MUN0.LSM

ALPHAP.ARORA EXP1.ARO MU0P.CVT MUP0.LSM

BETAN.ARORA EXN2.ARO MU1N.CVT MUN1.LSM

BETAP.ARORA EXP2.ARO MU1P.CVT MUP1.LSM

GAMMAN.ARORA EXN3.ARO CRN.CVT CRN.LSM

GAMMAP.ARORA EXP3.ARO CRP.CVT CRP.LSM

NCRITN.ARORA CN.ARORA CSN.CVT CSN.LSM

NCRITP.ARORA CP.ARORA CSP.CVT CSP.LSM

AN.CCS AN.IIS ALPHAN.CVT EXN1.LSM

AP.CCS AP.IIS ALPHAP.CVT EXP1.LSM

1214 ATLAS User’s Manual

MOBILITY Statements

Table 21-2 Mobility Model Aliases

Parameter Alias Parameter Alias

BN.CCS BN.IIS BETAN.CVT EXN2.LSM

BP.CCS BP.IIS BETAP.CVT EXP2.LSM

P1N.SHI EX1N.SHI MUMAXN.CVT MUN2.LSM

P1P.SHI EX1P.SHI MUMAXP.CVT MUP2.LSM

ECRITN EON GAMN.CVT EXN3.LSM

ECRITP EOP GAMP.CVT EXP3.LSM

KN.CVT EXN8.LSM DELN.CVT DN.LSM

KP.CVT EXP8.LSM DELP.CVT DP.LSM

PCP.CVT PC.LSM

Note: These aliases are negated with respect to the primary parameter.

Modified Watt Model Example

The following example sets the Modified Watt Surface mobility model for MOSFETs. The
MOBILITY statement is used to set the models and to specify the value of the depth of action
of the Modified Watt model.
MODELS CONMOB FLDMOB SRH MIN.SURF PRINT
MOBILITY WATT.N MOD.WATT.N YMAXN.WATT=0.01

1215 ATLAS User’s Manual

MODELS Statements

21.37 MODELS
MODELS specifies model flags to indicate the inclusion of various physical mechanisms,
models, and other parameters such as the global temperature for the simulation.
Syntax
MODELS <mf> <gp> <mdp>

Parameter Type Default Units

2DXY.SCHRO Logical False

A.BBT.SCHENK Real 8.9771020 cm2s-1V-2
A.TEMP Logical False
A0.TNLC Real 0.0 degrees
ACC.SF Real 0.87
ADACHI Logical False
ALBRCT Logical False
ALL Logical True
ALN1 Real -0.160
ALN2 Real -2.17
ALN3 Real 1.07
ALP1 Real -0.296
ALP2 Real -1.62
ALP3 Real 1.02
ALPHA.DOS Logical False
ALT.SCHRO Logical False
ANALYTIC Logical False
AR.MU1N Real 88
AR.MU2N Real 1252
AR.MU1P Real 54.3
AR.MU2P Real 407
ARORA Logical False
ASUB Real 1.0 Å
AUGER Logical False

1216 ATLAS User’s Manual

MODELS Statements

Parameter Type Default Units

AUGGEN Logical False

E.BENDING Logical False
E.TAUR.VAR Logical False
E0_TCOEF Real
E0.TCOEF Real
ECRIT Real 4000.0 V/cm
EF.TOL Real 1010-10
E.FGCGTUN Logical False

EIGENS Integer 15
ELECTRONS Logical True
ENERGY.STEP Real 0.0025 eV
ERASE Logical False
E.SC.FGCGTUN Logical False

ESIZE.NEGF Real 2001

ET.MODEL Logical False
ETA.FNONOS Real 1.0
ETAN Real 0.50
ETAP Real 0.33
ETA.NEGF Real 0.0066 eV
EQUIL.NEGF Logical False
EXCITONS Logical False
EVSATMOD Integer 0
F.AE Real 1.8210-7 CV-2s-1
F.BE Real 1.90108 V/cm

F.AH Real 1.8210-7 CV-2s-1

F.BH Real 1.90108 V/cm

F.EFEMIS Character
F.HFEMIS Character
F.GENSINGLET Character

1221 ATLAS User’s Manual

MODELS Statements

Parameter Type Default Units

F.GENTRIPLET Character
F.KSN Character
F.KSP Character
F.PCM Character
F.PDN Character
F.PDP Character
F.TRAP.COULOMBIC Character
F.TRAP.TUNNEL Character
FAST.EIG Logical False
FC.ACCURACY Real 10-6
FERMIDIRAC Logical False
FERRO Logical False
FETIS Logical False
FG.RATIO Real
FILE.LID Character
FIX.BQN Logical False
FIX.BQP Logical False
FIXED.FERMI Logical False
FLDMOB Logical False
FN.CUR Logical False
FNHOLES Logical False
FNHPP Logical False
FNONOS Logical False
FNORD Logical False
FNPP Logical False
FREQ.CONV Logical False
G.EMPIRICAL Logical False
G.STANDARD Logical False
GAINMOD Logical 0

1222 ATLAS User’s Manual

MODELS Statements

Parameter Type Default Units

GANFET Logical False

GATE1 Logical False
GATE2 Logical False
GENERATE.LID Real 0 cm-3s-1
GIAR Logical False
GR.HEAT Logical False
HANSCHQM Logical False
H.ATOM Logical False
H.BENDING Logical False
H.FGCGTUN Logical False
H.MOLE Logical False
H.SC.FGCGTUN Logical False
H.TAUR.VAR Logical False
HCTE Logical False
HCTE.EL Logical False
HCTE.HO Logical False
HEAT.FULL Logical False
HEAT.PETHOM Logical False
HEI Logical False
HHI Logical False
HNSAUG Logical False
HOLE Logical False
HOPMOB Logical False
HVSATMOD Integer 0
HW.BBT.SCHENK Real 0.0186 eV
IG.ELINR Real 6.1610-6 cm

IG.HLINR Real 6.1610-6 cm

IG.ELINF Real 9.210-7 cm

IG.HLINF Real 9.210-7 cm

1223 ATLAS User’s Manual

MODELS Statements

Parameter Type Default Units

IG.EBETA Real 2.5910-4 (V.cm)0.5

IG.HBETA Real 2.5910-4 (V.cm)0.5
IG.EETA Real 2.010-5 V1/3·cm2/3
IG.HETA Real 2.010-5 V1/3·cm2/3
IG.EEOX Real 9.0104 V/cm

IG.HEOX Real 9.0104 V/cm

IG.EB0 Real 3.2 V

IG.HB0 Real 4.0 V
IG.LRELE Real 3.010-6 [Q=1] cm
IG.LRELH Real 2.010-6 [Q=1] cm
II.NODE Logical False
II.VALDI Logical False
IMPACT Logical False
INC.ION Logical False
INCOMPLETE Logical False
INFINITY Real 0.001
IONIZ Logical False
INTERSUB.SPONT Logical False
INV.SF Real 0.75
ISHIKAWA Logical False
ITAT.PP.EL Logical False
ITAT.PP.HO Logical False
ITAT.SC.EL Logical False
ITAT.SC.HO Logical False
JOULE.HEAT Logical True
JTAT Logical False
JTAT.DENSITY Real 0.0 cm-3

JTAT.LEVEL Real 0.0 eV

1224 ATLAS User’s Manual

MODELS Statements

Parameter Type Default Units

JTAT.M2 Real 10-20 V2cm3

K.P Logical False

KAUGCN Real
KAUGCP Real
KAUGDN Real
KAUGDP Real
KC.HYDROGEN Logical False
KLA Logical False
KLAAUG Logical False
KLASRH Logical False
KOSTER Logical False
KSRHCN Real
KSRHCP Real
KSRHGN Real
KSRHGP Real
KSRHTN Real
KSRHTP Real
KSN Integer -1
KSP Integer -1
L.TEMP Logical False
LAMHN Real 9.210-7 cm

LAMHP Real 9.210-7 cm

LAMRN Real 6.1610-6 cm

LAMRP Real 6.1610-6 cm

LAS.ABSOR_SAT Logical False

LAS.ABSORPTION Logical False
LAS.EFINAL Real eV
LAS.EINIT Real eV
LAS.ESEP Real eV

1225 ATLAS User’s Manual

MODELS Statements

Parameter Type Default Units

Parameter Type Default Units

NEW.EIG Logical False

NEW.SCHRO Logical False
NEWIMPACT Logical False
NI.FERMI Logical False
NINV.DORT Logical False
NINV.EC Real 0.0 V/cm
NINV.BETA Real 5.9210-8
NINV.EXP Real 2/3
NPRED.NEGF Real 7
NS.DISS Integer 30
NT.FNONOS Real 0.0 m
NUM.BAND Real
NUM.DIRECT Real
NSPECIES Integer 0
NUMBER Integer 0
NUMCARR Real
OLDIMPACT Logical False
OPTR Logical False
ORTH.SCHRO Real 10-3
OX.BOTTOM Real 0.0 m
OX.LEFT Real m

OX.MARGIN Real 0.0003 m

OX.POISSON Logical False

OX.RIGHT Real m
OX.SCHRO Logical False
PFREQ.CONV Logical False
P.CONCANNON Logical False
P.DORT Logical False
P.HOTSONOS Logical False

1229 ATLAS User’s Manual

MODELS Statements

Parameter Type Default Units

P.KSI_VAR Logical False

P.KSI.VAR Logical False
P.NEGF_PL Logical False
P.NEGF_PL1D Logical False
P.SCHRODING Logical False
P.TEMP Logical False
PACC.DORT Logical False
PACC.EC Real 0.0 V/cm
PACC.BETA Real 4.9710-8
PACC.N Real 3.61019 cm-3
PACC.EXP Real 2/3
PASSLER Logical False
PATH.N Real 3.4107 microns

PATH.P Real 2.3810-7 microns

PATRIN Logical False

PCH.INS Logical False
PCM Logical False
PCM.LOG Logical True
PEFF.N Real 3.2 eV
PEFF.P Real 4.8 eV
PF.NITRIDE Logical False
PFMOB Logical False
PHONONDRAG Logical False
PHOTON.ENERGY Real eV
PHUMOB Logical False
PIC.AUG Logical False
PIEZOELECTRIC Logical False
PIEZO.SCALE Real 1.0
PINV.DORT Logical False

1230 ATLAS User’s Manual

MODELS Statements

Parameter Type Default Units

PINV.EC Real 105 V/cm

PINV.BETA Real 6.110-8
PINV.EXP Real 2/3
POISSON Logical False
POLAR.CHARGE Real 0.0 cm-2
POLAR.SCALE Real 1.0
POLARIZATION Logical False
POST.SCHRO Logical False
PRINT Logical False
PRINT.REGION Integer
PRINT.MATERIAL Character
PRPMOB Logical False
PROGRAM Logical False
PRT.BAND Logical False
PRT.COMP Logical False
PRT.FLAGS Logical False
PRT.IMPACT Logical False
PRT.RECOM Logical False
PRT.THERM Logical False
PSP.SCALE Real 1.0
PT.HEAT Logical False
QCRIT.NEGF Real 310-3
QMINCONC Real 1.010-10 cm-3
QTNL.DERIVS Logical False
QTNLSC.BBT Logical False
QTNLSC.EL Logical False
QTNLSC.HO Logical False
QTREGION Integer
QTUNN Logical False

1231 ATLAS User’s Manual

MODELS Statements

Parameter Type Default Units

QTUNN.BBT Logical False

QTUNN.DIR Real 0
QTUNN.EL Logical False
QTUNN.HO Logical False
QTUNNSC Logical False
QUANTUM Logical False
QWELL Logical False
QWNUM Integer -1
QX.MAX Real RHS of device m
QX.MIN Real LHS of device m
QY.MAX Real Bottom of device m
QY.MIN Real Bottom of device m
R.ELEC Real 1.1
R.HOLE Real 0.7
RAJKANAN Logical False
REGION Integer 0
RTAT.PP Logical False
RTAT.SC Logical False
S.DISSOC Logical False
SBT Logical False
SCHENK.BBT Logical False
SCHENK.EL Logical False
SCHENK.HO Logical False
SCHENK Logical False
SCHKSC.EL Logical False
SCHKSC.HO Logical False
SCHKSC Logical False
SCHSRH Logical False
SCHRO Logical False

1232 ATLAS User’s Manual

MODELS Statements

Parameter Type Default Units

SDISS.TYPE Integer 1
SHAPEOX Logical False
SHI Logical False
SHIRAMOB Logical False
SHJ.EL Logical False
SHJ.HO Logical False
SHJ.NLDERIVS Logical False
SINGLET Logical False
SIS.EL Logical False
SIS.HO Logical False
SIS.NLDERIVS Logical False
SIS.REVERSE Logical False
SIS.TAT Logical False
SISTAT.TN Real 1.0e-6 s
SISTAT.TP Real 1.0e-6 s
SL.GEOM Character 1DY

SL.NUMBER Integer
SLATT Logical False
SLATT.NUMBER Integer
SP.BAND Logical True
SP.CHECK Logical False
SP.DIR Integer 0
SP.EXCORR Logical False
SP.FAST Logical False
SP.GEOM Character 1DX
SP.SMOOTH Logical True
SPEC.NAME Character
SPECIES1.A Integer 0
SPECIES2.A Integer 0

1233 ATLAS User’s Manual

MODELS Statements

DEVDEG.A This causes the dangling bonds created by the DEVDEG.RD or DEVDEG.KN models
to be treated as amphoteric.
DEVDEG.B Activates the device degradation caused by both the hot electron and hot hole
injection currents. HEI and HHI should also be specified in this case.
DEVDEG.E Activates the device degradation caused by hot electron injection current. HEI
should also be specified in this case.
DEVDEG.H Activates the device degradation caused by hot hole injection current. HHI should
also be specified in this case.
DEVDEG.KN Enables the kinetic degradation model.

1236 ATLAS User’s Manual

MODELS Statements

DEVDEG.PL Enables the power law degradation model.

DEVDEG.RD Enable the Reaction-Diffusion degradation model.
KC.HYDROGEN Enables the hydrogen diffusion model in kinetic degradation model.

Mobility Model Flags

ALBRCT Enables the Albrecht mobility model for electrons and holes (see Equation 5-151).
ANALYTIC Specifies an analytic concentration dependent mobility model for silicon which
includes temperature dependence (see Equations 3-202 and 3-203).
ARORA Specifies an analytic concentration and temperature dependent model
according to Arora (see Equations 3-204 and 3-205).
BROOKS Enables the Brooks-Herring mobility model for both electrons and holes (Equation
3-217).
CCSMOB Specifies carrier-carrier scattering model according to Dorkel and Leturq
(see Equations 3-208-3-212).
CONMOB Specifies that a concentration dependent mobility model be used for silicon and
gallium arsenide. This model is a doping versus mobility table valid for 300K only.
CVT Specifies that the CVT transverse field dependent mobility model is used for the
simulation. The alias for this parameter is LSMMOB.
FLDMOB Specifies a lateral electric field-dependent model (see Equation 3-306 and Equation
5-57). The EVSATMOD parameter may be used to define which field dependent
equation is used.
HOPMOB Specifies that the hopping mobility model will be used for electrons and holes (see
Chapter 15 “Organic Display and Organic Solar: Organic Simulators”).
KLA Specifies that the Klaassen Mobility Model (see Equations 3-220-3-245) will be
used for electrons and holes.
LSMMOB This is an alias for CVT.
MASETTI Enables the Masetti mobility model for electrons and holes (see Equations 3-206
and 3-207).
MIN.SURF Specifies that the WATT, TASCH, or SHI mobility models should only apply to
minority carriers.
MOBMOD Specifies mobility degradation by longitudinal electric field only (MOBMOD=1), or by
both longitudinal and transverse electric fields (MOBMOD=2). When
MOBMOD=2, specify the ACC.SF, INV.SF, OX.BOTTOM, OX.LEFT, and OX.RIGHT
parameters. This parameter is only used with TFLDMB1 and TFLDMB2.
MOB.INCOMPL Causes the low field mobility models having a dependence on doping level to
depend on ionized dopant density rather than total dopant density. See “Incomplete
Ionization and Doping Dependent Mobility” on page 170.

1237 ATLAS User’s Manual

MODELS Statements

PF.NITRIDE Enables a model for steady-state conduction in Silicon-Nitride that behaves like
Poole-Frenkel emission at high fields.
PFMOB Specifies that the Poole-Frenkel mobility model will be used for electrons and holes
(see Chapter 15 “Organic Display and Organic Solar: Organic Simulators”).
PHUMOB This is an alias for KLA.
PRPMOB Enables the perpendicular field dependent mobility model as described in
Equations 3-298 and 3-299.
SHIRAMOB Specifies that the Shirahata Mobility Model (see Equation 3-293) will be used for
electrons and holes.
SURFMOB Invokes the effective field based surface mobility model (see Equations 3-300 and
WATT 3-301). SURFMOB parameters are used in the calculation of surface mobility
according to the Watt Model [299]. Do not specify this parameter unless S-PISCES
is installed on your system.
TFLDMB1 Specifies the use of transverse electric field-dependent mobility models. The
SCHWARZ electron model is based on the Schwarz and Russe equations [246] and is
implemented in [255]. The hole model, which is used only when MOBMOD=2, is
based on the Watt and Plummer equations [300].
TFLDMB2 Specifies a transverse electric field dependent mobility model for electrons and
TASCH holes based on a semi-empirical equation [255, 256].
YAMAGUCHI Specifies that the Yamaguchi transverse field dependent mobility model is used in
the simulation.

Note: See Section 21.36 “MOBILITY” for alternative ways to set mobility models.

Recombination Model Flags

AUGER Specifies Auger recombination (see Equation 3-369).

AUGGEN Specifies that the Auger recombination model will be used as a
generation and a recombination term.
AUTOBBT2 Alternative to AUTOBBT.
CDL and CONCDL Enables the Coupled defect level recombination model. CONCDL
additionally enables concentration dependence of the recombination
lifetimes used in this model.
CONSRH Specifies Shockley-Read-Hall recombination using concentration
dependent lifetimes (see Equations 3-332, 3-333, and 3-334).
CONSRH.LAW Enables the concentration dependent Shockley-Read-Hall
recombination model using the alternate parameter set given in
Table 3-71.

1238 ATLAS User’s Manual

MODELS Statements

HNSAUG Enables Auger recombination model that has co-efficients

depending on Lattice temperature and carrier concentration (see
Equations 3-378 and 3-380).
JTAT Enables Local Trap assisted tunneling model with parameters on
MODELS statement.

JTAT.DENSITY Trap density for JTAT model.

JTAT.LEVEL Trap energy for JTAT model.
JTAT.M2 Matrix element squared for JTAT model.
KLAAUG Enables Klaassen’s model for concentration dependent auger
coefficients (see Equations 3-367, 3-368, and 3-369).
KAUGCN, KAUGCP, KAUGDN, and Specify parameters of Klaassen's Auger recombination model (see
KAUGDP Equations 3-203 and 3-204).
KLASRH Enables Klaassen’s model for concentration dependent lifetimes for
Shockley Read Hall recombination (Equations 3-335 and 3-336).
KOSTER Enables a modification to the Langevin recombination rate whereby
the rate is dependent on the minimum mobility as described in
Equation 15-48.
KSRHCN, KSRHCP, KSRHGN, Specify parameters of Klaassen's SRH recombination model
KSRHGP, KSRHTN, and KSRHTP (see Equations 3-335 and 3-336).
NI.FERMI Includes the effects of Fermi statistice into the calculation of the
intrinsic concentration in expressions for SRH recombination.
OPTR Selects the optical recombination model (see Equation 3-365). When
this parameter is specified, the COPT parameter of the MATERIAL
statement should be specified.
PIC.AUG Enables the carrier and doping concentration dependent Auger rate
coefficient model given by Equations 3-378 and 3-379.
S.DISSOC Enables the singlet dissociation model described in Section 15.3.3
“Exciton Dissociation”.
SCHSRH Enables Scharfetter concentration dependent lifetime model.
SRH Specifies Shockley-Read-Hall recombination using fixed lifetimes
(see Equation 3-331).
SRH.EXPTEMP Enables the alternative model for the Lattice Temperature
dependence of the recombination lifetimes.
TAT.LOCAL When used with the ITAT/RTAT models, this causes the tunnel
model to take into account tunnelling into localized defects in the
semiconductor.

1239 ATLAS User’s Manual

MODELS Statements

TAT.NLCURRS When used with the self-consistent ITAT/RTAT models, this puts
non-local couplings into the current continuity equations to try to
improve convergence.
TAT.NLDERIVS When used with the self-consistent ITAT/RTAT models, this puts
non-local couplings into the Poisson equation to try to improve
convergence.
TRAP.AUGER Enables the dependence of recombination lifetime on carrier density
as given by Equations 3-347 and 3-348.
TRAP.COULOMBIC Specifies that the Poole-Frenkel barrier lowering for columbic wells
will be used for traps.
TRAP.HURKX Specifies that the trap-assisted tunneling enhancement factor used in
the CONTINUOUS DEFECTS and INTDEFECTS models applied only to
the recombination/generation model.
TRAP.JTAT Enables Local Trap assisted tunneling model with parameters on
TRAP/DOPING statements.

TRAP.TUNNEL Specifies that the trap-assisted tunneling model will be used for
traps.
TAT.NLDEPTH Allows you to specify the trap energy used in the TAT.NONLOCAL
model. Its units are in eV and how it is interpreted depends on
whether the flag TAT.RELEI is specified. If TAT.NLDEPTH is not
specified, the trap energy is the same as the Intrinsic Fermi energy.
TAT.NONLOCAL Enables the non-local tunneling model in the calculation of the field
effect enhancement factors. Setting this flag will also set the
TRAP.TUNNEL flag.

TAT.RELEI This flag affects how the TAT.NLDEPTH parameter is used to

calculate the trap energy. See the “Trap-Assisted Tunneling” on
page 121 for details.
ZAMDMER Enables the electron lifetime model for TRAP states in LT-GaAs.

Direct Tunneling Flags

An alternative set of direct quantum tunneling models for calculating gate current can be set
using DT.CUR on the MODELS statement. If you set this parameter, then select the particular
model using DT.METH. The allowed values of DT.METH are 1, 2, and 3. The default value of
DT.METH is 1.

DT.METH=1 Selects a Fowler-Nordheim type model, with a correction for the trazepoidal, rather
than the triangular shape of the potential barrier.
DT.METH=2 Selects a WKB model with the assumption of a linearly varying potential. It involves
a calculation of the classical turning points and integration between those limits.
DT.METH=3 Selects an exact formula for a trapezoidal barrier that involves Airy function
solutions.

1240 ATLAS User’s Manual

MODELS Statements

You may also need to specify the type of tunneling taking place by using at least one of the
following flags.

DT.CBET Specifies electron tunneling from conduction band to conduction band.

DT.CUR Enables the self-consistent version of Direct Tunneling.
DT.VBHT Specifies holes tunneling from valence band to valence band.
DT.VBET Specifies Band-to-Band tunneling.
DTPP Specify this along with DT.CUR to enable the Post-Processing version of Direct
tunneling.

Note: Specifying DT.CUR on the MODELS statement will invoke the self-consistent version of these models. You
can also specify DT.CUR and associated parameters on the SOLVE statement, which will invoke the post
processing version of this model for the particular SOLVE statement.

Note: The effective masses used in the tunneling calculations are the default or specified effective masses for the
materials in question, unless you set the ME.DT and MH.DT parameters on the MATERIAL statement.

Generation Model Flags

A0.TNLC Specifies the initial “twist” angle for the twisted neumatic liquid crystal model.
A.BBT.SCHENK This is the schenk band-to-band tunneling parameter (see Equation 3-445).
AUTOBBT Causes band-to-band tunneling parameters to be calculated from first principals
(see Equations 3-456 and 3-457).
B.BBT.SCHENK This is the Schenk band-to-band tunneling parameter (see Equation 3-464).
BBT.ALPHA Specifies the statistical factor scaling used in the Hurkx band-to-band tunneling
model (see Equation 3-463).
BBT.DEHURKX Specifies that the statistical factor used in the Hurkx band-to-band tunneling
model will be modified according to Equation 3-461.
BBT.DJURKX Specifies that the statistical factor used in the Hurkx band-to-band tunneling
model will be modified according to Equation 3-462.
BBT.FORWARD Avoids convergence problems (if using BBT.NONLOCAL in a forward biased
junction).
BBT.HURKX Enables the Hurkx model (see Equation 3-451).
BBT.KANE Specifies a Band-to-Band tunneling model according to Kane.
BBT.KL Specifies a band-to-band tunneling model according to Klaassen.
BBT.NONLOCAL Enables the non-local band-to-band tunneling model. No other band-to-band
tunneling models should be enabled in the same region.

1241 ATLAS User’s Manual

MODELS Statements

BBT.NLDERIVS Enables the inclusion of non-local derivatives in the Jacobian matrix (if you set
BBT.NONLOCAL).

BBT.SHURKX Specifies that the Schenk statistical factor will be used in the band-to-band
tunneling model (see Equation 3-455).
BBT.STD Specifies a standard band-to-band tunneling model (see Equation 3-455). The
alias for this parameter is BTBT.
BBT1 This is an alias for BBT.KL.
BBT2 This is an alias for BBT.STD.
BBT3 This is an alias for BBT.HURKX.
BH.FNONOS This is the barrier height parameter for the FNONOS model.
BTBT This is an alias for BBT.KANE.

Note: The specification of either DEVDEG.E, DEVDEG.H, or DEVDEG.B will initialize the density of electron
(hole) like traps on the interface and will clear out the trapped electron (hole) density on the interface if any.

E.BENDING Specifies that electron band bending will be taken into account for electron
injection.
E.FGCGTUN Enables the Floating Gate to Control Gate (FGCG) tunneling model for
electrons.
E.SC.FGCGTUN Enables the self-consistent version of the Floating Gate to Control Gate
(FGCG) tunneling model for electrons.
ETA.FNONOS This is the trap capture efficiency for the FNONOS model.
FERRO Enables the ferroelectric permittivity model (see Section 3.6.9 “The
Ferroelectric Permittivity Model”).
FG.RATIO Specifies the ratio of the floating gate extent in the Z direction to the channel
width in Z direction for 2D simulation.
FN.CUR Enables both FNORD and FNHOLES.
FNHOLES Enables self-consistent analysis of hole Fowler-Nordheim currents.
FNONOS Enables model for gate tunneling in SONOS structures.
F.TRAP.TUNNEL Specifies the name of a file containing a C-interpreter function to simulate
trap assisted tunneling. The alias for this parameter is F.TRAP.DIRAC.
F.TRAP.COULOMBIC Specifies the name of a file containing a C-interpreter function to simulate
trap assisted tunneling with Coulombic wells.
FNORD Selects a self-consistent Fowler-Nordheim tunneling model for electrons (see
Equation 3-476). This is the recommended approach for calculating Fowler-
Nordheim current.

1242 ATLAS User’s Manual

MODELS Statements

FNHPP Selects a post-processing Fowler-Nordheim tunneling model for holes (see

Equation 3-477). Generally, the Fowler-Nordheim current does not cause
convergence problems so this approach is not required.
FNPP Selects a post-processing Fowler-Nordheim tunneling model for electrons
(see Equation 3-476). Generally, the Fowler-Nordheim current does not cause
convergence problems so this approach is not required.
GATE1 Enables both HEI and HHI.
GATE2 This is an alias for GATE1.
GIAR Enables the graded-index anti-reflective model for the real index of the
region.
H.BENDING Specifies that hole band bending will be taken into account for hole injection
(see Equation 3-484).
H.FGCGTUN Enables the Floating Gate to Control Gate (FGCG) tunneling model for holes.
H.SC.FGCGTUN Enables the self-consistent version of the Floating Gate to Control Gate
(FGCG) tunneling model for holes.
HEI Specifies the calculation of hot electron injection into oxides. This
parameter can be used to simulate MOS gate current or EPROM
programming. In transient mode, the oxide current is self-consistently added
to the floating gate charge (see Equation 3-479).
HHI Specifies the calculation of hot hole current into an oxide in a similar manner
to HEI (see Equation 3-480).
HW.BBT.SCHENK This is the schenk band-to-band tunneling parameter (see Equation 3-445).
II.NODE This is an alias for OLDIMPACT.
II.VALDI This is an alias for VALDINOCI.
IMPACT Invokes the empirical impact ionization model with ionization coefficients
taken from [91]. More rigorous impact ionization models may be specified
with the IMPACT statement.
ITAT.PP.EL Enables the Inelastic Trap Assisted Tunneling model for electrons with no
coupling to current continuity equations.
ITAT.PP.HO Enables the Inelastic Trap Assisted Tunneling model for holes with no
coupling.
ITAT.SC.EL Enables the Inelastic Trap Assisted Tunneling model for electrons with
coupling to current continuity equations.
ITAT.SC.HO Enables the Inelastic Trap Assisted Tunneling model for holes with
coupling to current continuity equations.
MASS.TUNNEL Specifies the effective mass for band to band tunneling (see Equations 3-103
and 3-104).

1243 ATLAS User’s Manual

MODELS Statements

MIM.ITAT.EL Enables the Ielmini Trap assisted Tunneling model for electrons in a Metal-
Insulator-Metal structure.
MIM.ITAT.HO Enables the Ielmini Trap assisted Tunneling model for holes in a Metal-
Insulator-Metal structure.
MIM.RTAT Enables the Ielmini Trap assisted Tunneling model for electrons and holes in a
Metal-Insulator-Metal structure.
MIMSLICEPTS Sets the number of points in the interpolation mesh using for tunneling
calculation in MIMTUM and MIM.ITAT models.
MIMTUN Enables the interelectrode tunneling model for electrons and holes in a Metal-
Insulator-Metal structure.
MIMTUN.BBT Enables a model for the interelectrode band-to-band tunneling in a Metal-
Insulator-Metal structure.
MIMTUN.EL Enables the interelectrode tunneling model for electrons in a Metal-Insulator-
Metal structure.
MIMTUN.HO Enable the interelectrode tunneling model for holes in a Metal-Insulator-Metal
structure.
N.ANISO Specifies that the region can be modeled with an anisotropic real index of
refraction.
N.CONCANNON Enables the Concannon gate current model for electrons.
NEARFLG Specifies the model used for oxide transport when HEI, HHI or FNSONOS are
used. The default is false which sets a purely drift based model assigning
gate current to the electrodes where the electric field lines through the oxide
terminate. Setting NEARFLG replaces this model with one assuming the gate
current is flowing to the electrode physically nearest the point of injection.
This assumes a purely diffusion transport mechanism. Setting NEARFLG for
devices with only one gate or a coarse mesh is advised.
NEWIMPACT Specifies that impact ionization should use the form described in Equation 3-
389.
N.HOTSONOS Alternative hot electron gate current model to be used with DYNASONOS
model.
N1.GIAR Specifies the starting isotropic index for a graded-index anti-reflective layer.
N2.GIAR Specifies the ending isotropic index for a graded-index anti-reflective layer.
NE.TNLC Specifies the constant extra-ordinary index for a twisted nematic liquid crystal
region.
NE1.GIAR Specifies the starting extra-ordinary index for a twisted anisotropic graded-
index anti-reflective layer.
NE2.GIAR Specifies the ending extra-ordinary index for a twisted anisotropic graded-
index anti-reflective layer.

1244 ATLAS User’s Manual

MODELS Statements

NE2.GIAR Specifies the ending extra-ordinary index for a twisted anisotropic graded-
index anti-reflective layer.
NO1.GIAR Specifies the starting extra-ordinary index for a twisted anisotropic graded-
index anti-reflective layer.
NO.TNLC Specifies the constant ordinary index for a twisted nematic liquid crystal
region.
NT.FNONOS This is the trap area density for saturation mode of the FNONOS model.
OLDIMPACT Specifies that the impact ionization should use the form described in Equation
3-388. The alias for parameter is II.NODE.
P.CONCANNON Enables the Concannon gate current model for holes.
P.HOTSONOS Alternative hot electron gate current model to be used with DYNASONOS
model.
QTNL.DERIVS Enables non-local couplings in the Jacobian matrix in the self-consistent
direct quantum tunneling models.
QTNLSC.BBT Enables band-to-band mode of the self-consistent direct quantum
tunneling model.
QTNLSC.EL Enables the self-consistent direct quantum tunneling model for electrons.
QTNLSC.HO Enables the self-consistent direct quantum tunneling model for holes.
QTREGION Restricts the particular non-local tunneling model to the specified
QTREGION. If omitted, then the tunneling model is enabled for all
QTREGIONs
QTUNN.BBT Enables the band-to-band mode of the direct quantum tunneling model. This
does not work with the quantum confined variant of quantum
tunneling.
QTUNN.DIR specifies the tunneling current direction if you specify a rectangular mesh
using the QTX.MESH and QTY.MESH statements. A value of 0 means the Y
direction (default) and a value of 1 means the X direction.
QTUNN.EL Enables direct quantum tunneling model for electrons (see Section 3.6.7
“Gate Current Models”).
QTUNN.HO Enables direct quantum tunneling model for holes.
QTUNN This is the same as specifying QTUNN.EL, QTUNN.HO, and QTUNN.BBT. These
models calculate the quantum tunneling current through an oxide.
QTUNNSC Enables the self-consistent versions of the direct quantum tunneling
models for gate current. These include the tunneling current self-consistently
in the carrier continuity equations. QTUNNSC is the same as specifying
QTNLSC.EL, QTNLSC.HO, and QTNLSC.BBT.

RTAT.PP Enables the Inelastic Trap Assisted Tunneling model for electrons and holes,
including recombination, with no coupling to current continuity equations.

1245 ATLAS User’s Manual

MODELS Statements

RTAT.SC Enables the Inelastic Trap Assisted Tunneling model for electrons and holes,
including recombination, with coupling to current continuity equations.
SBT This is an alias for UST.
SCHENK.BBT This is the schenk band-to-band tunneling model.
SCHENK.EL This is the schenk post-processing oxide tunneling current for electrons.
SCHENK.HO This is the schenk post-processing oxide tunneling current for holes.
SCHENK This is the schenk post-processing oxide tunneling current for bipolar.
SCHKSC.EL This is the schenk self-consistent oxide tunneling current for electrons.
SCHKSC.HO This is the schenk self-consistent oxide tunneling current for holes.
SCHKSC This is the schenk self-consistent oxide tunneling current for bipolar.
SHAPEOX Specifies which algorithm will be used for calculating quantum tunneling
current on an non-rectangular mesh. If this parameter is not specified, then the
current is calulated in a direction normal to each segment of the
semiconductor/oxide interface. If SHAPEOX is specified, then the minimum
distance is found between each segment and a contact or a polysilicon region.
The quantum tunneling current is calculated along the direction that gives this
minimum distance.
SHJ.EL Enables the non-local heterojunction tunneling model for electrons. See
Section 5.2.3 “Non-local Heterojunction Tunneling Model ”.
SHJ.HO Enables the non-local heterojunction tunneling model for holes. See Section
5.2.3 “Non-local Heterojunction Tunneling Model ”.
SHJ.NLDERIVS Enables non-local derivatives for the Newton solver for SHJ models.
SIS.EL Enables non-local quantum barrier tunneling model for electrons. See Section
5.2.4 “Non-local Quantum Barrier Tunneling Model”.
SIS.HO Enables non-local quantum barrier tunneling model for holes. See Section
5.2.4 “Non-local Quantum Barrier Tunneling Model”.
SIS.NLDERIVS Enables non-local derivatives into Newton solver for SIS models.
SIS.REVERSE Causes the SIS.EL and SIS.HO models to use the quasi-Fermi levels at the
start and end of the mesh slices. This may increase stability in some cases.
SIS.TAT Enables Trap-Assisted-Tunnelling for structures that have an insulator layer
between two semiconductors.
SISTAT.TN Basic electron recombination time for SIS.TAT model.
SISTAT.TP Basic hole recombination time for SIS.TAT model.
TAT.POISSON When used with ITAT/RTAT model, this causes the trap ionized densities to be
included in the Poisson equation solution.

1246 ATLAS User’s Manual

MODELS Statements

TAT.SLICEPTS When used with ITAT/RTAT model, this controls the number of discrete
points used in each slice of the Quantum Tunnel region.
TNLC Enables the twisted nematic liquid crystal model.
TR.TNLC Specifies the twist rate in degrees per micron for a twisted nematic liquid
crystal region.
TSQ.ZAIDMAN Parameter for the Zaidman model.
UST Enables the universal Schottky tunneling model (see Section 3.5.2 “Schottky
Contacts”). The alias for this parameter is SBT.
VALDINOCI Enables the Valdinoci impact ionization model. See Equations 3-394,
3-407, and 3-415. The alias for this parmeter is II.VALDI.
ZAIDMAN Specifies the tunneling model for surface field emission.

Carrier Statistics Model Flags

BGN Specifies bandgap narrowing model (see Equation 3-46) with Slotboom default
values.
BGN2 Specifies bandgap narrowing model (see Equation 3-46) with Klaassen default
values.
BGN.ALAMO Enables del Alamo Bandgap Narrowing model (see Section 3.2.11 “Bennett-
Wilson and del Alamo Bandgap Models”).
BGN.BENNETT Enables Bennett-Wilson Bandgap Narrowing model (see Section 3.2.11 “Bennett-
Wilson and del Alamo Bandgap Models”).
BGN.KLAASSEN Has same effect as BGN2.
BGN.LIND Enables the Lindefelt Bandgap narrowing model for Si and SiC. See Section 3.2.13
“Lindefelt Bandgap Narrowing Model”.
BGN.SLOTBOOM Has same effect as BGN.
BGN.SCHENK Enables the bandgap narrowing model described in Section 3.2.12 “Schenk
Bandgap Narrowing Model”.
BOLTZMANN Specifies that Boltzmann carrier statistics be used (see Equations 3-25 and 3-26).
FERMIDIRAC Specifies that Fermi-Dirac carrier statistics be used (see Equations 3-25–3-99).
INCOMPLETE Accounts for incomplete ionization of impurities in Fermi-Dirac statistics (see
Equations 3-73 and 3-74).

1247 ATLAS User’s Manual

MODELS Statements

INC.ION Specifies that Equations 3-78 and 3-79 should be used for calculations of
incomplete ionization.
IONIZ Accounts for complete ionization of heavily doped silicon when using
INCOMPLETE.

UBGN Enables the universal bandgap narrowing model given in Equation 3-50.

Quantum Model Parameters

2DXY.SCHRO This is the flag that tells ATLAS to solve a 2D Schrodinger equation in XY plane
when you also specify the SCHRODINGER or P.SCHRODINGER flag. When it is off,
1D Schrodinger equations will be solved at each slice perpendicular to X axis.
ALT.SCHRO Specifies usage of an alternate Schrodinger solver.
BC.QUANT Specifies Dirichlet boundary conditions for all insulator interfaces for the
quantum transport equation.
BQP.N Enables the BQP model for electrons.
BQP.P Enables the BQP model for holes.
BQP.NGAMMA Sets the value of BQP  factor for electrons.
BQP.NALPHA Sets the value of BQP  factor for electrons.
BQP.NEUMANN Uses explicitly enforced Neumann boundary conditions instead of the implicit
implementation.
BQP.PALPHA Sets the value of BQP  factor for holes.
BQP.PGAMMA Sets the value of BQP  factor for holes.
BQP.QDIR Sets the direction of principal quantization.
CALC.FERMI Specifies that the Fermi level used in calculation of the quantum carrier
concentration in the Schrodinger-Poisson model is calculated from the classic
carrier concentrations.
DEC.C1 This is a conduction band shift for 1st pair of electron valleys.
DEC.C2 This is a conduction band shift for 2nd pair of electron valleys.
DEC.C3 This is a conduction band shift for 3rd pair of electron valleys.
DEC.D2 This is a conduction band shift for 2 electron valleys if DEC.C1 is not set.
DEC.D4 This is a conduction band shift for 4 electron valleys if DEC.C2 is not set.
DEC.ISO This is a conduction band shift for isotropic electron band, when NUM.DIRECT=1.
DEV.HH This is a valence band shift for heavy holes.
DEV.LH This is a valence band shift for light holes.

1248 ATLAS User’s Manual

MODELS Statements

DEV.ISO This is a valence band shift for isotropic hole band, when NUM.BAND=1.
DEV.SO This is a valence band shift for split-off holes.
DG.N This is an alias for QUANTUM (see Section 13.3 “Density Gradient (Quantum
Moments Model)”).
DG.P This is an alias for P.QUANTUM (see Section 13.3 “Density Gradient (Quantum
Moments Model)”).
DGN.GAMMA Specifies the fit factor for electrons for the electron density gradient model (see
Equation 13-17).
DGP.GAMMA Specifies the fit factor for holes for the hole density gradient model (see Equation
13-18).
DGLOG Specifies that the gradient of the log of concentration is to be used in the
calculation of the quantum potential in the density gradient model (see Equation 13-
17).
DGROOT Specifies that the gradient of the root of the concentration is to be used in the
calculation of the quantum potential in the density gradient model (see Equation 13-
18).
EIGENS Specifies the number of eigenstates solved for by the Poisson-Schrodinger solver
(see Section 13.2 “Self-Consistent Coupled Schrodinger Poisson Model”).
EF.TOL Controls the eigenvector tolerance on numerical accuracy for the alternate
eigenvalue solver enabled by the NEW.EIG parameter.
EQUIL.NEGF Specified that the region will be trated as quasi-equilibrium in planar NEGF model.
ESIZE.NEGF Sets the number energy grid points in the NEGF solver.
ETA.NEGF An imaginary optical potential, added to Hamiltonian of quasi-equilibrium regions
in planar NEGF model.
EV.TOL Controls eigenvalue tolerance on numerical accuracy for the alternate
eigenvalue solver enabled by the NEW.EIG parameter.
FAST.EIG Enables a computationally efficient version of the eigen-value solver.
FIX.BQN Specifies that solutions will use the last calculated or loaded values of the
electron quantum potential even if BQP.N is not in effect.
FIX.BQP Specifies that solutions will use the last calculated or loaded values of the
electron quantum potential even if BQP.P is not in effect.
FIXED.FERMI Specifies that a constant Fermi level is used along each individual slice in the Y
direction for carrier concentration calculation in the Schrodinger-Poisson Model.
HANSCHQM Turns on the Hansch quantum effects approximation model for N channel MOS
devices (see Section 13.5.1 “Hansch’s Model”).

1249 ATLAS User’s Manual

MODELS Statements

LED Specifies that the region is to be treated as a light emitting region and included in
postprocessing for LED analysis. See Chapter 11 “LED: Light Emitting Diode
Simulator”.
N.DORT Turns on the Van Dort quantum effects approximation model for N channel MOS
devices (see Section 13.5.2 “Van Dort’s Model”).
N.NEGF_PL Activates planar NEGF solver for electrons. Solution is done in all 1D slices.
N.NEGF_PL1D Activates planar NEGF solver for electrons. Solution is done for the first 1D slice
and copied to all other slices.
N.QUANTUM This is an alias for QUANTUM (see Section 13.3 “Density Gradient (Quantum
Moments Model)”).
NACC.BETA Specifies a default parameter value for the Van Dort MOS quantum correction
model for electron accumulation described in Section 13.5.2 “Van Dort’s Model”.
NACC.DORT Enables the Van Dort MOS quantum correction model for electron accumulation
described in Section 13.5.2 “Van Dort’s Model”.
NACC.EC Specifies a default parameter value for the Van Dort MOS quantum correction
model for electron accumulation described in Section 13.5.2 “Van Dort’s Model”.
NACC.EXP Specifies a default parameter value for the Van Dort MOS quantum correction
model for electron accumulation described in Section 13.5.2 “Van Dort’s Model”.
NACC.N Specifies a default parameter value for the Van Dort MOS quantum correction
model for electron accumulation described in Section 13.5.2 “Van Dort’s Model”.
NEGF_CMS Enforces coupled mode space NEGF approach.
NEGF_MS Activates a mode-space NEGF solver. See Chapter 13 “Quantum: Quantum Effect
Simulator” for more information.
NEGF_UMS Enforces uncoupled mode space NEGF approach.
NPRED.NEGF Sets a number of predictor iterations in the predictor-corrector scheme of
Schrodinger and NEGF solvers.
NEW.SCHRO Specifies that the new non-rectilinear self-consistent Schrodinger-Poisson solver
will be used for an ATLAS mesh (see Section 13.2 “Self-Consistent Coupled
Schrodinger Poisson Model”). When this is specified, the Schrodinger solver mesh
should also be specified using the SX.MESH and SY.MESH statements.
NEW.EIG Enables an alternate eigenvalue solver that obtains eigenvectors using the INVERSE
ITERATION method.

NINV.BETA Specifies a default parameter value for the Van Dort MOS quantum correction
model for electron inversion described in Section 13.5.2 “Van Dort’s Model”.
NINV.DORT Enables the Van Dort MOS quantum correction model for electron inversion
described in Section 13.5.2 “Van Dort’s Model”.
NINV.EC Specifies a default parameter value for the Van Dort MOS quantum correction
model for electron inversion described in Section 13.5.2 “Van Dort’s Model”.

1250 ATLAS User’s Manual

MODELS Statements

NINV.EXP Specifies a default parameter value for the Van Dort MOS quantum correction
model for electron inversion described in Section 13.5.2 “Van Dort’s Model”.
NUM.BAND Specifies the number of valence bands to retain for Schrodinger Poisson
solutions.
NUM.DIRECT Specifies the number of conduction band directions in the k space to retain for
Schrodinger Poisson solutions.
ORTH.SCHRO Specifies the critical value used for determination of the orthogonality of
eigenvectors (wavefunctions) resulting from Schrodinger equation solutions. The
orthogonality measurement is done by taking the sum of the product of individual
samples from two eigenvectors in question and dividing the sum by the number of
samples. If that measure exceeds the value specified by ORTH.SCHRO, then an error
message will be issued during any subsequent output of structure files. This
message indicates that you should carefully check any wavefunction data contained
in the structure file and make adjustments as necessary.
OX.MARGIN This is the maximum electron penetration into oxide. You can use it with OX.SCHRO
in non-planar semiconductor-oxide interface or nonrectangular channel crossection
(ATLAS3D) to reduce the number of nodes used in the solution of the Schrodinger
equation.
OX.POISSON Calculates the quasi-Fermi levels in insulators for Schrodinger Poisson
solutions.
OX.SCHRO Specifies that the band edges of insulators are included in the solution of
Schrodinger's Equation in self-consistent solutions of the Schrodinger-Poisson
Model (see Section 13.2 “Self-Consistent Coupled Schrodinger Poisson Model”).
P.DORT Turns on the Van Dort quantum effects approximation model for P channel MOS
devices (see Section 13.5.2 “Van Dort’s Model”).
P.NEGF_PL Activates planar NEGF solver for holes. Solution is done in all 1D slices.
P.NEGF_PL1D Activates planar NEGF solver for holes. Solution is done for the first 1D slice and
copied to all other slices.
P.SCHRODING Computes the self consistent Schrodinger Poisson for holes.
PACC.BETA Specifies a default parameter value for the Van Dort MOS quantum correction
model for hole accumulation described in Section 13.5.2 “Van Dort’s Model”.
PACC.DORT Enables the Van Dort MOS quantum correction model for hole accumulation
described in Section 13.5.2 “Van Dort’s Model”.
PACC.EC Specifies a default parameter value for the Van Dort MOS quantum correction
model for hole accumulation described in Section 13.5.2 “Van Dort’s Model”.
PACC.EXP Specifies a default parameter value for the Van Dort MOS quantum correction
model for hole accumulation described in Section 13.5.2 “Van Dort’s Model”.
PACC.N Specifies a default parameter value for the Van Dort MOS quantum correction
model for hole accumulation described in Section 13.5.2 “Van Dort’s Model”.

1251 ATLAS User’s Manual

MODELS Statements

PIEZO.SCALE This is an alias for POLAR.SCALE.

PIEZOELECTRIC This is an alias for POLARIZATION.
PINV.BETA Specifies a default parameter value for the Van Dort MOS quantum correction
model for hole inversion described in Section 13.5.2 “Van Dort’s Model”.
PINV.DORT Enables the Van Dort MOS quantum correction model for hole inversion described
in Section 13.5.2 “Van Dort’s Model”.
PINV.EC Specifies a default parameter value for the Van Dort MOS quantum correction
model for hole inversion described in Section 13.5.2 “Van Dort’s Model”.
PINV.EXP Specifies a default parameter value for the Van Dort MOS quantum correction
model for hole inversion described in Section 13.5.2 “Van Dort’s Model”.
POLARIZATION Enables the automatic calculation of added interface charge due to spontaneous and
piezoelectric polarization. The alias for this parameter is PIEZOELECTRIC. See
Section 3.6.12 “Polarization in Wurtzite Materials”.
POLAR.CHARG Specifies polarization charge densities to replace the density calculated using
Equations 3-370 and 3-371.
POLAR.SCALE Specifies a constant scale factor multiplied by the calculated spontaneous and
piezoelectric polarization charges when you enable polarization by setting the
POLARIZATION parameter of the REGION statement. The alias for this parameter is
PIEZO.SCALE. See Section 3.6.12 “Polarization in Wurtzite Materials”.

POST.SCHRO Specifies to calculate the Schrodinger Equation (see Section 13.2 “Self-Consistent
Coupled Schrodinger Poisson Model”) for quantum effects as a post processing
step.
PSP.SCALE Specifies a constant scale factor similar to POLAR.SCALE but applied only to
spontaneous polarization.
QMINCONC Specifies the minimum carrier concentration considered in the gradient
calculations of the density gradient model or in Poisson's Equation in the self-
consistent Schrodinger-Poisson Model (see Sections 13.2 “Self-Consistent Coupled
Schrodinger Poisson Model” and 13.3 “Density Gradient (Quantum Moments
Model)”).
QCRIT.NEGF Sets a convergence criterium for potential in case of NEGF or Schrodinger solver.
QUANTUM Enables the density gradient (see Section 13.3 “Density Gradient (Quantum
Moments Model)”).
QWELL Specifies that the region is treated as a quantum well for calculation of
radiative recombination or gain or both for certain optoelectronic models.
QWNUM Optionally sets the number of the quantum well region.
QX.MAX, QX.MIN Specify the minimum and maximum extent of the Poisson-Schrodinger solver along
the X axis direction (see Section 13.2 “Self-Consistent Coupled Schrodinger
Poisson Model”).

1252 ATLAS User’s Manual

MODELS Statements

QY.MAX, QY.MIN Specify the minimum and maximum extent of the domain of the Schrodinger solver
(see Section 13.2 “Self-Consistent Coupled Schrodinger Poisson Model”).
SCHRO Enables the Poisson-Schrodinger solver (see Section 13.2 “Self-Consistent Coupled
Schrodinger Poisson Model”). This can be used for zero carrier solutions only
(specified by METHOD CARRIERS=0). This is typically combined with the EIGENS
parameter to control the number of eigenstates calculated. This is a 1D solver that is
solved within the mesh between limits set by QX.MIN and QX.MAX.
SL.GEOM Specifies the superlattice geometry as either 1DY (default) or 1DX.
SL.NUMBER Assigns the index for a regions membership in a indexed superlattice.
SLATT Enables the superlattice model for a region.
SLATT.NUMBER This is an alias for SLATT.NUMBER.
SP.BAND This is a flag that turns on a banded matrix solver used to solve the 2D
Schrodinger equation in ATLAS3D. Otherwise, a full matrix solver is used.
SP.CHECK Enables checking of orthagonality of wavefunctions for Schrodinger - Poisson
solutions.
SP.DIR Specifies direction of quantum confinement in Schrodinger equation and used in
materials with single electron band but with anistropic effective mass. It is active
only when you set SCHRODINGER and NUM.DIRECT=1 on the MODELS statement. The
values of effective mass(es) in quantization direction(s) and DOS effective mass are
give below.
In ATLAS, quantization is in Y direction:
SP.DIR=0 my = MC, dos_mass= MC

SP.DIR=1 my = ML, dos_mass= MT1  MT2

SP.DIR=2 my = MT1, dos_mass= ML  MT2

SP.DIR=3 my = MT2, dos_mass= MT1  MT2

In ATLAS, quantization is in XY plane (with SCHRO.2DXY on MODELS
statement):
SP.DIR=0 (mx,my)= (MC,MC), dos_mass= MC
SP.DIR=1 (mx,my)= (MT1,ML), dos_mass= MT2
SP.DIR=2 (mx,my)= (MT2,MT1), dos_mass= ML
SP.DIR=3 (mx,my)= (ML,MT2), dos_mass= MT1
In ATLAS3D, quantization is in YZ plane:
SP.DIR=0 (my,mz)= (MC,MC), dos_mass= MC
SP.DIR=1 (my,mz)= (ML,MT2), dos_mass= MT1
SP.DIR=2 (my,mz)= (MT1,ML), dos_mass= MT2
SP.DIR=3 (my,mz)= (MT2,MT1), dos_mass= ML

1253 ATLAS User’s Manual

MODELS Statements

SP.EXCORR Activates exchange-correlation correction in Schrodinger solvers.

SP.FAST Activates a fast product-space approach in a 2D Schrodinger solver.
SP.GEOM Sets a dimensionality and direction of a Schrodinger solver. See Chapter 13
“Quantum: Quantum Effect Simulator” for more information.
SP.SMOOTH Enables smoothing of quantum carrier calculations. When you set the SP.SMOOTH
parameter on the MODELS statement, the SP solver will use the degeneracy factor
and effective mass evaluated at the location of the maximum value of the primary
wave function for calculation of carrier concentration. This removes any
discontinuities due to abrupt changes in mass or degeneracy.
WELL.CAPT Enables the quantum well capture escape model through multiple phonon emission.
WELL.DENERGY Energy spacing for numerical integration of spontaneous emission spectrum to
radiative recombination.
WELL.INPLANE Enables in-plane 2D carrier drift diffusion transport for the quantum well capture
escape model.
WELL.NX Specifies the number of grid points in x, y, and z directions for Schrodinger
WELL.NY equation in a quantum well region.
WELL.NZ

WELL.SEPARATE Forces each subband to have its own Fermi level for the quantum well capture
escape model.
WELL.WBTUNN Enables tunneling between confined states and bulk carriers for the quantum well
capture secape model.
WELL.WWSPONT Enables raditative transitions between neighboring wells in the quantum well
capture escape model.
WELL.WWTUNN Enables interwell tunneling between confined states for the quantum well capture
escape model.

1254 ATLAS User’s Manual

MODELS Statements

Energy Balance Simulation Flags

A.TEMP This is an alias for HCTE.

E.TAUR.VAR Specifies that electron temperature dependent energy relaxation time is used.
Use the TRE.T1, TRE.T2, TRE.T3, TRE.W1, TRE.W2, and TRE.W3 parameters
on the material statement to specifies the energy relaxation time.
ET.MODEL This is an alias for HCTE.
H.TAUR.VAR Specifies that hole temperature dependent energy relaxation time is used. Use
TRH.T1, TRH.T2, TRH.T3, TRH.W1, TRH.W2, and TRH.W3 parameters on the
MATERIAL statement to specifies the energy relaxation time.

HCTE Specifies that both electron and hole temperature will be solved. The aliases for
this parameter are A.TEMP and ET.MODEL.
HCTE.EL Specifies that electron temperature will be solved.
HCTE.HO Specifies that hole temperature will be solved.
F.KSN Specifies the name of a file containing a C-INTERPRETER function specifying
the electron Scattering Law Exponent as a function of electron energy. This is
the value of Scattering Law Exponent. Values other than 0 or -1 are
permitted.
F.KSP Specifies the name of a file containing a C-INTERPRETER function specifying
the hole Scattering Law Exponent as a function of hole energy. This is the
value of Scattering Law Exponent. Values other than 0 or -1 are permitted.
GR.HEAT Includes generation/recombination heat source terms in the lattice heat flow
equation.
JOULE.HEAT Includes the Joule heat source terms in the lattice heat flow equation.
KSN Specifies which hot carrier transport model will be used for electrons. KSN=0
selects the hydrodynamic model and KSN=-1 selects the energy balance
model.
KSP Specifies which hot carrier transport model will be used for holes. KSP=0
selects the hydrodynamic model and KSP=-1 selects the energy balance model.
TAUMOB Specifies the dependence of relaxation times with carrier temperature in the
mobility definition. If TAUMOB is specified, the values of MATERIAL statement
parameters TAUMOB.EL and TAUMOB.HO are dependent on the carrier
temperature.
TAUTEM Specifies the dependence of relaxation times with carrier temperature. If
TAUTEM is specified, the values of MATERIAL statement parameters
TAUREL.EL and TAUREL.HO are dependent on carrier temperature.

N.TEMP or HCTE.EL Specifies that the electron temperature equation will be solved.

1255 ATLAS User’s Manual

MODELS Statements

P.TEMP or HCTE.HO Specifies that the hole temperature equation will be solved.
PASSLER Enables Passler's bandgap versus temperature model.
PT.HEAT Includes Peltier-Thomson heat sources in the lattice heat flow equation.

Lattice Heating Simulation Flags

F.PDN Uses C-Interpreter function for phonon drag contribution to electron

thermopower.
F.PDP Uses C-Interpreter function for phonon drag contribution to hole
thermopower.
HEAT.FULL Enables all thermal sources and sinks (Joule heat, generation-
recombination heat, and Peltier Thomson heat).
HEAT.PETHOM This can be used to turn off the Peltier-Thomson heat source in the
HEAT.FULL option.

LAT.TEMP Specifies that the lattice temperature equation will be solved. For lattice
L.TEMP heating simulation there must be at least one thermal contact defined using
the THERMCONTACT statement.
PCM Specifies that the phase change material temperature dependent
resistivity is to be used.
PCM.LOG Specifies that for phase change materials the interpolation for resistivity
between the low temperature value (PCM.CTC and PCM.CRHO) and the high
value (PCM.ATC and PCM.ARHO) is logarithmic.
PHONONDRAG Enables the phonon drag contribution to thermopower.

Note: These parameters should only be specified if GIGA or GIGA3D is enabled on your system.

Defect Generation Model

BEAM.LID Specifies that the generation rate for the light induced defect generation
model is calculated only from active BEAM statements, which have the LID
parameter specified.
FILE.LID Specifies the file name where the dangling bond density of states
distribution, as a function of energy, will be stored.
GENERATE.LID Specifies the generation rate for the light induced defect generation model.
TIME.EP Specifies the bias stress time value used in the amphoteric defect
generation model.
TIME.LID Specifies the time value used in the light induced defect generation model.
TRAP.LID Specifies the light induced defect generation model will be used.

1256 ATLAS User’s Manual

MODELS Statements

Magnetic Field Model

BX This is the X Component of Magnetic Field Vector (Tesla).

BY This is the Y Component of Magnetic Field Vector (Tesla).
BZ This is the Z Component of Magnetic Field Vector (Tesla).
MAGMIN.N Stability parameter for electron equation (see Section 3.11 “Carrier Transport in an
Applied Magnetic Field”).
MAGMIN.P Stability parameter for hole equation (Section 3.11 “Carrier Transport in an Applied
Magnetic Field”).
R.ELEC This is the Hall scattering factor for electrons .
R.HOLE This is the Hall scattering factor for holes.

Model Macros

BIPOLAR Selects a default set of models which are used when simulating bipolar devices. The
bipolar models are CONMOB, FLDMOB, BGN, CONSRH, and AUGER. If LAT.TEMP is also
specified on the MODELS statement, or the TEMPERATURE parameter differs from 300
Kelvin by more than 10 Kelvin, then the ANALYTIC model is used instead of CONMOB.
BIPOLAR2 Selects an alternative default set of models which are used when simulating bipolar
devices. The bipolar models are FERMI, KLA, KLASRH, KLAAUG, FLDMOB, and BGN.
ERASE Specifies a default set of models which are used to simulate EEPROM erasure. When
it’s specified, the MOS, FNORD, IMPACT, and BBT.KL models will be used.
GANFET Enables models convenient for simulation for GaN based FETs. These models include
POLAR, CALC.STRAIN, SRH, ALBRCT.N, and POLAR.SCALE=1.0.

MOS Specifies a default set of models for MOS devices. The MOS models are CVT, SRH
and FERMI.
MOS2 Specifies a default set of models for MOS devices. The MOS2 models are KLA, SHI,
SRH, FERMI and BGN.

PROGRAM Specifies a default set of models used when writing to EEPROMS. When PROGRAM is
specified, the MOS, HEI, and IMPACT models will be used.

1257 ATLAS User’s Manual

MODELS Statements

General Parameters

ADACHI Forces the material parameters for InGaAsP (see Section 5.4.5 “In(1-
x)Ga(x)As(1-y)P(y) System”) for InGaP.
ALPHA.DOS Specifies that density of states data should be extracted from the absorption
spectra during output of complex index of refraction using OUT.INDEX on the
MATERIAL statement.

ALL Enables solution for both electrons and holes.

ASUB Specifies the substrate lattice constant for strain calculations as described in
Section 3.6.11 “Epitaxial Strain Tensor Calculation in Wurtzite”.
BB.GAMMA Specify the band-to-band tunneling parameters (see Equation 3-455).
BBT.GAMMA This is an alias for BB.GAMMA.
BOUND.TRAP Enables modeling of traps coincident with ohmic boundaries.
CARRIERS Specifies the number of carriers to simulate, which should be 0, 1 or 2. The alias is
NUMCARR.

CHUANG This is an alias for the WZ.KP parameter.

CUBIC35 Specifies to use the material dependent band parameter model discussed in
Section 5.4.1 “Cubic III-V Semiconductors”.
DEVICE Specifies which device in MIXEDMODE simulation the MODELS statement should
apply to. The synonym for this parameter is STRUCTURE.
DRIFT.DIFF Specifies that the drift-diffusion transport model is to be used. This implies that
the electron and hole carrier temperature equations will not be solved.
ELECTRONS Specifies that the electron continuity equation is included in simulation.
FC.ACCURACY This is the maximum ratio of the change in integrated photogeneration to the total
that is acceptable during iterations for calculation of frequency conversion
material emission (see Section 10.11 “Frequency Conversion Materials (2D
Only)”).
FETIS Specifies that the default material parameters for GaN/AlN/IaN, AlGaN, and
InGaN are what is used by the FETIS [263] simulator.
FREQ.CONV Enables the frequency conversion material model.
F.PCM Specfies the file name of a C-Interpreter function that models the resistivity of a
conductor as a function of time, temperature, and history.
G.EMPIRICAL Enables the empirical gain model from Section 3.9.4 “The Empirical Gain Model”
(GAINMOD=2).
G.STANDARD Enables the gain model from Section 3.9.3 “The Standard Gain Model”
(GAINMOD=1).

1258 ATLAS User’s Manual

MODELS Statements

H.ATOM Enables the atomic hydrogen transport model for use with multistate trap model.
H.MOLE Enables the molecular hydrogen transport model for use with multistate trap
model.
HOLE Specifies that the hole continuity equation is included in simulation.
ISHIKAWA Enables Ishikawa's strain model for InGaAs/InGaAsP/InGaP (see Section 3.9.12
“Ishikawa's Strain Effects Model”).
K.P Enables using the k*p model effective masses and band edge energies for drift-
diffusion simulation.
LI This is an alias for the ZB.TWO parameter.
LORENTZ Enables Lorentz gain broadening.
MATERIAL Specifies which material from Table B-1 will apply to the MODELS statement. If a
material is specified, then all regions defined as being composed of that material
will be affected.
MS.DISS Specifies the upper limit for the integration of the geminate pair distance
distribution function in terms of integer multiples of the geminate pair
distance parameter A.SINGLET (i.e., the upper limit is given by
MS.DISS*A.SINGLET).

NAME Specifies which region will be applied to the MODELS statement. Note that the
name must match the name specified in the NAME parameter of the REGION
statement or the region number.
NS.DISS Specifies the number of samples to take in evalution of the integral of the
geminate pair distance distribution function.
NSPECIES Number of ionic species to be simulated up to a maximum of 3 (or 2 if GIGA is
enabled).
PATRIN Enables the strain dependence of bandgap in the Raykanan silicon absorption
model in Equation 3-624.
PCH.INS Specifies that polarization charges will be included along interfaces with
insulators or with the outside domain.
PFREQ.CONV Specifies that the ray trace for emissions from a frequency conversion material are
periodic with respect to the boundaries at the minimum and maximum x
coordinates.
PRT.BAND Enables run-time output of regional band parameter model information.
PRT.COMP Enables run-time output of regional composition information.
PRT.FLAGS Enables run-time output of regional model flag information.
PRT.IMPACT Enables run-time output of regional impact ionization model information.
PRT.RECOM Enables run-time output of regional recombination model information.

1259 ATLAS User’s Manual

MODELS Statements

PRT.THERM Enables run-time output of regional thermal model information.

PRINT Prints the status of all models, a variety of coefficients, and constants. We
recommended that you include this parameter in all ATLAS runs.
PRINT.REGION Prints the model status and coefficients for the specified region.
PRINT.MATERIAL Prints the model status and coefficients for the specified material.
REGION Specifies the region number for this MODELS statement. The alias is NUMBER.
RAJKANAN Enables the analytic absorption model for silicon given by Equation 3-624.
SDISS.TYPE Specifies the geminate pair distribution type as follows: 1 - delta function,
2 - Gaussian, 3 - exponential, and 4 - r-squared exponential. See Section 15.3.3
“Exciton Dissociation”.
SPEC.NAME Specifies the name of a file to collect the emission spectrum from the LASER
simulator.
SPECIES1.A Composition number of ionic species 1 (see Section 3.16 “Generic Ion Transport
Model”).
SPECIES2.A Composition number of ionic species 2 (see Section 3.16 “Generic Ion Transport
Model”).
SPECIES3.A Composition number of ionic species 3 (see Section 3.16 “Generic Ion Transport
Model”).
SPECIES1.Z Charge multiple of ionic species 1 (see Section 3.17 “Generic Ion Reaction
Model”).
SPECIES2.Z Charge multiple of ionic species 2 (see Section 3.17 “Generic Ion Reaction
Model”).
SPECIES3.Z Charge multiple of ionic species 3 (see Section 3.17 “Generic Ion Reaction
Model”).
SPONTANEOUS Includes the radiative recombination rates derived from the LI, YAN or CHUANG
model into the drift diffusion.

Note: You need additional computational resources since the radiative rate must be numerically integrated for
each grid point.

STRUCTURE This is a synonym for DEVICE.

STRESS Enables the silicon stress dependent bandgap model from Section 3.6.13
“Stress Effects on Bandgap in Si”.
SUBSTRATE This is a logical parameter indicating that the specified region should be
considered as the substrate for strain calcualtions described in Section 3.6.11
“Epitaxial Strain Tensor Calculation in Wurtzite”.
TAYAMAYA Enables the model from Section 3.9.5 “Takayama's Gain Model” (GAINMOD=3).

1260 ATLAS User’s Manual

MODELS Statements

TEMPERATURE Specifies the temperature in Kelvin.

TLU.INDEX Specifies that the Tauc-Lorentz dielectric function (see Section 3.10.3 “Tauc-
Lorentz Dielectric Function with Optional Urbach Tail Model for Complex
Index of Refraction”) should be used for calculation of the complex index of
refraction.
UNSAT.FERRO Enables unsaturated loop modeling for ferroelectrics.
WZ.KP Enables the WZ.KP gain/radiative recombination model (GAINMOD=5).
YAN This is an alias for the ZB.ONE parameter.
ZB.ONE Enables the ZB.ONE gain/radiative recombination model (GAINMOD=5).
ZB.TWO Enables the ZB.TWO gain/radiative recombination model (GAINMOD=5).

Model Dependent Parameters

ARORA Model Parameters
The parameters that can only be used with the ARORA model (see Equations 3-204 and 3-205)
are AR.MU1N, AR.MU2N, AR.MU1P, and AR.MU2P.
BBT.KL and BBT.STD Model Parameters
The parameters used with the BBTKL model are BB.A1 and BB.B. The BB.A2 and BB.B
parameters are used with the BBT.STD model. The BB.A1 and BB.A2 parameters are the pre-
exponential coefficients in the band-to-band tunneling models. The BB.B parameter is the
exponential coefficient used in both models. The default value of BB.B depends on which
model is chosen (see Equation 3-455).
CCSMOB Model Parameters
The CCS.EA, CCS.EB, CCS.HA, and CCS.HB parameters describe the dependence of mobility
on doping, carrier concentration, and temperature (see Equations 3-208 – 3-212).
FLDMOB Model Parameters

B.ELECTRONS This is used in the field-dependent mobility (see Equation 3-306). See also BETAN on
the MOBILITY statement.
B.HOLES This is used in the field-dependent mobility expression (see Equation 3-307). See
also BETAP on the MOBILITY statement.
E0 This is used in the field dependent mobility model for EVSATMOD=1 (see Equation 5-
57).

1261 ATLAS User’s Manual

MODELS Statements

EVSATMOD Specifies which parallel field dependent mobility model (see Equation 3-306 and
Equation 5-59) should be used for electrons. The value of EVSATMOD should be
assigned as follows:
• 0 - Use the standard saturation model (Equation 3-306). You also
can apply the option parameter, MOBTEM.SIMPL (see the “Carrier
Temperature Dependent Mobility” on page 198 for more
information).
• 1 - Use the negative differential velocity saturation model (Equation
5-59).
• 2 - Use a simple velocity limiting model.
In most cases, the default value of 0 should be used.
HVSATMOD Specifies which parallel field dependent mobility model (see Equation 3-307 and
Equation 5-60) should be used for holes. The value of HVSATMOD should be assigned
as follows:
• 0 - Use the standard saturation model (Equation 3-307).
• 1 - Use the negative differential velocity saturation model (Equation
5-60).
• 2 - Use a simple velocity limiting model.
In most cases, the default value of 0 should be used.

Fowler-Nordheim Tunneling Model Parameters

The parameters used in this model are F.AE and F.BE. F.AE is the pre-exponential factor
and F.BE is the exponential coefficient (see Equation 3-476).
WATT or SURFMOB Model Parameters
The parameters that can be used with the WATT model, include: ALN1, ALN2, ALN3, ALP1,
ALP2, ALP3, ETAN, ETAP, MREFN1, MREFN2, MREFN3, MREFP1, MREFP2, and MREFP3
(see Equations 3-300 and 3-301).
TFLDMB1 and TFLDMB2 Model Parameters

ACC.SF Specifies the accumulation saturation factor which describes the ratio of the
majority carrier concentration in the accumulation layer before and after bending of
conductivity and valence bands.
INV.SF Specifies the inversion saturation factor which describes the ratio of the
majority carrier concentration in the inversion layer before and after the bending of
conductivity and valance bands.
OX.BOTTOM Specifies the coordinate of the bottom edge of the gate oxide for a MOSFET
transistor.
OX.LEFT Specifies the coordinate of the left edge of the gate oxide for a MOSFET transistor.
OX.RIGHT Specifies the coordinate of the right edge of the gate oxide for a MOSFET
transistor.

1262 ATLAS User’s Manual

MODELS Statements

CONCANNON Model Parameters

CGATE.N Specifies an empirical tuning factor for electrons in the Concannon gate current model.
CGATE.P Specifies an empirical tuning factor for holes in the Concannon gate current model.
PEFF.N Specifies an effective barrier height for electrons in the Concannon gate current model.
An alias for IG.EB0.
PEFF.P Specifies an effective barrier height for holes in the Concannon gate current model. An
alias for IG.HB0.
THETA.N Specifies the critical rejection angle for electrons in the Concannon gate current model.
THETA.P Specifies the critical rejection angle for electrons in the Concannon gate current model.
C0 Specifies the electron distribution weight factor in the Concannon gate current model.
CHIA Specifies the electron distribution function constant in the Concannon gate current
model.
CHIB Specifies the electron distribution function constant in the Concannon gate current
model.
CHI.HOLES Specifies the hole distribution function constant in the Concannon gate current model.
ENERGY.STEP Specifies the energy step for numeric integration in the Concannon gate current model.
INFINITY Specifies the highest energy in numeric integration in the Concannon gate current
model.
PATH.N Specifies the mean free path in the oxide for electrons in the Concannon gate
current model.
PATH.P Specifies the mean free path in the oxide for holes in the Concannon gate current
model.

HEI Model Parameters

The parameters that may be used with the HEI model include: IG.ELINR, IG.HLINR,
IG.ELINF, IG.HLINF, IG.EBETA, IG.HBETA, IG.EETA, IG.HETA, IG.EEOX, IG.HEOX,
IG.EB0, IG.HB0.
Table 21-3 lists aliases for parameters of the HEI model.

Table 21-3 HEI model aliases

Parameter Alias Parameter Alias

IG.ELINF LAMHN IG.EBETA BARLN

IG.ELINR LAMRN IG.HBETA BARLP

IG.HLINF LAMHP IG.EETA TUNLN

IG.HLINR LAMRP IG.HETA TUNLP

1263 ATLAS User’s Manual

MODELS Statements

N.DORT and P.DORT Model Parameters

B.DORT User-specifiable model parameter for the Van Dort quantum effects approximation
model.
D.DORT User-specifiable model parameter for the Van Dort quantum effects approximation
model.

BBT.KANE Model Parameters

The parameters used with the BBT.KANE model are BBT.A_KANE (aliases BB.A_KANE or
A.BTBT), BBT.B_KANE (aliases BB.B_KANE or B.BTBT), and BBT.GAMMA (alias BB.GAMMA).
LASER Simulation Parameters

CAVITY.LENGTH Specifies the cavity length in the longitudinal direction (in mm).
GAINMOD Specifies the local optical gain model to be used. GAINMOD=0 specifies
that no optical gain model will be used. GAINMOD=1 specifies that the
simple gain model will be used (see Section 3.9.3 “The Standard Gain
Model”). GAINMOD=2 specifies that the complex frequency-dependent
gain model will be used (see Section 3.9.4 “The Empirical Gain Model”).
GAINMOD=3 specifies that Takayama’s gain model will be used (see
Section 3.9.5 “Takayama's Gain Model”). GAINMOD=5 specifies that the
Yam’s, Li’s, or Chuang’s gain model will be used (see Sections 3.9.7
“Unstrained Zincblende Models for Gain and Radiative Recombination”
and 3.9.10 “Strained Wurtzite Three-Band Model for Gain and Radiative
Recombination”).
LAS.ABSOR_SAT Enables the non-linear absorption loss saturation model.
LAS.ABSORPTION Enables the absorption loss model in LASER.
LAS.EINIT Specify the lower and upper photon energies. LASER will calculate
LAS.EFINAL multiple longitudinal photon rates within this range. Using wide ranges
can slow down simulation.
LAS.ESEP Specifies the photon energy separation. If this is not specified, LASER will
automatically calculate the number of longitudinal modes based on the
cavity length and the energy range.
LAS.ETRANS Specifies that the Helmholtz solver should use the transverse mode with
the eigen-energy closes to the energy specified by the
LAS.TRANS_ENERGY parameter.

LAS.FCARRIER Enables the free carrier loss model in LASER.

LAS.GAIN_SAT Enables the non-linear gain saturation model.
LAS.ITMAX Specifies the maximum number of iterations allowed for LASER
simulation at each bias point.

1264 ATLAS User’s Manual

MODELS Statements

LAS.LOSSES Specifies the total losses in Equation 8-28.

LAS.MIRROR Specifies the percentage facet reflectivity (both facets are assumed to
have this value of reflectivity) for the mirror loss in Laser. 100%
reflectivity is equivalent to no mirror loss.
LAS.NMODE Specifies the number of transverse modes simulated.
LAS.NX Specifies the number of discrete samples in the laser mesh in the X
direction.
LAS.NY Specifies the number of discrete samples in the laser mesh in the Y
direction.
LAS.OMEGA Specifies the lasing frequency to be used in Equation 8-1. If model 2 is
used for simulation, then this parameter specifies an estimate of the lasing
frequency. Instead of using this parameter, PHOTON.ENERGY can be used
to specify photon energy.
LAS.REFLECT Specifies that the Helmholtz equation should be solved with the Y axis
(X=0) as an axis of symmetry.
LAS.RF Specifies the front mirror reflectivity in percent.
LAS.RR Specifies the rear mirror reflectivity in percent.
LAS.SIN Specifies an initial photon density in the fundamental lasing mode. This
value provides an initial guess for subsequent iterations. This parameter is
used only when the single frequency model has been selected.
LAS.SPECSAVE The spectrum file will be saved after every LAS.SPECSAVE LASER
solution steps.
LAS.SPONTANEOUS Enables the physically based model for spontaneous emission to be used.
LAS.TIMERATE Specifies that the time dependent photon rate equation will be used in a
transient laser simulation.
LAS.TOLER Specifies the desired accuracy in photon areas.
LAS.TRANS_ENERGY Specifies the transverse mode energy for mode selection with
LAS.ETRANS set.

LAS.XMAX Specify the maximum and minimum limits on the laser Helmholtz
LAS.XMIN solution mesh.
LAS.YMAX
LAS.YMIN

LAS_MAXCH Specifies the maximum allowable relative change in photon densities

between iterations. Rapid changes of the photon densities can cause
convergence problems.
LASER Enables the LASER simulation.

1265 ATLAS User’s Manual

MODELS Statements

LMODE Specifies that multiple longitudinal models are to be taken into account
during the LASER simulation.
PHOTON.ENERGY Specifies the energy of photons. If model 2 is used for simulation, this
parameter specifies only an initial estimate of the photon energy. Instead
of using this parameter, LAS.OMEGA can be used to specify the lasing
frequency.
SPEC.NAME Specifies the name of a spectrum file, which LASER will produce for each
bias point if the LMODES parameter has been specified.

Exciton Model Flags

EXCITONS Specifies that singlet and triplet exciton continuity equations will be solved.
F.GENSINGLET Specifies the file name of a C-Interpreter function that models the generation of
singlet excitons
F.GENTRIPLET Specifies the file name of a C-Interpreter function that models the generation of
triplet excitons.
SINGLET Specifies that the singlet exciton continuity equation will be solved.
TRIPLET Specifies that the triplet exciton continuity equation will be solved.

Model Selection Example

This example selects concentration and field dependent mobilities, SRH recombination, and
Auger recombination. This is a typical model set for bipolar simulation. The simulation
temperature is 290K.
MODELS CONMOB FLDMOB SRH AUGER TEMP=290
Confirming Model Selection
To echo back model selections, parameters and material constants use
MODELS PRINT

Note: For the best model selection for different applications, see the Standard Examples set (see also Section 2.4
“Accessing The Examples” for more information about to how to access these examples ).

1266 ATLAS User’s Manual

MQW Statements

21.38 MQW
MQW defines multiple quantum well structures in an existing device structure. This statement
should follow all other structure definition statements (i.e., MESH, REGION, ELECTRODE, and
DOPING).
Syntax
MQW <parameters>

Parameter Type Default Units

ACCEPTORS Real 0.0 cm-3

ALT.SCHRO Logical False

ASTR Real 0.0
BSTR Real 0.0
CSTR Real 0.0
DELTA Real 0.341 eV
DONORS Real 0.0 cm-3

DSTR Real 0.0

EPSILON Real 0.0
ESTR Real 0.0
FSTR Real 0.0
GAIN0 Real 1.0
GAMMA0 Real 2.010-3 eV
GSCALE Real 1.0
ISHIKAWA Logical False
LI Logical False
LORENTZ Logical False
MATERIAL Character
MC Real 0.067
MHH Real 0.62
MLH Real 0.087
MSTAR Real 0.053
NBS Integer 15

1267 ATLAS User’s Manual

MQW Statements

NWELL Integer 1
NX Integer 10
NY Integer 10
RESOLVE Logical True
SPONTANEOUS Logical False
STRAIN Real 0.0
STABLE Integer 0
SY Real m

TAU.IN Real 3.310-13 seconds

TE.MODES Logical True

WB Real m

WW Real m

XCOMP Real 0.0

XMAX Real m

XMIN Real m

YAN Logical False

YCOMP Real 0.0
YMAX Real m

YMIN Real m

Description
The MQW regions will be added to ATLAS when you specify the MQW statement. This
statement is necessary to correctly model the recombination and gain effects of these MQW
regions.
The multiple quantum well structure is restricted to place the quantum wells only parallel to
the X axis (perpendicular to the Y axis). Also, if more than one well is specified, then each
well will have the same width defined by the WW parameter. Each barrier between wells will
then have the same width defined by the WB parameter.
The MQW structure is located within the rectangle defined by the XMIN, XMAX, YMIN, and YMAX
parameters. In the Y direction, the wells are centered within the span of YMIN to YMAX. Make
sure that the difference between YMIN and YMAX is larger than the sum of the well width (WW)
and the barrier width (WB) multiplied by the number of wells (NWELL). In the X direction, the
wells extend the entire length from XMIN to XMAX.
The quantum well material is specified by the MATERIAL parameter, modified by the
composition fractions: XCOMP and YCOMP. The barrier material is given by the material
specified in the region, where the wells are defined.

1268 ATLAS User’s Manual

MQW Statements

This implementation solves the Schrodinger's equation (see Equations 13-2 and 13-3) for
each quantum well, in order to calculate the quantum well-bound state energies and carrier
distributions in the wells. You can specify the number of bound states to use for this
calculation with the NBS parameter. For laser gain calculations, only the lowest bound state
energy in the conduction band and the highest bound state energy in the valence band are
used. These bound state energies feed directly into the gain model and spontaneous emission
models are shown in Section 3.9 “Optoelectronic Models”
MQW Parameters

ACCEPTORS Specifies a uniform density of ionized acceptors in the quantum wells.

ALT.SCHRO Specifies that the alternative Schrodinger solver should be used in the
calculations.
ASTR, BSTR, CSTR, DSTR, Specify coefficients to calculate the effective masses in the strained
ESTR, and FSTR quantum well.
DELTA Specifies the spin orbital splitting energy in the quantum well.
DONORS Specifies the uniform density of ionized donors in the quantum wells.
EPSILON Specifies the high frequency permittivity to used in calculating the gain
and radiative recombinations.
GAIN0 Scaling factor for the MQW gain models.
GAMMA0 Specifies the width in eV (Note: GAMMA0=h/TAU.IN).
GSCALE Specifies a scale factor to be multiplied by the optical gain.
ISHIKAWA Specifies that the Ishikawa model [123] is used to account for strain
effects (see Section 3.9.12 “Ishikawa's Strain Effects Model” for more
information).
LI Specifies that the Li model [42] be used for spontaneous emission in
multiple quantum wells. The DELTA, MSTAR, MC and MHH parameters
can be used with this model.
LORENTZ This is the Lorentzian line broadening model for MQW gain. The half
width of the Lorentzian shape function is controlled by the GAMMA0
parameter of the MQW statement. You can also set the TAU.IN
parameter to specify the intraband relaxation time in seconds. For more
information about the Lorentzian line broadening model, see Section
3.9.11 “Lorentzian Gain Broadening”.
MATERIAL Specifies the name of the material to be used in the wells. The barrier
material is taken from whatever material is already in the structure at the
MQW location.
MC This is the conduction band effective mass.
MHH This is the heavy hole effective mass.

1269 ATLAS User’s Manual

MQW Statements

MLH This is the light hole effective mass.

MSTAR Disperses energy in the conduction band. This doesn’t necessarily equate
to the conduction band effective mass, see [318].
NBS Defines the number of bound states used in solving the Schrodinger
equation.
NWELL Specifies the number of quantum wells.
NX and NY Specify the number of mesh points in the X and Y direction for
resolving the MQW structure. Make sure that these are specified large
enough to resolve the individual quantum wells. NY is defined as:

YMAX – YMIN
NY = -------------------------------------------------------- + 1 21-5
 YGridSpacing  m 

RESOLVE Specifies whether the band edge discontinuities at the well edges are
resolved by the device equations.
SPONTANEOUS Specifies whether the quantum well spontaneous recombination is
included in the device continuity equations.
STABLE Specifies the table row to be used for coefficients to calculate the
effective masses in strained quantum wells.
STRAIN Specifies the strain percentage in the quantum well.
SY Specifies the maximum mesh spacing to be applied in the Y direction
through the quantum wells.
TAU.IN Specifies the intraband relaxation time in seconds (GAMMA0=h/TAU.IN).
TE.MODES Specifies whether TE or TM modes will be used in calculating the
asymmetrical factors in the LI model.
WB Specifies the individual barrier width between wells in microns.
WW Specifies the individual quantum well width in microns.
XCOMP Defines the X composition fraction for the wells for ternary and
quaternary materials.
XMIN, XMAX, YMIN, YMAX Specify a rectangular box where the quantum wells are to be
simulated. The units of these parameters are in microns.
YCOMP Defines the Y composition fraction for the wells for quaternary materials.
YAN Specifies that the Yan model [318] be used for spontaneous emission in
multiple quantum wells. The DELTA, MSTAR, MC, and MHH
parameters can be used with this model (GAINMOD=5).

1270 ATLAS User’s Manual

NITRIDECHARGE Statements

21.39 NITRIDECHARGE
NITRIDECHARGE specifies the parameters describing the properties of traps in Silicon Nitride
in the DYNASONOS model and the PF.NITRIDE model.
The parameters apply to all Silicon Nitride regions.
Syntax
NITRIDECHARGE [<params>]

Parameter Type Default Units

NT.N Real 0.0 cm3

NT.P Real 0.0 cm3
TAU.N Real 1.0 10300 s

TAU.P Real 1.0 10300 s

ELEC.DEPTH Real -999.0 eV

HOLE.DEPTH Real -999.0 eV
SIGMAN.P Real 1.0 10-15 cm2
SIGMAP.N Real 1.0 10-14 cm2
SIGMAT.N Real 1.0 10-16 cm2
SIGMAT.P Real 1.0 10-14 cm2
PF.BARRIER Real eV
PF.A Real 1.0 105 s

PF.B Real 1.0e13 Hertz

Description
The NITRIDE charge allows you to set up the properties of electron and hole traps in Siliocn
Nitride. It is used in the DYNASONOS model and also the steady state version of the
PF.NITRIDE model.
Trap Density

NT.N Sets the value of the uniform density of electron traps (acceptor states) in the
Silicon Nitride.
NT.P Sets the value of the uniform density of hole traps (donor states) in the Silicon
Nitride.

1271 ATLAS User’s Manual

NITRIDECHARGE Statements

Cross-Section

SIGMAT.N A factor in determining the rate at which the acceptor states capture electrons from
the conduction band.
SIGMAT.P A factor in determining the rate at which the donor states capture holes from the
valence band.
SIGMAN.P A factor in determining the rate at which the acceptor states capture holes from the
valence band.
SIGMAP.N A factor in determining the rate at which the donor states capture electrons from
the conduction band.

Trap Depth

ELEC.DEPTH Energy below conduction band edge of acceptor traps.

HOLE.DEPTH Energy above valence band of donor traps.

Lifetimes

TAU.N Lifetime for electron detrapping from acceptor states to conduction band.
TAU.P Lifetime for hole detrapping from donor states to valence band.
PF.A Constant of proportionality for the ohmic part of the electron generation rate in
the PF.NITRIDE model.
PF.B Constant of proportionality for electron detrapping rate when Poole-Frenkel
model is enabled.
PF.BARRIER If specified, it overrides ELEC.DEPTH in calculation of electron detrapping rate
when Poole-Frenkel model is enabled.

1272 ATLAS User’s Manual

ODEFECTS Statements

21.40 ODEFECTS
ODEFECTS actives the bandgap defect model for organic polymer materials and sets the
parameter values. This model can be used when simulating organic polymer devices using the
ORGANIC DISPLAY AND ORGANIC SOLAR modules (see Chapter 15 “Organic Display and
Organic Solar: Organic Simulators”).
Syntax
ODEFECTS [<parameters>]

Parameter Type Default Units

AFILE Character
CONTINUOUS Logical False
DEVICE Character
DFILE Character
DOPANT Logical False
EA Real 0 eV
ED Real 0 eV
F.OTFTACC Character
F.OTFTDON Character
HA Real 0 cm-3
HD Real 0 cm-3
HOPPING Logical False
INT_LIM1 Real 0.0 eV
INT_LIM2 Real 0.0 eV
MATERIAL Character
NA Real 0 cm-3
ND Real 0 cm-3
NIA Real 0 cm-3
NID Real 0 cm-3
NUMA Real 12
NUMD Real 12
REGION Real All

1273 ATLAS User’s Manual

ODEFECTS Statements

Parameter Type Default Units

SIGAE Real 1.010-16 cm2

SIGAH Real 1.010-14 cm2
SIGDE Real 1.010-14 cm2
SIGDH Real 1.010-16 cm2
SIGMA Real 0 eV
SIGMD Real 0 eV
SIGMAIA Real 0 eV
SIGMAID Real 0 eV
STRUCTURE Character
TCA Real 0.0 K
TCD Real 0.0 K
TFILE Character
X.MIN Real m

X.MAX Real m

Y.MIN Real m

Y.MAX Real m

Z.MIN Real m

Z.MAX Real m

Description

AFILE Specifies the acceptor state density output file.

CONTINUOUS Specifies that the continuous defect model will be used.
DEVICE Specifies which device in MIXEDMODE simulation will apply to the
ODEFECTS statement. The synonym for this parameter is STRUCTURE.

DFILE Specifies the donor state density output file.

DOPANT Specifies that the defects defined using NA and ND are to be used in the dopant
singlet exciton calculation. Only one dopant is allowed per region.
EA Specifies the energy shift between the intrinsic and dopant states for
acceptor-like traps.

1274 ATLAS User’s Manual

ODEFECTS Statements

ED Specifies the energy shift between the intrinsic and dopant states for donor-
like traps.
F.OTFTACC Specifies the C-Interpreter file for the acceptor states energy distrubution
function.
F.OTFTDON Specifies the C-Interpreter file for the donor states energy distrubution function.
HA Specifies the total acceptor-like trap density.
HD Specifies the total donor-like trap density.
HOPPING Specifies the effective transport hopping energy model is to be used.
INT_LIM1 Specifies the lower limit for the numerical integration of the CONTINUOUS model.
INT_LIM2 Specifies the upper limit for the numerical integration of the CONTINUOUS model.
MATERIAL Specifies which material will apply to the ODEFECTS statement. If a
material is specified, then all regions defined as being composed of that material
will be affected.
NA Specifies the total dopant density for acceptor-like traps.
ND Specifies the total dopant density for donor-like traps.
NIA Specifies the total intrinsic dopant density for acceptor-like traps.
NID Specifies the total intrinsic dopant density for donor-like traps.
NUMA Specifies the number of discrete levels that will be used to simulate the
continuous distribution of acceptor states.
NUMD Specifies the number of discrete levels that will be used to simulate the
continuous distribution of donor states.
REGION Specifies the region to which the ODEFECTS statement applies.
SIGAE Specifies the electron capture cross-section for acceptor traps.
SIGAH Specifies the hole capture cross-section for acceptor traps.
SIGAE Specifies the electron capture cross-section for donor traps.
SIGAH Specifies the hole capture cross-section for donor traps.
SIGMAA Specifies the Gaussian width for the acceptor-like traps dopant distribution.
SIGMAD Specifies the Gaussian width for the donor-like traps dopant distribution.
SIGMAIA Specifies the Gaussian width for the acceptor-like traps intrinsic
distribution.
SIGMAID Specifies the Gaussian width for the donor-like traps intrinsic distribution.
STRUCTURE This is a synonym for DEVICE.
TCA Specifies the characteristic temperature for the acceptor-like trap states.

1275 ATLAS User’s Manual

ODEFECTS Statements

TCD Specifies the characteristic temperature for the acceptor-like trap states.
TFILE Specifies the file name where the acceptor and donor state density
distributions will be stored.
X.MIN Specify the bounding box for the ODEFECTS statement.
X.MAX
Y.MIN
Y.MAX
Z.MIN
Z.MAX

1276 ATLAS User’s Manual

OINTDEFECTS Statements

21.41 OINTDEFECTS
OINTDEFECTS actives the bandgap interface defect model for organic polymer materials and
sets the parameter values. This model can be used when simulating organic polymer devices
using the OTFT module (see Chapter 15 “Organic Display and Organic Solar: Organic
Simulators”).
Syntax
OINTDEFECTS [<parameters>]

Parameter Type Default Units

AFILE Character
CONTINUOUS Logical False
DEVICE Character
DFILE Character
EA Real 0 eV
ED Real 0 eV
F.OTFTACC Character
F.OTFTDON Character
HA Real 0 cm-2
HD Real 0 cm-2
HOPPING Logical False
INT_LIM1 Real 0.0 eV
INT_LIM2 Real 0.0 eV
INTMATERIAL Character
INTNAME Character
INTNUMBER Character
INTREGION Character
MATERIAL Character
NA Real 0 cm-3
ND Real 0 cm-3
NIA Real 0 cm-3
NID Real 0 cm-3

1277 ATLAS User’s Manual

OINTDEFECTS Statements

Parameter Type Default Units

NUMA Real 12
NUMD Real 12
R1MATERIAL Character

R1NAME Character
R1NUMBER Integer
R1REGION Character
R2MATERIAL Character
R2NAME Character
R2NUMBER Integer
R2REGION Character
REGION Real All
SIGAE Real 1.010-16 cm2
SIGAH Real 1.010-14 cm2
SIGDE Real 1.010-14 cm2
SIGDH Real 1.010-16 cm2
SIGMA Real 0 eV
SIGMD Real 0 eV
SIGMAIA Real 0 eV
SIGMAID Real 0 eV
STRUCTURE Character
TCA Real 0.0 K
TCD Real 0.0 K
TFILE Character
X.MIN Real m

X.MAX Real m

Y.MIN Real m

Y.MAX Real m

Z.MIN Real m

Z.MAX Real m

1278 ATLAS User’s Manual

OINTDEFECTS Statements

Description

AFILE Specifies the acceptor state density output file.

CONTINUOUS Specifies that the continuous defect model will be used.
DEVICE Specifies which device in MIXEDMODE simulation will apply to the OINTDEFECTS
statement. The synonym for this parameter is STRUCTURE.
DFILE Specifies the donor state density output file.
EA Specifies the energy shift between the intrinsic and dopant states for acceptor-like
traps.
ED Specifies the energy shift between the intrinsic and dopant states for donor-like traps.
F.OTFTACC Specifies the C-Interpreter file for the acceptor states energy distrubution function.
F.OTFTDON Specifies the C-Interpreter file for the donor states energy distrubution function.
HA Specifies the total acceptor-like trap density.
HD Specifies the total donor-like trap density.
HOPPING Specifies the effective transport hopping energy model is to be used.
INT_LIM1 Specifies the lower limit for the numerical integration of the CONTINUOUS model.
INT_LIM2 Specifies the upper limit for the numerical integration of the CONTINUOUS model.
INTMATERIAL Specifies that the interface models defined on this command should be applied to
every interface between the two given materials.
For example, INTMATERIAL="oxide/silicon".
INTNAME Specifies that the interface models defined on this command should be applied to
every interface between the two given regions (where the regions are defined by
their region name). For example, INTNAME="channel/substrate".
INTNUMBER Specifies that the interface models defined on this command should be applied to
every interface between the two given regions (where the regions are defined by their
region number). For example, INTNUMBER="2/3".
INTREGION Specifies that the interface models defined on this command should be applied to
every interface between the two given regions (where the regions are defined by
either their region name or their region number).
For example, INTREGION="channel/substrate".
MATERIAL Specifies which material will apply to the OINTDEFECTS statement. If a material is
specified, then all regions defined as being composed of that material will be
affected.
NA Specifies the total dopant density for acceptor-like traps.
ND Specifies the total dopant density for donor-like traps.

1279 ATLAS User’s Manual

OINTDEFECTS Statements

1289 ATLAS User’s Manual

OUTPUT Statements

HEI Specifies that the hot electron current density will be included in the
standard structure file.
HHI Specifies that the hot hole current density will be included in the standard
structure file.
HX.VELOCITY Specifies that the x-component of hole velocity will be included in the
standard structure file.
HY.VELOCITY Specifies that the y-component of hole velocity will be included in the
standard structure file.
HZ.VELOCITY Specifies that the z-component of the hole velocity will be included in the
standard structure file.
IMPACT Specifies that the impact ionization rate will be included in the standard
structure file.
J.CONDUC Specifies that the total conduction current density will be included in the
standard structure file.
J.DISP Specifies that the total displacement current density will be included in the
standard structure file.
J.ELECTRON Specifies that the total electron current density will be included in the
standard structure file.
J.HOLE Specifies that the total hole current density will be included in the standard
structure file.
J.TOTAL Specifies that the total current density will be included in the standard structure
file.
JX.CONDUC Specifies that the x-component of the total conduction current density will be
included in the standard structure file.
J.DRIFT Specifies that the drift current density will be included in the standard structure
file.
J.DIFFUSION Specifies that diffusion current density will be included in the standard structure
file.
JX.ELECTRON Specifies that the x-component of electron current density will be included in
the standard structure file.
JX.HOLE Specifies that the x-component of hole current density will be included in the
standard structure file.
JX.TOTAL Specifies that the x-component of total current density will be included in the
standard structure file.
JY.CONDUC Specifies that the y-component of the total conduction current density will be
included in the standard structure file.
JY.ELECTRON Specifies that the y-component of electron current density will be included in
the standard structure file.

1290 ATLAS User’s Manual

OUTPUT Statements

JY.HOLE Specifies that the y-component of hole current density will be included in the
standard structure file.
JY.TOTAL Specifies that the y-component of total current density will be included in the
standard structure file.
JZ.CONDUC Specifies that the z-component of the conduction current density will be
included in the standard structure file.
JZ.ELECTRON Specifies that the z-component of the electron current density will be included
in the standard structure file.
JZ.HOLE Specifies that the z-component of the hole current density will be included in
the standard structure file.
JZ.TOTAL Specifies that the z-component of the total current density will be included in
the standard structure file.
KSN Specifies that electron Scattering Law Exponent is to be written to any saved
structure file.
KSP Specifies that hole Scattering Law Exponent is to be written to any saved
structure file.
L.TEMPER Specifies that lattice temperature will be included in the standard structure file.
MINSET This is the minimum set of data (potential, carrier concentration, and
electric field) that will be included in the standard structure file.
NLTAT.GAMMA Specifies that the Nonlocal trap-assisted-tunneling Gamma factors are written
to any saved structure file. The values written have been interpolated
onto the device mesh.
OPT.INTENS Specifies that optical intensity is included in the standard structure file.
OX.CHARGE Specifies that fixed oxide charge is include in the standard structure file.
P.QUANTUM Specifies that the Bohm quantum potential is included in the solution file.
PERMITTIVITY Specifies the dielectric permittivity is saved.
PHOTOGEN Specifies that the photogeneration rate will be included in the standard structure
file.
POLAR.CHARGE Specifies that polarization charge will be included in the structure file.
QFN Specifies that the electron quasi-fermi level will be included in the standard
structure file.
QFP Specifies that the hole quasi-fermi level will be included in the standard
structure file.
QSS Specifies that the surface charge will be included in the standard structure file.
QTUNN.BBT Specifies that the direct quantum tunneling band-to-band current density at each
interface node is written to any saved structure file.

1291 ATLAS User’s Manual

OUTPUT Statements

QTUNN.EL Specifies that the direct quantum tunneling electron current density at each
interface node is written to any saved structure file.
QTUNN.HO Specifies that the direct quantum tunneling hole current density at each
interface node is written to any saved structure file.
RECOMB Specifies that the recombination rate will be included in the standard structure
file.
SCHOTTKY Specifies that recombination velocities and barrier lowering will be included in
the standard structure file.
SONOS.RATES This causes the output of four quantities to the structure file. These are the
Generation and Recombination rates for the trapped electron states and the
trapped hole states in the DYNSAONOS model. This requires the
DYNASONOS or BESONOS models.
T.QUANTUM Specifies that quantum temperatture from the density gradient model is
included in the solution file.
TAURN Specifies that electron relaxation times are to be written to any saved structure
file.
TAURP Specifies that hole relaxation times are to be written to any saved structure file.
TOT.DOPING Specifies that total doping will be included in the standard structure file.
TRAPS Specifies that trap density information will be included in the standard structure
file.
TRAPS.FT Specifies that the trap probability of occupation will be included in the standard
structure file.
U.AUGER Specifies that the Auger component of recombination is to be written to
solution files.
U.BBT Specifies that the band to band tunneling rate will be included in the standard
structure file.
U.LANGEVIN Specifies that the Langevin recombination rate will be included in the standard
structure file.
U.RADIATIVE Specifies that the radiative component of recombination is to be written to
solution files.
U.SRH Specifies that the SRH component of recombination is to be written to
solution files.
U.TRANTRAP Specifies that both electron recombination rate and the hole recombination rate
into transient traps will be included in the SILVACO standard structure file.
U.TRAP Outputs the reaction rates of the quantities specified in the REACTION
statements. The quantities are output in units of cm-3 s-1 and a positive value
means that the quantity is consumed by the reaction. For electrons and holes,
this is the same as a recombination rate.

1292 ATLAS User’s Manual

OUTPUT Statements

VAL.BAND Specifies that the valence band edge will be included in the standard
structure file.
VECTORS Specifies that only vector components will be included in the standard structure
file.
X.COMP Specifies that the composition fraction, x , is to be written to solution files.
Y.COMP Specifies that the composition fraction, y, is to be written to solution files.

Ionization Integral Parameters

INAME Specifies the name of a contact for which electric field lines are calculated.
CONTACT Specifies a contact number for which electric field lines are calculated.
LRATIO Specifies the spacing ratio between adjacent electric field lines. Defaults to 1.0 for
uniform spacing.
N.LINES Specifies the number of electric field lines.
DELTAV Since the electric field is near zero at the contact, the electric field line
calculations begin at a distance from the contact at which the contact voltage has
changed by DELTAV. Defaults to 0.1 V.

Note: See Section 21.57 “SOLVE” and the on-line examples for instructions on using ionization integrals.

NOISE Parameters

NOISE Selects the total local noise source for output.

NOISE.IMP Selects the impedance fields for output.
NOISE.ALL Selects everything for output. Currently, the local noise source, the impedance
fields, the short-circuit current Green's function, the individual microscopic noise
sources, and the individual local noise sources.

1293 ATLAS User’s Manual

OUTPUT Statements

Averaging Parameters for Vector Quantities

OLD.AVG Specifies that the current and field quantities will be averaged using an older
algorithm (from version 3.0.0.R and back). By default the new method is used.
ANGLE Specifies that averaging of current and fields will be weighted by the size of the
angle of triangles intersecting at the node.
INV.ANGLE Specifies that averaging of current and fields will be weighted by the inverse of the
size of the angle of triangles intersecting at the node.
AREA Specifies that averaging of current and fields will be weighted by the areas of
triangles intersecting at the node.
INV.AREA Specifies that averaging of current and fields will be weighted by the inverse of areas
of triangles intersecting at the node.

Note: Certain quantities that can be output into the structure file and subsequently displayed using TONYPLOT
need special mention. These quantities are evaluated within ATLAS along the links between grid points.
They are represented in the structure file at the grid points themselves. As such these quantities are subject
to averaging. In particular, electric field and currents are averaged so as to take into account the vector
nature of these values. Mobility is simply summed up over all the links surrounding the grid point and divided
by the total number of links. Carrier velocities are derived by dividing the averaged current by the carrier
density at the grid point and the fundamental electron charge, q.

An Example of Combining OUTPUT with SOLVE and SAVE

The OUTPUT statement is often used in conjunction with the SAVE statement. The following
statement lines specify that current flowlines and electron velocity components are saved in
all subsequent standard structure solution files.
OUTPUT FLOWLINES EX.VELO EY.VELO
SOLVE PREVIOUS V5=2 OUTF=data1.str MASTER
SAVE OUTF=data2.str

1294 ATLAS User’s Manual

PHOTOGENERATE Statements

21.44 PHOTOGENERATE
PHOTOGENERATE statement specifies photogenetion of carriers inside the device. It ensures
TMA compatibility with the PHOTOGEN statement.
Syntax
PHOTOGENERATE <parameters>

Parameter Type Default Units

A1 Real 0.0 cm-3

A2 Real 0.0 cm-3/um
A3 Real 0.0 cm-3
A4 Real 0.0 um-1
CONSTANT Real 0.0 cm-3
EXPONENT Real 0.0 um-1
FACTOR Real 0.0 cm-3
LINEAR Real 0.0 cm-3/um
RADIAL Real  um
R.CHAR Real  um
X.END Real Min x of device um
X.MAX Real Max x of device um
X.MIN Real Min x of device um
X.ORIGIN Real Min x of device um
X.START Real Min x of device um
Y.END Real Max y of device um
Y.MAX Real Max y of device um
Y.MIN Real Min y of device um
Y.ORIGIN Real Min y of device um
Y.START Real Min y of device um

1295 ATLAS User’s Manual

PHOTOGENERATE Statements

Description

A1 This is the synonym for CONSTANT. Added for TMA compatibility.

A2 This is the synonym for LINEAR. Added for TMA compatibility.
A3 This is the synonym for FACTOR. Added for TMA compatibility.
A4 This is the synonym for EXPONENT. Added for TMA compatibility.
CONSTANT Specifies the constant photogeneration rate.
EXPONENT Specifies the exponential coefficient for depth dependent photogeneration.
FACTOR Specifies the pre-exponential factor for depth dependent photogeneration.
LINEAR Specifies the linear factor for depth dependent photogeneration.
RADIAL Specifies the characteristic radial distance for Gaussian dependence of
photogeneration rate in the direction perpendicular to the line along which the
carriers are injected.
R.CHAR This is the synonym for RADIAL. Added for TMA compatibility.
X.END Specifies the X coordinate of the end of the line along which photogenerated
carriers are injected.
X.MAX This is the maximum X coordinate at which photogeneration occurs.
Generation of carriers is zero for x>X.MAX.
X.MIN This is the minimum X coordinate at which photogeneration occurs.
Generation of carriers is zero for x<X.MIN.
X.ORIGIN Specifies the X coordinate of the origin of the line along which
photogenerated carriers are injected.
X.START This is the synonym of X.ORIGIN. Added for TMA compatibility.
Y.END Specifies the Y coordinate of the end of the line along which photogenerated
carriers are injected.
Y.MAX This is the maximum Y coordinate at which photogeneration occurs.
Generation of carriers is zero for y>Y.MAX.
Y.MIN This is the minimum Y coordinate at which photogeneration occurs.
Generation of carriers is zero for y<Y.MIN.
Y.ORIGIN Specifies the Y coordinate of the origin of the line along which
photogenerated carriers are injected.
Y.START This is the synonym of Y.ORIGIN. Added for TMA compatibility.

1296 ATLAS User’s Manual

PML Statements

21.45 PML
The PML statement allows definition of perfectly matched layers for finite-difference time-
domain (FDTD) analysis.
Syntax
PML [parameters]

Parameter Type Default Units

ALL Logical False

ALPHA Real 0.0 1/cm
BACK Logical False
BEAM Integer All
BOTTOM Logical True
DEGREE Integer 2
FRONT Logical False
IMAG.INDEX Real 0.0
LEFT Logical False
MATERIAL Character
NSAMP Integer
NXX Real 0.0
NYY Real 0.0
NZZ Real 0.0
PERMEABILITY Real 1.0
PERMITTIVITY Real 1.0
R90 Real 0.0
REAL.INDEX Real 1.0
TOP Logical False
RIGHT Logical False
WIDTH Real 0.0 m

XDIR Logical False

YDIR Logical False

1297 ATLAS User’s Manual

PML Statements

Description

ALL Specifies that the PML is applied to all sides (TOP, BOTTOM, LEFT, and RIGHT) of the
FDTD domain.
ALPHA Specifies the maximum absorption coefficient of the PML.
BACK Specifies that the PML is added at the XY plane at the minimum value of z.
BEAM Specifies the index for the beam that the PML will be applied. If BEAM is
unspecified, then PML will be applied to all previously defined beams.
BOTTOM Specifies that the PML is added at the XZ plane at the maximum value of y.
DEGREE Specifies the degree of Equation 10-68
FRONT Specifies that the PML is added at the XY plane at the maximum of z.
IMAG.INDEX Specifies the imaginary part of the complex index of refraction to use in the
direction transverse to the PML.
LEFT Specifies that the PML is added at the YZ plane at the minimum X coordinate.
MATERIAL Specifies the material name to use to determine the real part of the complex index of
refraction versus wavelength for the PML.
NSAMP Specifies the number of samples in the PML in the direction of the coordinate axis
specified by X.DIR or Y.DIR.
NXX Specify the axial indices for an anisotropic PML. These parameters are only used if
NYY ANISO is specified on the BEAM statement.
NZZ

PERMEABILITY Specifies the relative permeability to use within the PML if NSAMP is unspecified it
is calculated from the WIDTH.
PERMITTIVITY Specifies the relative permittivity to use within the PML.
R90 Specifies the desired value of the normal reflection coefficient. This is R() at
=90° in Equation 10-69.

REAL.INDEX Specifies the real part of the complex index of refraction of the PML.
RIGHT Specifies that the PML is added at the YZ plane at the maximum X coordinate.
TOP Specifies that the PML is added at the XZ plane at the minimum value of y.
WIDTH Specifies the thickness of the PML in microns.
XDIR Specifies that the PML will be added at the maximum and minimum x
coordinates.
YDIR Specifies that the PML will be added at the maximum and minimum y
coordinates.

1298 ATLAS User’s Manual

PROBE Statements

J.CONDUCTION Logical False
J.DISP Logical False
J.ELECTRON Logical False
J.HOLE Logical False
J.PROTON Logical False
J.TOTAL Logical False
JOULE.HEAT Logical False
KX Logical False
KY Logical False
KZ Logical False
LAYER.ABSO Logical False
LAYER.REFL Logical False
LAYER.TRAN Logical False
LMAX Real m

LMIN Real m

LANGEVIN Logical False

LASER.INTENSITY Logical False
LASER.GAIN Logical False
LASER.MODE Real
LAT.TEMP Logical False
LEFT Real minimum x coord µm
LUMINOUS.INTENSITY Logical False
MATERIAL Character
MAX Logical False
MIN Logical False
MAX.XITION Logical False
MIN.XITION Logical False
MSC.STATE Integer 1

1301 ATLAS User’s Manual

PROBE Statements

Parameter Type Default Units

N.CONC Logical False

N.MOB Logical False
N.TEMP Logical False
NAME Character
NWELL Logical False
P.CONC Logical False
P.MOB Logical False
P.TEMP Logical False
PERMITTIVITY Logical False
PHOTOGEN Logical False
POLARIZATION Logical False
POTENTIAL Logical False
PT.HEAT Logical False
PWELL Logical False
QFN Logical False
QFP Logical False
R.TRAP Logical False
R.BBT Logical False
RADIATIVE Logical False
RAUGER Logical False
REACTION.ELEC Logical False
REACTION.HOLE Logical False
REACTION.SP1 Logical False
REACTION.SP2 Logical False
REACTION.SP3 Logical False
RECOMBIN Logical False
REGION Integer
RESISTIVITY Logical False
RIGHT Real maximum x coord µm

1302 ATLAS User’s Manual

COMPLIANCE Specifies the probed quantity compliance value. When the probed quantity
reaches the specifies value, the SOLVE statement (or .DC or .TRANS statement
in MIXEDMODE) will terminate and will execute the next statement.
CON.BAND Probes the conduction band.
CONCACC.CTRAP Specifies that the continuous acceptor trap concentration is to be probed.
CONCACC.TRAP Specifies that the acceptor trap concentration is to be probed. You can use the
ENERGY parameter to specify the trap energy level to be probed. If ENERGY is
not specified, then the total value of all acceptor energy levels will be probed.
CONCDON.CTRAP Specifies that the continuous donor trap concentration is to be probed.
CONCDON.TRAP Specifies that the donor trap concentration is to be probed. You can use the
ENERGY parameter to specify the trap energy level to be probed. If ENERGY is
not specified, then the total value of all donor energy levels will be probed.
CONDUCTIVITY Specifies that the conductivity of metals and semiconductors is probed. The
conductivity in insulators is returned as 0.
CURDVSE Stores current density versus energy for each solution of NEGF solver.
DENSVSE Stores carrier density versus energy for each solution of NEGF solver.
DEVICE Specifies which device in MIXEDMODE simulation the PROBE statement
should apply to. The synonym for this parameter is STRUCTURE.
DIR Specifies the direction relative to the X axis in degrees associated with
certain directed quantities. These quantities include FIELD, N.MOB, P.MOB,
and POLARIZATION.
DON.CTRAP Specifies that the continuous donor trap density of states is to be probed.
DON.TRAP Specifies that the ionized donor trap density is to be probed. If multiple trap
levels are present the ENERGY parameter should be set to the desired energy
level that is to be probed.
DOPANT.EXCITON Specifies that the singlet dopant exciton density is to be probed.
DOPRAD.EXCITON Specifies that the singlet dopant exciton radiative rate is to be probed.
DOSVSE Stores density of states (DOS) versus energy for each solution of NEGF
solver.
E.TRANTRAP Specifies that the electron recombination rate into transient traps is to be
probed.
EMAX and EMIN Specifies the range of energies to search for peak emission wavelength when
the WAVELENGTH parameter is set on the PROBE statement.
FILENAME Specifies the name of file where Transmission. DOS, carrier, and current
density versus energy are stored.

1306 ATLAS User’s Manual

PROBE Statements

Note: The algorithm used finds the triangle in the mesh containing the specified X and Y values. Then the value of
the DIR parameter is used to find which edge of the triangle lies in the direction nearest that value.

EXCITON Specifies that the exciton density is probed.

FIELD Specifies that a value of electric field is probed. The DIR parameter should
also be specified if FIELD is used.
FT.MSC Probes the occupation fraction of the internal state of the multistate traps
given by MSC.STATE.
FTACC.CTRAP Specifies that the continuous trap probability of occupation is to be
probed.
FTACC.TRAP Specifies that the acceptor trap probability of occupation is to be probed.
You can use the ENERGY parameter to specify the trap energy level to be
probed. If ENERGY is not specified, then the total value of all acceptor
energy levels will be probed.
FTDON.CTRAP Specifies that the continuous donor trap probability of occupation is to be
probed.
FTDON.TRAP Specifies that the donor trap probability of occupation is to be probed. You
can use the ENERGY parameter to specify the trap energy level to be
probed. If ENERGY is not specified, then the total value of all donor energy
levels will be probed.
GENERATION Specifies that the generation rate due to impact ionization is probed.
GR.HEAT Requests a PROBE of Generation-Recombination heat in GIGA.
H.CONC Specifies that proton concentration is probed.
H.ATOM Specifies that the atomic hydrogen density is to be probed.
H.MOLE Specifies that the molecular hydrogen density is to be probed.
H.TRANTRAP Specifies that the hole recombination rate into transient traps is to be
probed.
INTEGRATED Specifies that the probed value is to be integrated over all mesh points
inside a box defined with the parameters X.MIN, X.MAX, Y.MIN, Y.MAX,
Z.MIN and Z.MAX.

INTENSITY Evaluates optical intensity of an optical beam in LUMINOUS. If this is a

multispectral source, this value accounts for the subsampling, while the
intensity already in the logfile will only reflect the integrated input
spectrum.
J.CONDUCTION Requests a PROBE of Conduction current (J.ELEC+J.HOLE).
J.DISP Requests a PROBE of Displacement current.
J.ELECTRON Requests a PROBE of Electron Current density.

1307 ATLAS User’s Manual

PROBE Statements

J.HOLE Requests a PROBE of Hole Current density.

J.PROTON Specifies that the proton current density is probed.
J.TOTAL Requests a PROBE of Total Current density (J.TOTAL+J.DISP).
JOULE.HEAT Requests a PROBE of Joule-Heating in GIGA.
KX, KY, and KZ Specify directions in k space associated with probing conduction and
valence band bound state energies. See also NWELL and PWELL.
LANGEVIN Specifies that the Langevin recombination rate is probed.
LAYER.ABSO LAYER.REFL Specify that the absorption, reflection, or transmission coefficient is
LAYER.TRAN probed for LUMINOUS ray trace or transfer matrix. When specifying one of
these values, you also need to specify the BEAM and REGION parameters.
LASER.GAIN Specifies that the probe will operate on the LASER optical gain. For
more information about LASER, see Chapter 8 “Laser: Edge Emitting
Simulator”.
LASER.INTENSITY Specifies that the probe will operate on the LASER optical intensity. For
more about information LASER, see Chapter 8 “Laser: Edge Emitting
Simulator”.
LASER.MODE Specifies the optical mode for LASER.INTENSITY and LASER.GAIN.
LAT.TEMP Specifies that the probe will operate on lattice temperature.
LMAX, LMIN Specifies the range of wavelengths to search for peak emission
wavelength when the WAVELENGTH parameter is set on the PROBE
statement.
LUMINOUS.INTENSITY Specifies that the probe will operate on the LUMINOUS optical intensity.
For more information about LUMINOUS, see Chapter 10 “Luminous:
Optoelectronic Simulator”.
MATERIAL This is assigned to a specific material name that will act to direct the probe
to only address regions composed of the specified material.
MAX Specifies that the probe will find the maximum value on the mesh.
MIN Specifies that the probe will find the minimum value on the mesh.
MAX.XITION Specifies that the maximum allowable transition between bound state
energies is to be probed. If there are no allowable transitions between
bound state energies, the band gap is probed.
MIN.XITION Specifies that the minimum allowable transition between bound state
energies is to be probed. If there are no allowable transitions between
bound state energies, the band gap is probed.
MSC.STATE Specifies which of the 5-internal states of the multistate traps are probed.
NAME Sets a character string that allows you to specify the description
displayed by TONYPLOT.

1308 ATLAS User’s Manual

PROBE Statements

N.CONC Specifies that the probe will operate on electron concentration.

N.MOB Specifies that the probe will operate on the electron mobility. The DIR
parameter should also be specified if N.MOB is used.
N.TEMP Specifies that the probe will operate on electron temperature.
NWELL Specifies that the a specified conduction bound state energy is to be
probed. To specify the bound state, you must also specify which bound
state is to be probed with the STATE parameter (where 0 specifies the
lowest bound state) and the direction in k space using the KX, KY or KZ
parameters.
P.CONC Specifies that the probe will operate on hole concentration.
P.MOB Specifies that the probe will operate on the hole mobility. The DIR
parameters should also be specified if P.MOB is used.
P.TEMP Specifies that the probe will operate on hole temperature.
PT.HEAT Requests a PROBE of Peltier-Thomson Heat in GIGA.
PERMITTIVITY Specifies that material permittivity is probed.
PHOTOGEN Specifies that photogeneration or SEU generation rate is probed.
POLARIZATION Specifies that the probe will operate on ferroelectric polarization.The DIR
parameter should also be specified if POLARIZATION is used.
POTENTIAL Specifies that the probe will operate on electrostatic potential.
PWELL Specifies that the a specified valence bound state energy is to be probed.
To specify the bound state, you must also specify which bound state is to
be probed with the STATE parameter (where 0 specifies the highest bound
state) and the direction in k space using the KX, KY or KZ parameters.
QFN Specifies that the probe will operate on the electron quasi-Fermi level.
QFP Specifies that the probe will operate on the hole quasi-Fermi level.
R.TRAP Specifies that the trap recombination is to be probed.
R.BBT Specifies tha the band-to-band tunneling rate is to be probed.
RADIATIVE Specifies that the probe will operate on radiative recombination rate.
REACTION.ELEC Specifies that the PROBE will give the consumption rate of electrons in a
REACTION.

REACTION.HOLE Specifies that the PROBE will give the consumption rate of holes in a
REACTION.

REACTION.SP1 Specifies that the PROBE will give the consumption rate of ionic species 1
in a REACTION.
REACTION.SP2 Specifies that the PROBE will give the consumption rate of ionic species
2 in a REACTION.

1309 ATLAS User’s Manual

PROBE Statements

REACTION.SP3 Specifies that the PROBE will give the consumption rate of ionic species 3
in a REACTION.
RECOMBIN Specifies that the probe will operate on net recombination rate.
REGION Limits the application of the probe to a specific region of a specified
index.
RESISTIVITY Specifies that the resistivity of metals and semiconductors is probed. The
resistivity in insulators is returned as 0.
SONOS.CHARGE Specifies that the net trapped carrier density in a Silicon-Nitride region is
to be probed. The net density is defined as the electron density minus
the hole density.
SPECIES1 Specifies that the PROBE will operate on ionic species 1.
SPECIES2 Specifies that the PROBE will operate on ionic species 2.
SPECIES3 Specifies that the PROBE will operate on ionic species 3.
SRH Specifies that the probe will operate on SRH recombination rate.
STATE This is the number of the bound state. It can take values from 1 to the
maximum number of bound states, which is specified by the EIGENS
parameter on the MODELS statement.
STRUCTURE This is a synonym for DEVICE.
TOTAL.HEAT Requests a PROBE of Total heat in GIGA
(GR.HEAT+PT.HEAT+JOULE.HEAT).
TRANSMISSION Stores Transmission versus energy for each solution of NEGF solver. See
Chapter 13 “Quantum: Quantum Effect Simulator” for more information.
U.SP1TRAP Gives the overall recombination rate into multistate traps of ionic species
1, or atomic hydrogen if the atomic hydrogen transport model (H.ATOM)
has been enabled on the MODELS statement.
U.SP2TRAP Gives the overall recombination rate into multistate traps of ionic species
2, or molecular hydrogen if the atomic hydrogen transport model
(H.ATOM) has been enabled on the MODELS statement.
U.SP3TRAP Gives the overall recombination rate into multistate traps of ionic species
3.
VAL.BAND Probes the valence band.
VEL.ELECTRON Specifies that the probe will operate on electron velocity.
VEL.HOLE Specifies that the probe will operate on hole velocity.
WAVE.FUN Probes the wavefunction characterized by the BAND and STATE
parameters.
WAVELENGTH Specifies that the radiative emission wavelength is probed.

1310 ATLAS User’s Manual

PROBE Statements

Region Parameters

BACK This is a synonym for Z.MAX.

BOTTOM This is a synonym for Y.MAX.
FRONT This is a synonym for Z.MIN.
LEFT This is a synonym for X.MIN.
RIGHT This is a synonym for X.MAX.
TOP This is a synonym for Y.MIN.
X, Y Specifies the location of a point probe.
X.MAX Specifies the maximum X coordinate for the minimum, maximum, or integrated probe.
X.MIN Specifies the minimum X coordinate for the minimum, maximum, or integrated probe.
Y.MAX Specifies the maximum Y coordinate for the minimum, maximum, or integrated probe.
Y.MIN Specifies the minimum Y coordinate for the minimum, maximum, or integrated probe.
Z.MAX Specifies the maximum Z coordinate for the minimum, maximum, or integrated probe.
Z.MIN Specifies the minimum Z coordinate for the minimum, maximum, or integrated probe.

Example of Probing the Maximum Value

The following line will cause the maximum electron concentration on the grid to be output to
the log file:
PROBE NAME=peak_electrons MAX N.CONC
Probing at a location Example
This syntax will cause the potential at the location X=0.5 Y=0.1 to be output to the log file.
PROBE NAME=mypotential X=0.5 Y=0.1 POTENTIAL
Vector Quantity Example
For vector quantities, the direction parameter, DIR, must be specified. These two lines allow a
lateral mobility and vertical field in a MOSFET.
PROBE NAME=channel_mobility x=1 y=0.001 N.MOB DIR=0
PROBE NAME=channel_field x=1 y=0.001 FIELD DIR=90
In ATLAS3D, you can add a parameter POLAR to the PROBE statement. Its default value is
90°, which means the direction lies within the XY plane. POLAR gives the angle respect to the
Z axis. A value of zero means that the probed direction is along the Z axis. A value of POLAR
between 0 and 90° will PROBE along a direction vector with a z-component of COS(POLAR).
PROBE NAME=electroncurrent dir=45 polar=60 j.electron x=0.5 y=0.5
z=0.0
This probes the electron current at the specified position along the direction (0.61237,
0.61237, 0.5).

1311 ATLAS User’s Manual

QTREGION Statements

21.47 QTREGION
This allows you to specify one or more polygonal regions as regions where non-local band-
to-band tunneling will occur. Each QTREGION statement sets up a quadrilateral region. The
polygon is built up from one or more quadrilaterals. ATLAS deletes the mesh inside the
polygonal regions and replaces it with another, which is more suitable for the non-local band-
to-band tunneling calculation. The modified mesh is also the mesh to be outputted by any
SAVE statements.
Syntax
QTREGION NUMBER [X1 Y1] X2 Y2 X3 Y3 [X4 Y4]
PTS.TUNNEL| STNL.BEG STNL.END | F.RESOLUTION
PTS.NORMAL | SNRM.BEG SNRM.END

Parameter Type Default Units

DEVICE Character
F.RESOLUTION Character
NUMBER Real 1
PTS.NORMAL Real -1
PTS.TUNNEL Real -1
SNRM.BEG Real 0 m

SNRM.END Real 0 m

STNL.BEG Real 0 m

STNL.END Real 0 m

STRUCTURE Character
X1 Real 0 m

X2 Real 0 m

X3 Real 0 m

X4 Real 0 m

Y1 Real 0 m

Y2 Real 0 m

Y3 Real 0 m

Y4 Real 0 m

1312 ATLAS User’s Manual

QTREGION Statements

Description

DEVICE Specifies which device the statement applies in the mixed mode simulation. The
synonym for this parameter is STRUCTURE.
F.RESOLUTION Specifies a text file that contains a column of numbers between 0 and 1. This
determines the mesh spacing in the tunneling direction and is an alternative to
PTS.TUNNEL. The numbers give the mesh points relative to the corners, 0.0 maps to
the starting corner, and 1.0 maps to the ending corner. Intermediate mesh points are
calculated according to a linear mapping from the supplied coordinates in [0,1].
NUMBER This is the index of the polygonal region that the QTREGION statement applies to.
PTS.NORMAL This is the number of points in quadrilateral in direction perpendicular to
tunneling direction. This dictates a regular mesh spacing along the sides from
corner 1 to corner 2 and from corner 3 to corner 4. See Figure 3-9 for an example.
PTS.TUNNEL This is the number of points in quadrilateral in direction parallel to tunneling
direction. This dictates a regular mesh spacing along the sides from corner 1 to
corner 4 and from corner 2 to corner 3.
SNRM.BEG Along with SNRM.END, is one way of specifying the mesh spacing in the direction
normal to tunneling. SNRM.BEG specifies the mesh spacing at the corner 1 (4).
SNRM.END specifies it at the corner 2 (3).

SNRM.END See SNRM.BEG.

STNL.BEG Along with STNL.END, is one way of specifying the mesh spacing in the direction
parallel to tunneling. STNL.BEG specifies the mesh spacing at the corner 1 (2).
STNL.END specifies it at the corner 4 (3).

STNL.END See STNL.BEG.

STRUCTURE This is a synonym for DEVICE.
X1 This is the X coordinate of the first corner of the quadrilateral.
X2 This is the X coordinate of the second corner of the quadrilateral.
X3 This is the X coordinate of the third corner of the quadrilateral.
X4 This is the X coordinate of the fourth corner of the quadrilateral.
Y1 This is the Y coordinate of the first corner of the quadrilateral.
Y2 This is the Y coordinate of the second corner of the quadrilateral.
Y3 This is the Y coordinate of the third corner of the quadrilateral.
Y4 This is the Y coordinate of the fourth corner of the quadrilateral.

Note: The corners must be specified in counter-clockwise order.

1313 ATLAS User’s Manual

QTREGION Statements

Examples
The following four QTREGION statements make up a polygonal region similar in shape to that
in Figure 3-10. You need to specify all four corners in the first statement.
Subsequent statements may omit (x1, y1) and (x4, y4). In the tunneling direction, only one
specification of mesh spacing in the tunneling direction is allowed per polygonal region. In
this case, a regular spacing with 51 points is requested.
QTREGION NUMBER=1 PTS.TUNNEL=51 X1=0.0 Y1=0.45 X2=0.9 Y2=0.45
X3=0.9 Y3=0.55 X4=0.0 Y4=0.55 SNRM.BEG=0.2 SNRM.END=0.02

QTREGION NUMBER=1 PTS.NORMAL=9 X1=0.9 Y1=0.45 X2=0.94 Y2=0.38

X3=1.03 Y3=0.45 X4=0.9 Y4=0.55

QTREGION NUMBER=1 PTS.NORMAL=9 X2=0.975 Y2=0.25

X3=1.075 Y3=0.25

QTREGION NUMBER=1 PTS.NORMAL=3 X2=0.975 Y2=0.0

X3=1.075 Y3=0.0 SNRM.BEG=0.05 SNRM.END=0.1
The following example uses a data file, mesh.dat, to specify the mesh spacing in the
tunneling direction.
QTREGION NUMBER=2 X1=0.0 Y1=1.55 X2=0.8 Y2=1.55 \
X3=0.8 Y3=1.65 X4=0.0 Y4=1.65 SNRM.BEG=0.2 SNRM.END=0.02
F.RESOLUTION=MESH.DAT
The mesh.dat file should have a format such as
0.0
0.2
0.4
0.45
0.5
0.55
0.6
0.8
1.0
In the direction normal to tunneling, the spacing is 0.2 at one end and 0.02 at the other.

1314 ATLAS User’s Manual

QTX.MESH, QTY.MESH Statements

21.48 QTX.MESH, QTY.MESH

QTX.MESH and QTY.MESH allow you to specify a rectangular mesh in order to calculate
quantum tunneling current in either the X direction or Y direction. It is required in some
situations for the direct quantum tunneling model for gate insulators. It is also required for
non-local band-to-band and trap assisted tunneling models.
Syntax
QTX.MESH NODE=<n> LOCATION=<n>
QTX.MESH SPACING=<n> LOCATION=<n>
QTY.MESH NODE=<n> LOCATION=<n>
QTY.MESH SPACING=<n> LOCATION=<n>

Parameter Type Default Units

LOCATION Real m
NODE Integer
SPACING Real m

Description

NODE Specifies the mesh line index. These mesh lines are assigned consecutively.
LOCATION Specifies the location of the grid line.
SPACING Specifies the mesh spacing at the mesh locations specified by the LOCATION
parameter. If the SPACING parameter is specified, then the NODE parameter should
not be specified.

Example
qtx.mesh loc=0.0 spac=0.25
qtx.mesh loc=1.0 spac=0.25

qty.mesh loc=1.5 spac=0.05

qty.mesh loc=2.0 spac=0.01
qty.mesh loc=3.0 spac=0.01
qty.mesh loc=3.5 spac=0.05
will set up a mesh in the rectangle bounded by x=0.0, x=1.0, y=1.5 and y=3.5. The mesh is
finer in the Y direction because it is intended to calculate the tunneling current in this
direction.

1315 ATLAS User’s Manual

QUIT Statements

21.49 QUIT
QUIT stops execution of ATLAS.
Syntax
QUIT
Synonyms
STOP
END
EXIT
Description
The QUIT statement may be placed anywhere in an input file. ATLAS will stop execution
upon encountering the QUIT statement. All input lines after the occurrence of the QUIT
statement will be ignored for that execution of ATLAS.

Note: To quit and immediately restart ATLAS inside of DECKBUILD, use the GO ATLAS statement. Full
details on the GO syntax statement are found in the DECKBUILD USER ’S MANUAL.

1316 ATLAS User’s Manual

REACTION Statements

21.50 REACTION
REACTION specifies the desired reactions between ionic species and between ionic species
and free carriers. The reaction rates are also specified.\\
Syntax
REACTION [BEFORE.<X>AFTER.<X>] [Rate parameters]
See Sections 3.16 “Generic Ion Transport Model” and 3.17 “Generic Ion Reaction Model”for
full description.

Parameter Type Default Units

AFTER.N Integer 0
AFTER.P Integer 0
AFTER.SP1 Integer 0
AFTER.SP2 Integer 0
AFTER.SP3 Integer 0
BEFORE.N Integer 0
BEFORE.P Integer 0
BEFORE.SP1 Integer 0
BEFORE.SP2 Integer 0
BEFORE.SP3 Integer 0
EQUIRATE Real 0.0 cm-3s-1
FORWARD.EA Real 0.0 eV
FORWARD.RATE Real 0.0 See description
REVERSE.EA Real 0.0 eV
REVERSE.RATE Real 0.0 See description

Description
You specify reactants and products with the parameters BEFORE.<X> and AFTER.<X>
described in first description table below. You specify reaction rates with the parameters
described in the second table below.

AFTER.N Number of electrons produced by reaction.

AFTER.P Number of holes produced by reaction.
AFTER.SP1 Number of SPECIES1 ions produced by reaction.
AFTER.SP2 Number of SPECIES2 ions produced by reaction.
AFTER.SP3 Number of SPECIES3 ions produced by reaction.

1317 ATLAS User’s Manual

REACTION Statements

BEFORE.N Number of electrons consumed by reaction.

BEFORE.P Number of holes consumed by reaction.
BEFORE.SP1 Number of SPECIES1 ions consumed by reaction.
BEFORE.SP2 Number of SPECIES2 ions consumed by reaction.
BEFORE.SP3 Number of SPECIES3 ions consumed by reaction.
EQUIRATE Fixed reaction rate in cm-3s-1.
FORWARD.EA Activation energy in eV of forward reaction rate.
FORWARD.RATE Forward reaction rate constant. The forward rate itself can depend on any
product of concentrations, and the rate constant should give overall rate
dimensions of cm-3s-1.
REVERSE.EA Activation energy in eV of reverse reaction rate.
REVERSE.RATE Reverse reaction rate constant. The reverse rate itself can depend on any
product of concentrations, and the rate constant should give overall rate
dimensions of cm-3s-1.

Examples
The set of reactions
2H  H 2
- -
H+e H
- +
H +h H
can be represented by the following set of REACTION statements
REACTION BEFORE.SP1=2 AFTER.SP2=1
REACTION BEFORE.SP1=1 BEFORE.N=1 AFTER.SP3=1
REACTION BEFORE.SP3=1 BEFORE.P=1 AFTER.SP1=1
where SPECIES1 is H, SPECIES2 is H2, and SPECIES3 is H-.

1318 ATLAS User’s Manual

REGION Statements

21.51 REGION
REGION specifies the location of materials in a previously defined mesh. Every triangle must
be defined as a material.
Syntax
REGION NUMBER=<n> <material> [<position>]

Parameter Type Default Units

ABS Logical False

A.MAX Real degrees
A.MIN Real degress
A-SILICO Logical False

ACCEPTORS Real 0.0 cm-3

ALGAAS Logical False

ALINAS Logical False

ASUB Real 1.0 Å

BOTTOM Logical False

CALC.STRAIN Logical False

COMPX.BOTTOM Real 0.0
COMPY.BOTTOM Real 0.0
COMPX.TOP Real 0.0
COMPY.TOP Real 0.0
CONDUCTOR Logical False
CONVERT Logical False
DEC.C1 Real 0.0 eV
DEC.C2 Real 0.0 eV
DEC.C3 Real 0.0 eV
DEC.D2 Real 0.0 eV
DEC.D4 Real 0.0 eV
DEC.ISO Real 0.0 eV
DEV.HH Real 0.0 eV
DEV.LH Real 0.0 eV

1319 ATLAS User’s Manual

REGION Statements

Parameter Type Default Units

DEV.SO Real 0.0 eV

DEV.ISO Real 0.0 eV
DIAMOND Logical False
DONORS Real 0.0 cm-3

ETA.NEGF Real 0.0066 eV

EQUIL.NEGF Logical False
FIXED.FERMI Logical False
FN.SIS Logical False
HELMHOLTZ Logical True
GAAS Logical False
GAASP Logical False
GERMANIU Logical False
GRAD.12 Real m
GRAD.23 Real mm
GRAD.34 Real m
GRAD.41 Real m
GRADED Logical False
HGCDTE Logical False
INGAAS Logical False
INAS Logical False
INASP Logical False
INGAP Logical False
INP Logical False
INSULATO Logical False
IX.LOW Integer left of structure
IX.HIGH Integer right of structure
IX.MAX Integer right of structure
IX.MIN Integer left of structure
IY.LOW Integer top of structure

1320 ATLAS User’s Manual

REGION Statements

Parameter Type Default Units

IY.HIGH Integer bottom of structure

IY.MAX Integer top of structure
IY.MIN Integer bottom of structure
IZ.HIGH Real
IZ.LOW Real
IZ.MAX Integer
IZ.MIN Integer
LED Logical False
MATERIAL Character
MODIFY Logical False
NAME Character
NA.TOP Real 0.0 cm-3
ND.TOP Real 0.0 cm-3
NA.BOTTOM Real 0.0 cm-3
ND.BOTTOM Real 0.0 cm-3
NITRIDE Logical False
NUMBER Integer
NX Real 2

NY Real 2

OXIDE Logical False

OXYNITRI Logical False
P1.X Real m

P1.Y Real m

P2.X Real m

P2.Y Real m

P3.X Real m

P3.Y Real m

P4.X Real m

P4.Y Real m

1321 ATLAS User’s Manual

REGION Statements

Parameter Type Default Units

PIEZOELECTRIC Logical False

PIEZO.SCALE Real 1.0
POLAR.CHARGE Real 0.0 cm-2
POLAR.SCALE Real 1.0
POLARIZATION Logical False
POLYSILICON Logical False
PSP.SCALE Real 1.0
QTREGION Integer 0
QWELL Logical False
QWNUM Integer -1
R.MAX Real m
R.MIN Real m
S.OXIDE Logical False
SAPPHIRE Logical False
SCHRO Logical True
SEMICOND Logical False
SELF Logical False
SI3N4 Logical False
SIC Logical False
SIGE Logical False
SILICON Logical False
SIO2 Logical False
SL.GEOM Character 1DY

SL.NUMBER Integer
SLATT Logical False
SLATT.NUMBER Integer
STAY Logical False
STRAIN Real 0.0

STR.BOT Real 0.0

1322 ATLAS User’s Manual

REGION Statements

Parameter Type Default Units

STR.TOP Real 0.0

SUBSTRATE Logical False
SX Real 0.0

SY Real 0.0

THICKNESS Real 0.0 m

TOP Logical False

TRAPPY Logical False

USER.GROUP Character
USER.MATERIAL Character
WELL.CNBS Integer 1
WELL.DEBUG Logical False
WELL.FIELD Logical True
WELL.GAIN Real 1.0
WELL.NX Integer 10
WELL.NY Integer 10
WELL.NZ Integer 10
WELL.OVERLAP Real
WELL.VNBS Integer 1
X.COMP Real 0.0
X.MAX Real right of structure m
X.MIN Real left of structure m
X.MOLE Real 0.0
Y.MOLE Real 0.0
Y.COMP Real 0.0
Y.MIN Real top of structure m
Y.MAX Real bottom of structure m
Z.MAX Real
Z.MIN Real

1323 ATLAS User’s Manual

REGION Statements

Parameter Type Default Units

ZEROBP Logical False

ZNSE Logical False
ZNTE Logical False

Description

n Specifies a region number from 1 to 200.

material This is one or more of the material names described below.
position This is one or more of the position parameters described below.
SL.GEOM Specifies the superlattice geometry as either 1DY (default) or 1DX.
SL.NUMBER Assigns the index for a regions membership in a indexed superlattice.
SLATT Enables the superlattice model for a region.
SLATT.NUMBER This is an alias for SLATT.NUMBER.

Material Parameters

ABS Specifies that quantum confinement is included in the calculation of

absorption or imaginary index of refraction. To use this model, you must also
specify QWELL on the REGION statement. See Section 10.8.3 “Quantum Well
Absorption”.
ACCEPTORS Specifies a uniform density of ionized acceptors in the region.
ASUB Specifies the substrate lattice constant for strain calculations as described in
Section 3.6.11 “Epitaxial Strain Tensor Calculation in Wurtzite”.
CALC.STRAIN Specifies that the strain in the region is calculated from the lattice mismatch with
adjacent regions.
COMPX.BOTTOM Specify the composition fractions at the top and bottom of a region when
COMPY.BOTTOM linearly graded. Linear grading is specified by the GRADED parameter. See
COMPX.TOP Section 2.6.4 “Automatic Meshing (Auto-meshing) Using The Command
COMPY.TOP Language”.
CONDUCTOR Specifies that the region is to be simulated as a conductor. This means that the
conduction equation is solved for the region.
CONVERT This is an alias for MODIFY.
DEC.C1 This is a conduction band shift for 1st pair of electron valleys.
DEC.C2 This is a conduction band shift for 2nd pair of electron valleys.

1324 ATLAS User’s Manual

REGION Statements

DEC.C3 This is a conduction band shift for 3rd pair of electron valleys.
DEC.D2 This is a conduction band shift for 2 electron valleys if DEC.C1 is not set.
DEC.D4 This is a conduction band shift for 4 electronvalleys if DEC.C2 is not set.
DEC.ISO This is a conduction band shift for isotropic electron band, when
NUM.DIRECT=1.

DEV.HH This is a valence band shift for heavy holes.

DEV.LH This is a valence band shift for light holes.
DEV.ISO This is a valence band shift for isotropic hole band, when NUM.BAND=1.
DEV.SO This is a valence band shift for split-off holes.
DONORS Specifies a uniform density of ionized donors in the region.
ETA.NEGF An imaginary optical potential, added to Hamiltonian of quasi-equilibrium
regions in planar NEGF model.
EQUIL.NEGF Specified that the region will be trated as quasi-equilibrium in planar NEGF
model.
GRAD.<n> Specifies the compositional gradings for heterojunctions along each side of the
region rectangle or quadrilateral. The value of the GRAD parameters specifies the
distance at which the composition fraction reduces to zero. A value of 0.0
specifies that the heterojunction is abrupt. The value of <n> can be the numbers:
12, 23, 34, and 41. The meanings of these numbers is down below.
• 12 = top surface
• 23= right hand side
• 34= bottom surface
• 41 = left hand side
These correspond to the point indices around the rectangular region working
clockwise from top left.
GRADED Specifies that the region has linear compositional or doping variation or both.
See also COMPX.BOTTOM, COMPY.BOTTOM, COMPX.TOP, COMPY.TOP, ND.BOTTOM,
ND.TOP, NA.BOTTOM and NA.TOP.

HELMHOLTZ Allows (default) or prevents a solution of vector Helmholtz equation in the

region.
LED Specifies that the region is to be treated as a light emitting region and included in
postprocessing for LED analysis. See Chapter 11 “LED: Light Emitting Diode
Simulator”.
MATERIAL Specifies the material used for the region. Valid material names are listed in
Table B-1. All materials are divided into three classes: semiconductors,
insulators and conductors.

1325 ATLAS User’s Manual

REGION Statements

Note: You can specify the following logical parameters to indicate the region material instead of assigning the
MATERIAL parameter: SILICON, GAAS, POLYSILI, GERMANIU, SIC, SEMICOND, SIGE,
ALGAAS, A-SILICO, DIAMOND, HGCDTE, INAS, INGAAS, INP, S.OXIDE, ZNSE, ZNTE,
ALINAS, GAASP, INGAP, INASP, OXIDE, SIO2, NITRIDE, SI3N4, INSULATO, SAPPHIRE, and
OXYNITRI.

MODIFY This is used to modify the characteristics of regions imported into the simulator.
See Section 2.6.5 “Modifying Imported Regions”. The alias for this parameter is
CONVERT.

NAME Specifies the name of the region. The name can be used in the MODELS,
MATERIAL, and IMPACT statements to provide regionally dependent models. This
name is just a label. It doesn’t imply any material parameter settings.
ND.BOTTOM Specify the doping concentrations at the top and bottom of a region when
ND.TOP linearly graded. Linear grading is specified by the GRADED parameter. See
NA.BOTTOM Section 2.6.4 “Automatic Meshing (Auto-meshing) Using The Command
NA.TOP Language”.
NUMBER Assigns a region number. Multiple REGION lines with the same number can be
used to define region shapes made from several rectangles.
PIEZO.SCALE This is an alias for POLAR.SCALE.
PIEZOELECTRIC This is an alias for POLARIZATION.
POLARIZATION Enables the automatic calculation of added interface charge due to
spontaneous and piezoelectric polarization. The alias for this parameter is
PIEZOELECTRIC. See Section 3.6.12 “Polarization in Wurtzite Materials”.

POLAR.CHARGE Specifies polarization charge densities to replace the density calculated using
Equations 3-601 and 3-602.
POLAR.SCALE Specifies a constant scale factor multiplied by the calculated spontaneous and
piezoelectric polarization charges when you enable polarization by setting the
POLARIZATION parameter of the REGION statement. The alias for this parameter
is PIEZO.SCALE. See Section 3.6.12 “Polarization in Wurtzite Materials”.
PSP.SCALE Specifies a constant scale factor similar to POLAR.SCALE but applied only to
spontaneous polarization.
QTREGION This designates the region as a quantum barrier for the quantum barrier tunnelling
models SIS.EL and SIS.HO. The index of the QTREGION is also set with this
parameter. For further information, see Section 5.2.4 “Non-local Quantum Barrier
Tunneling Model”.
QWELL Specifies that the region is treated as a quantum well for calculation of
radiative recombination or gain or both for certain optoelectronic models.
QWNUM Optionally sets the number of the quantum well region.
SCHRO Allows (default) or prevents a solution of Schrodinger in the region. Switch it off,
if the region is to be treated classically.

1326 ATLAS User’s Manual

REGION Statements

STRAIN Specifies the strain in the region. (Negative strain is compressive).

STR.BOT Specify the strain at the bottom and top of the region for calculations of the k.p
STR.TOP model. The strain values between the top and bottom of the region are linearly
interpolated.
SUBSTRATE This is a logical parameter indicating that the specified region should be
considered as the the substrate for strain calculations as described in Section
3.6.11 “Epitaxial Strain Tensor Calculation in Wurtzite”.
TRAPPY Used only with the BESONOS model. This indicates that a Silicon-Nitride region
will be able to trap electrons and holes.
USER.GROUP Specifies the material group for the user-defined material. USER.GROUP can be
either SEMICONDUCTOR, INSULATOR, or CONDUCTOR.
USER.MATERIAL Specifies a user-defined material name. The specified material name can be any
name except that of a default material, such as Silicon. You should define each
material with an accompanying definition in the MATERIAL statement. You can
define up to 50 user-defined materials.
WELL.CNBS Specify the number of bound states retained for calculation of radiative
WELL.VNBS recombination or gain if the region is treated as a quantum well as specified by the
QWELL parameter.

WELL.FIELD Specifies that the calculations of bound state energies should include the effects of
the local field.
WELL.GAIN Specifies a constant scale factor multiplied by the calculated gain to give the net
gain used for certain optoelectronic calculations.
WELL.NX Specifies the number of grid points in x, y, and z directions for Schrodinger
WELL.NY equation in a quantum well region.
WELL.NZ

WELL.OVERLAP Specifies the wavefunction overlap integral. If WELL.OVERLAP is not specified,

then the overlap integral is calculated from the wavefunctions.

Note: The highest region number takes precedence if REGION definitions overlap.

REGRID Statements

21.52 REGRID
REGRID allows you to refine a crude mesh. A triangle is refined when the chosen variable
changes by more than one specified criteria.
Syntax
REGRID RATIO=<n> <var> [<lp>] [<cp>] [<io>]

Parameter Type Default Units

ABSOLUTE Logical False

ASCII Logical False
CHANGE Logical True
COS.ANG Real 2.0
DOPFILE Character
DOPING Logical cm-3
E.TEMP Logical False
EL.FIELD Logical V/cm
ELECTRON Logical cm-3
H.TEMP Logical False
HOLE Logical cm-3
IGNORE Integer
IN.GREEN Character
LOCALDOP Logical False
LOGARITHM Logical False
MAX.LEVEL Integer 1+ maximum level of grid
MIN.CARR Logical cm-3
NET.CARR Logical cm-3
NET.CHRG Logical cm-3
PISCES.OUT Logical False
OUT.GREEN Character value of OUTFILE + tt
OUTFILE Character
POTENTIAL Logical V

1331 ATLAS User’s Manual

REGRID Statements

Parameter Type Default Units

QFN Logical V
QFP Logical V
REGION Integer All
SMOOTH.K Integer
STEP Real
X.MAX Real Right m
X.MIN Real Left m
Y.MAX Real Bottom m
Y.MIN Real Top m

Description

RATIO Specifies the maximum allowed variance across one element.

STEP

var This is one of the variable parameters described below. The selected parameter is
used as the basis for regridding.
lp This is one of more of the location parameters described below. These parameters are
used to select the areas which are to be refined.
cp This is one or more of the control parameters. These parameters are used to
control the plotted output.
io This is one or more of the File I/O parameters.

Variable Parameters

DOPING Selects net doping.

E.TEMP Select electron temperature.
EL.FIELD Selects electric field.
ELECTRON Selects electron concentration.
H.TEMP Selects hole temperature.
HOLE Selects hole concentration.
MIN.CARR Selects minority carrier concentration.
NET.CARR Selects net carrier concentration.

1332 ATLAS User’s Manual

REGRID Statements

NET.CHRG Selects net charge.

POTENTIAL Selects mid-gap potential.
QFN Selects the electron quasi-Fermi level.
QFP Selects the hole quasi-Fermi level.

Location Parameters
If no location parameters are specified, refinement will include:
• All regions for potential and electric field regrids
• All semiconductor regions for regrids which depend on the other variables

IGNORE Specifies regions that are not to be refined.

REGION Specifies regions which are refined according to the specified criterion.
Other regions may be refined to maintain well-shaped triangles.
X.MAX Uses device coordinates to specify the maximum x-value for refinement.
X.MIN Uses device coordinates to specify the minimum x-value for refinement.
Y.MAX Uses device coordinates to specify the maximum y-value for refinement.
Y.MIN Uses device coordinates to specify the minimum y-value for refinement.

Control Parameters

ABSOLUTE Specifies that the absolute value of the variable be used.

CHANGE Determines whether to use the magnitude or the difference of a triangle variable as
the refinement criterion. This parameter defaults to “difference”.
COS.ANGLE Limits the creation of obtuse angles in the mesh by specifying obtuse criterion. If this
parameter is used, nodes are added to the mesh so that the number of obtuse triangles
is reduced.

Note: Be careful when using the COS.ANGLE parameter. Recommended values are from 0.8 to 0.95. Smaller
values may dramatically increase the number of nodes.

1333 ATLAS User’s Manual

REGRID Statements

LOCALDOP Specifies that if minority carrier concentration exceeds local doping, the grid will be
refined. This parameter is used in conjunction with minority carrier regrids.
LOGARITHM Specifies a logarithmic refinement scale. Since many of the quantities may become
negative, numerical problems are avoided by using log(x)=sign(x)·log10(1+|x|). If you
wish to obtain the true logarithm of a quantity, the ABSOLUTE parameter must be
specified before the LOGARITHM parameter is specified. The absolute value of a
quantity is computed first, thereby eliminating negative arguments.
MAX.LEVEL Specifies the maximum level of any triangle relative to the original mesh. This
parameter defaults to one more than the maximum level of the grid, but can be set to a
smaller value to limit refinement. Values less than or equal to zero are interpreted
relative to the current maximum grid level.
SMOOTH.KEY Specifies a smoothing index. The digits of the index are read in reverse order and
interpreted as follows:
1. Triangle smoothing: All region boundaries remain fixed.
2. Triangle smoothing: Only material boundaries are maintained.
3. Node averaging.
4. Improved triangle smoothing method: This method uses diagonal
flipping to reduce the number of obtuse triangles.
5. Aligns triangles with electric field gradient.
Usually option 1 is sufficient. Option 2 is useful only if a device has several regions of
the same material and the border between different regions is unimportant. Option 3 is
not recommended when the initial mesh is basically rectangular, such as mesh
information usually obtained from SSUPREM4. Option 4 is similar to option 1, but
option 4 usually creates less obtuse triangles.

File I/O Parameters

ASCII Specifies that mesh files and triangle trees will be written in an ASCII rather than a
binary format. This parameter has no effect on the device doping file (see the
DOPFILE parameter).

DOPFILE Specifies the name of a file, which contains device doping information. This file is
created on the DOPING statement. Specifying DOPFILE avoids linear interpolation
of doping values at newly created grid points by using the initial doping
specification to apply doping to the new grid points.
IN.GREEN Specifies a triangle tree for the mesh which will be used in this regrid. If this
parameter is not specified, the program will look for a file with the same name as
the current mesh plus, tt, at the end. If no such file exists, the program will not use a
triangle tree for the previous mesh.
MASTER.OUT Saves mesh and doping information in a standard structure file format.

1334 ATLAS User’s Manual

REGRID Statements

OUTFILE Specifies the name of a standard structure output file where mesh information will
be stored. This parameter must be specified if the mesh is to be used for subsequent
runs.
OUT.GREEN Specifies the name of the file that holds the history of the triangle tree. This history
is used in further regrid steps.
PISCES.OUT Saves mesh and doping information in a binary PISCES-II format. Files in this
format cannot be displayed in TONYPLOT.

Doping Regrid Examples

Starting with an initial grid, we refine twice so that all triangles with large doping steps are
refined.
REGRID LOG DOPING RATIO=6 OUTF=grid1 DOPF=dopxx SMOOTH=4
REGRID LOG DOPING RATIO=6 OUTF=grid2 DOPF=dopxx SMOOTH=4
A similar effect could be obtained with just one regrid statement.
REGRID LOG DOPING RATIO=6 OUTF=grid2 DOPF=dopxx MAX.LEVEL=2
In both cases, two levels of refinement are performed. The first example is recommended
because new doping information is introduced at each level of refinement. This produces
better refinement criterion and fewer triangles.
Potential Regrid Example
Next, an initial solution is produced and triangles which exhibit large potential steps are
refined.
SOLVE INIT
REGRID POTENTIAL RATIO=0.2 OUTF=grid3.str SMOOTH=4
Re-initializing after Regrid example
Often you need to re-solve the same bias point after a REGRID using the following style of
syntax.
SOLVE VDRAIN=3.0
REGRID POTENTIAL RATIO=0.25 SMOOTH=4 OUTF=mygrid.str
SOLVE PREV
Occasionally, you need to quit and restart ATLAS with the new mesh. To do this, type syntax
such as:
SOLVE VDRAIN=3.0
REGRID POTENTIAL RATIO=0.25 SMOOTH=4 OUTF=mygrid.str
go atlas
MESH INF=mygrid.str
After this MESH statement all models, material parameters and numerical methods have to be
re-specified before any SOLVE statement.

1335 ATLAS User’s Manual

SAVE Statements

21.53 SAVE
SAVE saves all node point information into an output file.

Note: In all cases the region boundaries, electrodes, mesh and doping are saved. If a SOLVE statement has
preceded the SAVE statement all electrical data from the last solution is stored.

Syntax
SAVE OUTFILE=<filename> [MASTER]

Parameter Type Default Units

ABSORB Logical False

AMBIEN.IMAG Real 0.0
AMBIEN.REAL Real 1.0
ANGLE.OUTPUT Real 30.0 degrees
ANGPOWER Character
COMPARE Logical False
COUPLING3D Logical False
CUTPLANE Logical False
DIPOLE Logical False
DDMS.LOG Logical False
EDGE_ONLY Logical False
EDGE_REFLECT Logical False
EMAX Real eV
EMIN Real eV
FAST Logical False
HORIZONTAL Real 2/3
INCOHERENT Logical False
INCOH.TOP Logical False
INTERFERE Logical False
L.WAVE Real 0.8 m

LMAX Real m

LMIN Real m

MASTER Character True

1336 ATLAS User’s Manual

SAVE Statements

Parameter Type Default Units

MIR.BOTTOM Logical False

MIN.POWER Real 110-4
MIR.TOP Logical False
MQWELLS Character
NEAR.FIELD Character
NEGF.LOG Logical False
NSAMP Integer 100
NUMRAYS Integer 180
OUT.FILE Character
OUTFILE Character
PATTERNS Character
PISCES Logical False
POLAR Real 0.0
RAYPLOT Character
REFLECTS Integer 0
SIDE Logical False
SMOOTHNESS Real 8.0
SOURCE.TERM Logical False
SOURCE_TERM Logical False
SPEC.INTERSUB Character
SPECT.ANGLE Character
SPECTRUM Character
STRUCTURE Character False
SURFACE_ONLY Logical False
TEMPER Real 300.0
TR.MATRIX Logical False
TRAP.FILE Character
USER.SPECT Character
X Real 0.0 m

1337 ATLAS User’s Manual

SAVE Statements

Parameter Type Default Units

X.CUTPLANE Real 0.0 m

XMAX Real 0.0 m

XMIN Real 0.0 m

XNUM Integer 0
Y Real 0.0 m

Y.CUTPLANE Real 0.0 m

YMAX Real 0.0 m

Y.MAX Real 0.0 m

YMIN Real 0.0 m

Y.MIN Real 0.0 m

YNUM Integer 0
Z Real 0.0 m

Z.CUTPLANE Real m

ZMAX Real 0.0 m

Z.MAX Real 0.0 m

ZMIN Real 0.0 m

Z.MIN Real 0.0 m

Description

ABSORB Takes absorption into account in reverse ray-tracing (see Section 11.5 “Reverse
Ray-Tracing”). The absorption is assumed to be constant specified for each
material by the imaginary part of the refractive index.
AMBIEN.IMAG Specifies the imaginary index of refraction for the ambient outside the device
domain for LED reverse ray tracing.
AMBIEN.REAL Specifies the real index of refraction for the ambient outside the device domain
for LED reverse ray tracing.
ANGLE.OUTPUT Specifies the angular interval in degrees where spectrum versus angle are written
to the file specified by the SPECT.ANGLE parameter.
ANGPOWER Enables the reverse ray-tracing algorithm (see Section 11.5 “Reverse Ray-
Tracing”) for analysis of output coupling of light from the structure of a Light
Emitting Diode. ANGPOWER specifies the name of the output file for the angular
power density vs. output angle dependence.

1338 ATLAS User’s Manual

SAVE Statements

COUPLING3D Calculates an output coupling coefficient, assuming a cylindrically

symmetric light output with the light source located on the axis of
symmetry (see Section 11.5 “Reverse Ray-Tracing”). Takes into account the 3D
nature of the problem with output coupling calculation.
CUTPLANE Specifies that a 2D cutplane will be saved from a 3D structure to the file
specified by the OUTFILE parameter. The direction of the cutplane depends on
the value specified on either the X.CUTPLANE, Y.CUTPLANE or Z.CUTPLANE
parameter. If you specify the X.CUTPLANE parameter, the output cutplane will be
perpendicular to the X axis. If you specify the Y.CUTPLANE parameter, the output
cutplane will be perpendicular to the Y axis. If you specify the Z.CUTPLANE
parameter, the output cutplane will be perpendicular to the Z axis.
DIPOLE Specifies a particular angular distribution of the internal radiating field that
corresponds to a preferred in-plane orientation of dipoles often relevant to OLED
devices (see Section 11.5 “Reverse Ray-Tracing”).
DDMS.LOG Specifies that in case of DDMS model (see Section 13.8 “Drift-Diffusion Mode-
Space Method (DD_MS)”), an additional log file with subband-resolved
quantities will be stored along with structure file.
EDGE_ONLY Specifies that for LED reverse ray tracing only the rays exiting from the edges
are considered. Rays exiting at the top and bottom surfaces are not shown.
EDGE_REFLECT Specifies the rays in reverse ray tracing reaching the edges of the device are
totally reflected.
EMAX and EMIN Specify the energy range for saving a spectrum file.
FAST Specifies that a fast algorithm should be used for LED reverse ray tracing. In
this case, the ray tracing is performed from region boundary to region boundary
rather than the default element by element trace.
HORIZONTAL Specifies the proportion of horizontally oriented dipoles for the dipole
emission model.
INCOHERENT Specifies that an incoherent material in LED/OLED structure is used for the
purpose of combined transfer matrix method/reverse ray tracing (TTM+RRT)
optical analysis.
INCOH.TOP When specified with the SOURCE.TERM parameter, this allows a combination of
transfer matrix method in thin coherent layers and ray tracing in thick incoherent
layers for LED reverse ray tracing.
INTERFERE Specifies that rays originating in the same point are fully coherent in reverse ray-
tracing (see Section 11.5 “Reverse Ray-Tracing”). In this case, the phase
information upon propagation is preserved. Phase change upon reflection is also
considered. Thus, interference of rays exiting the device at the same angle are
taken into account.
L.WAVE Specifies the wavelength for reverse ray-tracing (see Section 11.5 “Reverse Ray-
Tracing”).
LMAX and LMIN Specify the range of wavelength for saving spectrum files.

1339 ATLAS User’s Manual

SAVE Statements

MASTER Specifies that the output file will be written in a standard structure format. Files
in this format can be plotted in TONYPLOT.
MIN.POWER Specifies the minimum relative power of a ray (see Section 11.5 “Reverse Ray-
Tracing”). The ray is not traced after its power falls below MIN.POWER value.
This is useful to limit the number of rays traced. The default value is
MIN.POWER=1e-4.

MIR.TOP Specifies that the top surface of the device be treated as an ideal mirror in reverse
ray-tracing (see Section 11.5 “Reverse Ray-Tracing”).
MIR.BOTTOM Specifies that the bottom surface of the device be treated as an ideal mirror in
reverse ray-tracing (see Section 11.5 “Reverse Ray-Tracing”).
NEAR.FIELD Specifies the output file name where LED near field distribution is saved. A file
for each surface where light exits is created.
NEGF.LOG Specifies that spectra of transmission, DOS, density and and curent obtained
from NEGF model need to be saved.
NSAMP Specifies the number of samples to use for a spectrum plot.
NUMRAYS Specifies the number of rays starting from the origin in reverse ray-tracing (see
Section 11.5 “Reverse Ray-Tracing”). The default is 180. Acceptable range is 36-
3600.
OUT.FILE This is a synonym for OUTFILE.
OUTFILE Specifies the name of an output file name. The synonym for this parameter is
OUT.FILE.

PISCES Specifies that the output file will be written in the original PISCES-II format.
PATTERNS Specifies a character string representing the root of the file names where near and
far field patterns are written for LASER or VCSEL. The near field pattern file is
appended with the string .nfp and the far field pattern file is appended with the
string .ffp.
POLAR Specifies polarization of the emitted photons in degrees while linearly
polarized light is assumed (see Section 11.5 “Reverse Ray-Tracing”). Parallel
(TM-mode, POLAR=90.0) and perpendicular (TE-mode, POLAR=0.0)
polarizations result in significantly different output coupling values. Use
POLAR=45.0 if there is no preferred direction of polarization of emitted photons
(unpolarized light emission).
RAYPLOT Specifies the name of the output file containing the information on each ray
exiting the device in single origin reverse ray-tracing (see Section 11.5 “Reverse
Ray-Tracing”). This file is only created when the single origin for all rays is
assumed. The information includes ray output angle, relative ray power (TE and
TM-polarization and total), and initial internal angle at the origin (only if
INTERFERE parameter is unspecified). 0° angle corresponds to the rays in the X
axis direction. 90° angle corresponds to the rays in the Y axis direction.

1340 ATLAS User’s Manual

SAVE Statements

REFLECTS Specifies a number of reflections to be traced for each ray in reverse ray-tracing
(see Section 11.5 “Reverse Ray-Tracing”). The default value is REFLECTS=0.
The maximum allowed value is REFLECTS=10. Setting the number of reflections
to 3 or 4 is often a good choice.
SIDE Specifies that the rays reaching the sides of the device are terminated there and
do not contribute to the total light output (see Section 11.5 “Reverse Ray-
Tracing”).
SOURCE.TERM Indicates that dipole source terms are combined with transfer matrix
calculation of light output for LEDs. The alias is SOURCE_TERM.
SPEC.INTERSUB Specifies a file name for saving the spectrum of intersubband transitions.
SPECT.ANGLE Specifies the output file name where LED spectrum is written as a function of
emission angle.
SPECTRUM Specifies the name of a file for saving spectrum plots.
STRUCTURE This is a synonym for SAVE.
SURFACE_ONLY Specifies that for LED reverse ray tracing only the rays exiting from the top and
bottom surfaces are considered. Rays exiting at the edges are not shown.
TEMPER This is the temperature (needed for using appropriate refractive indexes of the
materials in reverse ray-tracing). The default setting of 300 K will be used if
TEMPER is unspecified.

TRAP.FILE Specifies a file name to which all trap densities at the location specified by X,Y,
and Z are to be saved.
TR.MATRIX Specifies that the transfer matrix method should be used to handle thin film
LEDs.
X This is the X coordinate of light origin for a single point reverse ray-tracing (see
Section 11.5 “Reverse Ray-Tracing”).
X.CUTPLANE Specifies the X coordinate for a YZ cutplane. The region and nodal data from the
nearest X plane to this X coordinate will be saved in the structure file.
XMAX This is the maximum X coordinate of a rectangular area containing multiple
origin points in reverse ray-tracing (see Section 11.5 “Reverse Ray-Tracing”).
XMIN This is the minimum X coordinate of a rectangular area containing multiple
origin points in reverse ray-tracing (see Section 11.5 “Reverse Ray-Tracing”).
XNUM Specifies the number of points along X axis within a rectuangular area in
multiple origin reverse ray-tracing (see Section 11.5 “Reverse Ray-Tracing”).
Y This is the Y coordinate of light origin for a single point reverse ray-tracing (see
Section 11.5 “Reverse Ray-Tracing”).
Y.CUTPLANE Specifies the Y coordinate for a XZ cutplane. The region and nodal data from the
nearest Y plane to this Y coordinate will be saved in the structure file.

1341 ATLAS User’s Manual

SAVE Statements

YMAX This is the maximum Y coordinate of a rectangular area containing multiple

origin points in reverse ray-tracing (see Section 11.5 “Reverse Ray-Tracing”).
YMIN This is the minimum Y coordinate of a rectangular area containing multiple
origin points in reverse ray-tracing (see Section 11.5 “Reverse Ray-Tracing”).
YNUM Specifies the number of points along Y axis within a rectuangular area in
multiple origin reverse ray-tracing (see Section 11.5 “Reverse Ray-Tracing”).
Z Specifies the Z coordinate.
Z.CUTPLANE Specifies the Z coordinate for a XY cutplane. The region and nodal data from the
nearest Z plane to this Z coordinate will be saved in the structure file.

Basic Save Example

SOLVE V1=5
SAVE OUTF=data1.str
is equivalent to
SOLVE V1=5 OUTF=data1.str MASTER
Save Example with User-Defined Output
In the second example, the SAVE and OUTPUT commands are used to produce two output files
for the same bias. The OUTPUT statement selects which data will be stored in each file. The
first file (data1.str) contains the default contents, total electric field, and components of
electron velocity. The second file (data2.str) contains components of hole velocity and
band edge potentials. Note that the EX.VELO and EY.VELO parameters are used to prevent
electron velocity components from being stored in file data2.
OUTPUT E.FIELD EX.VELO EY.VELO
SAVE OUTF=data1.str
OUTPUT HX.VELO HY.VELO CON.BAND VAL.BAND ^EX.VELO ^EY.VELO
SAVE OUTF=data2.str

Note: You can customize the contents of the saved file by using the OUTPUT statement.

1342 ATLAS User’s Manual

SET Statements

21.54 SET
The SET statement is used to define variables for substitution into ATLAS syntax. SET
commands are executed by DECKBUILD.

Note: Full documentation of the SET statement is found in the DECKBUILD USER ’S MANUAL.

Numeric Variable Example

Define a numerical variable. Use it in a calculation and substitute it into the ATLAS syntax
for REGION definition
SET MYLENGTH=0.1
SET HALFLENGTH= $”MYLENGTH”*0.5
...
REGION NUM=1 MATERIAL=SILICON X.MIN=$”HALFLENGTH” X.MAX=$”MYLENGTH”
String Variable Example
Define a string variable to use as part of a filename
SET LABEL=testcase1
..
LOG OUTF=$”LABEL”.log
..
SAVE OUTF=bias_$”LABEL”_25.str
This will produce files called testcase1.log and bias_testcase1_25.str.

1343 ATLAS User’s Manual

SINGLEEVENTUPSET Statements

21.55 SINGLEEVENTUPSET
SINGLEEVENTUPSET specifies the values of parameters used in Single Event Upset modeling.
Syntax
SINGLEEVENTUPSET <parameters>

Parameter Type Default Units

A1 Real
A2 Real
A3 Real
A4 Real
B1 Real
B2 Real
B3 Real
B4 Real
B.DENSITY Real 0 cm-3
BEAM.RADIUS Real 0 m
DENSITY Real cm-3
DEVICE Character
STRUCTURE Character
ENTRYPOINT Real vector m
EXITPOINT Real vector m
F.SEU Character
PCUNITS Logical False
RADIALGAUSS Logical False
RADIUS Real m
RESCALE Logical False
TFINAL.SEU Real
T0 Real s

1344 ATLAS User’s Manual

SINGLEEVENTUPSET Statements

Parameter Type Default Units

TC Real s
TF Real s
UNIFORM Logical False

Description

A1, A2, A3, and A4 The first set of parameters for the length dependence of the charge
generation pulse.
B1, B2, B3, and B4 The second set of parameters for the length dependence of the charge
generation pulse.
B.DENSITY Specifies the number of electron-hole pairs per unit volume or generated
charge in pico Coulumbs per micron if the PCUNITS parameter is specified.
BEAM.RADIUS This is the radius of the beam where the generation rate is maintained
constant. Beyond this point the generation will decay by either an
exponential or by a Gaussian function.
DENSITY Specifies the number of electron-hole pairs per unit volume generated along
the alpha particle track.
DEVICE Specifies which device the SINGLEEVENTUPSET statement applies to in
MixedMode. The synonym for this parameter is STRUCTURE.
ENTRYPOINT Specifies the X, Y, and Z coordinates of the beginning of the alpha particle
track. The specified point should belong to the semiconductor region.
EXITPOINT Specifies the X, Y, and Z coordinates of the end of the alpha particle track.
The specified point should belong to the semiconductor region.
F.SEU Specifies the name of C-Interpreter function describing the SEU
generation rate as a function of position and time.
PCUNITS Sets the units of B.DENSITY to be pC per micron.
RADIALGAUSS Specifies the Gaussian radial dependence of the charge generation pulse. By
default, the exponential dependence is used.
RADIUS Specifies the radius of the alpha particle track.
RESCALE This causes nodal generation rates to be scaled by the ratio of the integral of
the analytic generation rate divided by the numerically integrated value.
STRUCTURE This is a synonym for DEVICE.
TFINAL.SEU Specifies the finish time for the track, this defaults to the finish time of the
transient simulation if TFINAL.SEU is not specified.

1345 ATLAS User’s Manual

SINGLEEVENTUPSET Statements

T0 Specifies the peak in time of the charge generation pulse.

TC Specifies the width of the charge generation pulse.
UNIFORM Specifies that a uniform generation rate corresponding to that specified by
the DENSITY parameter is applied.

SEU Example
This statement specifies a track path, radius and density:
SINGLEEVENTUPSET ENTRYPOINT="1.5,2.0,0.0"\
EXITPOINT="1.0,1.0,4.0 RADIUS=0.05" \
DENSITY=1E18

Note: For user-defined Single Event Generation profiles, the C-I NTERPRETER function F.SEU on the
SINGLEEVENTUPSET statement can be used.

1346 ATLAS User’s Manual

SOLAR Statements

21.56 SOLAR
The SOLAR statement is used to simplify standard analysis of solar cells.
Syntax
SOLAR <parameters>

Parameter Type Default Units

ANODE Integer
AZIMUTH Real 0.0 Degrees
BEAM Integer
DV Integer 0.05 Volts
ELEVATION Real 90.0 Degrees
IA Character
IV Character
IW Character
LATTITUDE Real 35.0 Degrees
MAX.ANGLE Real 85.0 Degrees
MAX.WAVE Real microns
MIN.DV Real 0.002 Volts
MIN.WAVE Real microns
RTO Character
STEP.ANGLE Real 5.0 Degrees
STEP.WAVE Real microns
VANODE Real 0.5 Volts

1347 ATLAS User’s Manual

SOLAR Statements

Description

ANODE Specifies the electrode index of the "anode". The "anode" is the electrical contact
where positive voltages are applied to extract the standard current-voltage
characteristic of a solar cell. If the ANODE parameter is left unspecified, the simulator
will search for a contact named "anode".
AZIMUTH Specifies the solar cell azimuth angle (angle with respect to due south) for orbital
simulation of annual energy collection.
BEAM Specifies the beam index of the beam used as a solar source. If the BEAM parameter is
left unspecified, the simulator will automatically choose the first beam defined on a
BEAM statement.

DV Specifies the initial voltage step increment for extraction of the standard IV
characteristic.
ELEVATION Specifies the solar cell elevation angles (angle from the horizon) for orbital
simulation of annual energy collection.
IA Specifies the name of a file for collection of efficiency versus angle data. You must
specify the IA parameter to perform the efficiency versus angles extraction.
IV Specifies the name of a file for collection of the standard current-voltage
characteristics of the solar cell. You must specify the IV parameter to perform
standard current voltage extraction including the extraction of Voc, Isc, Pmax, and FF.
IW Specifies the name of a file for collection of solar cell efficiency versus wavelength.
If you specify this parameter, you must also specify the MIN.WAVE, MAX.WAVE, and
STEP.WAVE parameters.

LATTITUDE Specifies the lattitude of the solar cell for orbital simulation of the annual energy
collection.
MAX.ANGLE Specifies the maximum angle for angular efficiency collection (see IA). This value
must be less than 90° and greater than 0°.
MAX.WAVE Specifies the maximum wavelength for collection of spectral response of the solar
cell (see IW).
MIN.DV Specifies minimum voltage step at open circuit conditions for the extraction of
standard solar cell current-voltage characteristics (see IV).
MIN.WAVE Specifies the minimum wavelength for collection of spectral response of the solar cell
(see IW).
RTO Specifies the name of a file where the solar cell run-time output will separately be
written.

1348 ATLAS User’s Manual

SOLAR Statements

STEP.ANGLE Specifies the angle step in degrees for angular efficiency collection (see IA).
STEP.WAVE Specifies the wavelength step for collection of spectral response of the solar cell
(see IW).
VANODE Specifies the anode operating voltage for extraction of angular response (see IA) or
spectral response (see IW) or both. If the current-voltage characteristics are extracted
(see IV) the voltage at maximum power, Pmax, will be used as the operating voltage.

1349 ATLAS User’s Manual

SOLVE Statements

21.57 SOLVE
SOLVE instructs ATLAS to perform a solution for one or more specified bias points.
Syntax
SOLVE [<ion>] <dc> [<fp>][<ep>][<tp>][<ac>][<photo>] [<thermal>]

Parameter Type Default Units

AC.ANALYSIS Logical False

AFINAL Real 85.0 Degrees
ANAME Character
ANGLE Real 0.0 Degrees
ASCII Logical False
ASTEP Real 5.0 Degrees
AUTO Logical True
B<n> Real 0.0 W/cm2
BEAM Integer
BRANIN Logical True
C.IMAX Real 10-4 A

C.IMIN Real -10-4 A

C.VSTEP Real 0.0 V

C.VMAX Real 5.0 V
C.VMIN Real -5.0 V
CNAME Character
COMPLIANCE Real
CONTINUE Logical False
CURVETRACE Logical False
CYCLES Integer 1
CYCLIC.BIAS Logical False
CYCLIC.RELAX Real 0.2
CYCLIC.TOL Real 1.010-5
DDMS.LOG Logical False

1350 ATLAS User’s Manual

SOLVE Statements

NLAYERS Real 15
NOCONTACT Logical False
NOCURRENT Logical False
NOINTERFACE Logical False
NOISE.SS Logical False
NSAMP Integer 100
NSTEPS Integer 0
OFFSET.ANGLE Real 0.0 Degrees
ONEFILEONLY Logical See Description
OUT.FILE Character
OUTFILE Character
P.BIAS Real
PB1 Real V
PB2 Real V
PB3 Real V
PB4 Real V
PB5 Real V
PB6 Real V
PB7 Real V
PB8 Real V

1353 ATLAS User’s Manual

SOLVE Statements

Parameter Type Default Units

PIEZSCALE Real 1.0

POSITIVE Logical False
POWER<n> Real 

POWERFINAL Real 

PREVIOUS Logical False

PROJ.MOD Logical False
PROJECT Logical False
PULSE.WIDTH Real s
Q<n> Real C/m
QFACTOR Real 1.0 or previous value
QFINAL Real C/m
QSCV Logical False
QSTEP Real C/m
RAMPTIME Real s
RAMP.LIT Logical False
RELATIVE Logical False
SCAN.SPOT Integer
S.OMEGA Real 1.0
SINUAMP.COMP Real 0.0
SINUVAR.COMP Real 0.0
SONOS Logical False
SOR Logical False
SPECTRUM Character
SQPULSE Logical False
SP.LAST Logical False
SS.LIGHT Real 0.001 W/cm2
SS.PHOT Logical False
T<n> Real K
T.COMP Real 0 K

1354 ATLAS User’s Manual

SOLVE Statements

Parameter Type Default Units

T.SAVE Real 0.0 s

TABLE Character
TDELAY Real 0.0 s
TFALL Real 0.0 s
TRISE Real 0.0 s
TSAVE.MULT Real 1.0
TERMFINAL Real K
TERMINAL Integer all contacts
TIMESPAN Real  seconds
TOLERANCE Real 1.010-5
TRANS.ANALY Logical False
TSTOP Real
TWOFILESONLY Logical False
V<n> Real V
VFINAL Real V
VSTEP Real 0.0 V
VSS Real 0.1 KT/Q
WAVE<n> Logical False
WFINAL Real microns
WAVEFORMS Character
WSTEP Real microns

Description
Each SOLVE statement must specify an initial bias condition. Once any DC condition has been
solved, either a transient or AC analysis may be performed. You may also solve for carrier
generation due to incident light under DC, or AC analysis transient conditions.

dc This is one or more of the DC bias parameters

fp This is one or more of the file parameters
ep This is one or more of the initial guess or estimate parameters. Estimate parameters
are used to specify how the initial approximation for the solution is to be obtained.

1355 ATLAS User’s Manual

SOLVE Statements

EMAX and EMIN Specify the energy range for output of spectral data in the file named by the
SPECTRUM parameter.

IMPACT.I Specifies that impact ionization rate should be saved to the structure file. If
impact ionization is not enabled, these results will not be self-consistent.
INT.INCREMENT Specifies that the last character of OUTFILE will not be incremented for each
structure file. Instead an integer corresponding to the bias point number will be
appended to the OUTFILE along with .str.
IT.SAVE Specifies the name of an output file where the structure information will be
stored after every GUMMEL or NEWTON iteration. The ASCII code of the last
non-blank character of the supplied file name will be incremented by one for
each iteration, resulting in an unique file per iteration.
LMAX and LMIN Specify the wavelength range for spectral output in the file named by the
SPECTRUM parameter.

MASTER Specifies that the output file selected by the OUTFILE parameter will be
written in a standard structure file rather than in the older PISCES-II binary
format.
ONEFILEONLY Specifies that only one filename will be used to save solutions during a bias
ramp. If this parameter is specified, filename incrementing described under
OUTFILE is not applied. This parameter is true by default unless the NAME or
ELECTRODE parameters or transient simulations specified. When CURVETRACE
is used, you need to explicitly set this parameter to false using
^ONEFILEONLY to save a different filename at each bias point.

OUT.FILE This is an alias for OUTFILE.

OUTFILE Specifies the name of the output file where bias point solution information will
be stored. If an electrode is stepped so that more than one solution is generated
by the SOLVE statement, the ASCII code of the last non-blank character of the
supplied file name will be incremented by one for each bias point in
succession, resulting in a unique file per bias point. The alias for this
parameter is OUT.FILE.
SPECTRUM Specifies the name of an output file where LED spectral data are stored after
each step.

Note: The output file specified by OUTFILE has a limit of 132 characters on a UNIX system and eight characters
on a PC system.

1359 ATLAS User’s Manual

SOLVE Statements

Initial Guess Parameters

INITIAL Sets all voltages to zero. If this parameter is not specified for the first bias point in a
given structure a SOLVE INIT statement is automatically inserted. The SOLVE INIT
statement is always solved in zero carrier mode with no external elements attached. It is
used to provide a good initial guess to subsequent solutions.
LOCAL Specifies that the initial approximation should use local values of quasi-Fermi
levels. See Section 20.6 “Initial Guess Strategies” for more detailed information.
MLOCAL Specifies the modified local initial guess. This parameter is used when solved for carrier
temperatures in the energy balance models. If any energy balance model is specified,
MLOCAL defaults to true.

NOCURRENT Specifies an initial guess strategy that solves the non-linear Poisson equation and uses
this solution as the initial guess. This will typically give good results for situations
where the quasi-fermi levels are almost flat and the current is consequently low. Under
these conditions, you can solve almost any desired bias point directly.
PREVIOUS Specifies that the previous solution as the initial approximation.
PROJ.MOD Gives an alternative to the PROJ initial guess strategy in the case of one carrier type
solution by making a different update for the unsolved carrier concentration.
PROJ Specifies that an extrapolation from the last two solutions will be used as an initial
approximation. This parameter may be used if there are two or more existing
solutions and equivalent bias steps are taken on any electrodes that are changed.

Note: If no initial guess parameters are specified, ATLAS will use PROJ wherever possible.

QFACTOR Specifies as scaling factor to improve initial guesses when the QUANTUM model is used.
This parameter should be ramped slowly from zero to unity at the start of quantum effect
simulations.

Compliance Parameters
Compliance parameters define a current limit for a DC bias ramp or transient simulation. You
can terminate the simulation by monitoring the current and checking against a user-defined
limit.

COMPLIANCE Sets a limit on the current from the electrode which has been specified by the
CNAME or E.COMPLIANCE parameter. When the COMPLIANCE value is reached,
any bias ramp is stopped and the program continues with the next line of the
input file. The COMPLIANCE parameter is normally specified in A. If the GRAD
parameter is specified, COMPLIANCE is specified in A/V.
E.COMPLIANCE Specifies the electrode number to be used by the COMPLIANCE parameter. See
also CNAME.
CNAME Specifies the name of the electrode used by the COMPLIANCE.

1360 ATLAS User’s Manual

SOLVE Statements

GRAD Specifies that the compliance value is a current/voltage gradient, and not a
current value.
E.CHECK Specifies the electrode number for compliance testing for monotonicity,
positiveness or negativeness.
MONOTONIC Specifies that current in the electrode associated with the index specified on the
E.CHECK parameter must remain monotonic during a static ramp. If this
condition is violated, the solution ramp is immediately exited.
POSITIVE Specifies that current in the electrode associated with the index specified on the
E.CHECK parameter must remain positive during a static ramp. If this codition is
violated, the solution ramp is immediately exited.
NEGATIVE Specifies that current in the electrode associated with the index specified on the
E.CHECK parameter must remain negative during a static ramp. If this codition
is violated, the solution ramp is immediately exited.

Transient Parameters

CYCLES Specifies the number of periods to be simulated (both FREQUENCY and

TRANS.ANALY must be specified when this parameter is used). The
synonym for this parameter is PERIODS.
CYCLIC.BIAS Specifies that a cyclic bias [270] simulation is being performed. Cyclic
biasing allows the effects of repeated trapezoidal/square transient pulses (see
SQPULSE) on the output characteristics to be determined, without having to do
the extremely large numbers of cycles. When CYCLIC.BIAS is specified,
ATLAS will calculate a series of trapezoidal/square pulses using the SQPULSE
parameters. At the end of the third cycle (and at all subsequent cycles) the
values of the potential (V), electron concentration (n), hole concentration (p)
and trap probability of occupation (fT) will be modified according to the
equation:

x  k + 1  = x  k  – CYLIC · RELAX   x  k  – x  k – 1   dx  k    d x  h  – dx  k – 1   21-6

where x is a parameter such as V, n, p, or fT at every node point, x(k) is the updated value
calculated from the equation, k is the cycle number (with k+1 being the new update for
parameter x), CYCLIC.RELAX is the relaxation factor, x(k) is the difference between the
simulated values of x at the start and beginning of the current cycle, while x(k-1) is the
equivalent for the previous cycle.
Steady state cyclic convergence is determined by comparing the normalizing sum of the
updated values of V, n, p, and fT with a tolerance value (CYCLIC.TOL).

1361 ATLAS User’s Manual

SOLVE Statements

CYCLIC.RELAX Specifies the CYCLIC.BIAS relaxation factor (the recommended range for this
parameter to ensure stable convergence is between 0.2 and 1).
CYCLIC.TOL Specifies the CYCLIC.BIAS tolerance factor.
DECAY Specifies the time constant used for defining an exponential change in bias for
transient simulation.
ENDRAMP Applies any bias changes as linear ramps. ENDRAMP specifies the exact end of the
ramp in running time (i.e., the ramp will start at t=t0 and end at t=ENDRAMP).
NSTEPS This can be used to signal the end of the run (i.e., the final time would be t = t +
NSTEPS*DT). You can use this parameter instead of the TSTOP parameter.

PULSE.WIDTH Specifies the time constant used for sinusoidal non-linear time domain simulation.
If SQPULSE is specified, PULSE.WIDTH is the width of the trapezoidal/square pulse
not including the rise and fall times.
RAMPTIME Applies any bias changes as linear ramps. RAMPTIME specifies a ramp interval in
seconds (i.e., the ramp will begin at t=t0 and ends at t=t0 +RAMPTIME).

RELATIVE Specifies that the TSTOP value is relative to the current time value.
SINUAMP.COMP Specifies that sinusoidal amplitude compliance will be used. The transient
simulation will stop if the amplitude of a sinusoidal waveform is less than the
value of SINUAMP.COMP.
SINUVAR.COMP Specifies that sinusoidal variance compliance will be used. The transient
simulation will stop if the change in amplitude of a sinusoidal waveform is less
than SINUVAR.COMP.
SQPULSE Specifies that multiple trapezoidal/square transient pulses will be applied. The
pulse is controlled by the TDELAY, TRISE, PULSE.WIDTH, TFALL, and FREQUENCY
parameters.
T.COMP Specifies a temperature for temperature compliance. Once the specified
temperature is obtained at some location on the mesh the solution process is
discontinued.
T.SAVE Specifies a time increment at which the device structure files are saved.

1362 ATLAS User’s Manual

SOLVE Statements

Note: Actual time steps may not correspond to the user-specified increment.

TABLE Specifies the filename of a table of time and bias values for the
specified electrode. The following format must be used for the TABLE file.
time1 bias1
time2 bias2
time3 bias3
.... ....
end
The time value should be in seconds. The value of the bias should be in the
same units as displayed in ATLAS (i.e., Volts for a voltage controlled electrode,
A/um for a current controlled electrode, and A for a current controlled
electrode in DEVICE3D or ATLAS with the WIDTH parameter specified on the
MESH statement). There’s no limit to the number of time/bias pairs that can be
defined in the table file. The keyword, END, must be placed after the last time/
bias pair. Linear interpolation will be used to determine the bias for time points
that aren’t specified in the table file.

Note: An electrode must also be specified when using the TABLE parameter. The electrode can be specified
either by name (e.g., NAME=anode) or by number (e.g., ELECTRODE=1).

TDELAY Specifies the time delay before the first cycle of multiple trapezoidal/square
transient pulses (SQPULSE) will be applied.
TFALL Specifies the fall time for trapezoidal/square transient pulses (SQPULSE).
TRISE Specifies the rise time for trapezoidal/square transient pulses (SQPULSE).
TRANS.ANALY Specifies that transient analysis is being performed and that a sinusoid with a
frequency, specified by the FREQUENCY parameter, should be applied.
TSAVE.MULT Specifies a multiplier the save time increment, T.SAVE, is multiplied by after each
time a structure is saved.
TSTEP or DT Specifies the time-step to be used. For automatic time-step runs, DT is used to
select only the first time step (see Section 21.35 “METHOD”).
TSTOP Specifies the end of the time interval. If simulation begins at t=t0 and
RELATIVE is specified, it will end at t=TSTOP+t0.

WAVE<n> Specifies that the WAVEFORM, n (where n is between 1 and 10), will be enabled for
the current SOLVE statement. This parameter should appear on every SOLVE
statement where the specified WAVEFORM is to be applied.

1363 ATLAS User’s Manual

SOLVE Statements

Note: You can solve more than one WAVEFORM at the same time.

WAVEFORMS Specifies multiple WAVEFORMs to be enabled for the current SOLVE statement. The
WAVEFORM numbers should be specified as a comma separated list. For example,
SOLVE WAVEFORMS= “1, 2, 12” TFINAL=1e-6. This will enable WAVEFORM
numbers 1, 2, and 12.

AC Parameters

AC.ANALYSIS Specifies that AC sinusoidal small-signal analysis should be performed after

solving for the DC condition. The full Newton method must be used for this
analysis. This is typically specified with the statement:
METHOD NEWTON CARRIERS=2

ANAME Specifies the name of the electrode to which the AC bias will be applied. See also
TERMINAL. If no ANAME is specified, all electrodes have AC bias applied in turn.

AUTO Selects an automatic AC analysis method. This method initially uses SOR. The
DIRECT method will be used if convergence problems occur. We strongly
recommend the use of the AUTO parameter.
DIRECT Selects the direct AC solution method. This method is robust at all frequencies, but
slow.
FREQUENCY Specifies the AC analysis frequency. Analysis may be repeated at a number of
different frequencies (without solving for the DC condition again) by specifying
FSTEP. FREQUENCY can also be used to specify a sinusoidal or square pulse
frequency for transient simulations.
FSTEP Specifies a frequency increment which is added to the previous frequency. If
MULT.FREQ is specified, the frequency will be multiplied by FSTEP.

MAX.INNER Specifies the maximum number of SOR iterations.

SONOS Parameters

NIT.N Specifies the trapped electron concentration in a Silicon-Nitride region for the
steady-state SOLVE. This can be used in conjunction with the SONOS models or
independently from them.
NIT.P Specifies the trapped hole concentration in a Silicon-Nitride region for the steady-
state SOLVE. This can be used in conjunction with the SONOS models or
independently from them.
NIT.XMAX Maximum x-ccordinate for applying density specified by NIT.N or NIT.P.
NIT.XMIN Minimum x-coordinate for applying density specified by NIT.N or NIT.P.
NIT.YMAX Maximum y-ccordinate for applying density specified by NIT.N or NIT.P.
NIT.YMIN Minimum y-coordinate for applying density specified by NIT.N or NIT.P.
SONOS Only used in the case of SOLVE INIT with a SONOS structure. Helps prevent SOLVE
INIT introducing large carrier concentrations in a charged Silicon Nitride region.

Thermal3D Parameters

POWER<n> Specifies the power in watts on the n-th heat source.

POWERFINAL Specifies the final power for a linearly ramped heat source.
T<n> Specifies the temperature in K on the n-th heat sink.
TEMPFINAL Specifies the final temperature for a linearly ramped heat sink.

For more information about these parameters, see also Section 17.4 “Obtaining Solutions In
THERMAL3D”.
DC Conditions Example
The following statement solves for a defined bias point and saves the solution to output file,
OutA. The voltages on electrodes other than the gate will keep the value from the previous
SOLVE statement.
SOLVE VGATE=0.1 OUTFILE=OutA

1368 ATLAS User’s Manual

SOLVE Statements

Bias Stepping Example

In the next example, the bias stepping is illustrated. The two SOLVE statements produce the
following bias conditions.

Bias Point Vgate Vdrain Vsub

1 0.0 0.5 -0.5

2 1.0 0.5 -0.5
3 2.0 0.5 -0.5
4 3.0 0.5 -0.5
5 4.0 0.5 -0.5
6 5.0 0.5 -0.5

The solutions for these bias points will be saved to the files: OUT1, OUTA, OUTB, OUTC, OUTD,
and OUTE. The initial approximation for each bias point is obtained directly from the
preceding solution.
For bias points 4, 5, and 6, the program will use a PROJ to obtain an initial approximation.
Since, starting with bias point 4, both of its preceding solutions (bias points 2 and 3) only had
the same electrode bias (number 1) altered.
SOLVE Vdrain=.5 Vsub=–.5 OUTF=OUT1
SOLVE Vgate=1 VSTEP=1 VFINAL=5 NAME=gate OUTF=OUTA
Transient Simulation Example
The following sequence is an example of a time dependent solution. The METHOD statement
specifies second-order discretization, automatic time-step selection, and an automated
Newton-Richardson procedure.
The first SOLVE statement then computes the solution for a device with 1V on the base
electrode and 2V on the collector in steady-state. The second SOLVE statement specifies that
the base electrode is to be ramped to 2V over a period of 10 ns and is left on until 25 ns. Each
solution is written to a file. The name of the file is incremented in a manner similar to that
described for a DC simulation (UP1, UP2, and so on). Note that an initial time step had to be
specified in this statement.
The third SOLVE statement ramps ther base down from 2V to - 0.5V in 20 ns (end of ramp is
at =45 ns). The device is then solved at this bias for another 55 ns (out to 100 ns). Each
solution is again saved in a separate file (DOWN1, DOWN2, and so on).
No initial timestep was required since one had been estimated from the last transient solution
from the previous SOLVE statement.
Finally, the fourth SOLVE statement performs the steady-state solution at Vbe = -0.5V and
Vce = 2V.
METHOD 2ND TAUTO AUTONR
SOLVE Vbase=1 Vcollector=2

1369 ATLAS User’s Manual

SOLVE Statements

SOLVE Vbase=2 DT=1E-12 TSTOP=25E-9 RAMPTIME=10E-9 OUTF=UP1

SOLVE Vbase= –0.5 TSTOP=100E-9 RAMPTIME=20E-9 OUTF=DOWN1
SOLVE Vbase= –0.5 Vcollector=2
AC Analysis Example
The following example illustrates an application of the SOLVE statement for AC analysis. It is
assumed that the device has three electrodes. This analysis is being performed after DC
conditions are solved at V1 = 0V, 0.5V, 1.0V, 1.5V, and 2.0V. A series of 10mV AC signals
with frequencies of 1 MHz, 10 MHz, 100 MHz, 1 GHz, 10 GHz, and 100 GHz are applied to
each electrode in the device.
Ninety AC solutions will be performed (5 X 6 X 3 = 90).
SOLVE V1=0 V2=0 V3=0 VSTEP=0.5 NSTEPS=4 ELECT=1 \
AC FREQ=1E6 FSTEP=10 MULT.F NFSTEP=5 VSS=0.01
Photogeneration Examples
The following statement simulates two DC optical beams incident on the device with optical
spot powers of 15 and 25 W/cm2.
SOLVE B1=15 B3=25
The next example shows how DC spot power of the two optical beams can be stepped
simultaneously. Beam #1 will increase to 35 W/cm2 and beam #3 will increase to 45 W/cm2.
SOLVE LIT.STEP=5.0 NSTEP=4
In the next example, the beams are ramped in the time domain. Beam #1 is ramped down to 0
W/cm2 and beam #3 is ramped up to 100 W/cm2. The duration of the ramp is 1 ns. After the
ramp, simulation will continue for 4 ns.
SOLVE B1=0 B3=100 RAMP.LIT TSTEP=2E-11 RAMPTIME=1E-9 TSTOP=5E-9
Next, the small signal response of a single beam is analyzed. First, ATLAS will solve the DC
characteristics at the specified optical spot powers. Then, the AC response of beam #1 will be
calculated at a frequency of 10 MHz.
SOLVE B1=10 B3=20 BEAM=1 FREQUENCY=1e7 SS.PHOT SS.LIGHT=0.01
Frequency stepping is used to look at the small signal AC frequency response of one of the
beams. AC response is calculated at frequencies from 1kHz to 100MHz at each decade. The
MULT.F parameter is used to geometrically increase the frequency.
SOLVE B1=10 B3=5 BEAM=1 SS.PHOT SS.LIGHT=0.01 \
MULT.F FREQUENCY=1E3 FSTEP=10 NFSTEP=6
The following illustrates the specification of analysis of angular response of beam number 2
at angles from 0 to 80° in 1° increments.
SOLVE BEAM=2 ANGLE=0 AFINAL=80 ASTEP=1
Similarly, the following illustrates the specification of analysis of spectral response over the
wavelength range from 1 to 3 microns in steps of 0.05 microns.
SOLVE BEAM=1 LAMBDA=1 WFINAL=3 WSTEP=0.05

1370 ATLAS User’s Manual

SOLVE Statements

Ionization Integral Example

Ionization integrals are used to estimate the breakdown voltage from analysis of the electric
field. They can be used in zero carrier mode providing much faster simulation that
conventional breakdown analysis. The ionization syntax uses the parameter IONIZ to enable
the ionization integral calculation. An equipotential contour is calculated at a potential
DELTAV from the contact NAME (or INAME). Electric field lines are started at distances along
this potential contour. IONLINES sets the number of lines and LRATIO is the ratio of the
distance between the starting points of the lines. The following syntax is for a PMOS
transistor. See the on-line examples for other cases using ionization integrals.
IMPACT SELB
METHOD CARR=0
SOLVE VDRAIN=-1 VSTEP=-1 VFINAL=-20.0 NAME=DRAIN \
IONIZ IONLINES=50 LRATIO=0.9 DELTAV=1.2

Note: There are over 300 on-line examples supplied with ATLAS to provide examples of sequences of SOLVE
statements applied to practical problems for a variety of device technologies.

1371 ATLAS User’s Manual

SPX.MESH, SPY.MESH, SPZ.MESH Statements

21.58 SPX.MESH, SPY.MESH, SPZ.MESH

S<n>.MESH specifies the location of grid lines along the <n>-axis in a rectangular mesh used
in self-consistent Schrodinger-Poisson Model simulation (see Section 13.2 “Self-Consistent
Coupled Schrodinger Poisson Model”). The syntax is equivalent for all directions. SPX.MESH
and SPY.MESH should always be set or nor set together, while SPZ.MESH is always optional.
Syntax
SPX.MESH NODE=<n> LOCATION=<n>
SPX.MESH SPACING=<n> LOCATION=<n>
SPY.MESH NODE=<n> LOCATION=<n>
SPY.MESH SPACING=<n> LOCATION=<n>

Parameter Type Default Units

LOCATION Real m
NODE Integer
SPACING Real m

Description

Note: The mesh defined in these statements for the Self-Consistent Schrodinger-Poisson Model is entirely
separate from the electrical device simulation mesh defined on the MESH statement.

Setting Locally Fine Grids Example

This example shows how to space grid lines closely around a topological feature such as a
junction at y=0.85 microns.
SPY.MESH LOC=0.0 SPAC=0.2
SPY.MESH LOC=0.85 SPAC=0.01
SPY.MESH LOC=2 SPAC=0.35

1372 ATLAS User’s Manual

SYMBOLIC Statements

21.59 SYMBOLIC
Note: In versions 3.0 and greater, the SYMBOLIC statement is no longer needed. All functions have been moved
to the METHOD statement.

1373 ATLAS User’s Manual

SPREAD Statements

21.60 SPREAD
SPREAD distorts rectangular grids defined by ATLAS in the vertical direction to follow
surface and junction contours.

Note: The use of this parameter is not recommended by SILVACO.

Syntax
SPREAD LEFT|RIGHT WIDTH=<r> UPPER=<i> LOWER=<i>
Y.LOWER|THICKNESS [<options>]

Parameter Type Default Units

ENCROACH Real 1.0

GRADING Real 1.0
GR1 Real 1.0
GR2 Real 1.0
LEFT Logical False
LOWER Integer
MIDDLE Integer Halfway between UPPER and LOWER
RIGHT Logical False
THICKNESS Real m
UPPER Integer
VOL.RATIO Real 0.44
WIDTH Real m
Y.LOWER Real m
Y.MIDDLE Real 0.50 m

Description
SPREAD can reduce the grid complexity of specific simulations. Since the SPREAD statement is
somewhat complicated, we suggest that you follow the supplied examples carefully until
you’re confident to understand the workings of this statement.

1374 ATLAS User’s Manual

SPREAD Statements

Mandatory Parameters

LEFT Distorts the left side of the grid. If LEFT is specified, RIGHT must not be specified.
LOWER Specifies the lower y-grid line above which distortion will take place.
RIGHT Distorts the right side of the grid. If RIGHT is specified, LEFT must not be specified.
THICKNESS Specifies the thickness of the distorted region. Unless VOL.RATIO is set to 0 or 1,
THICKNESS will usually move the positions of both the UPPER and LOWER grid lines.
The Y.LOWER and THICKNESS parameters define the distorted grid region. Only one of
these parameters should be specified.
UPPER Specifies the upper y-grid line under which distortion will take place.
WIDTH Specifies the width from the left or right edge (depending on whether LEFT or RIGHT is
selected) of the distorted area. The actual X coordinate specified by WIDTH (min[x]+
WIDTH for LEFT, max[x] + WIDTH for RIGHT) will lie in the middle of the transition
region between distorted and undistorted grid regions.
Y.LOWER Specifies the physical location in the distorted region to which the line specified by
LOWER will be moved. The line specified by UPPER is not moved. The Y.LOWER and
THICKNESS parameters define the distorted grid region. Only one of these parameters
should be specified.

Optional Parameters

ENCROACH Defines the abruptness of the transition between a distorted and non-distorted grid.
The transition region becomes more abrupt with smaller ENCROACH factors (the
minimum is 0.1).

Note: Depending on the characteristics of the undistorted grid, long, thin, or obtuse triangles may result if too low
an ENCROACH value is used.

GRADING Specifies a grid ratio which produces a non-uniform grid in the distorted region.
This parameter is identical to the RATIO parameter in the X.MESH and Y.MESH
statements.
GR1 and GR2 These may be used instead of the GRADING parameter. GR1 and GR2 may be
specified in conjunction with MIDDLE (y grid line) and Y.MIDDLE (location) so
that GR1 specifies grading in the spread region from UPPER to MIDDLE, and GR2
specifies grading from MIDDLE to LOWER.

1375 ATLAS User’s Manual

SPREAD Statements

MIDDLE Specifies the y-grid line that serves as a boundary between the grading specified
by GR1 and the grading specified by GR2.
VOL.RATIO Specifies the ratio of the downward displacement of the lower grid line to the
net increase in thickness. The default (0.44) should be used for oxide-silicon
interfaces. VOL.RATIO is ignored if Y.LOWER is specified.
Y.MIDDLE Specifies the physical location in the distorted grid to which the line specified by
MIDDLE will be moved.

Examples
This example spreads a uniform 400 Å of oxide to 1000 Å on the left side of the device. This
increases oxide thickness by 600 Å. Because VOL.RATIO is not specified, the default (0.44) is
used. Therefore, 0.44 X 600 = 264 Å of the net increase will lie below the original 400 Å and
0.56600 = 336 Å of the net increase will lie above the original 400 Å. The width of the
spread region is 0.5 m, and the oxide taper is quite gradual because of the high
encroachment factor. The grid is left uniform in the spread region.
# *** Mesh definition ***
MESH NX=30 NY=20 RECT
X.M N=1 L=0
X.M N=30 L=5
Y.M N=1 L=-.04
Y.M N=5 L=0
Y.M N=20 L=1 R=1.4
# *** Thin oxide ***
REGION IY.H=5 OXIDE NUM=1
# *** Silicon substrate ***
REGION IY.L=5 SILICON NUM=2
# *** Spread ***
SPREAD LEFT W=0.7 UP=1 LO=4 THICK=0.1 MID=2 Y.MID=0.05
In the second example, the right side of the grid is distorted in order to follow a junction
contour. The initial grid is assumed to be above. Y.LOWER is used so that there is no increase
in the size of the device, just grid redistribution. When Y.LOWER is set to the junction, choose
the ENCROACH parameter so that the lower grid line (LOWER=10) follows the junction as
closely as possible. The grid is graded so that grid lines are spaced closer together as they
approach the junction. Because the point specified by WIDTH lies in the middle of the
transition region, it should be chosen to be slightly larger than the width of the doping box.
# *** Doping ***
DOPING UNIFORM N.TYPE CONC=1E15
DOPING GAUSS P.TYPE X.LEFT=1.5 X.RIGHT=2 \
PEAK= CONC=1.e19 RATIO=.75 JUNC=0.3
# *** Spread ***
SPREAD RIGHT W=0.7 UP=1 LO=4 THICK=0.3 MID=2 Y.MID=0.10

1376 ATLAS User’s Manual

SYSTEM Statements

21.61 SYSTEM
SYSTEM allows execution of any UNIX command within an input file

Note: The SYSTEM statement is executed by DECKBUILD and is fully documented in the DECKBUILD
USER ’S MANUAL. To enable SYSTEM command, select Category: Options in the DeckBuild Main
Control menu.

Examples
The following command will remove all files test*.str before a SOLVE statement where
the OUTF parameter is used.
system \rm -rf test*.str
SOLVE .... OUTF=test0
The system command and the UNIX commands are case sensitive.
UNIX commands may be concatenated on a single line using the semicolon (;) operator. For
example to run a third party program that reads and writes Silvaco format files with fixed
names input.str and output.str.
SAVE OUTF=mysave.str
system mv mysave.str input.str; source myprog.exe; mv output.str
myrestart.str
EXTRACT INIT INF=myrestart.str
The UNIX re-direct symbol > is not supported by the system command. The UNIX echo
and sed syntax can be used instead to output values or variables to a given filename. For
example to save the extracted value of variable $myvariable to the file myfile.
system echo $”myvariable” | sed -n “ w myfile”

1377 ATLAS User’s Manual

THERMCONTACT Statements

21.62 THERMCONTACT
THERMCONTACT specifies the position and properties of thermal contacts. This statement must
be used when lattice heating solutions are specified using GIGA or GIGA3D.
Syntax
THERMCONTACT NUMBER=<n> <position> [EXT.TEMPER=<n>] [ALPHA=<n>]

Parameter Type Default Units

ALPHA Real W/(cm2·K)

BOUNDARY Logical True
DEVICE Character
ELEC.NUMBER Integer
EXT.TEMPER Real 300 K
F.CONTEMP Character
NAME Character
NUMBER Integer 1
STRUCTURE Character
X.MAX Real Right side of structure m
X.MIN Real Left side of structure m
Y.MAX Real Bottom of structure m
Y.MIN Real Top of structure m
Z.MAX Real Back m

Z.MIN Real Front m

1378 ATLAS User’s Manual

THERMCONTACT Statements

Description
At least one THERMCONTACT statement must be specified when simulating lattice heating
effects (MODELS LAT.TEMP). The THERMCONTACT statement must appear in the input deck
before any METHOD statement.

position This is a set of the position parameters described below. Use either the X.MIN,
X.MAX, Y.MIN, and Y.MAX parameters to specify the exact position of the contact or
the ELEC.NUMBER parameter to specify an electrode number that coincides with the
thermal contact.
NUMBER Specifies a thermal contact number from 1 to 20. Contact numbers should be
specified in increasing order. This parameters must be specified on all
THERMCONTACT statements.

ALPHA Specifies the reverse value of thermal resistance (a=1/RTH).

EXT.TEMPER Specifies the external temperature. The synonym for this parameter is
TEMPERATURE.

F.CONTEMP Specifies the name of a file containing a C-Interpreter function describing the
contact temperature as a function of time.

Position Parameters

BOUNDARY Specifies which parts of the thermal contact the thermal boundary conditions are
applied at. See Section 7.2.7 “Thermal Boundary Conditions” for a full
description.
NAME Specifies an electrode name that the thermal contact is coincident with. See
Section 21.15 “ELECTRODE” for more information.
DEVICE Specifies which device in MIXEDMODE simulation the THERMCONTACT
statement applies to. The synonym for this parameter is STRUCTURE.
ELEC.NUMBER Specifies an electrode number that the thermal contact is coincident with.
STRUCTURE This is a synonym for DEVICE.
X.MAX Specifies the right edge of the contact.
X.MIN Specifies the left edge of the contact.
Y.MAX Specifies the bottom edge of the contact.
Y.MIN Specifies the top edge of the contact.
Z.MAX Specifies the location of the rear edge of the thermal contact.
Z.MIN Specifies the location of the front edge of the thermal contact.

1379 ATLAS User’s Manual

THERMCONTACT Statements

Coordinate Definition Example

A thermal contact is located where Y coordinate values range from 10 m to the bottom side
of the structure and X coordinate values range from the left edge of the structure to the right
edge of the structure (be default). The external temperature is set to 300K and a thermal
resistance of 1 is added. Thus, the temperature at y = 10 m will be greater than 300K once
lattice heating effects occur.
THERMCONTACT NUM=1 Y.MIN=10 EXT.TEMP=300 ALPHA=1
Setting Thermal and Electrical Contacts Coincident Example
The next statement line creates a thermal contact at the location of electrode #4. An external
temperature of 400K is specified.
THERMCONTACT NUM=2 ELEC.NUM=4 EXT.TEMP=400

Note: Location and Parameters of thermal contacts are not stored in the ATLAS solution files. Therefore,
THERMCONTACT statements must be defined in each ATLAS run involving lattice heating.

1380 ATLAS User’s Manual

TITLE Statements

21.63 TITLE
TITLE specifies the title (up to 29 characters) that will appear in the standard output. If used,
the TITLE command should be the first statement in the ATLAS input file.
Syntax
TITLE <string>
Example
This example causes the text, *** CMOS p-channel device ***, to be printed at the top of
all ATLAS printouts and screen displays.
TITLE *** CMOS p-channel device ***

Note: TITLE cannot be used with the automatic ATHENA to ATLAS interface feature of DECKBUILD

1381 ATLAS User’s Manual

TONYPLOT Statements

21.64 TONYPLOT
TONYPLOT starts the graphical post-processor TONYPLOT

Note: The TONYPLOT statement is executed by DECKBUILD and is fully described in the DECKBUILD
USER ’S MANUAL.

Examples
All graphics in ATLAS is performed by saving a file and loading the file into TONYPLOT. The
TONYPLOT command causes ATLAS to automatically save a structure file and plot it in
TONYPLOT. The TonyPlot window will appear displaying the material boundaries. Use the
Plot:Display menu to see more graphics options.
This command will display the file, myfile.str.
tonyplot -st myfile.str
This command will overlay the results of myfile1.str and myfile2.str.
tonyplot -overlay myfile1.str myfile2.str

Note: For documentation of the extensive features of TONYPLOT for graphical display and analysis, see
TONYPLOT USER ’S MANUAL.

1382 ATLAS User’s Manual

TRAP Statements

21.65 TRAP
TRAP activates bulk traps at discrete energy levels within the bandgap of the semiconductor
and sets their parameter values.
Syntax
TRAP DONOR|ACCEPTOR E.LEVEL=<r> DENSITY=<r> DEGEN=<v>
<capture parameters> REGION|NUMBER

Parameter Type Default Units

ACCEPTOR Logical False

DEGEN.FAC Real undefined
DENSITY Real cm-3
DEVICE Character
DONOR Logical False
E.LEVEL Real eV
F.DENSITY Character
FAST Logical False
MIDGAP Logical False
MATERIAL Character
NUMBER Integer 0
REGION Integer 0
SIGN Real cm2
SIGP Real cm2
STRUCTURE Character
TAT.TRAP Logical False
TAUN Real s
TAUP Real s
X.MAX Real Maximum X coord µm
X.MIN Real Minimum X coord µm
Y.MAX Real Maximum Y coord µm

1383 ATLAS User’s Manual

TRAP Statements

Parameter Type Default Units

Y.MIN Real Minimum Y coord µm

Z.MAX Real Maximum X coord µm
Z.MIN Real Minimum Z coord µm

Description

ACCEPTOR Specifies an acceptor-type trap level.

DEGEN.FAC Specifies the degeneracy factor of the trap level used to calculate the density.
DENSITY Sets the maximum density of states of the trap level.
DEVICE Specifies which device the statement applies to in MIXEDMODE simulation. The
synonym for this parameter is STRUCTURE.
DONOR Specifies a donor-type trap level.
E.LEVEL Sets the energy of the discrete trap level. For acceptors, E.LEVEL is relative to the
conduction band edge. For donors, it depends on the valence band edge.
F.DENSITY Specifies the name of a file containing a C-Interpreter function describing the
density of donor/acceptor traps as a function of position.
FAST Specifies that the trap densities are static even during transient simulation (i.e.,
they reach equilibrium instantaneously).
MATERIAL Specifies which material from the table in Appendix B “Material Systems” will
apply to the TRAP statement. If a material is specified, then all regions defined as
being composed of that material will be affected.
MIDGAP Specifies that the energy of the discrete trap be set to the middle of the
bandgap.
NUMBER This is the synonym for REGION.
REGION Specifies which region the traps apply to. If unspecified, the traps apply to all
regions.
STRUCTURE This is a synonym for DEVICE.
TAT.TRAP Causes the trap level to be used in the ITAT/RTAT model. The trap must be located
in an insulator or wide bandgap semiconductor quantum barrier.
X.MAX Specifies the maximum X coordinate of a box where traps are to be applied.
X.MIN Specifies the minimum X coordinate of a box where traps are to be applied.
Y.MAX Specifies the maximum Y coordinate of a box where traps are to be applied.

1384 ATLAS User’s Manual

TRAP Statements

Y.MIN Specifies the minimum Y coordinate of a box where traps are to be applied.
Z.MAX Specifies the maximum Z coordinate of a box where traps are to be applied.
Z.MIN Specifies the minimum Z coordinate of a box where traps are to be applied.

Capture Parameters
Either the cross section or lifetime parameters should be used to define the capture
parameters.

EC
E.level for acceptor trap
Eta

Etd
E.level for donor trap
EV

Figure 21-8: Acceptor and Donor Trap Energy Levels

Multiple Trap Level Definition Example
The following example sets three discrete trap levels within the silicon bandgap. These trap
levels will capture carriers, slowing the switching speed of any device. In this example the
capture cross sections are used to define the properties of each trap.
trap e.level=0.49 acceptor density=2.e15 degen=12 \
sign=2.84e-15 sigp=2.84e-14
trap e.level=0.41 acceptor density=1.e15 degen=12 \
sign=7.24e-16 sigp=7.24e-15
trap e.level=0.32 donor density=1.e15 degen=1 \
sign=1.00e-16 sigp=1.00e-17

1385 ATLAS User’s Manual

TRAP Statements

Note: For distributed trap levels, see Section 21.10 “DEFECTS”. For interface traps, see Section 21.25
“INTTRAP”. Also, spatial TRAP distributions in X, Y, or Z directions must be defined in the DOPING
statement.

1386 ATLAS User’s Manual

UTMOST Statements

21.66 UTMOST
UTMOST converts data from a ATLAS logfile format into an UTMOST logfile format.

Note: UTMOST 12.3.4 or later can read ATLAS format logfiles directly in batch mode. This statement is obsolete
and its use is not recommended.

Syntax
UTMOST BIP|DIODE|MOS <input> OUTFILE=<fn> WIDTH=<n> 
[<elec>] <cntrl> 
[TRANSLATE INFILE1=<filename>][<analysis>] [<parasites>]

Parameter Type Default Units

AC Logical False
ANODE Integer 1
APPEND Logical False
BASE Integer
BIP Logical False
BULK Integer
CATHODE Integer
COLLECTOR Integer
DEVICE Integer 1
DIODE Logical False
DRAIN Integer
EMITTER Integer
GATE Integer
INFILE Character
INTERNAL Logical False
LENGTH Real 1.0 m
MESFET Logical False
MINSTEP Real 0.01 V
MOS Logical False
OUTFILE Character
POLARITY Integer 1 (n-type)

1387 ATLAS User’s Manual

UTMOST Statements

Parameter Type Default Units

ROUTINE Integer 1
SOURCE Integer
TEMPERATURE Real 300 K
WELL Integer
WIDTH Real 1.0 m

Description
Specify a technology parameter (BIP, DIODE, or MOS), an input file (INFILE1=), and an output
file (OUTFILE=). You can also specify one or more of the optional electrode parameters.

INFILE This is used to convert up to nine ATLAS logfiles into a single UTMOST
logfile. The ATLAS logfiles must be specified in the form:
INFILE1=<fn> INFILE2=<fn> INFILE3=<fn> . . .
where fn is the name of the logfile that you wish to convert.
OUTFILE Specifies the name of a file where I-V data will be stored in an UTMOST file
format.
APPEND Specifies that I-V data should be appended to the output file specified by
OUTFILE2.

WIDTH This is used to specify the width of the device. Electrode current is multiplied by
the value of WIDTH before being saved in the logfile.

Technology Parameters

BIP Creates an UTMOST bipolar transistor log file.

DIODE Creates an UTMOST diode log file.
MOS Creates an UTMOST MOSFET log file.
MESFET Creates an UTMOST MESFET log file.

Electrode Parameters
Different electrode parameters may be specified with different technologies.

BIP Technology
BASE Specifies the base electrode number.
COLLECTOR Specifies the collector electrode numb.

1388 ATLAS User’s Manual

UTMOST Statements

EMITTER Specifies the emitter electrode numb.

POLARITY Indicates the device type. POLARITY=1 specifies npn. POLARITY=-1
specifies pnp.
MOS and MESFET Technologies

Note: If you have used the NAME parameter of the ELECTRODE statement to assign standard electrode names
(bulk, drain, gate, and source), you don’t need to re-specify these names in the UTMOST statement.

BULK or SUBSTRATE Specifies the bulk electrode number.

DRAIN Specifies the drain electrode number.
GATE Specifies the gate electrode number.
POLARITY Indicates the device type. POLARITY=1 specifies nmos. POLARITY=-1
specifies pmos.
SOURCE Specifies the source electrode number.

Control Parameters
The optional control parameters are used to specify what type of information will be
converted to an UTMOST logfile.

AC Specifies that input logfiles contain AC parameters and that the UTMOST
routine numbers refer to the UTMOST capacitance routines.
DEVICE Specifies the device (or row) number used to identify different devices in
UTMOST. The synonym for this parameter is ROW.
INTERNAL Specifies that internal contact voltage will be used instead of applied bias.
LENGTH Specifies the length of the device.
MINSTEP Specifies the minimum voltage step between data points.
TEMPERATURE Specifies the simulation temperature.
ROUTINE Specifies which UTMOST routine number data will be saved for. The following
routines are supported.

1389 ATLAS User’s Manual

UTMOST Statements

BIP Technology

ROUTINE=1 Specifies IC vs. VCE curves. Each input file holds I-V data for a solution with a
fixed base current and VCE stepped. (UTMOST IC/VCE routine).

ROUTINE=9 Specifies a BF vs. VBE plot. Each input file holds I-V data for a solution with
constants VCE and VBE stepped. (UTMOST BF vs. IC routine)

ROUTINE=10 Specifies a BF vs. VCE plot. Each input file holds I-V data for a solution with
constants VCE and VBC stepped. (UTMOST BR routine)

ROUTINE=14 Specifies a IC,IB vs. VBE (Gummel) plot. Each input file holds I-V data for a
solution with constant VCE and VBE stepped. (UTMOST gummel routine)

ROUTINE=15 Specifies a IE,IB vs. VBC (Reverse Gummel) plot. Each input file holds I-V data
for a solution with constants VCE and VBC stepped. (UTMOST rgummel routine)

ROUTINE=29 Specifies IE vs. VEC plot. Each input file holds I-V data for a solution with a fixed
base current and VEC stepped. (UTMOST IE/VEC routine)

MOS Technology

ROUTINE=1 Specifies a ID vs. VDS plot. VB is constant and VG is stepped. (UTMOST ID/VD-
VG routine)
ROUTINE=2 Specifies a ID vs. VGS plot. VD is constant and VB is stepped. (UTMOST ID/VG-
VB routine)
ROUTINE=3 Specifies the LEFF, RSD routine. Two devices of different lengths are needed. The
ID/VGS data should be used.

ROUTINE=5 Specifies a ID vs. VGS plot. VD is constant and VB=0. (UTMOST VTH routine)

ROUTINE=9 Specifies a ID vs. VGS plot. VD is constant and VB is stepped. (UTMOST NSUB
routine).
ROUTINE=10 Specifies the IS routine. The forward diode characteristics of the drain to bulk
junction are required.
ROUTINE=11 Specifies the LAMBDA routine. An ID/VD curve is required.

ROUTINE=12 Specifies the ETA routine. A set of ID/VGS data for different VDS is required.

1390 ATLAS User’s Manual

UTMOST Statements

ROUTINE=13 Specifies the VMAX routine. A set of ID/VGS curves for small changes in VDS are
required.
ROUTINE=14 Specifies the U0 routine. A set of ID/VGS data is required.

ROUTINE=26 Specifies a ID vs. VGS plot. VB is constant and VD is stepped. (UTMOST ID/VG-
VD routine).

MOS Capacitances
The following routines may be specified if both MOS and AC parameters are selected.

ROUTINE=1 Specifies the CGSO routine. CGS vs. VGS data is required.

ROUTINE=2 Specifies the CGDO routine. CGD vs. VGS data is required.

ROUTINE=5 Specifies the CJ routine. CBD vs. VDS data is required.

ROUTINE=6 Specifies the CJSW routine. CBD vs. VDS data is required.

ROUTINE=12 Specifies the CJ/CJSW routine. CGD vs. VGS data is required for two different size
drain areas.

Diode Technology

ROUTINE=1 Specifies an I vs. V plot. V is stepped. (UTMOST ID vs. VD routine).

Parameter Type Default Units

AMPLITUDE Real 0
BEAM Real
CONTINUE Logical True
DECAY Real 0
ELEC.NAME Character
ELEC.NUMBER Real 0
F.WAVE Character
FREQUENCY Real 0 Hz
NUMBER Real 0
OFFSET Logical False
PERIODS Real
PHASE Real 0 Deg
PWL Character
SINUSOID Logical False

Description

AMPLITUDE Specifies the sinusoidal amplitude.

BEAM Specifies the optical BEAM number to which the WAVEFORM parameters will
apply.
CONTINUE Specifies that the waveform will continue (i.e. retain the phase) across mulitple
SOLVE statements. This is the default.

DECAY Specifies an optional decay value.

ELEC.NAME Specifies the electrode name.
ELEC.NUMBER Specifies the electrode number (either ELEC.NAME or ELEC.NUMBER must be
specified).

1395 ATLAS User’s Manual

WAVEFORM Statements

F.WAVE Specifies the C-Interpreter file. The function prototype is:

int wave(double time, double dt, double
*bias)
where time is the transient time, dt is the time step and bias is the bias
calculated by the C-Interpreter function.
FREQUENCY Specifies the frequency.
NUMBER Specifies the waveform number. This must be specified on each WAVEFORM
statement.
OFFSET Specifies that a phase offset is to be used. If you use a non-zero phase in a
sinusoidal definition, then SOLVE statement will go immediately to this bias.
For example, if PHASE=90, AMPLITUDE=1 and the DC bias is 1V, the transient
SOLVE statement will start at 2V and vary from 2V to 0V. If you specify
OFFSET, then the bias will vary from 1V to -1V.

PERIODS Specifies the maximum number of sinusoidal periods (default is unlimited).

PHASE Specifies the phase.
PWL Specifies a list of time and bias value points. The time values must be in
ascending order (e.g., PWL=”1e-9,1 2e-9, 2 3e-9, 3”). This specifies that
the bias with change to 1V at 1ns, 2V at 2ns, and 3V at 3ns.
SINUSOID Specifies that a sinusoidal waveform will be generated according to:
AMPLITUDE*exp(-time*DECAY)
*sin(2*PI*(FREQUENCY*time+PHASE/360.0))

1396 ATLAS User’s Manual

WAVEGUIDE Statements

21.69 WAVEGUIDE
WAVEGUIDE defines geometry and physical models for a stand-alone optical mode simulation,
which is done by solving the Helmholtz equation. For information about WAVEGUIDE, see
Section 8.3 “WAVEGUIDE Statement”.
Syntax
WAVEGUIDE <parameters>

Parameter Type Default Units

DISP.NAME Character dispersion.log

DISPERSION Logical False
E.INIT Real -999 eV
E.FINAL Real -999 eV
FAR.NX Integer 100
FAR.NY Integer 100
LX.MAX Real 999 m

LX.MIN Real -999 m

LY.MAX Real 999 m

LY.MIN Real -999 m

NDISP Integer 10
NMODE Integer 1
NB.ITEREIG Integer 4
NEAR.NX Integer 100
NEAR.NY Integer 100
OMEGA Real -999 1/s
OMEGA.FINAL Real -999 1/s
OMEGA.INIT Real -999 1/s
PHOTON.ENERGY Real -999 eV
PRT.ITEREIG Logical False
REFLECT Logical False
V.HELM Logical False

1397 ATLAS User’s Manual

WAVEGUIDE Statements

Description

DISP.NAME Specifies the log file name for waveguide dispersion.

E.INIT Sets the lower photon energy boundary for dispersion calculation.
E.FINAL Sets the upper photon energy boundary for dispersion calculation.
FAR.NX Sets the number of X-samples for far-field pattern.
FAR.NY Sets the number of Y-samples for far-field pattern.
LX.MAX Sets the upper X boundary of solution domain for vector Helmholtz solver.
LX.MIN Sets the lower X boundary of solution domain for vector Helmholtz solver.
LY.MAX Sets the upper Y boundary of solution domain for vector Helmholtz solver.
LY.MIN Sets the lower Y boundary of solution domain for vector Helmholtz solver.
NB.ITEREIG Sets the ratio of basis size of iterative Helmholtz eigensolver to the number of
requested transverse modes, given by parameter NMODES.
NDISP Sets the number of photon energy samples for dispersion calculation.
NEAR.NX Sets the number of X-samples for near-field pattern.
NEAR.NY Sets the number of Y-samples for near-field pattern.
NMODE Sets the number of requested optical modes in vector Helmholtz solver.
OMEGA Sets the photon frequency in vector Helmholtz solver.
OMEGA.INIT Sets the lower frequency boundary for dispersion calculation.
OMEGA.FINAL Sets the upper frequency boundary for dispersion calculation.
PHOTON.ENERGY Sets the photon energy in vector Helmholtz solver.
PRT.ITEREIG Specifies that iterative Helmholtz eigensolver prints information about its
internal parametrs and convergence.
REFLECT Specifies that the structure is mirror-symmetrical around x=0 axis.
V.HELM Specifies that vector Helmholtz solver will be used for solution for transverse optical
modes.

1398 ATLAS User’s Manual

Appendix A
C-Interpreter Functions
Overview C-Interpreter Functions

A.1 Overview
ATLAS has a built-in C language interpreter that allows many of the models contained in
ATLAS to be modified. To use the SILVACO C-INTERPRETER, (SCI), you must write C
language functions containing analytic descriptions of the model. The C-INTERPRETER uses
the ANSI standard definition of C. If you are not familiar with the C language, we suggest that
you refer to any of the popular C language references such as the one written by Kernighan
and Ritchie [134]. Additional information about the C-INTERPRETER can be found in the
SILVACO C-INTERPRETER MANUAL.
The function arguments of the C-INTERPRETER functions and return values are fixed in
ATLAS. Make sure that the arguments and the return values for the functions match those
expected by ATLAS. To help you, this release of ATLAS includes a set of templates for the
available functions. The C-INTERPRETER function template can be obtained by typing:
atlas -T filename
where filename is the name of a file where you want the template to be copied. You can also
get the C-INTERPRETER function template by accessing the template file through DECKBUILD.
This template file should be copied and edited when implementing user-defined C-
INTERPRETER functions. To use the C-INTERPRETER function, specify the file name containing
this function in the appropriate ATLAS statement. The relevant statement names and
parameters are listed in the template file. The template function prototypes are defined in the
include file template.in.
The following example shows how the C-INTERPRETER function munsat can be used to
describe velocity saturation for electrons. First, examine the template file and find the
template for munsat. The template should look something like this:
/*
* Electron velocity saturation model.
* Statement: MATERIAL
* Parameter: F.MUNSAT
* Arguments:
* tl lattice temperature (K)
* na acceptor concentration (per cc)
* nd donor concentration (per cc)
* e electric field (V/cm)
* v saturation velocity (cm/s)
* mu0 low field mobility (cm^2/Vs)
* *mu return: field dependent mobility (cm^2/Vs)
* *dmde return: derivative of mu with e
*/
int munsat(double tl, double na, double nd, double e,
double v, double mu0, double *mu, double *dmde)
{
return(0); /* 0 - ok */
}

1400 ATLAS User’s Manual

Overview C-Interpreter Functions

Here, we see that the function is passed the lattice temperature, the acceptor and donor
concentrations, the electric field, the saturation velocity, and the low field mobility. The
function returns the field dependent mobility and the derivative of mobility with respect to
electric field.

Note: It is important to properly calculate and return derivative values when specified, since the convergence
algorithms in ATLAS require these derivatives.

The return value is an error flag. In this case, it is set to zero indicating that the function was
successful (0=OK, 1=fail).
In this example, the C-INTERPRETER implements a standard model for velocity saturation. This
particular model is already built into ATLAS. If you want to compare the user-defined version
with the standard version, you can activate the built-in model by setting the B.ELECTRONS
parameter to 1.
To implement the model, two lines additional lines of C code in the body of the function are
specified as follows:
/*
* Electron velocity saturation model.
* Statement: MATERIAL
* Parameter: F.MUNSAT
* Arguments:
* tl lattice temperature (K)
* na acceptor concentration (per cc)
* nd donor concentration (per cc)
* e electric field (V/cm)
* v saturation velocity (cm/s)
* mu0 low field mobility (cm^2/Vs)
* *mu return: field dependent mobility (cm^2/Vs)
* *dmde return: derivative of mu with e
*/

int munsat(double tl, double na, double nd, double e,

double v, double mu0, double *mu, double *dmde)
{
*mu = mu0/(1.0+mu0*e/v);
*dmde = (*mu)*(*mu)/v;
return(0); /* 0 - ok */
}

1401 ATLAS User’s Manual

Overview C-Interpreter Functions

The function will then need to be stored in a file. For example, the function can be stored in
the file: test.lib. The function is then introduced into a specific example by specifying the
file name on the F.MUNSAT parameter in the MATERIAL statement as follows:
MATERIAL F.MUNSAT="test.lib"
In this case, you will also need to activate the calculation of the field dependent mobility by
adding the FLDMOB parameter to a MODELS statement. For example:
MODELS FLDMOB
When the input deck is executed, your C-INTERPRETER functions will be used in place of the
built in function. When trying this example, place PRINT statements (using the printf C
command) in the function to check that the function is working correctly.
Table A-1 shows a complete list of all the interpreter functions available in ATLAS.

Table A-1 Complete list of available C-Interpreter functions in ATLAS

Statement Parameter Template Description

BEAM F.IMAGE image () Specifies 2D relative intensity

versus offset at the source
perpendicular to the direction of
propogation for ray tracing in
LUMINOUS3D.

BEAM F.RADIATE radiate () Generation rate as a function of

position.
BEAM F3.RADIATE radiate3 () Generation rate as a function of
position (3D).
BEAM F.REFLECT reflect () Reflection coefficient.
DEFECT F.TFTACC tftacc () TFT acceptor trap density as a
function of energy.
CONTACT F.ETUNNEL fetunnel () Schottky contact electron
tunneling.
CONTACT F.HTUNNEL fhtunnel () Schottky contact hole tunneling.
CONTACT F.WORKF workf () Electrical Field dependent
workfunction.
DEFECT F.TFTDON tftdon () TFT donor trap density as a
function of energy.
DEGREDAT F.NTA devdeg_nta () Interface acceptor trap density as
a function of location.
DEGREDAT F.NTD devdeg_ntd () Interface donor trap density as a
function of location.
DEGREDAT F.SIGMAE devdeg_sigmae () Interface electron capture cross-
section.

1402 ATLAS User’s Manual

Overview C-Interpreter Functions

Table A-1 Complete list of available C-Interpreter functions in ATLAS

Statement Parameter Template Description

DEGREDAT F.SIGMAH devdeg_sigmah () Interface hole capture cross-

section.
DOPING F.COMPOSIT composition () Position dependent composition
fractions.
DOPING F.DOPING doping () Position dependent net doping.
DOPING F3.DOPING doping3 () Position dependent net doping.
IMPACT F.IIAN iian() Electron impact ionization
coefficient as a function of
composition, temperature, and
electric field.
IMPACT F.IIAP iiap() Hole impact ionization
coefficient as a function of
composition, temperature, and
electric field.
INTERFACE F.QF int_fixed_charge () Interface fixed charge as a
function of position.
INTERFACE F.HTE.N hten() Electron heterojunction
thermionic emission current.
INTERFACE F.HTE.P htep() Hole heterojunction thermionic
emission current.
INTERFACE F.HTFE.N htfen() Electron heterojunction
thermionic field emission delta
term.
INTERFACE F.HTFE.P htfep() Hole heterojunction thermionic
field emission delta term.
INTTRAP F.DENSITY int_acc_trap () Interface ACCEPTOR trap density
as a function of position.
INTTRAP F.DENSITY int_don_trap () Interface DONOR trap density as a
function of position.
INTRAP F.MSCRATES msctransitionrates() Enables you to specify the
transition rates between all
internal states of a multistate trap.
IMPACT F.EDIIN ediin () Electron temperature coefficients
for impact ionization.
IMPACT F.EDIIP ediip () Hole temperature coefficients for
impact ionization.

1403 ATLAS User’s Manual

Overview C-Interpreter Functions

Table A-1 Complete list of available C-Interpreter functions in ATLAS

Statement Parameter Template Description

LASER F.MIRROR mirror Mirror facet reflectivities on a

function of wavelength.
LENS F.LENS lenslet() Lenslet surface as a function of
position.
MATERIAL F.BANDCOMP bandcomp () Temperature and composition
dependent band parameters.
MATERIAL F.BBT bbt () Electric field and carrier
concentration dependent band-to-
band generation rate.
MATERIAL F.BGN bgn () Composition, temperature, and
doping dependent band gap
narrowing.
MATERIAL F.CBDOSFN cbdosfn Electron temperature dependent
effective conduction band density
of states.
MATERIAL F.CONMUN conmun () Composition, temperature,
position, and doping dependent
electron mobility.
MATERIAL F.CONMUP conmup () Composition, temperature,
position, and doping dependent
hole mobility.
MATERIAL F.COPT copt () Radiative recombination rate as a
function of composition and
temperature.
MATERIAL F.EPSILON epsilon () Composition and temperature
dependent permittivity.
MATERIAL F.FERRO ferro () Position and field dependent
permittivity.
MATERIAL F.GAUN gaun () Electron Auguer rate as a
function of composition and
temperature.
MATERIAL F.GAUP gaup () Hole Auger rate as a function of
composition and temperature.
MATERIAL F.INDEX index () Wavelength dependent complex
index of refraction.
MATERIAL F.MNSNDIFF mnsndiff () Microscopic noise source for
electron diffusion noise.

1404 ATLAS User’s Manual

Overview C-Interpreter Functions

Table A-1 Complete list of available C-Interpreter functions in ATLAS

Statement Parameter Template Description

MATERIAL F.MNSPDIFF mnspdiff () Microscopic noise source for

hole diffusion noise.
MATERIAL F.MNSNFLICKER mnsnflicker () Microscopic noise source for
electron flicker noise.
MATERIAL F.MNSPFLICKER mnspflicker () Microscopic noise source for
hole flicker noise.
MATERIAL F.MUNSAT munsat () Electron velocity saturation
model.
MATERIAL F.MUPSAT mupsat () Hole velocity saturation model.
MATERIAL F.NKEG nkeg () Specifies the complext index of
refraction as a function of band
gap.
MATERIAL F.PDN fphonondragn() Adds a phonon drag term to the
electron thermopower in GIGA.
Function of lattice temperature
and electron density.
MATERIAL F.PDP fphonondragp() Adds a phonon drag term to the
hole thermopower in GIGA.
Function of lattice temperature
and hole density.
MATERIAL F.RECOMB recomb () Position, temperature, and
concentration dependent
recombination.
MATERIAL F.TARUP taurp () Energy dependent hole relaxation
time.
MATERIAL F.TAUN taun () Position, temperture, and doping
dependent electron SRH lifetime.
MATERIAL F.TAUP taup () Position, temperture, and doping
dependent hole SRH lifetime.
MATERIAL F.TAURN taurn () Energy dependent electron
relaxation time.
MATERIAL F.TCAP TCAP() Lattice Thermal Capacity as a
function of lattice temperature,
position, doping, and fraction
composition.

1405 ATLAS User’s Manual

Overview C-Interpreter Functions

Table A-1 Complete list of available C-Interpreter functions in ATLAS

Statement Parameter Template Description

MATERIAL F.TCOND TCOND() Lattice Thermal Conductivity as

a function of lattice temperature,
position, doping, and fraction
composition.
MATERIAL F.VSATN vsatn () Composition and temperature
dependent electron saturation
velocity.
MATERIAL F.VSATP vsatp () Composition and temperature
dependent hole saturation
velocity.
MATERIAL F.VBDOSFN vbdosfn Hole temperature dependent
effective valence band density
of states.
MATERIAL/ F.ENMUN endepmun Electron mobility as a function
MOBILITY of electron temperature and a
few other choice variables.
MATERIAL/ F.ENMUP endepmup Hole mobility as a function of
MOBILITY hole temperature and a few other
choice variables.
MATERIAL/ F.LOCALMUN localmun Electron mobility with an explicit
MOBILITY position dependence.
MATERIAL/ F.LOCALMUP localmup Hole mobility with an explicit
MOBILITY position dependence.
MATERIAL/ F.TOFIMUN tofimun Electron mobility with a
MOBILITY dependence on both parallel and
perpendicular electric field.
MATERIAL/ F.TOFIMUP tofimup Hole mobility with a dependence
MOBILITY on both parallel and
perpendicular electric field.
MODELS F.ALPHAA bulk_absorb() Bulk absorption coefficient
versus carrier density and photon
energy.
MODELS F.BFIELD magneticfield() Position dependent magnetic
field vector (BX,BY,BZ).
MODELS F.GENSINGLET gensinglet() Independent generation of singlet
excitons.

1406 ATLAS User’s Manual

Overview C-Interpreter Functions

Table A-1 Complete list of available C-Interpreter functions in ATLAS

Statement Parameter Template Description

MODELS F.GENTRIPLET gentriplet() Independent generation of triplet

excitons.
MODELS F.KSN fksn() Energy dependent electron
Scattering Law Exponent.
MODELS F.KSP fksp() Energy dependent hole Scattering
Law Exponent.
MODELS F.PDN fphonondragn() Adds a phonon drag term to the
electron thermopower in GIGA.
Function of lattice temperature
and electron density.
MODELS F.PDP fphonondragp() Adds a phonon drag term to the
hole thermopower in GIGA.
Function of lattice temperature
and hole density.
MODELS F.TRAPTUNNEL ftraptunnel Energy and electric field
dependent Dirac enhancement
factor.
MODELS F.TRAP.COULOMBIC ftrapcoulombic Energy and electric field
dependent Coulombic
enhancement factor.
SINGLEEVENT F.SEU seu() Time and position dependent
SEU generation.
TRAP F.DENSITY acc_trap() ACCEPTOR trap density as a
function of position.
TRAP F.DENSITY don_trap() DONOR trap density as a function
of position.
LASER F.LGAIN lgain() Laser gain as a function of
position, composition, carrier
concentration, quasi-Fermi level
and current.
LASER F.OMODE omode() Laser transverse optical mode
distribution as a function of
position.
MODELS F.PCM pcm() Phase change material resistivity
as a function of time step,
temperature, previous
temperature, and previous
resistivity.

1407 ATLAS User’s Manual

Appendix B
Material Systems
Overview Material Systems

B.1 Overview
ATLAS understands a library of materials for reference to material properties and models of
various regions in the semiconductor device. These materials are chosen to represent those
most commonly used by semiconductor physicists today. BLAZE or DEVICE3D users will have
access to all of these materials. S-PISCES users will have only access to Silicon and
Polysilicon.
S-PISCES is designed to maintain backward compatibility with the standalone program,
SPISCES2 Version 5.2. In the SPISCES2 syntax, certain materials can be used in the REGION
statement just by using their name as logical parameters. This syntax is still supported.

1409 ATLAS User’s Manual

Semiconductors, Insulators, and Conductors Material Systems

B.2 Semiconductors, Insulators, and Conductors

All materials in ATLAS are strictly defined into three classes: semiconductor materials,
insulator materials, or conductors. Each class of materials has particular properties.

B.2.1 Semiconductors
All equations specified by your choice of models are solved in semiconductor regions. All
semiconductor regions must have a band structure defined in terms of bandgap, density of
states, affinity, and so on. The parameters used for any simulation can be echoed to the run-
time output using MODELS PRINT. For complex cases with mole fraction dependent models,
these quantities can be seen in TONYPLOT by specifying OUTPUT BAND.PARAM and saving a
solution file.
Any semiconductor region that is defined as an electrode is then considered to be a conductor
region. This is typical for polysilicon gate electrodes.

B.2.2 Insulators
In insulator materials, only the Poisson and lattice heat equations are solved. Therefore for
isothermal simulations, the only parameter required for an insulator is dielectric permittivity
defined using MATERIAL PERM=<n>.
Materials usually considered as insulators (e.g., SiO2) can be treated as semiconductors using
BLAZE, however all semiconductor parameters are then required.

B.2.3 Conductors
All conductor materials must be defined as electrodes, and all electrode regions are defined as
conductor material regions. If a file containing regions of a material known to be a conductor
are read in, these regions will automatically become un-named electrodes. As noted below, if
the file contains unknown materials, these regions will become insulators.
During electrical simulation, only the electrode boundary nodes are used. Nodes that are
entirely within an electrode region are not solved. Any quantities seen inside a conductor
region in TONYPLOT are spurious. Only optical ray tracing and absorption for LUMINOUS and
lattice heating are solved inside of conductor/electrode regions.

B.2.4 Unknown Materials

If a mesh file is read containing materials not in Table B-1, these will automatically become
insulator regions with a relative permittivity of 3.9. All user-defined materials from
ATHENA, regardless of the material name chosen, will also become such insulator materials.

1410 ATLAS User’s Manual

Semiconductors, Insulators, and Conductors Material Systems

B.2.5 Specifying Unknown or User-Defined Materials in ATLAS

As mentioned previously, all materials in ATLAS are classified as a semiconductor, an
insulator, or a conductor. These classes are termed user groups. In order to correctly define a
new material in ATLAS, you must specify the name of the material, the user group it belongs
to, and the known ATLAS material it is to take as a default material. Once you set these three
elements up in their appropriate places in the input deck, you can change the specific
properties of them using MATERIAL statements as you would normally do.
A user groups takes one of the following definitions:
user.group=semiconductor
user.group=insulator
user.group=conductor
The name of the material is specified using:
user.material=material_name
The default material that the new material is to be based on is specified using:
user.default=known_atlas_material_name
As an example of the syntax required to define a new material, we will define a new material
called my_oxynitride. For simplicity, we will omit all the other code for the complete
ATLAS deck and only include code for deck structure purposes.
Defining a new insulator called my_oxynitride.
go atlas

mesh
.
.
.
region num=1 x.min=1 x.max=2 y.min=0 y.max=2
user.material=my_oxynitirde

electrode ...

doping ...

material material=my_oxynitride user.group=insulator

user.default=oxide permittivity=9

.
.
.

1411 ATLAS User’s Manual

Semiconductors, Insulators, and Conductors Material Systems

After creating your new material in ATLAS, it is advised to include a print on your MODELS
statement to echo the parameter values used for the material.

1412 ATLAS User’s Manual

ATLAS Materials Material Systems

B.3 ATLAS Materials

This section lists the materials in ATLAS. The superscripted numbers (e.g., Silicon(1)) on
these tables refer to the notes on the next page.

Table B-1 The ATLAS Materials

Single Element Semiconductors

Silicon(1) Poly(2) Germanium Diamond

Binary Compound Semiconductors

GaAs(3) GaP CdSe SnTe

SiGe InP CdTe ScN

SiC-6H InSb HgS GaN
SiC-3C InAs HgSe AlN
SiC-4H
AlP ZnS HgTe InN
AlAs ZnSe PbS BeTe
AlSb ZnTe PbSe ZnO
GaSb CdS PbTe IGZO
Ternary Compound Semiconductors

AlGaAs GaSbP InAlAs GaAsP

InGaAs GaSbAs InAsP HgCdTe
InGaP InGaN AlGaN CdZnTe
InAlP InGaSb InAlSb AlGaSb
InAsSb GaAsSb AlAsSb InPSb
AlPSb AlPAs AlGaP AlInN
Quaternary Compound Semiconductors

InGaAsP AlGaAsP AlGaAsSb InAlGaN

InGaNAs InGaNP AlGaNAs AlGaNP
AlInNAs AlInNP InAlGaAs InAlGaP
InAlAsP InPAsSb InGaAsSb InAlAsSb
CuInGaSe2
Insulators

Vacuum Oxide Nitride Si3N4

1413 ATLAS User’s Manual

ATLAS Materials Material Systems

Table B-1 The ATLAS Materials

Air SiO2 SiN Sapphire
Ambient
HfO2 HfSiO4 OxyNitride
Al2O3 BSG BPSG TiO
Ta2O5 TiO2 ZrO2 SnO2

La2O3

Conductors(4)

Polysilicon(2) Palladium TiW TaSi

Aluminum Cobalt Copper PaSi

Gold Molybdenum Tin PtSi
Silver Lead Nickel MoSi
AlSi Iron WSi ZrSi
Tungsten Tantalum TiSi AlSi
Titanium AlSiTi NiSi Conductor
Platinum AlSiCu CoSi Contact
ITO In2O3
TiON GST GaSbTe PCM
Organics

Organic Pentacene Alq3

TPD PPV Tetracene
NPB IGZO

Notes 
1: The material models and parameters of Silicon are identical to those of S-Pisces2 Version 5.2. Be aware that although
these band parameters may be physically inaccurate compared to bulk silicon measurements, most other material
parameters and models are empirically tuned using these band parameters.

2: Polysilicon is treated differently depending on how it is used. In cases where it’s defined as an electrode, it’s treated as
a conductor. It can also be used as a semiconductor such as in a polysilicon emitter bipolars. 

3: The composition of SiGe is the only binary compound that can be varied to simulate the effects of band gap variations.

4: Conductor names are only associated with electrodes. They are used for the specification of thermal conductivities and
complex index of refraction and for display in TONYPLOT.

1414 ATLAS User’s Manual

ATLAS Materials Material Systems

B.3.1 Specifying Compound Semiconductors Rules

The rules for specifying the order of elements for compound semiconductors are derived from
the rules used by the International Union of Pure and Applied Chemistry [195]. These rules
are:
1. Cations appear before anions.
2. When more than one cation is present the order progresses from the element with the
largest atomic number to the element with the smallest atomic number.
3. The order of anions should be the in order of the following list: B, Si, C, Sb, As, P, N, H,
Te, Se, S, At, I, Br, Cl, O, and F.
4. The composition fraction x is applied to the cation listed first.
5. The composition y is applied to the anion listed first.
To accommodate popular conventions, there are several exceptions to these rules (see Table
B-2).

Table B-2 Exceptions to the Specifying Compound Semiconductor Rules

Material Description

SiGe The composition fraction x applies to the Ge component. SiGe is then

specified as Si(1-x)Ge(x), an exception to rule #4.

AlGaAs This is specified as Al(x)Ga(1-x)As. This is an exception to rule #2.

InGaAsP (system) The convention In(1-x)Ga(x)As(y)P(1-y) as set forth by Adachi [1] is used.
This is an exception to rule #4.
InAlAs The convention In(1-x)Al(x)As as set forth by Ishikawi [123] is used.
This is an exception to rule #4.
InAlGaN Here we use the convention InyAlxGa(1-x-y)N. This is an exception to rule
#4.
HgCdTe Here we use the convention Hg(1-x)C(x)dTe. This is an exception to rule
#4.
InAlN Here we use the convention In(1-x)Al(x)N. This is an exception to rule #4.

1415 ATLAS User’s Manual

Silicon and Polysilicon Material Systems

B.4 Silicon and Polysilicon

The material parameters defaults for Polysilicon are identical to those for Silicon. The
following paragraphs describe some of the material parameter defaults for Silicon and
Polysilicon.

Note: A complete description is given of each model Chapter 3 “Physics”. The parameter defaults listed in that
chapter are all Silicon material defaults.

Table B-3 Band parameters for Silicon and Poly

Eg300  Nc300 Nv300 

Material a b
eV per cc per cc eV

Silicon 1.08 4.7310-4 636.0 2.81019 1.041019 4.17

Poly 1.08 4.7310-4 636.0 2.81019 1.041019 4.17

Table B-4 Static dielectric constants for Silicon and Poly

Material Dielectric Constant

Silicon 11.8
Poly 11.8

The default mobility parameters for Silicon and Poly are identical in all cases. The defaults
used depend on the particular mobility models in question. A full description of each mobility
model and their coefficients are given in Section 3.6.1 “Mobility Modeling”.
Table B-5 contains the silicon and polysilicon default values for the low field constant
mobility model.

Table B-5 Lattice Mobility Model Defaults for Silicon and Poly

Material MUN MUP TMUN TMUP

cm2/Vs cm2/Vs

Silicon 1000.0 500.0 1.5 1.5

Poly 1000.0 500.0 1.5 1.5

1416 ATLAS User’s Manual

Silicon and Polysilicon Material Systems

Table B-6 shows the silicon and polysilicon default values for the field dependent mobility
model.

Table B-6 Parallel Field Dependent Mobility Model Parameters for Silicon and Poly

Material BETAN BETAP

Silicon 2 1
Poly 2 1

Table B-7 shows the default values used in the bandgap narrowing model for Silicon and
Polysilicon.

Table B-7 Bandgap Narrowing Parameters for Silicon and Poly

Statement Parameter Defaults Units

MATERIAL BGN.E 6.9210-3 V

MATERIAL BGN.N 1.31017 cm-3

MATERIAL BGN.C 0.5 —

Table B-8 shows the default parameters for Schockley-Read-Hall (SRH) recombination.

Table B-8 SRH Lifetime Parameter Defaults for Silicon and Poly

Material TAUN0 (s) TAUP0 (s) NSRHN (cm-3) NSRHP (cm-3)

Silicon 1.010-7 1.010-7 5.01016 5.01016

Poly 1.010-7 1.010-7 5.01016 5.01016

The default parameters for Auger recombination are given in Table B-9.

Table B-9 Auger Coefficient Defaults for Silicon and Poly

Material AUGN AUGP

Silicon 2.810-31 9.910-32

Poly 2.810-31 9.910-32

1417 ATLAS User’s Manual

Silicon and Polysilicon Material Systems

Table B-10 shows the default values for the SELB impact ionization coefficients.

Table B-10 Impact Ionization Coefficients for Silicon and Poly

Parameter Value

EGRAN 4.0105
BETAN 1.0
BETAP 1.0
AN1 7.03105
AN2 7.03105
BN1 1.231106
BN2 1.231106
AP1 6.71105
AP2 1.582106
BP1 1.693106
BP2 2.036106

Table B-11 shows the default Richardson cofficients.

Table B-11 Effective Richardson Coefficients for Silicon and Poly

Material ARICHN (A/cm2/K2) ARICHP (A/cm2/K2)

Silicon 110.0 30.0

Poly 110.0 30.0

1418 ATLAS User’s Manual

The Al(x)Ga(1-x)As Material System Material Systems

B.5 The Al(x)Ga(1-x)As Material System

Table B-12 shows the default recombination parameters for AlGaAs.

Table B-12 Default Recombination Parameters for AIGaAs

Parameter Value Equation

TAUN0 1.010-9 3-331

TAUP0 1.010-8 3-331
COPT 1.510-10 3-365

AUGN 5.010-30 3-369

AUGP 1.010-31 3-369

Table B-13 shows the default values for the SELB impact ionization coefficients used for
GaAs. AlGaAs uses the same values as GaAs.

Table B-13 Impact Ionization Coefficients for GaAs

Parameter Value

EGRAN 0.0
BETAN 1.82
BETAP 1.75
EGRAN 0.0
AN1 1.889105
AN2 1.889105
BN1 5.75105
BN2 5.75105
AP1 2.215105
AP2 2.215105
BP1 6.57105
BP2 6.57105

The default values for the effective Richardson coefficients for GaAs are 6.2875 A/cm2/K2 for
electrons and 105.2 A/cm2/K2 for holes.

1419 ATLAS User’s Manual

The In(1-x)Ga(x)As(y)P(1-y) System Material Systems

B.6 The In(1-x)Ga(x)As(y)P(1-y) System

The default material thermal models for InGaAsP assumes lattice-matching to InP. The
material density is then given by
 = 4.791 + 0.575y.composition + 0.138y.composition
The specific heat for InGaAsP is given by
C p = 0.322 + 0.026y.composition – 0.008y.composition

The thermal resistivities of InGaAsP are linearly interpolated from Table B-14.

Table B-14 Thermal Resistivities for InGaAsP Lattice-Matched to InP

Composition Fraction y Thermal Resistivity (deg(cm/w)

0.0 1.47
0.1 7.05
0.2 11.84
0.3 15.83
0.4 19.02
0.5 21.40
0.6 22.96
0.7 23.71
0.8 23.63
0.9 22.71
1.0 20.95

The default thermal properties for the ternary compounds in the InGaAsP system:
In(1-x)Ga(x)As, In(1-x)Ga(x)P, InAs(y)P(1-y), and GaAs(y)P(1-y), are given as a function of
composition fraction by linear interpolations from these binary compounds.

1420 ATLAS User’s Manual

Silicon Carbide (SiC) Material Systems

B.7 Silicon Carbide (SiC)

Table B-15 shows the default values for the SELB impact ionization coefficients used for SiC.

Table B-15 Impact Ionization Coefficients for SiC

Parameter Value

EGRAN 0.0
BETAN 1.0
BETAP 1.0
AN1 1.66106
AN2 1.66106
BN1 1.273107
BN2 1.273107
AP1 5.18106
AP2 5.18106
BP1 1.4107
BP2 1.4107

Table B-16 shows the default thermal parameters used for both 6H-SiC and 3C-SiC.

Table B-16 Default Thermal Parameters for SiC

Parameter Value

3C-SiC 6H-SiC

TCA 0.204 0.385

1421 ATLAS User’s Manual

Material Defaults for GaN/InN/AlN System Material Systems

B.8 Material Defaults for GaN/InN/AlN System

Tables B-17 through B-20 show the default values for the band structure parameters for InN,
GaN and AlN for each of the user-defined parameter sets. To choose the parameters sets,
specify KP.SET1 (Table B-17), KP.SET2 (Table B-18), KP.SET3 (Table B-19), and KP.SET4
(Table B-20). By default, the parameters in Table B-17 are used.

Table B-17 Electronic Band-Structure Parameters for InN, GaN and AlN Using KP.SET1
(All Parameters Values are from [216] unless otherwise noted.)

Parameter Symbol Syntax Units InN GaN AlN

Electron eff. mass mcz MZZ m0 0.11 0.20 0.33

(z)
Electron eff. mass mct MTT m0 0.11 0.18 0.25
(t)
Hole eff. mass A1 A1 -9.24 -7.24 -3.95
param.
Hole eff. mass A2 A2 -0.60 -0.51 -0.27
param.
Hole eff. mass A3 A3 8.68 6.73 3.68
param.
Hole eff. mass A4 A4 -4.34 -3.36 -1.84
param.
Hole eff. mass A5 A5 -4.32 -3.35 -1.92
param.
Hole eff. mass A6 A6 -6.08 -4.72 -2.91
param.
Valence band Ev0 eV -1.59 -2.64 -3.44
reference
Direct band gap Eg(300) eV 1.89 3.42 6.28
(300K)
Spin-orbit split D1 DELTA1 eV 0.041 0.019 -0.164
energy
Crystal-field split D2 DELTA2 eV 0.0013 0.01413 0.01913
energy
Lattice  a0 ALATTICE Å 3.548 3.189 3.112
constant
Elastic  C33 C33 GPa 200 392 382
constant

1422 ATLAS User’s Manual

Material Defaults for GaN/InN/AlN System Material Systems

Table B-17 Electronic Band-Structure Parameters for InN, GaN and AlN Using KP.SET1
(All Parameters Values are from [216] unless otherwise noted.)
Elastic  C13 C13 GPa 94 100 127
constant
Hydrostatic ac AC eV -4.08 -4.08 -4.08
deformation
potential
Shear deform. D1 D1 eV -0.89 -0.89 -0.89
potential
Shear deform. D2 D2 eV 4.27 4.27 4.27
potential
Shear deform. D3 D3 eV 5.18 5.18 5.18
potential
Shear deform. D4 D4 eV -2.59 -2.59 -2.29
potential

Table B-18 Electronic Band-Structure Parameters for InN, GaN and AlN Using KP.SET2
(All Parameters Values are from [293] unless otherwise noted.)

Parameter Symbol Syntax Units InN GaN AlN

Electron eff. mass mcz MZZ m0 0.12 0.20 0.32

(z)
Electron eff. mass mct MTT m0 0.12 0.20 0.28
(t)
Hole eff. mass A1 A1 -8.21 -6.56 -3.95
param.
Hole eff. mass A2 A2 -0.68 -0.91 -0.27
param.
Hole eff. mass A3 A3 7.57 5.65 3.68
param.
Hole eff. mass A4 A4 -5.23 -2.83 -1.84
param.
Hole eff. mass A5 A5 -5.11 -3.13 -1.92
param.
Hole eff. mass A6 A6 -5.96 -4.86 -2.91
param.
Valence band Ev0 eV -1.59 -2.64 -3.44
reference

1423 ATLAS User’s Manual

Material Defaults for GaN/InN/AlN System Material Systems

Table B-18 Electronic Band-Structure Parameters for InN, GaN and AlN Using KP.SET2
(All Parameters Values are from [293] unless otherwise noted.)
Direct band gap Eg(300) eV 1.994 3.507 6.23
(300K)
Spin-orbit split D1 DELTA1 eV 0.041 0.019 -0.164
energy
Crystal-field split D2 DELTA2 eV 0.0013 0.01413 0.01913
energy
Lattice  a0 ALATTICE Å 3.548 3.189 3.112
constant
Elastic  C33 C33 GPa 224 398 373
constant
Elastic  C13 C13 GPa 92 106 108
constant
Hydrostatic ac AC eV -4.08 -4.08 -4.08
deformation
potential
Shear deform. D1 D1 eV -0.89 -3.0 -0.89
potential
Shear deform. D2 D2 eV 4.27 3.6 4.27
potential
Shear deform. D3 D3 eV 5.18 8.82 5.18
potential
Shear deform. D4 D4 eV -2.59 -4.41 -2.29
potential

Table B-19 Electronic Band-Structure Parameters for InN, GaN and AlN Using KP.SET3
(All Parameters Values are from [48] unless otherwise noted.)

InN
Parameter Symbol Syntax Units GaN AlN
[293]

Electron eff. mass mcz MZZ m0 0.12 0.20 0.32

(z)
Electron eff. mass mct MTT m0 0.12 0.20 0.28
(t)
Hole eff. mass A1 A1 -8.21 -6.56 -3.95
param. [293]

1424 ATLAS User’s Manual

Material Defaults for GaN/InN/AlN System Material Systems

Tables B-21 through B-23 show the default polarization parameters for InN, GaN and AlN for
each of the user-defined parameter sets. To choose the parameter sets, specify POL.SET1
(Table B-21), POL.SET2 (Table B-22), or POL.SET3 (Table B-23) on the MATERIAL
statement. By default, the parameter values in Table B-24 are used.

Table B-21 Polarization Parameters for InN, GaN and AlN Using POL.SET1
(All Parameters Values are from [216] unless otherwise noted.)

Parameter Symbol Syntax Units InN GaN AlN

Lattice  a0 ALATTICE Å 3.548 3.189 3.112

constant
Spontaneous Psp PSP C/m2 -0.042 -0.034 -0.09
polarization
Piezoelectric e33 E33 C/m2 0.81 0.67 1.5
const.(z)
Piezoelectric e31 E31 C/m2 -0.41 -0.34 -0.53
const.(x,y)

Table B-22 Polarization Parameters for InN, GaN and AlN Using POL.SET2
(All Parameters Values are from [293] unless otherwise noted.)

Parameter Symbol Syntax Units InN GaN AlN

Lattice  a0 ALATTICE Å 3.545 3.189 3.112

constant

1427 ATLAS User’s Manual

Material Defaults for GaN/InN/AlN System Material Systems

Table B-22 Polarization Parameters for InN, GaN and AlN Using POL.SET2
(All Parameters Values are from [293] unless otherwise noted.)
Spontaneous Psp PSP C/m2 -0.032 -0.029 -0.081
polarization
Piezoelectric e33 E33 C/m2 0.97 1.27 1.79
const.(z)
Piezoelectric e31 E31 C/m2 -0.57 -0.35 -0.5
const.(x,y)

Table B-23 Polarization Parameters for InN, GaN and AlN Using POL.SET3
(All Parameters Values are from [49] unless otherwise noted.)

Parameter Symbol Syntax Units InN GaN AlN

Lattice  a0 ALATTICE Å 3.54 3.189 3.112

constant
Spontaneous Psp PSP C/m2 -0.032 -0.029 -0.081
polarization
Piezoelectric e33 E33 C/m2 0.97 0.73 1.46
const.(z)
Piezoelectric e31 E31 C/m2 -0.57 -0.49 -0.6
const.(x,y)

1428 ATLAS User’s Manual

Material Defaults for Compound Semiconductors Material Systems

B.9 Material Defaults for Compound Semiconductors

Tables B-24a and B-24b, and B-24c show the material defaults for binary compound
semiconductors. Table B-25 shows Material defaults for ternary compound semiconductors.

Note: x.comp is set to 0.3. See Section 5.4 “Material Dependent Physical Models” for more information.

Table B-24a. Materials Defaults for binary compound semiconductors

Material GaAs GaP CdSe SnTe InP CdTe ScN
Epsilon 13.2 11.1 10.6 0 12.5 10.9 0
Eg (eV) 1.42 2.26 1.74 0.18 1.35 1.5 2.15
Chi (eV) 4.07 4.4 0 0 4.4 4.28 0
Nc (per cc) 4.35E+17 1.76E+18 1.18E+18 1 5.68E+17 1 1
Nv (per cc) 8.16E+18 8.87E+18 7.57E+18 1 8.87E+18 1 1
ni (per cc) 2.12E+06 0.409 7.22E+03 0.0308 1.03E+07 2.51E-13 8.70E-19
Gc 2 2 2 2 2 2 2
Gv 4 4 4 4 4 4 4
Ed (eV) 0.044 0.044 0.044 0.044 0.044 0.044 0.044
Ea (eV) 0.045 0.045 0.045 0.045 0.045 0.045 0.045
Lifetime (el) 1.00E-09 1 1 1 1 1 1
Lifetime (ho) 2.00E-08 1 1 1 1 1 1
Auger cn 5.00E-30 0 0 0 0 0 0
Auger cp 1.00E-31 0 0 0 0 0 0
Auger kn 0 0 0 0 0 0 0
Auger kp 0 0 0 0 0 0 0
Copt 1.50E-10 0 0 0 0 0 0
An** 6.29 20.4 15.6 1.40E-11 9.61 1.40E-11 1.40E-11
Ap** 105 60.1 54.1 1.40E-11 60.1 1.40E-11 1.40E-11
betan 1.82 1 1 1 1 1 1
betap 1.75 1 1 1 1 1 1
egran 0 0 0 0 0 0 0
an1 1.90E+05 7.30E+04 7.30E+04 7.03E+05 7.30E+04 7.03E+05 7.03E+05
bn1 5.75E+05 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06
an2 1.90E+05 7.30E+04 7.30E+04 7.03E+05 7.30E+04 7.03E+05 7.03E+05
an2 5.75E+05 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06
ap1 2.22E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05
bp1 6.57E+05 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06
ap2 2.22E+05 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06
bp2 6.57E+05 2.40E+05 2.40E+05 2.04E+06 2.40E+05 2.04E+06 2.04E+06
Vsatn (cm/s) : 7.70E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06
Vsatp (cm/s) : 7.70E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06
mun (cm^2/Vs) 8000 300 800 4600 1050
tmun 1 1.5 1.5 1.5 1.5
mup (cm^2/Vs) 400 100 150 100
tmup 2.1 1.5 1.5 1.5

1429 ATLAS User’s Manual

Material Defaults for Compound Semiconductors Material Systems

Table B-24b. Materials Defaults for binary compound semiconductors

M aterial GaAs GaP CdSe SnTe InP CdTe ScN
Epsilon 13.2 11.1 10.6 0 12.5 10.9 0
Eg (eV) 1.42 2.26 1.74 0.18 1.35 1.5 2.15
Chi (eV ) 4.07 4.4 0 0 4.4 4.28 0
Nc (per cc) 4.35E +17 1.76E +18 1.18E+ 18 1 5.68E+ 17 1 1
Nv (per cc) 8.16E +18 8.87E +18 7.57E+ 18 1 8.87E+ 18 1 1
ni (per cc) 2.12E +06 0.409 7.22E+ 03 0.0308 1.03E+ 07 2.51E -13 8.70E-19
Gc 2 2 2 2 2 2 2
Gv 4 4 4 4 4 4 4
Ed (eV) 0.044 0.044 0.044 0.044 0.044 0.044 0.044
Ea (eV) 0.045 0.045 0.045 0.045 0.045 0.045 0.045
Lifetim e (el) 1.00E -09 1 1 1 1 1 1
Lifetim e (ho) 2.00E -08 1 1 1 1 1 1
Auger cn 5.00E -30 0 0 0 0 0 0
Auger cp 1.00E -31 0 0 0 0 0 0
Auger kn 0 0 0 0 0 0 0
Auger kp 0 0 0 0 0 0 0
Copt 1.50E -10 0 0 0 0 0 0
An** 6.29 20.4 15.6 1.40E -11 9.61 1.40E -11 1.40E-11
Ap** 105 60.1 54.1 1.40E -11 60.1 1.40E -11 1.40E-11
betan 1.82 1 1 1 1 1 1
betap 1.75 1 1 1 1 1 1
egran 0 0 0 0 0 0 0
an1 1.90E +05 7.30E +04 7.30E+ 04 7.03E+ 05 7.30E+ 04 7.03E+ 05 7.03E+ 05
bn1 5.75E +05 1.23E +06 1.23E+ 06 1.23E+ 06 1.23E+ 06 1.23E+ 06 1.23E+ 06
an2 1.90E +05 7.30E +04 7.30E+ 04 7.03E+ 05 7.30E+ 04 7.03E+ 05 7.03E+ 05
an2 5.75E +05 1.23E +06 1.23E+ 06 1.23E+ 06 1.23E+ 06 1.23E+ 06 1.23E+ 06
ap1 2.22E +05 6.71E +05 6.71E+ 05 6.71E+ 05 6.71E+ 05 6.71E+ 05 6.71E+ 05
bp1 6.57E +05 1.69E +06 1.69E+ 06 1.69E+ 06 1.69E+ 06 1.69E+ 06 1.69E+ 06
ap2 2.22E +05 1.58E +06 1.58E+ 06 1.58E+ 06 1.58E+ 06 1.58E+ 06 1.58E+ 06
bp2 6.57E +05 2.40E +05 2.40E+ 05 2.04E+ 06 2.40E+ 05 2.04E+ 06 2.04E+ 06
Vsatn (cm /s) : 7.70E +06 1.00E +06 1.00E+ 06 1.00E+ 06 1.00E+ 06 1.00E+ 06 1.00E+ 06
Vsatp (cm /s) : 7.70E +06 1.00E +06 1.00E+ 06 1.00E+ 06 1.00E+ 06 1.00E+ 06 1.00E+ 06
m un (cm ^2/V s) 8000 300 800 4600 1050
tm un 1 1.5 1.5 1.5 1.5
m up (cm ^2/V s) 400 100 150 100
tm up 2.1 1.5 1.5 1.5

1430 ATLAS User’s Manual

Material Defaults for Compound Semiconductors Material Systems

Table B-24c. Material Defaults for binary compound semiconductors

Material ZnS HgTe AlAs ZnSe PbS BeTe AlSb ZnTe PbSe GaSb PbTe SiGe
Epsilon 8.3 20 12 8.1 170 0 11 9.7 250 15.7 412 11.8
Eg (eV) 3.8 0 2.16 2.58 0.37 2.57 1.52 2.28 0.26 0.72 0.29 0.78
Chi (eV) 0 0 3.5 4.09 0 0 3.65 3.5 0 4.06 4.6 4.17
Nc (per cc) 6.35E+18 1 4.35E+17 1 3.14E+18 1 6.79E+18 1 1 5.68E+18 1.76E+18 1.92E+19
Nv (per cc) 1 1 8.16E+18 1 3.14E+18 1 6.35E+18 1 1 2.95E+18 2.24E+18 8.20E+18
ni (per cc) 3.02E-23 1 1.35 2.13E-22 2.45E+15 2.58E-22 1.12E+06 7.04E-20 0.00655 3.66E+12 7.28E+15 3.52E+12
Gc 2 2 2 2 2 2 2 2 2 2 2 2
Gv 4 4 4 4 4 4 4 4 4 4 4 4
Ed (eV) 0.044 0.044 0.044 0.044 0.044 0.044 0.044 0.044 0.044 0.044 0.044 0.044
Ea (eV) 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045

Recombination Parameters
Lifetime (el) 1 1 1.00E-09 1 1 1 1 1 1 1 1 3.00E-05
Lifetime (ho) 1 1 2.00E-08 1 1 1 1 1 1 1 1 1.00E-05
Auger cn 0 0 5.00E-30 0 0 0 0 0 0 0 0 8.30E-32
Auger cp 0 0 1.00E-31 0 0 0 0 0 0 0 0 1.80E-31
Auger kn 0 0 0 0 0 0 0 0 0 0 0 0
Auger kp 0 0 0 0 0 0 0 0 0 0 0 0
Copt 0 0 1.50E-10 0 0 0 0 0 0 0 0 0
An** 48.1 1.40E-11 8.05 1.40E-11 30 1.40E-11 50.3 1.40E-11 1.40E-11 44.6 20.4 110
Ap** 1.40E-11 1.40E-11 56.8 1.40E-11 30 1.40E-11 48.1 1.40E-11 1.40E-11 28.8 24 30

Impact Ionization Parameters

betan 1 1 1.82 1 1 1 1 1 1 1 1 1
betap 1 1 1.75 1 1 1 1 1 1 1 1 1
egran 0 0 0 0 0 0 0 0 0 0 0 4.00E+05
an1 7.03E+05 7.03E+05 1.90E+05 7.03E+05 7.30E+04 7.03E+05 7.30E+04 7.03E+05 7.03E+05 7.30E+04 7.30E+04 7.03E+05
bn1 1.23E+06 1.23E+06 5.75E+05 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06
an2 7.03E+05 7.03E+05 1.90E+05 7.03E+05 7.30E+04 7.03E+05 7.30E+04 7.03E+05 7.03E+05 7.30E+04 7.30E+04 7.03E+05
an2 1.23E+06 1.23E+06 5.75E+05 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06
ap1 6.71E+05 6.71E+05 2.22E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05
bp1 1.69E+06 1.69E+06 6.57E+05 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06
ap2 1.58E+06 1.58E+06 2.22E+05 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06
bp2 2.04E+06 2.04E+06 6.57E+05 2.04E+06 2.40E+05 2.04E+06 2.40E+05 2.04E+06 2.04E+06 2.40E+05 2.40E+05 2.04E+06

Saturation Velocities
Vsatn (cm/s) 1.00E+06 1.00E+06 7.70E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+07
Vsatp (cm/s) 1.00E+06 1.00E+06 7.70E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 7.98E+06

Mobility Parameters
mun (cm2/Vs) 165 22000 1000 100 600 200 7 1020 4000 6000 1430
tmun 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5
mup (cm2/Vs) 5 100 100 700 550 930 1400 4000 480
tmup 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5

Note: x.comp is set to 0.3. See Section 5.4 “Material Dependent Physical Models” for more information.

1431 ATLAS User’s Manual

Material Defaults for Compound Semiconductors Material Systems

Table B-25 Material Default for ternary compound semiconductors

Material AlGaAs GaSbP InAlAs GaAsP InGaAs GaSbAs InAsP HgCdTe InGaP
Epsilon 12.3 0 0 11.7 14.2 0 13.1 16.4 12.1

Band Parameters
Eg (eV) 1.8 2.01 2.1 2.31 0.571 0.665 1.03 0.291 1.61
Chi (eV) 3.75 0 0 4.32 4.13 4.06 4.32 2.54 4.4
Nc (per cc) 4.35E+17 1 1 1.04E+18 1.15E+17 1 2.86E+17 7.58E+16 7.18E+17
Nv (per cc) 8.16E+18 1 1 8.54E+18 8.12E+18 1 8.54E+18 1.02E+19 8.87E+18
ni (per cc) 1.37E+03 1.24E-17 2.29E-18 0.125 1.56E+13 2.61E-06 3.56E+09 3.18E+15 7.43E+04
Gc 2 2 2 2 2 2 2 2 2
Gv 4 4 4 4 4 4 4 4 4
Ed (eV) 0.044 0.044 0.044 0.044 0.044 0.044 0.044 0.044 0.044
Ea (eV) 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045

Recombination Parameters
Lifetime (el) 1.00E-09 1 1 1 1 1 1 1 1
Lifetime (ho) 2.00E-08 1 1 1 1 1 1 1 1
Auger cn 5.00E-30 0 0 0 0 0 0 0 0
Auger cp 1.00E-31 0 0 0 0 0 0 0 0
Auger kn 0 0 0 0 0 0 0 0 0
Auger kp 0 0 0 0 0 0 0 0 0
Copt 1.50E-10 0 0 0 0 0 0 0 0
An** 8.05 1.40E-11 1.40E-11 14.4 3.32 1.40E-11 6.08 2.51 11.2
Ap** 56.8 1.40E-11 1.40E-11 58.6 56.6 1.40E-11 58.6 66.1 60.1

Impact Ionization Model Parameters (Selberherr Model)

betan 1.82 1 1 1 1 1 1 1 1
betap 1.75 1 1 1 1 1 1 1 1
egran 0 0 0 0 0 0 0 0 0
an1 1.90E+05 7.03E+05 8.60E+06 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05
bn1 5.75E+05 1.23E+06 3.50E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06
an2 1.90E+05 7.03E+05 8.60E+06 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05
an2 5.75E+05 1.23E+06 3.50E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06
ap1 2.22E+05 6.71E+05 2.30E+07 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05
bp1 6.57E+05 1.69E+06 4.50E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06
ap2 2.22E+05 1.58E+06 2.30E+07 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06
bp2 6.57E+05 2.04E+06 4.50E+06 2.04E+06 2.04E+06 2.04E+06 2.04E+06 2.04E+06 2.04E+06

Saturation Velocities
Vsatn (cm/s) 7.70E+06 1.00E+06 1.00E+06 1.00E+06 7.70E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06
Vsatp (cm/s) 7.70E+06 1.00E+06 1.00E+06 1.00E+06 7.70E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06

Note: x.comp is set to 0.3. See Section 5.4 “Material Dependent Physical Models” for more information.

1432 ATLAS User’s Manual

Material Defaults for Compound Semiconductors Material Systems

Table B-26 shows material defaults for quaternary compound semiconductors.

Table B-26 Material Defaults for quaternary compound semiconductors
Material : InGaAsP AlGaAsP AlGaAsSb InGaNAs InGaNP AlGaNAs AlGaNP AlInNAs AlInNP InAlGaAs InAlGaP InAlAsP
Epsilon : 12.7 0 0 0 0 0 0 0 0 0 0 0

Band Parameters
Eg (eV) : 1.27 0 0 0 0 0 0 0 0 1.34 0 0
Chi (eV) : 4.32 0 0 0 0 0 0 0 0 1.46 0 0
Nc (per cc) : 3.61E+17 1 1 1 1 1 1 1 1 1 1 1
Nv (per cc) : 8.54E+18 1 1 1 1 1 1 1 1 1 1 1
ni (per cc) : 3.72E+07 1 1 1 1 1 1 1 1 5.04E-12 1 1
Gc : 2 2 2 2 2 2 2 2 2 2 2 2
Gv : 4 4 4 4 4 4 4 4 4 4 4 4
Ed (eV) : 0.044 0.044 0.044 0.044 0.044 0.044 0.044 0.044 0.044 0.044 0.044 0.044
Ea (eV) : 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045

Recombination Parameters
Lifetime (el): 1 1 1 1 1 1 1 1 1 1 1 1
Lifetime (ho): 1 1 1 1 1 1 1 1 1 1 1 1
Auger cn : 0 0 0 0 0 0 0 0 0 0 0 0
Auger cp : 0 0 0 0 0 0 0 0 0 0 0 0
Auger kn : 0 0 0 0 0 0 0 0 0 0 0 0
Auger kp : 0 0 0 0 0 0 0 0 0 0 0 0
Copt : 0 0 0 0 0 0 0 0 0 0 0 0
An** : 7.11 1.40E-11 1.40E-11 1.40E-11 1.40E-11 1.40E-11 1.40E-11 1.40E-11 1.40E-11 1.40E-11 1.40E-11 1.40E-11
Ap** : 58.6 1.40E-11 1.40E-11 1.40E-11 1.40E-11 1.40E-11 1.40E-11 1.40E-11 1.40E-11 1.40E-11 1.40E-11 1.40E-11

Impact Ionization Parameters (Selberherr model)

betan : 1 1 1 1 1 1 1 1 1 1 1 1
betap : 1 1 1 1 1 1 1 1 1 1 1 1
egran : 0 0 0 0 0 0 0 0 0 0 0 0
an1 : 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05
bn1 : 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06
an2 : 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05 7.03E+05
an2 : 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06 1.23E+06
ap1 : 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05 6.71E+05
bp1 : 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06 1.69E+06
ap2 : 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06 1.58E+06
bp2 : 2.04E+06 2.04E+06 2.04E+06 2.04E+06 2.04E+06 2.04E+06 2.04E+06 2.04E+06 2.04E+06 2.04E+06 2.04E+06 2.04E+06

Saturation Velocities
Vsatn (cm/s) : 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06
Vsatp (cm/s) : 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06

1433 ATLAS User’s Manual

Material Defaults for Compound Semiconductors Material Systems

Table B-27 shows material defaults for organic materials.

Table B-27 Material Defaults for organic materials
Material Organic Pentacene Alq3 TPD PPV Tetracene
Epsilon 11.8 11.8 3.4 11.8 11.8 11.8

Band Parameters
Eg (eV) 1.08 1.08 2.4 1.08 1.08 1.08
Chi (eV) 4.17 4.17 4.17 4.17 4.17 4.17
Nc (per cc) 2.80E+19 2.80E+19 1.44E+20 2.80E+19 2.80E+19 2.80E+19
Nv (per cc) 1.04E+19 1.04E+19 1.44E+20 1.04E+19 1.04E+19 1.04E+19
ni (per cc) 1.45E+10 1.45E+10 0.998 1.45E+10 1.45E+10 1.45E+10
Gc 2 2 2 2 2 2
Gv 4 4 4 4 4 4
Ed (eV) 0.044 0.044 0.044 0.044 0.044 0.044
Ea (eV) 0.045 0.045 0.045 0.045 0.045 0.045
An** 129 129 385 129 129 129
Ap** 66.8 66.8 385 66.8 66.8 66.8

Recombination Parameters
taun0 1 1 1 1 1 1
taup0 1 1 1 1 1 1
etrap 0 0 0 0 0 0
nsrhn -1.00E+03 -1.00E+03 -1.00E+03 -1.00E+03 -1.00E+03 -1.00E+03
nsrhp -1.00E+03 -1.00E+03 -1.00E+03 -1.00E+03 -1.00E+03 -1.00E+03
ksrhtn 0.0025 0.0025 0.0025 0.0025 0.0025 0.0025
ksrhtp 0.0025 0.0025 0.0025 0.0025 0.0025 0.0025
ksrhcn 3.00E-13 3.00E-13 3.00E-13 3.00E-13 3.00E-13 3.00E-13
ksrhcp 1.18E-12 1.18E-12 1.18E-12 1.18E-12 1.18E-12 1.18E-12
ksrhgn 1.77 1.77 1.77 1.77 1.77 1.77
ksrhgp 0.57 0.57 0.57 0.57 0.57 0.57
nsrhn -1.00E+03 -1.00E+03 -1.00E+03 -1.00E+03 -1.00E+03 -1.00E+03
nsrhp -1.00E+03 -1.00E+03 -1.00E+03 -1.00E+03 -1.00E+03 -1.00E+03
augn 0 0 0 0 0 0
augp 0 0 0 0 0 0
augkn 0 0 0 0 0 0
augkp 0 0 0 0 0 0
kaugcn 1.83E-31 1.83E-31 1.83E-31 1.83E-31 1.83E-31 1.83E-31
kaugcp 2.78E-31 2.78E-31 2.78E-31 2.78E-31 2.78E-31 2.78E-31
kaugdn 1.18 1.18 1.18 1.18 1.18 1.18
kaugdp 0.72 0.72 0.72 0.72 0.72 0.72
copt 0 0 0 0 0 0

Band-to-Band Tunneling Parameters

bb.a 4.00E+14 4.00E+14 4.00E+14 4.00E+14 4.00E+14 4.00E+14
bb.b 1.90E+07 1.90E+07 1.90E+07 1.90E+07 1.90E+07 1.90E+07
bb.gamma 2.5 2.5 2.5 2.5 2.5 2.5
mass.tunnel 0.25 0.25 0.25 0.25 0.25 0.25
me.tunnel 1.08 1.08 3.21 1.08 1.08 1.08
mh.tunnel 0.556 0.556 3.21 0.556 0.556 0.556

Saturation Velocities
Vsatn (cm/s) : 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06
Vsatp (cm/s) : 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06 1.00E+06

Note: x.comp and y.comp are both set to 0.3. See Section 5.4 “Material Dependent Physical Models” for more
information.

1434 ATLAS User’s Manual

Material Defaults for Compound Semiconductors Material Systems

Table B-28 lists the default material parameters for metal oxides and various other
semiconductors.

Table B-28 Material Defaults for Conductive Oxides and Others

Material CMO IMO CdO MgO CdZnO MgZnO MgCdO MgCdZnO CNT Se IGZO

Epsilon 9.7 28.0 9.0 9.8 8.49 8.49 8.49 8.49 9.7 11.56 10.0

Band
Parameters

Eg(eV) 0.5 5.0 0.8 2.22 3.05

Chi(eV) 5.0 3.5 3.2 2.02 4.16

Nc(percc) 1.0e20 1.0e20 2.2e18 2.2e18 2.2e18 2.2e18 2.2e18 2.2e18 6.65e17 5.0e19 5.0e18

Nv(percc) 1.0e20 1.0e20 1.8e19 1.8e19 1.8e19 1.8e19 1.8e19 1.8e19 6.65e17 5.0e19 5.0e18

mun0 0.1 0.1 100.0 0.003 15

(cm2/Vs)

mup0 0.1 0.1 25.0 0.1 0.1

(cm2/Vs)

Table B-29 Lattice Parameters for Binary Zincblende Semiconductors [216]

Parameter a0 da0/dT C11 C12 b av ac

Unit (A) (10-5 A/K) (GPa) (GPa) (eV) (eV) (eV)

GaAs 5.65325 3.88 1221 566 -2.0 1.16 -7.17

InP 5.8697 2.79 1011 561 -2.0 0.6 -6.0
AlAs 5.6611 2.90 1250 534 -2.3 2.47 -5.64
GaSb 6.0959 4.72 884.2 402.6 -2.0 0.8 -7.5
AlSb 6.1355 2.60 876.9 434.1 -1.35 1.4 -4.5
InAs 6.0583 2.74 832.9 452.6 -1.8 1.0 -5.08
GaP 5.4505 2.92 1405 620.3 -1.6 1.7 -8.2
AlP 5.4672 2.92 1330 630 -1.5 3.0 -5.7
InSb 6.4794 3.48 684.7 373.5 -2.0 0.36 -6.94

1435 ATLAS User’s Manual

Material Defaults for Compound Semiconductors Material Systems

Table B-30 Luttinger Parameters for Zincblende Semiconductors

Parameter 1 2 3

GaAs 6.98 2.06 2.93

InP 5.08 1.60 2.10
AlAs 3.76 0.82 1.42
GaSb 13.4 4.7 6.0
AlSb 5.18 1.19 1.97
InAs 20.0 8.5 9.2
GaP 4.05 0.49 2.93
AlP 3.35 0.71 1.23
InSb 34.8 15.5 16.5

1436 ATLAS User’s Manual

Miscellaneous Semiconductors Material Systems

B.10 Miscellaneous Semiconductors

The remainder of the semiconductors available have defined default parameter values to
various degrees of completeness [1]. The following tables describe those parameter defaults
as they exist. Since many of the material parameters are currently unavailable, we
recommended that you should be careful when using these materials. Make sure the proper
values are used

Note: You can use the MODEL PRINT statement to echo the parameters used to the run-time output.

Table B-31 Band Parameters for Miscellaneous Semiconductors

Material Eg(0)eV Eg(300)eV a b mc mv eV

Silicon

Polysilicon

Ge 0.7437 4.7710-4 235.0 0.2225 0.2915 4.0

Diamond 5.45 4.7710-4 0.0 (a) (b) 7.2
6H-SiC 2.9 2.9 0.0 0.0 0.454 0.33 3.5
4H-SiC 3.26 3.26 0.0 0.0 0.76 1.20 4.0
3C-SiC 2.2 2.2 0.0 0.0 0.41 0.165 3.2
A1P 2.43 2.43 0.0 0.0
A1As 2.16 2.16 0.0 0.0 3.5
A1Sb 1.6 2.6910-4 2.788 (c) 0.4
GaSb 0.81 3.32910-4 -27.6622 (c) 0.24 3.65
InSb 0.235 2.81710-4 90.0003 0.014 0.4 4.06
ZnS 3.8 3.8 0.0 0.0 0.4 4.59
ZnSe 2.58 2.58 0.0 0.0 0.1 0.6
ZnTe 2.28 0.0 0.0 0.1 0.6 4.09
CdS 2.48 2.48 0.0 0.0 0.21 0.8 4.18
CdSe 1.74 1.74 0.0 0.0 0.13 0.45 4.5
CdTe 1.5 1.5 0.0 0.0 0.14 0.37
HgS 2.5 2.5 0.0 0.0 4.28
HgSe

HgTe

1437 ATLAS User’s Manual

Miscellaneous Semiconductors Material Systems

Table B-31 Band Parameters for Miscellaneous Semiconductors

Material Eg(0)eV Eg(300)eV a b mc mv eV

PbS 0.37 0.37 0.0 0.0 0.25 0.25

PbSe 0.26 0.26 0.0 0.0 0.33 0.34
PbTe 0.29 0.29 0.0 0.0 0.17 0.20 4.6
SnTe 0.18 0.18 0.0 0.0
ScN 2.15 2.15 0.0 0.0
BeTe 2.57 2.57 0.0 0.0

Notes
• Nc300 = 5.01018
• Nv300 = 1.81019
• mc(X) = 0.39
mc(G) = 0.09
Nc = Nc(X) + Nc(G)
• mc(G) = 0.047
mc(L) = 0.36
Nc = Nc(G) + Nc(L)

Table B-32 Static Dielectric Constants for Miscellaneous Semiconductors

Material Dielectric Constant

Ge 16.0
Diamond 5.5
6H-SiC 9.66
4H-SiC 9.7
3C-SiC 9.72
AlP 9.8
AlAs 12.0
AlSb 11.0
GaSb 15.7
InSb 18.0
ZnS 8.3

1438 ATLAS User’s Manual

Miscellaneous Semiconductors Material Systems

Table B-32 Static Dielectric Constants for Miscellaneous Semiconductors

Material Dielectric Constant

ZnSe 8.1
CdS 10
CdSe 10.6
CdTe 9.4
HgS
HgSe 25.0
HgTe 20.
PbS 170.0
PbSe 250.0
PbTe 412.0
SnTe 1770.0
ScN
BeTe
ZnO 2.0

Table B-33 Mobility Parameters for Miscellaneous Semiconductors

Material MUNO (cm2/Vs) MUPO (cm2/Vs) VSATN(cm/s) VSAT(cm/s)

Ge 3900.0(a) 1900.0(b)
Diamond 500.0 300.0 2.0107
6H-SiC 330.0 300.0 2.0107 1107
4H-SiC 460 124 2.2107 1107
3C-SiC 1000.0 50.0 2.0107 1107
AlP 80.0
AlAs 1000.0 100.0
AlSb 200.0 550.0
GaSb 4000.0 1400.0
InSb 7800.0 750.0
ZnS 165.0 5.0

1439 ATLAS User’s Manual

Miscellaneous Semiconductors Material Systems

Table B-33 Mobility Parameters for Miscellaneous Semiconductors

Material MUNO (cm2/Vs) MUPO (cm2/Vs) VSATN(cm/s) VSAT(cm/s)

ZnSe 100.0 16
CdS 340.0 50.0
CdSe 800.0
CdTe 1050.0 100.0
HgS

HgSe 5500.0
HgTe 22000.0 100.0
PbS 600.0 700.0
PbSe 1020.0 930.0
PbTe 6000.0 4000.0
SnTe

ScN

BeTe

1440 ATLAS User’s Manual

Insulators Material Systems

B.11 Insulators
The following tables show the default material parameters for insulator materials. As noted in
the Section B.2 “Semiconductors, Insulators, and Conductors”, the only parameter required
for electrical simulation in insulator materials is the dielectric constant. Thermal and optical
properties are required in GIGA and LUMINOUS respectively.

Table B-34 Default Static Dielectric Constants of Insulators

Material Dielectric Constant

Vacuum 1.0
Air 1.0
Ambient 1.0
Oxide 3.9
Si02 3.9
BSG 3.9
BPSG 3.9
Nitride 7.5
SiN 7.5
Si3N4 7.5
OxyNitride 7.5
Sapphire 12.0
Al2O3 9.3
HfO2 22.0
HfSiO4 12.0
SnO2 9.0

CuPc 4.0
NPD 4.0
CBP 4.0
Irppy 4.0
BAlq 4.0
Ta2O5 26

1441 ATLAS User’s Manual

Insulators Material Systems

Table B-34 Default Static Dielectric Constants of Insulators

Material Dielectric Constant

TiO2 80
ZrO2 22
La2O3 27

Table B-35 Default Thermal Parameters for Insulators at 300K

Material Thermal Capacity (J/cm3/K) Thermal Conductivity (W/cm/K)

Vacuum 0.0 0.0

Air 1.0 0.02610-2
Ambient 1.0 0.02610-2
Oxide 3.066 0.014
Si02 3.066 0.014
BSG 3.066 0.014
BPSG 3.066 0.014
Nitride 0.585 5.4
SiN 0.585 5.4
Si3N4 0.585 5.4
OxyNitride 0.585 5.4
Sapphire 1.97 33
Al2O3 1.97 33
HfO2
HfSiO2

1442 ATLAS User’s Manual

Metals/Conductors Material Systems

B.12 Metals/Conductors
Table B-36 shows the default values for resistivities of metals and the temperature
coefficients of resistivity [see also 301].

Table B-36 Resistivity and Temperature Coefficient of Metals

Resistivity Temperature Coefficient

Material
(µcm) (µcm/K)

Aluminum 2.6540 0.00429

Gold 2.35 0.004
Silver 1.59 0.0041
Tungsten 5.65
Titanium 42.0
Platinum 10.6 0.003927
Palladium 10.8 0.00377
Cobalt 6.24 0.00604
Molybdinum 5.2
Lead 20.648 0.00336
Iron 9.71 0.00651
Tantalum 12.45 0.00383
Copper 6.24 0.00604
Tin 11.0 0.0047
Nickel 6.84 0.0069

1443 ATLAS User’s Manual

Optical Properties Material Systems

B.13 Optical Properties

The default values for complex index of refraction in LUMINOUS are interpolated from the
tables in [54, 107, 161, 207, 212, 320]. Rather than print the tables here, the ranges of optical
wavelengths for each material are listed in Table B-37.

Table B-37 Wavelength Ranges for Default Complex Index of Refraction

Temperature
Material Composition Fraction Wavelengths (microns)
(K)

AlAs 300 NA 0.2213 - 50.0

Alq3 300 NA 0.3 - 0.8
CdS 300 NA 0.2 - 1.6
CIGS 300 YES 0.27 - 1.65
GaAs 300 NA 0.0 - 0.9814
GaN 300 NA 0.372 - 1.167
InSb 300 NA 0.2296 - 6.5
InP 300 NA 0.1689 - 0.975
ITO 300 NA 0.3 - 0.8
NPB 300 NA 0.3 - 0.8
PPV 300 NA 0.3 - 0.8
Poly 300 NA 0.1181 - 18.33
SiO2 300 NA 0.1145 - 1.7614
Silicon 300 NA 0.0103 - 2.0
SnTe 300 NA 0.2 - 1.6
Zn0 300 NA .349 - .898

Table B-38 show the default parameters for the Sellmeier refractive index model (see
Equations 3-700) for various materials.

Table B-38 Default Parameter Values [216] for the Sellmeier Refractive Index Model

Parameter S0SELL S1SELL L1SELL S2SELL L2SELL

AlAs 2.616 5.56711 0.2907 0.49252 10.0

AlN 3.14 1.3786 0.1715 3.861 15.0333

1444 ATLAS User’s Manual

Optical Properties Material Systems

Table B-38 Default Parameter Values [216] for the Sellmeier Refractive Index Model
GaAs 3.5 7.4969 0.4082 1.9347 37.17526
GaN 3.6 1.75 0.256 4.1 17.86057
GaP 3.096 5.99865 0.30725 0.83878 17.32051
GaSb 13.1 0.75464 1.2677 0.68245 10.0
Ge 9.282 6.7288 0.66412 0.21307 62.20932
InAs 11.1 0.71 2.551 2.75 45.6618
InP 7.255 2.316 0.6263 2.756 32.93934
InSb 15.4 0.10425 7.15437 3.47475 44.72136
Si 3.129 8.54297 0.33671 0.00528 38.72983

Table B-39 shows the default parameters for the Adachi refractive index model [1] (see
Equations 3-701) for various materials.

Table B-39 Default Parameter Values [216] for the Adachi Refractive Index Model

Parameter AADACHI BADACHI DADACHI

GaAs 6.3 9.4 0.34

InP 8.4 6.6 0.11
AlAs 25.3 -0.8 0.28
GaSb 4.05 12.66 0.82
AlSB 59.68 -9.53 0.65
InAs 5.14 10.15 0.38
GaP 22.25 0.9 0.08
AlP 24.1 -2.0 0.07
InSb 7.91 13.07 0.81

1445 ATLAS User’s Manual

Optical Properties Material Systems

B.13.1 SOPRA Database

The SOPRA database is a collection of complex refractive index for various materials
sometimes as a function of temperature and composition. To access this data, you should
specify the file name from Table B-40 using the SOPRA parameter of the MATERIAL statement.
The SOPRA database is courtesy of SOPRA, a thin film metrology company located world
wide (www.sopra-sa.com). The ranges are given in their specified units. ATLAS will
automatically handle unit conversions.

Table B-40 Contents of SOPRA Database

File Description Default

7059.nk Glass 7059 x

Ag.nk Silver (0.6eV - 6.6eV) x
Againp.all (AlxGa1-x)0.5In0.5P vs composition fraction x x

Againp0.nk Ga0.5In0.5P (0.225m - 1m) x

Againp1.nk (Al0.1Ga0.9)0.5In0.5P (0.225m - 1.0m) x

Againp3.nk (Al0.3Ga0.7)0.5In0.5P (0.225m - 1.0m) x

Againp6.nk (Al0.6Ga0.4)0.5In0.5P (0.225m - 1.0m) x

Againp7.nk (Al0.7Ga0.3)0.5In0.5P (0.225m - 1.0m) x

Againp10.nk Al0.5In0.5P (0.225m - 1.0m) x

Al.nk Aluminum (0.5eV - 6.75eV)

Al2o3.nk Al2O3 (Sapphire) )(0.25m - 0.9m) x

Al2o3p.nk Al2O3 (Palik) (0.5eV - 6eV)

Alas.nk AlAs (0.5eV - 5.6eV)

Alas.all AlAs vs temperature from 28K to 626K x
Alas028t.nk AlAs 28K (1.255eV - 4.495eV) x
Alas052t.nk AlAs 52K (1.255eV - 4.495eV) x
Alas072t.nk AlAs 72K (1.255eV - 4.495eV) x
Alas098t.nk AlAs 98K (1.255eV - 4.495eV) x
Alas125t.nk AlAs 125K (1.255eV - 4.495eV) x
Alas152t.nk AlAs 152K (1.255eV - 4.495eV) x
Alas178t.nk AlAs 178K (1.255eV - 4.495eV) x
Alas204t.nk AlAs 204K (1.255eV - 4.495eV) x

1446 ATLAS User’s Manual

Optical Properties Material Systems

Table B-40 Contents of SOPRA Database

Alas228t.nk AlAs 228K (1.255eV - 4.495eV) x
Alas305t.nk AlAs 305K (1.255eV - 4.495eV) x
Alas331t.nk AlAs 331K (1.255eV - 4.495eV) x
Alas361t.nk AlAs 361K (1.255eV - 4.495eV) x
Alas390t.nk AlAs 390K (1.255eV - 4.495eV) x
Alas421t.nk AlAs 421K (1.255eV - 4.495eV) x
Alas445t.nk AlAs 445K (1.255eV - 4.495eV) x
Alas469t.nk AlAs 469K (1.255eV - 4.495eV) x
Alas499t.nk AlAs 499K (1.255eV - 4.495eV) x
Alas527t.nk AlAs 527K (1.255eV - 4.495eV) x
Alas552t.nk AlAs 552K (1.255eV - 4.495eV) x
Alas578t.nk AlAs 578K (1.255eV - 4.495eV) x
Alas602t.nk AlAs 602K (1.255eV - 4.495eV) x
Alas626t.nk AlAs 626K (1.255eV - 4.495eV) x
Alcu.nk Aluminum Copper (0.25m - 0.9m) x
Algaas.all AlxGa1-xAs vs composition fraction x x
Algaas0.nk GaAs (0.5eV - 6eV) x
Algaas1.nk Al.1Ga.9As (1.5eV - 6eV) x
Algaas2.nk Al.2Ga.8As (1.5eV - 6eV) x
Algaas3.nk Al.3Ga.7As (1.5eV - 6eV) x
Algaas4.nk Al.4Ga.6As (1.5eV - 6eV) x
Algaas5.nk Al.5Ga.5As (1.5eV - 6eV) x
Algaas6.nk Al.6Ga.4As (1.5eV - 6eV) x
Algaas7.nk Al.7Ga.3As (1.5eV - 6eV) x
Algaas8.nk Al.8Ga.2As (1.5eV - 6eV) x
Algaas9.nk Al.9Ga.1As (1.5eV - 6eV) x
Algaas10.nk AlAs (0.5eV - 5.6eV) x
Alon.nk Aluminum Oxynitride (0.5eV - 3.1eV)
Alsb.nk AlSb (0.5eV - 5.8eV) x

1447 ATLAS User’s Manual

Optical Properties Material Systems

Table B-40 Contents of SOPRA Database

Alsi.nk AlSi (0.25m - 0.8m) x
Alsiti.nk Aluminum Silicon Titanium (0.25m - 0.8m)
Asi.nk Aluminum Silicon (0.25m - 0.8m)
Au.nk Gold (0.25m - 0.8m)
Baf2.nk BaF2 (0.2m - 2m)
Bk7.nk Glass BK7 (1eV - 7eV)
Bk7_abs.nk (1eV - 6.5eV)
Caf2.nk CaF2 (1eV - 6.5eV)
Carbam.nk Amorphous Carbon (0.3m - 0.84m)
Ccl4.nk HexaChloromethane (0.25m - 0.9m)
Cdse.nk CdSe (0.5m - 6.5m) x
Cdte.nk CdTe (0.5m - 6.5m) x
Co.nk Co (0.25m - 0.9m) x
Co_2.nk Cobalt Hexagonal (Thin Flim) (0.65eV - 6.6eV)
Cor7059.nk Glass Corning 7059 (0.44m - 0.65m)
Cosi2-4.nk CoSi2 (1.0eV - 4.8eV)

Cr.nk Cr (1.0eV - 6.0eV)

Cr3si.nk CrSi3 (0.1eV - 5.9eV)

Cr5si3.nk Cr5Si3 (0.1eV - 5.9eV)

Crsi2el2.nk Chromimum Silicide (11.0eV - 23.1eV)

Cu.nk Cu (0.5eV - 6.5eV) x
Cu2o.nk Cu2O (0.5eV - 4.2eV)
Cuo.nk CuO (0.5eV - 4.2eV)
Diam.nk Diamond (0.5eV - 5.8eV) x
Fesi2el1.nk Iron Silicide (4.5eV - 9eV)
Fesi2el2.nk Iron Silicide (11.0eV - 23.1eV)
Fesi2epi.nk Iron Silicide (0.2eV - 5.0eV)
Gaas.nk GaAs (0.5eV - 6eV)
Gaas100.nk GaAs 100K

1448 ATLAS User’s Manual

Optical Properties Material Systems

Table B-40 Contents of SOPRA Database

Gaas111.nk GaAs 111K (0.234m -0.84m)
GaAsTemp.ALL GaAs vs temperature from 31K to 634K x
Gaas031t.nk GaAs 31K (1.24eV - 4.495eV) x
Gaas041t.nk GaAs 41K (1.24eV - 4.495eV) x
Gaas060t.nk GaAs 60K (1.24eV - 4.495eV) x
Gaas081t.nk GaAs 81K (1.24eV - 4.495eV) x
Gaas103t.nk GaAs 103K (1.24eV - 4.495eV) x
Gaas126t.nk GaAs 126K (1.24eV - 4.495eV) x
Gaas150t.nk GaAs 150K (1.24eV - 4.495eV) x
Gaas175t.nk GaAs 175K (1.24eV - 4.495eV) x
Gaas199t.nk GaAs 199K (1.24eV - 4.495eV) x
Gaas224t.nk GaAs 224K (1.24eV - 4.495eV) x
Gaas249t.nk GaAs 249K (1.24eV - 4.495eV) x
Gaas273t.nk GaAs 273K (1.24eV - 4.495eV) x
Gaas297t.nk GaAs 297K (1.24eV - 4.495eV) x
Gaas320t.nk GaAs 320K (1.24eV - 4.495eV) x
Gaas344t.nk GaAs 344K (1.24eV - 4.495eV) x
Gaas367t.nk GaAs 367K (1.24eV - 4.495eV) x
Gaas391t.nk GaAs 391K (1.24eV - 4.495eV) x
Gaas415t.nk GaAs 415K (1.24eV - 4.495eV) x
Gaas443t.nk GaAs 443K (1.24eV - 4.495eV) x
Gaas465t.nk GaAs 465K (1.24eV - 4.495eV) x
Gaas488t.nk GaAs 488K (1.24eV - 4.495eV) x
Gaas515t.nk GaAs 515K (1.24eV - 4.495eV) x
Gaas546t.nk GaAs 546K (1.24eV - 4.495eV) x
Gaas579t.nk GaAs 579K (1.24eV - 4.495eV) x
Gaas603t.nk GaAs 603K (1.24eV - 4.495eV) x
Gaas634t.nk GaAs 634K (1.24eV - 4.495eV) x
Gaaso.nk Gallium Arsenide Oxide (1.0eV - 6.0eV)

1449 ATLAS User’s Manual

Optical Properties Material Systems

Table B-40 Contents of SOPRA Database

Gaasox.nk Gallium Arsenide Oxide (1.0eV - 6.0eV)
Gan-uv.nk GaN (0.25m - 0.8m)
Gan01.nk (1eV - 5eV)
Gan02.nk (0.2404368m - 0.8m)
Gan03.nk (0.25m - 0.8m)
Gan60.nk (0.3m - 0.8m)
Gan70.nk (0.3m - 0.8m)
Gap.nk GaP (0.5eV - 6.0eV) x
Gap100.nk GaP (100)
Gapox.nk Gallium Phosphide Oxide (1.6eV - 6.0eV)
Gasb.nk GaSb (1.25eV - 6.0eV) x
Gasbox.nk Gallium Antimonide Oxide (1.6eV - 6.0eV)
Ge.nk Ge (0.5eV - 6.0eV)
Ge100.nk Ge (100)
H2o.nk H2O (1.5eV - 6.0eV)

Hfo2.nk HFO2 (0.25m - 0.9m)

Hfsi2.nk HFSi2 (0.05m - 9.5m)
Hgcdte.all Hg1-xCdxTe as a function of composition fraction x x
Hgcdte0.nk HgTe (0.2m - 0.9m) x
Hgcdte2.nk Hg.8Cd.2Te (0.2m - 0.9m) x
Hgcdte3.nk Hg.71Cd.29Te (0.2m - 0.9m) x
Inas.nk InAs (1eV - 6.0eV) x
Inasox.nk Indium Arsenide Oxide (1eV - 6eV)
Ingaas.nk In0.53Ga0.41As (1.5eV - 6.0eV)
Ingasb.all In1-xGaxSb as a function of composition fraction x x
Ingasb0.nk InSb (0.225m - 0.825m) x
Ingasb1.nk In.9Ga.1Sb (0.225m - 0.825m) x
Ingasb3.nk In.7Ga.3Sb (0.225m - 0.825m) x
Ingasb5.nk In.5Ga.5Sb (0.225m - 0.825m) x

1450 ATLAS User’s Manual

Optical Properties Material Systems

Table B-40 Contents of SOPRA Database

Ingasb7.nk In.3Ga.7Sb (0.225m - 0.825m) x
Ingasb9.nk In.1Ga.9Sb (0.225m - 0.825m) x
Ingasb10.nk GaSb (0.225m - 0.825m) x
Inp.nk InP (0.5eV - 6.0eV) x
Inpox.nk Indium Phosphide Oxide (1.6eV - 6.0eV)
Insb.nk InSb (1.0eV - 6.0eV) x
Insbox.nk Indium Antimonide Oxide (1.6eV - 6.0eV)
Ir.nk Ir (0.5eV - 6.5eV)
Ir3si5e.nk Indium Silicide (4.1eV - 8.2eV)
Ir3si5p.nk Indium Silicide (0.6eV - 55eV)
Ito2.nk Indium Tin Oxide (0.25m - 0.85m) x
Kcl.nk KCl (0.5eV - 6.0eV)
Lasf9.nk Glass LASF9
Li.nk Li
Lif.nk LiF (10.5eV - 6.75eV)
Mgf2.nk MgF2 (0.25m - 0.9m)
Mo.nk Mo (0.5eV - 6.5eV) x
Mosi2-a.nk Molybdenuin Silicide (Parallel) (1eV - 4.4eV)
Mosi2-b.nk Molybdenuin Silicide (Perpendicular)
(0.05eV - 6eV)
Nbsi-a.nk Niobuim Silicide (Parallel) (0.05eV - 6eV)
Nbsi-b.nk Niobuim Silicide (Perpendicular) (0.05eV - 6eV)
Ni.nk Ni (0.05eV - 6.0eV) x
Ni2si.nk Nickel Silicide (0.1eV - 5.8eV)
Ni3si.nk Nickel Silicide (0.1eV - 5.8eV)
Nisi.nk Nickel Silicide (0.5eV - 5.8eV) x
Os.nk Os (0.5eV - 6.5eV)
P_sias.nk Silicon Arsenide
P_siud.nk Polysilicon (heavy As doping)
Pbs.nk PbS (0.05eV - 6.2eV) x

1451 ATLAS User’s Manual

Optical Properties Material Systems

Table B-40 Contents of SOPRA Database

Pbse.nk PbSe (1.0eV - 6.0eV) x
Pd.nk Pd (0.5eV - 6.5eV) x
Pd2si-a.nk Palladium Silicide (Parallel) (0.1eV - 7eV)
Pd2si-b.nk Palladium Silicide (Perpendicular) (0.1eV - 7eV)
Pt.nk Pt (0.5eV - 6.0eV) x
Resi1-75.nk Si.75Ge.25 (0.1eV - 12eV) x
Resige.all SixGe1-x as a function of composition fraction x x
Resige0.nk Ge (0.5eV - 7.6eV) x
Resige1.nk Si (0.5eV - 6eV) x
Resige22.nk Si.22Ge.78 (1.7eV - 5.6eV) x
Resige39.nk Si.39Ge.61 (1.7eV - 5.6eV) x
Resige51.nk Si.51Ge.49 (1.7eV - 5.6eV) x
Resige64.nk Si.64Ge.36 (1.7eV - 5.6eV) x
Resige75.nk Si.75Ge.25 (1.7eV - 5.6eV) x
Resige83.nk Si.83Ge.27 (1.7eV - 5.6eV) x
Resige91.nk Si.91Ge.09 (1.7eV - 5.6eV) x
Rh.nk Rh (0.5eV - 6.0eV)
Ringas.all InxGa1-xAs as a function of composition fraction x x
Ringas0.nk GaAs (0.5eV - 6.0eV) x
Ringas10.nk In.1Ga.9As (0.24eV - 0.84eV) x
Ringas20.nk In.2Ga.8As (1.3eV - 5.1eV) x
Ringas24.nk In.24Ga.78As (1.5eV - 5.0eV) x
Sf11.nk Glass SF11
Si100_2.nk Si (100) (0.23m - 0.84m)
Si110.nk Si (110) (0.23m - 0.84m)
Si111.nk Si (111) (0.23m - 0.84m)
Si11ge89.nk Si0.11Ge.89
Si20ge80.nk Si0.2Ge0.8
Si28ge72.nk Si0.88Ge0.72

1452 ATLAS User’s Manual

Optical Properties Material Systems

Table B-40 Contents of SOPRA Database

Si3n4.nk Si3N4 (1.0eV - 8.0eV) x
Si65ge35.nk Si0.35Ge0.35
Si85ge15.nk Si0.85Ge0.15
Si98ge02.nk Si0.98Ge0.02
Siam1.nk Amorphous Si (0.6eV - 6.0eV)
Siam2.nk Amorphous Si (1.0eV - 6.0eV)
Sic.nk SiC (0.19m - 1.25m) x
Sicr-t.all Si (0-450C) (1.5eV - 4.7eV) x
Sicr-t02.nk Si (20C) (1.5eV - 4.7eV) x
Sicr-t10.nk Si (100C) (1.5eV - 4.7eV) x
Sicr-t15.nk Si (150C) (1.5eV - 4.7eV) x
Sicr-t20.nk Si (200C) (1.5eV - 4.7eV) x
Sicr-t25.nk Si (250C) (1.5eV - 4.7eV) x
Sicr-t30.nk Si (300C) (1.5eV - 4.7eV) x
Sicr-t35.nk Si (350C) (1.5eV - 4.7eV) x
Sicr-t40.nk Si (400C) (1.5eV - 4.7eV) x
Sicr-t45.nk Si (450C) (1.5eV - 4.7eV) x
Sicrir.nk (4.95940810-2eV - 1.241092eV)
Sige_ge.nk SiGe on Ge
Sige_si.nk SiGe on Si
Singas.all InGaAs (Strained)
Singas0.nk GaAs (Strained) (0.24m - 0.84m)
Singas10.nk In0.1Ga0.9AS (Strained) (0.24m - 0.84m)
Singas20.nk In0.2Ga0.8AS (Strained) (0.24m - 0.84m)
Singas24.nk In0.24Ga0.76AS (Strained) (0.24m - 0.84m)
Sio.nk SiO (0.6eV - 5.9eV)
Sio2.nk SiO2 (0.6eV - 7.6eV) x
Sio2ir.nk SiO2 (0.0495941eV - 1.386316eV)
Sion0.nk SiliconOxynitride (10% N) (0.25m - 0.8m)

1453 ATLAS User’s Manual

Optical Properties Material Systems

Table B-40 Contents of SOPRA Database

Sion20.nk SiliconOxynitride (20% N) (0.25m - 0.8m)
Sion40.nk SiliconOxynitride (40% N) (0.25m - 0.8m)
Sion60.nk SiliconOxynitride (60% N) (0.25m - 0.8m)
Sion80.nk SiliconOxynitride (80% N) (0.25m - 0.8m)
Siop.nk SiO (0.5eV - 6.5eV)
Sipoly.nk Polysilicon x
Sipoly10.nk (0.5eV - 6.5eV)
Sipoly20.nk (0.5eV - 6.5eV)
Sipoly30.nk (0.5eV - 6.5eV)
Sipoly40.nk (0.5eV - 6.5eV)
Sipoly50.nk (0.5eV - 6.5eV)
Sipoly60.nk (0.5eV - 6.5eV)
Sipoly70.nk (0.5eV - 6.5eV)
Sipoly80.nk (0.5eV - 6.5eV)
Sipoly90.nk (0.5eV - 6.5eV)
Sipore.nk Porous Silicon (1.0eV - 5.9eV)
Stsg.all strained SixGe1-x verses composition fraction x
Stsg0.nk Ge(strained) (2.0eV - 4.5eV)
Stsg064.nk Si.064Ge.936(strained) (2.0eV - 4.5eV)
Stsg123.nk Si.123Ge.877(strained) (2.0eV - 4.5eV)
Stsg169.nk Si.169Ge.831(strained) (2.0eV - 4.5eV)
Stsg229.nk Si.229Ge.771(strained) (2.0eV - 4.5eV)
Ta.nk Ta (0.18m - 2.48m)
Taox1.nk Tantalum Oxide (0.3m - 0.85m)
Taox2.nk Tantalum Oxide (0.25m - 0.85m)
Tasi2-a.nk Tantalum Silicide (Parallel) (0.05eV - 10eV)
Tasi2-b.nk Tantalum Silicide (Perpendicular) (0.05eV - 10eV)
Thf4.nk ThF4 (0.65eV - 5.6eV)
Ti.nk Ti (1.0eV - 6.0eV) x

1454 ATLAS User’s Manual

Optical Properties Material Systems

Table B-40 Contents of SOPRA Database

Tini.nk TiN (0.25m - 0.9m)
Tio2.nk TiO2 (0.18m - 1.5m)
Tio2b.nk TiO2 (0.22m - 2.42m)
Tisi-a.nk TiSi (0.1eV - 19.5eV)
V.nk V (0.5eV - 6.0eV)
Vsi2-a.nk VSi2 (Parallel) (0.05eV - 4.9eV)
Vsi2-b.nk VSi2 (Perpendicular) (0.05eV - 4.9eV)
W.nk W (0.05eV - 6.0eV)
Wsi2-a.nk WSi2 (Parallel) (0.1eV - 7.5eV)
Wsi2-b.nk WSi2 (Perpendicular) (0.1eV - 7.5eV)
Y2o3.nk Y2O3 (0.5eV - 6.0eV)
Zncdte.all Zn1-xCdxTe as a function of composition fraction x
Zncdte0.nk ZnTe (1.1eV - 5.4eV)
Zncdte1.nk Zn.9Cd.1Te (1.1eV - 5.4eV)
Zncdte10.nk CdTe (1.1eV - 5.4eV)
Zncdte3.nk Zn.7Cd.3Te (1.1eV - 5.4eV)
Zncdte5.nk Zn.5Cd.5Te (1.1eV - 5.4eV)
Zncdte7.nk Zn.3Cd.7Te (1.1eV - 5.4eV)
Zncdte9.nk Zn.1Cd.9Te (1.1eV - 5.4eV)
Znscub.nk ZnS (Cubic) (1.0eV - 6.0eV)
Znse.nk ZnSe (1.0eV - 6.0eV) x
Znsete.all ZnSexTe1-x as a function of composition fraction x
Znsete0.nk ZnTe x
Znsete1.nk ZnSe.1Te.9 (0.7eV - 6.3eV)
Znsete10.nk ZnSe (0.7eV - 6.3eV)
Znsete3.nk ZnSe.3Te.7 (0.7eV - 6.3eV)
Znsete5.nk ZnSe.5Te.5 (0.7eV - 6.3eV)
Znsete7.nk ZnSe.7Te.3 (0.7eV - 6.3eV)
Znsete9.nk ZnSe.9Te.1 (0.7eV - 6.3eV)

1455 ATLAS User’s Manual

Optical Properties Material Systems

Table B-40 Contents of SOPRA Database

Zro2.nk ZrO2
Zrsi2.nk ZrSi2
gan-tit.NK

sin_bf5.NK

Note: You can add the INDEX.CHECK parameter to the SOLVE statement to list the values of real and
imaginary index being used in each solution. The indices are only printed when you perform the ray trace at
the first reference to a given beam on the SOLVE statement.

1456 ATLAS User’s Manual

User Defined Materials Material Systems

B.14 User Defined Materials

The current version of ATLAS doesn’t directly support user-defined materials. A simple
workaround can be done using the already existing user specifications. This workaround is
based on the use of an already existing material name and modifying the material parameters
as appropriate.
In ATLAS, material names are defined to give you a reasonable set of default material
parameters. You can override any of these defaults using the MATERIAL, IMPACT, MODEL, and
MOBILITY statements.
The key to defining new materials is to choose a material name defined in ATLAS. Then,
modify the material parameters of that material to match the user material. Here, you should
choose a material that has default parameter values that might best match the user material,
while making sure to choose a material that is not already in the user device. Then, associate
this material name with the device regions where the new material is present. To do this,
either specify the chosen material name on the appropriate REGION statements (when the
device is defined in the ATLAS syntax) or choose the material name from the materials menu
when defining the region in DEVEDIT. Next, modify the material statements using MATERIAL,
IMPACT, MOBILITY, and MODEL statements. When doing this, the MATERIAL parameter of the
given statement should be assigned to the chosen material name.
For materials with variations in composition fraction, choose a defined material with X and/or
Y composition fractions (i.e., a ternary or quaternary material). You may also find it
convenient to use C interpreter functions to define the material parameters as a function of
composition.
The C interpreter functions that are useful for this approach are: F.MUNSAT, F.MUPSAT,
F.BANDCOMP, F.VSATN, F.VSATP, F.RECOMB, F.INDEX, F.BGN, F.CONMUN, F.CONMUP,
F.COPT, F.TAUN, F.TAUP, F.GAUN, and F.GAUP.
When defining new materials, there is a minimum set of parameters that should be defined.
This set includes bandgap (EG300), electron and hole density of states (NC300 and NV300),
dielectric permittivity (PERMITIVITY), and electron and hole mobilities (MUN and MUP). For
bipolar devices, certain recombination parameters should also be defined such as: lifetimes
(TAUN and TAUP), radiative recombination rates (COPT), and Auger coefficients (AUGN and
AUGP).
For devices with variations in material composition, certain band-edge alignment parameters
should also be defined: either electron affinity (AFFINITY) or edge alignment (ALIGN). See
Section 5.1.3 “Temperature Dependence of Electron Affinity” if you’re using F.BANDCOMP to
specify a lattice temperature dependent AFFINITY. If impact ionization is considered, the
impact ionization coefficients should also be defined.
As an example, consider the case where a device is simulating with an AlInGaP region. In
Table B-1, we see that this material system is not defined in ATLAS. We then choose a
material that’s defined in ATLAS, which has default material parameters that best
approximate the material parameters of the new material. In this case, we choose InGaAsP
since, at least for this example, we feel that this material is closest to the AlInGaP. Next, we
must specify InGaAsP as the material of the region(s) that is/are composed of AlInGaP. This
can be done either on the REGION statement if the structure is defined in ATLAS syntax or
from the material menu when the region is defined in DEVEDIT.

1457 ATLAS User’s Manual

User Defined Materials Material Systems

Supposing that we’re satisfied with the default values of the parameters from the minimum
set discussed above and that we’re principally concerned with the recombination and heat
flow parameters defaults, the following section of the input deck illustrates how these
parameter defaults can be modified:
# new material AlInGaP
MATERIAL MATERIAL=InGaAsP
# SRH
MATERIAL MATERIAL=InGaAsP TAUN0=1.1e-9 TAUP0=2.3e-8
# Auger
MATERIAL MATERIAL=InGaAsP AUGN=5.8e-30 AUGP=1.1e-31
# Optical
material material=InGaAsP COPT=1.7e-30
# Thermoconductivity
MATERIAL MATERIAL=InGaAsP TC.A=2.49
# Heat capacity
MATERIAL MATERIAL=InGaAsP HC.A=1.9

1458 ATLAS User’s Manual

Appendix C
RF and Small Signal AC
Parameter Extraction
RF and Small Signal AC Parameter Extraction

When performing small signal analysis, you can choose to capture various RF analysis
parameters [290] to the log file along with the default capacitance and conductance values.
These include Y, Z, S, ABCD, and H parameters and certain gain parameters including GUmax
(unilateral power gain), GTmax (maximum unilateral transducer power gain), k (stability
factor), and h21 (current gain in dB). In all cases, these are calculated as two port parameters.
To include any of these parameters, set the appropriate parameters of the LOG statement (see
Section 21.30 “LOG”).
All of the RF analysis parameters are calculated from the small signal capacitance and
conductance values provided during small signal analysis (see Sections 2.9.3 “Small-Signal
AC Solutions”, 2.10.3 “Parameter Extraction In DeckBuild”, and 20.10 “Small Signal and
Large Signal Analysis”).
An intermediate step in calculation of most of the RF parameters is the calculation of the Y
parameters. Equations C-1 through C-8 describe the relationship between small signal
capacitances and conductances and the complex Y parameters.
Re  Y 11  = G 11  IMPEDANCE  WIDTH C-1

Im  Y 11  = 2fC 11  IMPEDANCE  WIDTH C-2

Re  Y 12  = G 12  IMPEDANCE  WIDTH C-3

Im  Y 12  = 2fC 12  IMPEDANCE  WIDTH C-4

Re  Y 21  = G 21  IMPEDANCE  WIDTH C-5

Im  Y 21  = 2fC 21  IMPEDANCE  WIDTH C-6

Re  Y 22  = G 22  IMPEDANCE  WIDTH C-7

Im  Y 22  = 2fC 22  IMPEDANCE  WIDTH C-8

Here, IMPEDANCE is the matching impedance in ohms specified on the LOG statement. WIDTH
is the device width in microns defined on the MESH statement. f is the frequency in Hz. Cnm is
the small signal capacitance between ports n and m. Gnm is the small signal conductance
between ports n and m. Ynm is the complex Y parameter between ports n and m. Re() indicates
the real part of the argument. Im( ) indicates the imaginary part of the argument.
The complex S parameters are calculated from the complex Y parameters as given in
Equations C-9 through C-13.
1
 1 = ----------------------------------------------------------------------- C-9
 1 + Y 11   1 + Y 22  – Y 12 Y 21

1460 ATLAS User’s Manual

RF and Small Signal AC Parameter Extraction

S 11 =   1 – Y 11   1 + Y 22  + Y 12 Y 21  1 C-10

S 12 – 2Y 12  1 C-11

S 21 – 2Y 21  1 C-12

S 22 =   1 + Y 11   1 – Y 22  + Y 12 Y 21  1 C-13

During post-processing, you can choose to apply lumped resistances and inductances to the
calculation of the complex S parameters. In the first step, ATLAS is programmed to convert
the S parameters into complex Z parameters.
Equations C-14 through C-18 describe the conversion from S parameters to Z parameters.
1
 2 = --------------------------------------------------------------------- C-14
 1 – S 11   1 – S 22  – S 12 S 21

Z 11 =   1 + S 11   1 – S 22  + S 12 S 21  2 C-15

Z 12 = 2 S 12  2 C-16

Z 21 = 2 S 21  2 C-17

Z 22 =   1 – S 11   1 + S 22  + S 12 S 21  2 C-18

Once you have Z parameters, you can apply the lumped resistances and inductances as
described in Equations C-19 through C-26.

Re  Z 11  = Re  Z 11  +  RIN + RCOM   IMPEDANCE C-19


I m  Z 11  = I m  Z 11  + 2f  LIN + LCOM   IMPEDANCE C-20


Re  Z 12  = Re  Z 12  + RCOM  IMPEDANCE C-21


I m  Z 12  = I m  Z 12  + 2fLCOM  IMPEDANCE C-22


Re  Z 21  = Re  Z 21  + RCOM  IMPEDANCE C-23


I m  Z 21  = I m  Z 21  + 2fLCOM  IMPEDANCE C-24

1461 ATLAS User’s Manual

RF and Small Signal AC Parameter Extraction


Re  Z 22  = Re  Z 22  +  ROUT + RCOM   IMPEDANCE C-25


I m  Z 22  = I m  Z 22  + 2f  LOUT + LCOM   IMPEDANCE C-26

Here, RCOM, LCOM, RIN, LIN, ROUT and LOUT are the common, input and output resistances
and inductances defined on the LOG statement.
Once you get the updated Z parameters, you can convert them back into S parameters as
described in Equations C-27 through C-31.
1
 3 = ----------------------------------------------------------------------- C-27
 1 + Z 11   1 + Z 22  – Z 12 Z 21

S 11 =   Z 11 – 1   Z 22 + 1  – Z 12 Z 21  3 C-28

S 12 = 2Z 12  3 C-29

S 21 = 2Z 21  3 C-30

S 22 =   Z 11 + 1   Z 22 – 1  – Z 12 Z 21  3 C-31

When not involving the application of lumped resistance and inductance to the S parameters,
the Z parameters are calculated directly from the Y parameters as described by Equations C-
32 through C-36.
1
 4 = ----------------------------------------------
- C-32
 Y 11 Y 22 – Y 12 Y 21 

Z 11 = Y 22  4 C-33

Z 12 = – Y 12  4 C-34

Z 21 = – Y 21  4 C-35

Z 22 = Y 11  4 C-36

The H parameters can also be calculated directly from the Y parameters as described in
Equations C-37 through C-40.
H 11 = 1  Y 11 C-37

1462 ATLAS User’s Manual

RF and Small Signal AC Parameter Extraction

H 11 = – Y 12  Y 11 C-38

H 21 = Y 21  Y 11 C-39

H 22 =  Y 11 Y 22 – Y 12 Y 21   Y 11 C-40

The ABCD parameters are also directly calculated from the Y parameters as described by
Equations C-41 through C-44.
A = – Y 22  Y 21 C-41

B = – 1  Y 21 C-42

C = –  Y 11 Y 22 – Y 12 Y 21   Y 21 C-43

D = – Y 11  Y 21 C-44

The gain parameters k, GUmax, GTmax and h21 are calculated from the S parameters as given
by Equations C-45 through C-50.
 5 = S 11 S 22 – S 12 S 21 C-45

2 2 2
1 – S 11 – S 22 +  5
k = --------------------------------------------------------
- C-46
2 S 12 S 21

2
0.5 S 21  S 12 – 1
GUmax = -----------------------------------------------------------------
- C-47
k S 21  S 12 – Re  S 21  S 12 

2
S 21
GTmax = --------------------------------------------------------
- C-48
2 2
 1 – S 11   1 – S 22 

2
– 2 S 21
 6 = ---------------------------------------------------------------------
- C-49
 1 – S 11   1 + S 22  + S 12 S 21

h 21 = 20 log 10   6  C-50

1463 ATLAS User’s Manual

Appendix D
MC Device Files
The Default Configuration File: defaults.in MC Device Files

D.1 The Default Configuration File: defaults.in

#@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
# This is the default config file for the current version of moca.
# All parameters that make sense are assigned reasonable default
values
# (e.g., temp = 300K cut=(1,0,0) etc).
# All others are set to illegal values (such as -1 or 0).
#@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@

#-----------------------------------------------------------------
--------
# General simulation parameters
# If maxwell=yes and restart=yes, positions are read from the
# the restart file, but momentums are given a Maxwellian
distribution.
#-----------------------------------------------------------------
--------

algo mode=-1 carrier=e dt=2e-15 iter=0 trans=0 restart=no

seed=4212345 \
maxwell=no

reflect mindist=-1E0

debug npick=0

output scatoutfile=”scat.out” summaryoutfile=”summary.out” \

difflogfile=”diff.log” currentlogfile=”current.log” \
difframplogfile=”diff_ramp.log”
currentramplogfile=”current_ramp.log” \
solstrfile=”sol.str” \
tstep=-1 xstep=(1,1,1) estep=50 restart=-1 wpstep=pinf \
init=no hydro=no hist=yes qeta=no outfiles=no

#-----------------------------------------------------------------
--------

1465 ATLAS User’s Manual

The Default Configuration File: defaults.in MC Device Files

# dz [=] cm. If dz > 0.0, device width is fixed to this value,

# otherwise, if conc is set device width is set based on conc
# otherwise, width is set to achieve N particles using initial
guess
# for particle concentration
#-----------------------------------------------------------------
--------
particle n=0 nnorm=0 conc=0.0 dz=-1.0

# The threshold energy for DoS-dominated phonon scattering is fixed

at
# 1.5eV, i.e., sufficiently above the L valley to account properly
for
# XL scattering but still more or less within the validity of the
k.p
# nonparabolic approximation for the first valley.
#
# bf eth=1.70, where non-parabolic dos and full-band dos do not
differ
# bf eth=0.63, where non-parabolic dos and full-band dos do not
differ
# bf eth=0.57, to get agreement with average over measured ii-
coeff. (Kane)
# bf eth=0.42, to get agreement with lowest measured ii-coeff.
(Cartier)
phon eth=0.42 model=0 file=”dos_c.in”

#-----------------------------------------------------------------
--------
# Interpolation of energy from k, using bandstructure tables
#
# 0 = most accurate cubic interpolation
# For strain model (when using “mmat type=1”):
# Include the quad correction for energy and velocity WITHOUT
USING
# interpolated fitting parameters for energy/velocity as
# a function of mole fraction.
#
# 1 = less accurate linear interpolation
# For strain model (when using “mmat type=1”):

1466 ATLAS User’s Manual

The Default Configuration File: defaults.in MC Device Files

# Include the quad correction for energy and velocity USING

# interpolated fitting parameters for energy/velocity as
# a function of mole fraction.
#
# 2 = more accurate linear interpolation
# For strain model (when using “mmat type=1”):
# Include the quad correction for energy and velocity USING
# interpolated fitting parameters for energy/velocity as
# a function of mole fraction.
#
#-----------------------------------------------------------------
--------

bands interp=0 n=2 nk=40 file=”eder40_2.in”

# The setttings for the n, nk, and file parameters were moved
# from the bands2d.elec.in file to here.
# n=(2,3)
# nk=(40,40)
# interp=(0,0)
# file=(“eder40_2.in”,”eder40_3v.in”)

qbands interp=0 n=1 nk=100 file=”qbandtest”

# The settings for the n, nk, and file parameters were moved
# from the bands2d.elec.in file to here.
# The file parameter on the qbands statement is not presently used.

#-----------------------------------------------------------------
--------
# Bulk, 1-D, and 2-D simulation parameters
#-----------------------------------------------------------------
--------

mat cut=(1.0,0.0,0.0) temp=300.0

bulk field=(0.0,0.0) doping=0.0 conc=0.0 memory=no mmfrac=0.0

1467 ATLAS User’s Manual

The Default Configuration File: defaults.in MC Device Files

oned mode=1 size=(0.0,0.0,0.0) nx=(100,0,0) conc=(0.0,0.0)

reflect1d=(no,no) vbias=0.0

enhan mode=0 tstep=-1 estep=100 xstep=(0,0,0) \

damp=0.0 lothre=1e-20 hithre=1e20 relthre=0.05 ratio=2.0 n=10
nfrac=0.95

#-----------------------------------------------------------------
--------
#traj: If mode is not 0, then particles will be initialized to be
uniformly
# distributed in the BOUND box. If mode=2, detailed
information
# will be stored about particle trajectories. If mode=1,
detailed
# will not be stored. If ZEROSCAT is not 0, all scattering
# in the simulation will be disabled.
#-----------------------------------------------------------------
--------
traj mode=0 zeroscat=no bound=(0.0,0.0,0.0,0.0)
boundp=(0.0,0.0,0.0,0.0)

tunnel mode=0

excit energy=0.0 shape=0 width=0.0 rate=0.0 alpha=(0.0,0.0,0.0)

#-----------------------------------------------------------------
--------
#cutoff: maximum cutoff for doping if complete ionization is used.
#-----------------------------------------------------------------
--------
poisson tstep=5 iter=100 tol=(2e-4,1e-6) ioniz=no start=1
freeze=pinf \
linpoi=no cutoff=1e25 ipoly=no hideplasmons=no

#-----------------------------------------------------------------
--------
#negfield: If yes, getfield, will produce negative of the field
calculated

1468 ATLAS User’s Manual

The Default Configuration File: defaults.in MC Device Files

# from the potential. Useful in a hole frozen field

simulation,
# where potential is read in from an electron simulation.
#-----------------------------------------------------------------
--------
field maxfield=(-1.0,-1.0) ramp=(-1,-1) negfield=no usebarr=no

quantum restart=no option=(no,no) decay=(100e-15,2000e-15) \

minconc=(1e10,1e10) maxqv=-1.0 maxqf=-1.0 strength=1.0 \
q1d=(-1,-1,-1) freeze=pinf sigma=8e-8 radius=3 ladders=(1,0,0) \
energy=no ttemp=no

import type=8 volt=false nconc=false pconc=false dop=false \

avgvolt=false avgnconc=false avgpconc=false \
mesh=false regions=false \
dipsh=no adddop=no bound=(0,0,0,0) \
boundp=(0.0,0.0,0.0,0.0) \
voltfile=”volt.in” concfile=”conc.in” dopfile=”dop.in” \
gridfile=”grid_filename” solfile=”sol_filename” \
verfile=”ver_filename” trifile=”tri_filename” \
sdfile=”sd_filename” retrofile=”retro_filename” \
strfile=”str_filename”

#-----------------------------------------------------------------
--------
# Physical models
#
# Phonon-scattering parameters from Tang and Hess, JAP 54:5139
# bf threshold in bands.elec.in is set so that xl vanishes
# this fairly reproduces the drift velocity versus field charac-
# teristics at 300 and 77K.
#-----------------------------------------------------------------
--------

xxf temp=(220.0,550.0,685.0) defo=(3e7,2e8,2e8)

xxg temp=(140.0,215.0,720.0) defo=(5e7,8e7,11e8)

1469 ATLAS User’s Manual

The Default Configuration File: defaults.in MC Device Files

xl temp=(672.0,634.0,480.0,197.0) defo=(2e8,2e8,2e8,2e8)

ac defo=9.0

#-----------------------------------------------------------------
--------
# Parameters for all four impact-ionization models
# (default=Cartier, then Cartier et al, APL 62(25):3339).
# model=Cartier
# final=no: Sets all 3 particles after ii event to zero
energy.
# weight=yes: Doubles the weight of the primary impacting
particle.
# sec=yes: Generates a secondary particle equal in weight to
# primary.
# csec=yes: Generates a complementary secondary particle for
slave
# process.
# randsec=yes: Secondary particles are randomly placed in rspos
region.
# rspos=(min max): If randsec=yes, specifies minimum and maximum
radii (cm)
# of a ring-shaped region in which secondaries are
placed.
# The ring is centered about the location of ii.
#-----------------------------------------------------------------
--------

impact model=cartier param=(6.25e10,3e12,6.8e14)

eth=(1.2,1.8,3.45) \
final=yes weight=no sec=no csec=no start=1000 randsec=no \
rspos=(0.0,0.0)

#-----------------------------------------------------------------
--------
# default scale = 1
# bf scale = 1.9 to get agreement with the experimental doping
dependent
# low-field mobility of majority carriers

1470 ATLAS User’s Manual

The Default Configuration File: defaults.in MC Device Files

#-----------------------------------------------------------------
--------
ridley tstep=-1 scale=1.9

#-----------------------------------------------------------------
--------
# Parameters for analytic surface scattering model
# For more info, see
# J. Appl. Phys., vol. 79, Jan 1996, p. 911
#
# In some calibration tests, reducing SEXPL to 11.0e-10 reduced
rolloff
# in mobility at high fields and gave a better fit to the universal
# mobility curves from Tagaki: Trans. on ED, 41, december 1992, p.
2357
#-----------------------------------------------------------------
--------

ssparam sexpl=22.0e-10 sexpd=1.78e-10 ssfield=-1.0

file=”ssphonon.dat” \
surfphon=(0.0,0.0)
# surfphon=(7e10,0.34)
# Values above were used in a bulk example.

#-----------------------------------------------------------------
--------
# Extra crossing estimators (energy distributions as a function of
space)
# are calculated for the crossest region (ignored for crossreg = 0)
#-----------------------------------------------------------------
--------

crossest crossreg=0

#-----------------------------------------------------------------
--------
# Pipe sets up interprocess communication which lets an electron
and a hole
# simulation exchange secondary particles created by impact
ionization.

1471 ATLAS User’s Manual

The Default Configuration File: defaults.in MC Device Files

# If mode = master, the simulation proceeds as usual except that

secondaries
# are exchanged. If mode=slave, the particles move in fields which
are
# received from the master. The communcation occurs in the pipe
files
# designated by msfile and smfile. These files must be specified
exactly
# the same in both the master and slave command files. If initnlin
= yes,
# non-linear poisson is solved once at the beginning of the
simulation to
# determine the initial concentration of slave particles. If
initnlin=no,
# the concentration is set by charge neurality. initnlin must also
be
# specified the same in slave and master command files.
#-----------------------------------------------------------------
--------

bipolar comm=none msfile=”msfile” smfile=”smfile” nslave=0

nfracii=0.5 initnlin=yes

#-----------------------------------------------------------------
--------
#INJECT block controls injection from contact
#
# dn: fraction of mesh to inject in normal direction
# dp: fraction of mesh to inject in parallel direction
# hemimax: if no, inject carriers on a full maxwellian
distribution
# overinj: Maintain overinj fraction more carriers at interface
than
# required by charge neutrality
#-----------------------------------------------------------------
--------
inject dn=0.0 dp=1.0 hemimax=no overinj=1e4 tstep=-1

#-----------------------------------------------------------------
--------
# The treatment of broadening is a bit incomplete and inconsistent.

1472 ATLAS User’s Manual

The Default Configuration File: defaults.in MC Device Files

# Also, bulk simulations do not appear to be sensitive to cutoff,

so,
# play it safe and turn off broadening. (cutoff = 0.0)
#-----------------------------------------------------------------
--------
final cutoff=0.0 n=178152 nk=100 format=0 file=”eord100_2.in”
# The setting for n, nk, format, and file parameter were moved
# from # bands2d.elec.in to here.
# NK= (100,100)
# N=(178152,267228)
# FORMAT=(0,0)
# FILE=(“eord100_2.in”,”eord100_3v.in”)

#-----------------------------------------------------------------
--------
# Fullband numerical scattering rates
#-----------------------------------------------------------------
--------

escat type=analytic fbproc=7 fbband=4 anproc=24 anband=1 \

file=”scatfl_c_uni.in” \
phncplfile=”phcpl_c.in” \
phnfacfile=”phon_fac_c_uni_300k.in” \
plotfile=”scateng_c.in”

hscat type=analytic fbproc=7 fbband=3 anproc=24 anband=1 \

file=”scatfl_v.in” \
phncplfile=”phcpl_v.in” \
phnfacfile=”phon_fac_v.in” \
plotfile=”scateng_v.in”

#-----------------------------------------------------------------
--------
# Multimaterial options
#-----------------------------------------------------------------
--------

mmat type=no

1473 ATLAS User’s Manual

The Default Configuration File: defaults.in MC Device Files

#-----------------------------------------------------------------
--------
# Multimaterial for electrons for strained silicon
#-----------------------------------------------------------------
--------

ssielec band=”eder40_c_ssi_interp.in” \
rate=”rates_ssi_c_45_uni.in” phncpl=”coupl_ssi_c.in” \
latt=”latts_ssi_c.in”

units length=um

#-----------------------------------------------------------------
--------
# Material definition for 2-D simulation
#-----------------------------------------------------------------
--------

#Note: Hard-coded associations exist (in mat.*) for materials

# with n=1, 2, 3, 4, 5, 11, and 12. These should not be
changed
# to relate to different materials than their names below.
matdef n=1 name=”Air” eps= 1.0 barrier=0.0 rough=0.0 \
egap=1.1245 affinity=4.17
matdef n=2 name=”Si” eps=11.7 barrier=0.0 rough=0.0 \
egap=1.1245 affinity=4.17
matdef n=3 name=”Poly” eps=11.7 barrier=0.0 rough=0.0 \
egap=1.1245 affinity=4.17
# For SiO2, use barrier=4.0 eV instead of barrier=3.15 for hole
simulation.
# (For example, when simulating a p-MOSFET with “ALGO CARRIER=H”.)
#matdef n=4 name=”SiO2” eps= 3.9 barrier=3.15 rough=0.10 \
# egap=1.1245 affinity=4.17
matdef n=4 name=”SiO2” eps= 3.9 barrier=3.15 rough=0.25 \
egap=8.9000 affinity=1.02
matdef n=5 name=”Si3N4” eps= 7.5 barrier=3.15 rough=0.0 \
egap=1.1245 affinity=1.02

1474 ATLAS User’s Manual

The Default Configuration File: defaults.in MC Device Files

matdef n=6 name=”Al” eps= 1.0 barrier=0.69 rough=0.0 \

egap=1.1245 affinity=3.48
matdef n=7 name=”Pt” eps= 1.0 barrier=0.85 rough=0.0 \
egap=1.1245 affinity=3.32
matdef n=8 name=”W” eps= 1.0 barrier=0.65 rough=0.0 \
egap=1.1245 affinity=3.52

#use barrier=0.34? for hole simulation

matdef n=9 name=”Au” eps= 1.0 barrier=0.79 rough=0.0 \
egap=1.1245 affinity=3.38

# I will leave the two definitions below commented for now.

# Note that these already exist in the hard codes mentioned above.
#matdef n=11 name=”SSi” eps=11.7 barrier=0.0 rough=0.0 \
# egap=1.1245 affinity=4.17
#matdef n=12 name=”SiGe” eps=11.7 barrier=0.0 rough=0.0 \
# egap=1.1245 affinity=4.17

1475 ATLAS User’s Manual

File Formats MC Device Files

D.2 File Formats

This appendix provides a description of input and output file formats used by MC DEVICE. By
default, MC DEVICE writes diff.log files for bulk simulations and current.log and
sol.str files for 2D simulations. Section D.2.1 “Silvaco Output Files” describes these file
formats.
Many output files are written using the UIUC MOCA file formats. These output files use the
suffix .out are only written when the OUTFILES parameter on the OUTPUT statement is set to
1|TRUE|YES. By default, the OUTFILES parameter in the OUTPUT section is set to 0 and
these output files are suppressed. In addition to the OUTFILES parameter, some *.out files
have additional boolean control parameters (see Section 19.3.4 “Commonly-Used
Statements”). Sections D.2.2 “General Output Files” through Section D.2.6 “Miscellaneous
Input And Output Files” describe the UIUC MOCA output files.
Some input files use UIUC MOCA file formats. These input files use the suffix .in and have
the same format as the corresponding UIUC MOCA output (.out) file.
Many files written in the UIUC MOCA file format consist of three columns. The first two
columns are the x and y positions (in cm) and the third column is data field associated with
that file in cgs units. You can see from the x and y values that the rows (all values of y) are
scanned for each column (each value of x with y fixed) and then for each row (each value of
y). Note that data (x and y values and data field) may be written at the nodes of the mesh (for
potential and MC carrier concentration) or at the element centers (for the doping). Quantities
are averaged over iterations greater than or equal to ISTART (or time>=0). All output files are
in ASCII format.

D.2.1 Silvaco Output Files

*diff*.log files with this extended suffix contain several time-dependent time-averaged
quantities for a bulk simulation. They are as follows:
• Time = [s].
• Average Position = [cm].
• Variance of Position = [cm2].
• Diffusion Coefficient = [cm/s].
• Diffusion-Derived Mobility = [cm2/V/s].
• Drift-derived Mobility = [cm2/V/s].
Here is the format of the columns:
Time <X(t)> <X(t)2>-<X>2 Diff_Coeff(t) Mobility_D(t)
Mobility_E(t)
You can change this filename with the DIFFLOGFILE parameter on the OUTPUT statement (see
Section 19.3.4 “Commonly-Used Statements”). We recommend you use the extended file
suffix diff.log (as shown in the examples) to keep this filename consistent with this
documentation.

1476 ATLAS User’s Manual

File Formats MC Device Files

Note: The first aster isk in the file name above respresents the base file name. For example,
mcdeviceex01_diff.log. The second asterisk in the file name above represents a possible ramp
step. For example, mcdeviceex01_diff_field=1000.log.

*diff_ramp.log files with this extended suffix contain several time-averaged quantities for
a bulk simulation for different bias voltages, which typically represent the steady-state
solution for each bias voltage. The quantities are the same as those for *diff*.log but
exclude the Time. Instead, they include the Electric Field = [V/cm].
You can change this filename with the DIFFRAMPLOGFILE parameter on the OUTPUT
statement (see Section 19.3.4 “Commonly-Used Statements”). We recommend you use the
extended file suffix diff.log (as shown in the examples) to keep this filename consistent
with this documentation.
*current*.log files with this extended suffix contain time-dependent time-averaged and
space-averaged currents over CREGIONs defined in the input file. They are as follows:
• Time = [s].
• Current = [A/cm]. The sign convension is given at the end of the current regions section.
See “Current Regions” on page 840.
• Velocity = [cm/s].
• Concentration-weighted force = [V/cm].
• Mobility = [cm2/V/s].
Each current, velocity, force, and mobility has first an x-directed, then a y-directed column
denoted by dim=1 and dim=2 respectively. Each current, velocity, force, and mobility is also
written for each current region and denoted by creg=1 up to creg=n, where n is the number
of current regions.
Here is the format of the columns:
Time Current(t) Velocity(t) Force(t) Mobility(t)
You can change this filename with the CURRENTLOGFILE parameter on the OUTPUT statement
(see Section 19.3.4 “Commonly-Used Statements”). We recommend you use the extended file
suffix current.log (as shown in the examples) to keep this filename consistent with this
documentation.

Note: The first aster isk in the file name above respresents the base file name. For example,
mcdeviceex02_current.log. The second asterisk in the file name above represents a possible
ramp step. For example, mcdeviceex02_current_vdrain=1.log.

*current_ramp.log files with this extended suffix contain several time-averaged and
space-averaged currents over CREGIONs defined in the input file for different bias voltages,
which typically represent the steady-state solution for each bias voltage. The quantities are
the same as those for *current*.log but exclude the Time. Instead, they include the contact
voltage as vcontact_name = [V], where contact_name represents the name of the
contacted for which the voltage is ramped. For example, vdrain.

1477 ATLAS User’s Manual

File Formats MC Device Files

You can change this filename with the CURRENTRAMPLOGFILE parameter on the OUTPUT
statement (see Section 19.3.4 “Commonly-Used Statements”). We recommend you use the
extended file suffix current_ramp.log (as shown in the examples) to keep this filename
consistent with this documentation.
*sol*.str files contain quantities for a 2D simulation. They are as follows:
• Arsenic = [cm-3] (when DOP=0 on IMPORT or IONIZ=1 on POISSON)
• Band gap = [eV]
• Boron = [cm-3] (when DOP=0 on IMPORT or IONIZ=1 on POISSON)
• Conduction Band DOS = [cm-3]
• Conudction Band Energy = [eV]
• e- Energy = [eV] (when CARRIER=E on ALGO)
• e- Velocity X = [cm/s]
• e- Velocity Y = [cm/s]
• E Field X = [V/cm]
• E Field Y = [V/cm]
• (EHP) Generation Rate = [s1]
• Electron Conc = [cm-3]
• Electron QFL = [eV]
• h+ Energy = [eV] (when CARRIER=H on ALGO)
• h+ Velocity X = [cm/s]
• h+ Velocity Y = [cm/s]
• Hole Conc = [cm-3]
• Hole QFL = [eV]
• Impact (Ionization) Gen(eration) Rate = [cm-3/s]
• Indium = [cm-3] (when DOP=0 on IMPORT or IONIZ=1 on POISSON)
• Instant Electron Conc = [cm-3]
• Instant Hole Conc = [cm-3
• Instant Potential = [V]
• Net Doping = -Boron + Phosphorus - Indium + Arsenic = [cm-3]
• Phosphorus = [cm-3] (when DOP=0 on IMPORT or IONIZ=1 on POISSON)
• Potential = [V]
• Relative Permittivity = [-]
• Total Doping = B + P + In + As = [cm-3] (when DOP=0 on IMPORT or IONIZ=1 on
POISSON)
• Valence Band DOS = [cm-3]
• Valence Band Energy = [eV]
All quantitites above beginning with “Average” are time averages from time=0. These
quantities are also Monte Carlo carrier-ensemble averages from the MC carriers, which
resided in the rectangular box centered about each mesh node.

1478 ATLAS User’s Manual

File Formats MC Device Files

You can change this filename with the SOLSTRFILE parameter on the OUTPUT statement (see
Section 19.3.4 “Commonly-Used Statements”). We recommend you use the extended file
suffix sol.str (as shown in the examples) to keep this filename consistent with this
documentation.

Note: The first aster isk in the file name above respresents the base file name. For example,
mcdeviceex02_sol.str. The second asterisk in the file name above represents a possible ramp
step. For example, mcdeviceex02_sol_vdrain=1.str.

D.2.2 General Output Files

diff.out contains several time-dependent time-averaged quantities for a bulk simulation.
They are as follows:
• Time = [s].
• Average Position = [cm].
• Variance of Position = [cm2].
• Diffusion Coefficient = [cm2/s].
• Diffusion-Derived Mobility = [cm2/V/s].
• Drift-derived Mobility = [cm2/V/s].
Here is the format of the columns:
Time <X(t)> <X(t)2>-<X>2 Diff_Coeff(t) Mobility_D(t)
Mobility_E(t)
sol.out contains several time-averaged quantities for a 2D simulation. They are as follows:
• X/Y = [cm].
• Potential = [V].
• Force = [V/cm].
• Particle concentration = [cm3].
• Velocity = [cm/s].
• Energy = [eV].
• Impact Ionization Rate = [cm-3/s].
• Electron-Hole Pair (EHP) generation rate = [s1].
Force is the electric field for holes and its negative for electrons. Force and velocity each
have two components: the first is x-directed, the second is y-directed.
Here is the format of the columns:
X Y Potential Force Concentration Velocity Energy
Generation
xhist.out, yhist.out contain the energy distribution along slices in the X/Y directions.
They are as follows:
• Energy = [eV].
• X/Y slice = [cm].
• Distribution = [cm-3eV-1].

File Formats MC Device Files

D.2.5 Quantum Output Files

calcsp2moca.out contains the transformation between the MC DEVICE solution grid and
the Schrödinger grid. This file is only output if the quantum correction is turned on. The first
column is the MC DEVICE grid point. The second column is the quantum correction grid point
that falls closest to the MC DEVICE grid point. The third column is the relative distance of the
SP grid point between MC grid point i and i+1. The last two columns are the real-space
values of the quantum correction grid and the MC DEVICE grid.
Here is the format of the columns:
MC_grid SP_grid R SP_dist MC_dist
etabarr.out, etabarr1.out show the Eta barrier for each mesh area. The Eta barrier for
each mesh area is taken to be the average of the eta barriers of the materials on each of the 4
sides of the mesh. The file is written as
for y=ymin:ymax {
for x=xmin:xmax {
write etabarr
}
}
spgrid.out contains the definition of Schrödinger-Poisson grid. It is the same as with
xaxis.out. The first column is the grid point index. The second column is the real-space
coordinate in [cm].
grid_point X
qatten.out contains quantum correction attenuation factor through the whole device
traversing it one y-grid line at a time.
qcap.out contains quantum solution for simulation of a MOS capacitor with the same
bounds as in qregion. The first two columns are the X and Y coordinates respectively, in Å.
The next two columns are carrier concentration and energy respectively. Here is the format of
the columns:
X Y Concentration Energy
qhist.out contains quantum correction histogram. The first column is the energy bin in
[eV] and the second is the energy histogram of particles in the quantum region.
qrestart.out shows the transient quantum correction particle status. The quantum Eta for
each grid point in the device structure and traversed y-grid line by line.
qsol.out contains quantum correction throughout the entire device in [eV]. The results are
written for each increasing y-grid line.
qsol.trans.out has exactly the same format as qsol.out but is the transient solution.
qstatus.out contains the quantum correction particle status. The quantum Eta for each grid
point in the device structure, traversed y-grid line by line.
qvolt.out contains quantum corrected quantum potential. This may be used as an input by
copying it to qvolt.in and using the IMPORT statement.

1483 ATLAS User’s Manual

File Formats MC Device Files

ttemp.out contains transverse temperature in [K] for the transverse grid lines. First column
is the grid line number. The third column is the actual transverse temperature. Columns 2 and
4 are fitting parameters for the Marquardt fit method. Here is the format of the columns:
Grid_Line Param1 Transverse_Temp Param2

D.2.6 Miscellaneous Input And Output Files

inj_2.in contains parameters where you tell MC DEVICE how you want to simulate a
Schottky barrier device. bar_poisson must equal the value of the silicide barrier height set
in your input command (*.in) file. hqflreg is typically set to the substrate contact region. It
is best not to modify any of the parameters relating to the Airy functions.
balance.out contains the variance reduction balancing grid. The file contains the size of the
balancing grid and the number of energy bins.
oxide0x.out contains the oxide injection currents for the various current estimator regions.
crosshist.out, xcross.out, ycross.out contains tracked energies of particles that
change regions.
fieldeff.out contains the effective electric field in a surface scattering region. The field is
written along the x grid points only. The field is given in [V/cm].
geom.out contains the boundaries [cm] of all regions in the order they were loaded or
defined in the input file. Unlike the *summary.out file (see Section D.2.2 “General Output
Files”), which only contains region definitions in the input file, geom.out contains all region
definitions including those obtained by importing the structure.
mat.out contains the location of the cell centers [cm], the relative permittivity [-], the band
gap [eV], the density of states in the conduction band [cm-3], the density of states in the
valence band [cm-3], and the quasi-Fermi potential [V]. The quasi-Fermi potential is used to
calculate the density of non-MC carriers. It is also used to calculate the ionized doping
concentration when using incomplete ionization.
mesh.out contains the node location [cm], mesh areas ([cm2] for 2D) inside cells of type MC
(with a negative average area where it would be 0), the Poisson boundary conditions [V], and
the Poisson boundary conditions types where 0=Dirichlet, 1=Neumann, 2=Interior.

1484 ATLAS User’s Manual

File Formats MC Device Files

D.2.7 Band Structure Input Files

Three types of band structure files are used in building the data files used in the MC
simulation: band structure (used for high-resolution energy calculations), gradients, and
eigenstates. Figure D-1 shows the formats. The files are located in the $MCDEVICEDIR
directory and are named eder40_*.in.
a.

kx ky kz E0 E1 ... EN – 1
b

.
.
.

kx ky kz E 0 ... E N – 1 ... E 0 ... E N – 1

b b
x x z z

.
.
.

kx ky kz E0 ... EN – 1
b

0 1
c0
... Nc
c0 c0

.
.
.

Figure D-1: Formats of the band structure files: a. energies, b. gradients, c. eigenstates.
The final state is chosen based on final energy. To make this task efficient, the bandstructure
tables are reordered according to energy. Figure D-2 shows the format. The files are in the
$MCDEVICEDIR directory and are named reord100_*.in.

kx ky kz E n

.
.
.

Figure D-2: Format of the final-state table.

1485 ATLAS User’s Manual

File Formats MC Device Files

There is also a density of states table for electrons and holes. These files are located in the
same directory as the bandstructure files and are named dos_c.in for the conduction band
and dos_v.in for the valence band. The format of these files is a single column of
probabilities of occupation discretized over 100 pts.

1486 ATLAS User’s Manual

Examples MC Device Files

D.3 Examples
This appendix contains several example input files for MC DEVICE. These include Section
D.3.1 “Bulk n-type Silicon: mcdeviceex01”, Section D.3.2 “A 25-nm n-MOSFET:
mcdeviceex02”, Section D.3.3 “A 25-nm n-MOSFET with Quantum Correction:
mcdeviceex03”, Section D.3.4 “A 25-nm p-MOSFET: mcdeviceex04”, and Section D.3.5
“An imported 25-nm n-MOSFET: mcdeviceex05”. The example input files below and the
sample output and TONYPLOT set files (for each output file) may be accessed with
DECKBUILD (see Section 19.3.6 “Accessing The Examples”).

D.3.1 Bulk n-type Silicon: mcdeviceex01

go atlas
mcdevice
# Bulk simulation for electron transport in silicon at STP
algo mode=0 carrier=e iter=10000 dt=1e-15
# bulk simulation
bulk field=(5e2,5e2)
#bulk field=(1e3,1e3)
# write output
output tstep=1000 \
difflogfile =”mcdeviceex01_diff.log” \
difframplogfile=”mcdeviceex01_diff_ramp.log” \
summaryoutfile =”mcdeviceex01_summary.out”
# Set the number of particles (i.e. electrons).
particle n=10000

# Do a single solve at field=(5e2,5e2) V/cm.

solve
tonyplot mcdeviceex01_diff.log -set mcdeviceex01_diff_log.set

# Do a 1-step field ramp from field=5e2 V/cm to field=1e3 V/cm.

# with uniform spacing on a linear scale.
#solve fstep=5e2 ffinal=1e3
#tonyplot mcdeviceex01_diff_ramp.log -set
mcdeviceex01_diff_ramp_log.set

# Do a 3-step field ramp from field=1e3 V/cm to field=1e6 V/cm.

# with uniform spacing on a logarithmic scale.
#bulk field=(1e3,1e3)

1487 ATLAS User’s Manual

Examples MC Device Files

#solve nsteps=3 ffinal=1e6 uselog=1

#tonyplot mcdeviceex01_diff_ramp.log -set
mcdeviceex01_diff_ramp_log.set

# Do a second solve at field=(1e3,1e3) V/cm.

# Do a single solve for vgate=vdrain=1.0 V.

solve vgate=1.0 vdrain=1.0
tonyplot mcdeviceex03_current.log -set
mcdeviceex03_current_log.set
tonyplot mcdeviceex03_sol.str -set mcdeviceex03_sol_str.set

# Do a 1-step voltage ramp from vdrain=1.0 V to vdrain=2.0 V.

#solve vgate=1.0 vdrain=1.0 name=”drain” vfinal=2.0 vstep=1.0

# Do a second solve at Vgate=Vdrain=2.0 V.

#output currentlogfile=”mcdeviceex03_current2.log” \
# solstrfile =”mcdeviceex03_sol2.str”
#solve vgate=2.0 vdrain=2.0
#tonyplot mcdeviceex03_current2.log -set
mcdeviceex03_current_log.set
#tonyplot mcdeviceex03_sol2.str -set mcdeviceex03_sol_str.set

1494 ATLAS User’s Manual

Examples MC Device Files

# Do a singe solve at Vgate=vdrain=-1.0 V.

solve vgate=-1.0 vdrain=-1.0
tonyplot mcdeviceex04_current.log -set
mcdeviceex04_current_log.set
tonyplot mcdeviceex04_sol.str -set mcdeviceex04_sol_str.set

# Do a 1-step voltage ramp from vdrain=-1 V to vdrain=-2 V.

#solve vgate=-1.0 vdrain=-1.0 name=”drain” vfinal=-2.0 vstep=-1.0

# Do a second solve at Vgate=vdrain=-2.0 V.

#output currentlogfile=”mcdeviceex04_current2.log” \
# solstrfile =”mcdeviceex04_sol2.str”
#solve vgate=-2.0 vdrain=-2.0
#tonyplot mcdeviceex04_current2.log -set
mcdeviceex04_current_log.set
#tonyplot mcdeviceex04_sol2.str -set mcdeviceex04_sol_str.set

quit

1498 ATLAS User’s Manual

Examples MC Device Files

D.3.5 An imported 25-nm n-MOSFET: mcdeviceex05

go atlas
mcdevice
algo mode=2 iter=2500 dt=0.1e-15
import mesh=true regions=true dop=true \
strfile=”mcdeviceex05_atlas.str”
particle n=40000
region name=”source” type=contact
region name=”drain” type=contact usefermi=1
region name=”gate” type=contact
region name=”substrate” type=contact
cregion boundp=(0.0275,0.0675,0.0,0.08)
cregion boundp=(0.0325,0.0475,0.0,0.08)
cregion boundp=(0.0475,0.0625,0.0,0.08)
output tstep=500 \
currentlogfile =”mcdeviceex05_current.log” \
currentramplogfile=”mcdeviceex05_current_ramp.log” \
solstrfile =”mcdeviceex05_sol.str” \
summaryoutfile =”mcdeviceex05_summary.out”

# Do a single solve at Vgate=vdrain=1.0 V.

solve vgate=1.0 vdrain=1.0
tonyplot mcdeviceex05_current.log -set
mcdeviceex05_current_log.set
tonyplot mcdeviceex05_sol.str -set mcdeviceex05_sol_str.set

# Do a 1-step voltage ramp from vdrain=1.0 V to vdrain=2.0 V.

#solve vgate=1.0 vdrain=1.0 name=”drain” vfinal=2.0 vstep=1.0

# Do a second solve at Vgate=Vdrain=2.0 V.

#output currentlogfile=”mcdeviceex05_current2.log” \
# solstrfile =”mcdeviceex05_sol2.str”
#solve vgate=2.0 vdrain=2.0
#tonyplot mcdeviceex05_current2.log -set
mcdeviceex05_current_log.set
#tonyplot mcdeviceex05_sol2.str -set mcdeviceex05_sol_str.set

1499 ATLAS User’s Manual

Examples MC Device Files

quit

1500 ATLAS User’s Manual

Bibliography

1. Adachi, S. Physical Properties of III-V Semiconductor Compounds InP, InAs, GaAs, GaP,
InGaAs, and InGaAsP. New York: John Wiley and Sons, 1992.
2. Adachi, S., "Band gaps and refractive indices of AlGaAsSb, GaInAsSb, and InPAsSb:
Key properties for a variety of the 2-4 um optoelectronic device applicatios", J. Appl.
Phys., Vol. 61, No. 10, 15 (May 1987): 4896-4876.
3. Adamsone A.I., and B.S. Polsky, “3D Numerical Simulation of Transient Processes in
Semiconductor Devices”, COMPEL—The International Journal for Computation and
Mathematics in Electrical and Electronic Engineering Vol. 10, No. 3 (1991): 129-139.
4. Akira Hiroki, S. Odinaka et. al., “A Mobility Model for Submicrometer MOSFET
Simulations, including Hot-Carrier Induced Device Degradation”, IEEE Trans. Electron
Devices Vol. 35 (1988): 1487-1493.
5. Alamo, J. del, S. Swirhun and R.M. Swanson, “Simultaneous measuring of hole lifetime,
hole mobility and bandgap narrowing in heavily doped n-type silicon”, IEDM Technical
Digest, (December 1985): 290-293.
6. Albrecht, J.D. et al, "Electron transport characteristics of GaN for high temperature
device modeling", J. Appl. Phys. Vol. 83 (1998): 4777-4781.
7. Allegretto, W., A.Nathan and Henry Baltes, “Numerical Analysis of Magnetic Field
Sensitive Bipolar Devices”, IEEE Trans. on Computer-Aided Design, Vol. 10, No. 4
(1991): 501-511.
8. Ambacher, O. et. al., “Two Dimensional Electron Gases Induced by Spontaneous and
Piezoelectric Polarization in Undoped and Doped AlGaN/GaN Heterosturctures”, J.
Appl. Phys. Vol. 87, No. 1, (1 Jan. 2000): 334-344.
9. Anderson, D.G., “Iterative Procedure for Nonlinear Integral Equations”, ACM Journal
Vol. 12, No. 4 (n.d.): 547-560.
10. Andrews, J.M., and M.P. Lepselter, “Reverse Current-Voltage Characteristics of Metal-
Silicide Schottky Diodes”, Solid-State Electronics 13 (1970): 1011-1023.
11. Antogneti, P., and G. Massobrio. Semiconductor Device Modeling With SPICE.
McGraw-Hill Book Company, 1987.
12. Apanovich Y. et. al, “Numerical Simulation of Submicrometer Devices, Including
Coupled Non-local Transport and Non-isothermal Effects,” IEEE Trans. Electron
Devices Vol. 42, No. 5 (1995): 890-898.
13. Apanovich Y. et. al, “Stready-State and Transient Analysis of Submircon Devices Using
Energy Balance and Simplified Hydrodyamic Models,” IEEE Trans. Comp. Aided
Design of Integrated Circuits and Systems Vol. 13, No. 6 (June 1994): 702-710.
14. Arkhipov, V.I., P. Heremans, E.V. Emelianova, G.J. Adriaenssens, H. Bassler, “Charge
carrier mobility in doping disordered organic semiconductors”, Journal of Non-
Crystalline Solids, Vol. 338-340 (2004): 603-606.
15. Arkhipov, V.I., P. Heremans, E.V. Emelianova, G.J. Adriaenssens, H. Bassler, “Charge
carrier mobility in doped semiconducting polymers”, Applied Physics Letters, Vol. 82,
No. 19 (May 2003): 3245-3247.

1501 ATLAS User’s Manual

Bibliography

Semiconductor Device Modeling”, IEEE Trans., Computer-Aided Design Vol. 9, No. 11
(1990): 1141-1149.
295. Wada, M. et al. “A Two-Dimensional Computer Simulation of Hot Carrier Effects in
MOSFETs.” Proc. IEDM Tech. Dig. (1981): 223-225.
296. Wegener et al., "Metal-insulator-semiconductor transistor as a nonvolatile storage ",
IEEE IEDM Abstract, (1967): 58.
297. Walker, D., Zhang, X., Saxler, Z., Kung, P., Xj, J., Razeghi, M., "AlxGa(1-
x)N(0<=x,=1)Ultraviolet Photodetectors Grown on Sapphire by Metal-Organic
Chemical-Vapor Deposition", Appl. Phys. Lett. Vol. 70, No. 8 (1997): 949-951.
298. Wu, J., et. al., "Unusual properties of the fundamental band gap of InN", Applied
Physics Letters, Vol. 80, No. 21, 27 (May 2002): 3967-3969.
299. Watt, J.T., Ph.D., Thesis, Stanford University, 1989.
300. Watt, J.T., and J.D. Plummer. “Universal Mobility-Field Curves for Electrons and
Holes in MOS Inversion Layers.” 1987 Symposium on VLSI Technology, Karuizawa,
Japan.
301. Weast, R. Ed. CRC Handbook of Chemistry and Physics. CRC Press, 1976.
302. Welser, J., J. L. Hoyt, S. Takagi, J.F. Gibbons, “Strain dependence of the performance
enhancement in strained-Si n-MOSFETs”, IEDM Tech. Dig., 1994: 337-376.
303. Wenus, J., J. Rutkowski, and A. Rogalski, “Two-Dimensional Analysis of Double-
Layer Heterojunction HgCdTe Photodiodes”, IEEE Trans. Electron Devices Vol. 48,
No.7 (July 2001): 1326-1332.
304. Wenzel, H., and H.J. Wünsche, “The effective frequency method in the analysis of
vertical-cavity surface-emitting lasers,” IEEE J. Quantum Electron 33 (1997): 1156-
1162.
305. Wettstein, A., A. Schenk, and W. Fichtner, “Quantum Device-Simulation with the
Density Gradient Model on Unstructured Grids”, IEEE Trans. Electron Devices Vol. 48,
No. 2 (Feb. 2001): 279-283.
306. White, Marvin H., Dennis A. Adams, and Jiankang Bu, "On the go with SONOS ",
IEEE Circuits and Devices Magazine, Vol, 16, No. 4,(2000): 22-31.
307. White, W.T. et.al., IEEE Trans. Electron Devices. Vol. 37 (1990): 2532.
308. Wilt, D.P. and A. Yariv, “Self-Consistent Static Model of the Double-Heterostructure
Laser,” IEEE J. Quantum Electron. Vol. QE-17, No. 9 (1981): 1941-1949.
309. Wimp, T. Sequence Transformation and their Application. Academic Press, 1981.
310. Winstead, B., “Impact of source-side heterojunctions on MOSFET performance,” M.S.
thesis, University of Illinois at Urbana-Champaign, 1998.
311. Winstead, B., “Monte Carlo simulation of silicon devices including quantum correction
and strain”, Ph.D. Dissertation, University of Illinois at Urbana-Champaign, 2001.
312. Wünsche, H.J., H. Wenzel, U. Bandelow, J. Piprek, H. Gajewski, and J. Rehberg, “2D
Modelling of Distributed Feedback Semiconductor Lasers,” Simulation of
Semiconductor Devices and Processes Vol. 4 (Sept. 1991): 65-70.
313. Wu, C.M., and E.S. Yang, “Carrier Transport Across Heterojunction Interfaces”, Solid-
State Electronics 22 (1979): 241-248.
314. Xiao, H. et. al., "Growth and characterization of InN on sapphire substrate by FR-
MBE", Journal of Crystal Growth, Vol. 276 (2005): 401-406.

1517 ATLAS User’s Manual

Bibliography

315. Yamada, M., S. Ogita, M. Yamagishi, and K. Tabata, “Anisotropy and broadening of
optical gain in a GaAs/AlGaAs multi-quantum-well laser,” IEEE J. Quantum Electron.
QE-21 (1985): 640-645.
316. Yamada, T., J.R. Zhou, H. Miyata, and D.K. Ferry, “In-Plane Transport Properties of
Si/Si(1-x)Ge(x) Structure and its FET Performance by Computer Simulation”, IEEE
Trans. Electron Devices Vol. 41 (Sept. 1994): 1513-1522.
317. Yamaguchi, K. “A Mobility Model for Carriers in the MOS Inversion Layer”, IEEE
Trans. Electron Devices Vol. 30 (1983): 658-663.
318. Yan, R., S. Corzine, L. Coldren, and I. Suemune, "Corrections to the Expression for
Gain in GaAs", IEEE J. Quantum Electron. Vol. 26, No. 2 (Feb. 1990): 213-216.
319. Yang K., J. East, and G. Haddad, “Numerical Modeling of Abrupt Heterojunctions
Using a Thermionic-Field Emission Boundary Condition”, Solid-State Electronics Vol.
36, No. 3 (1993): 321-330.
320. Yariv, A. Optical Electronics. CBS College Publishing, 1985.
321. Yee, K. S., "Numerical Solution of Initial Boundary Value Problems Involving
Maxwell's Equations in Isotropic Media", IEEE Trans. on Antennas and Propagation, V.
AP-14, No. 3 (May, 1966): 302-307.
322. Yeo, Y., King, T., and Hu, C., “Metal-dielectric Band Alignment and its Implications
for Metal Gate Complementry Metal-Oxide-Semiconductor Technology”, J. Appl. Phys.,
Vol. 92, No. 12 (Dec. 15, 2002): 7266-7271.
323. Young, G. A. and J. R. Scully, “The diffusion and trapping of hydrogen in high purity
aluminum”, Acta Mater Vol. 46, No. 18 (1998): 6337-6349.
324. Yu G.C., and S.K. Ken, “Hydrogen Diffusion coefficient of silicon nitride thin films”,
Applied Surface Science, Vol. 201 (2002): 204-207.
325. Yu, Z., D. Chen, L. So, and R.W. Dutton, “Pisces-2ET, Two Dimensional Device
Simulation for Silicon and Heterostructures”, Integrated Circuits Laboratory, Stanford
University (1994): 27.
326. Yu, Z., and R.W. Dutton, “SEDAN III-A Generalized Electronic Material Device
Analysis Program”, Stanford Electronics Laboratory Technical Report, Stanford
University, July 1985.
327. Zabelin, V., Zakheim, D., and Gurevich, S., "Efficiency Improvement of AlGaInN
LEDs Advanced by Ray-Tracing Analysis", IEEE Journal of Quantum Electronics, Vol.
40, No. 12, (Dec. 2004): 1675-1686.
328. Zaidman, Ernest G., “Simulation of Field Emission Microtriodes”, IEEE Transactions
on Electron Devices, Vol 40 (1993):1009-1015.
329. Zakhleniuk, N. A., “Nonequilibrium drift-diffusion transport in semiconductors in
presence strong inhomogeneous electric fields”, Applied Physics Letters, Vol. 89,
(2006): 252112.
330. Zamdmer, N., Qing Hu, K. A. McIntosh and S. Verghese, “Increase in response time of
LT-GaAs photoconductive switches at high voltage bias”, Appl. Phys. Lett., Vol. 75, No.
15, (October 1999): 2313-2315.
331. Zappa, F., Lovati, P., and Lacaita, A., “Temperature Dependence of Electron and Hole
Ionization Coefficients in InP.” (Eighth International Conference Indium Phosphide and
Related Materials, IPRM '96, Schwabisch Gmund, Germany April 21-25, 1996), 628-
631.

1518 ATLAS User’s Manual

Bibliography

332. Zeman, M., R. van Swaaij, and J. Metselaar, "Optical Modeling of a-Si:H Solar Cells
with Rough Interfaces: Effect of Back Contact and Interface Roughness", J. Appl. Phys.,
Vol. 88, No. 11 (Dec. 2000): 6436-6443.
333. Zhou, B., Butcher, K., Li, X., Tansley, T., "Abstracts of Topical Workshop on III-V
Nitrides". TWN '95, Nagoya, Japan, 1995.
334. Zhou, J.R. and D.K. Ferry, “Simulation of Ultra-small GaAs MESFET Using Quantum
Moment Equations,” IEEE Trans. Electron Devices Vol. 39 (Mar. 1992): 473-478.
335. Zhou, J.R. and D.K. Ferry, “Simulation of Ultra-small GaAs MESFET Using Quantum
Moment Equations - II:Volocity Overshoot”, IEEE Trans. Electron Devices Vol. 39
(Aug. 1992): 1793-1796.
336. Zhu, W. and Ma, T., “Temperature Dependence of CHannel Mobility in HfO2-Gated
NMOSFETs”, IEEE Elec. Dev. Lett., Vol. 25, No. 2, (Feb. 2, 2004): 89-91.
337. Zundel, T., A.Mesli, J.C. Muller and P.Siffert, “Boron neutralization and hydrogen
diffusion in silicon subjected to low-energy hydrogen implantation”, Applied Physics A,
Vol. 48 (1989): 31-40.
338. Zundel, T., and J.Weber, “Trap-limited hydrogen diffusion in boron-doped silicon”,
Phys. Rev. B, Vol. 46 (1992): 2071-2077.

1519 ATLAS User’s Manual

Index

U
Uncoupled Mode Space (UMS) .............................................705
Universal Schottky Tunneling (UST) Model ......................148–149
User Defined Materials ............................................1457–1458
User-Defined Two-Terminal Elements
Input Parameters ............................................................639
Output Parameters .................................................639–640
User-Defined Model ........................................................638
Using MC Device ...............................................................817
UTMOST .............................................................1387–1391

V
VCSEL ................................................................1392–1394
Alternative Simulator .......................................................509
Far Field Patterns ...........................................................511
Material Parameters ........................................................507
Numerical Parameters .....................................................509
Solution ........................................................................508
VCSEL Physical Models ..............................................497, 507
Effective Frequency Model (EFM) ......................................497

1533 ATLAS User’s Manual

Pc200-6 Custom Parts Book
100% (21)
Pc200-6 Custom Parts Book
261 pages
BOQ Villa Dubai - Mech. - Ele
100% (1)
BOQ Villa Dubai - Mech. - Ele
3 pages
ClarionMag TheClarionIDE
No ratings yet
ClarionMag TheClarionIDE
35 pages
NET Prog Guide
No ratings yet
NET Prog Guide
309 pages
13S1 FE1073 E2-Magnetic Field
33% (3)
13S1 FE1073 E2-Magnetic Field
15 pages
Pe - 1969-11
No ratings yet
Pe - 1969-11
118 pages
Atlas Users1
No ratings yet
Atlas Users1
1,576 pages
Atlas Users1 PDF
100% (1)
Atlas Users1 PDF
1,270 pages
Atlas Users1
No ratings yet
Atlas Users1
1,254 pages
Athena Users1 PDF
No ratings yet
Athena Users1 PDF
436 pages
Atlas Users1
No ratings yet
Atlas Users1
1,691 pages
Silvaco TCAD Athena User Manual
No ratings yet
Silvaco TCAD Athena User Manual
438 pages
Schematic Gateway User
No ratings yet
Schematic Gateway User
330 pages
Atlas Users PDF
No ratings yet
Atlas Users PDF
878 pages
2014sman Users1
No ratings yet
2014sman Users1
48 pages
Athena Users12012
No ratings yet
Athena Users12012
436 pages
Tonyplot QT: User'S Manual
No ratings yet
Tonyplot QT: User'S Manual
126 pages
Atlaas Manual
No ratings yet
Atlaas Manual
290 pages
Atlas98-V1 Users PDF
No ratings yet
Atlas98-V1 Users PDF
318 pages
Athena Users Manual
No ratings yet
Athena Users Manual
448 pages
ATLAS User's Manual: Device Simulation Software
No ratings yet
ATLAS User's Manual: Device Simulation Software
1,062 pages
Tonyplot: User'S Manual
No ratings yet
Tonyplot: User'S Manual
160 pages
Atlas User Man
No ratings yet
Atlas User Man
740 pages
A 2 Quick Start Guide
No ratings yet
A 2 Quick Start Guide
30 pages
victoryextract_users1
No ratings yet
victoryextract_users1
147 pages
Sep Ed 303004 en
No ratings yet
Sep Ed 303004 en
164 pages
Delft3D-TIDE User Manual
No ratings yet
Delft3D-TIDE User Manual
96 pages
Tonyplot: User'S Manual
No ratings yet
Tonyplot: User'S Manual
161 pages
Tutorial
No ratings yet
Tutorial
32 pages
User Manual and Programmers' Reference
No ratings yet
User Manual and Programmers' Reference
1,442 pages
A3c5 PDF
No ratings yet
A3c5 PDF
335 pages
Python GTK 3 Tutorial PDF
No ratings yet
Python GTK 3 Tutorial PDF
123 pages
Download Essential MATLAB for engineers and scientists 6th edition Edition Hahn ebook All Chapters PDF
100% (4)
Download Essential MATLAB for engineers and scientists 6th edition Edition Hahn ebook All Chapters PDF
66 pages
Python GTK 3 Tutorial
No ratings yet
Python GTK 3 Tutorial
129 pages
Silos User Manuel
0% (1)
Silos User Manuel
286 pages
Scilab Tutorials For Computational Science - Chandler and Roberts
No ratings yet
Scilab Tutorials For Computational Science - Chandler and Roberts
82 pages
CGAL Tutorial
No ratings yet
CGAL Tutorial
63 pages
Manual Logipam en
No ratings yet
Manual Logipam en
164 pages
2014sflm Users1
No ratings yet
2014sflm Users1
89 pages
QM3.8 Rapid 7 - 090304
No ratings yet
QM3.8 Rapid 7 - 090304
77 pages
Aquatrans AT600 Flowmeter
No ratings yet
Aquatrans AT600 Flowmeter
190 pages
Kaye Valprobe User Guide
No ratings yet
Kaye Valprobe User Guide
93 pages
Atlas Olt 1000a Conf. Manual
No ratings yet
Atlas Olt 1000a Conf. Manual
79 pages
W Java16 PDF
No ratings yet
W Java16 PDF
58 pages
Atlas2000 Definicion Funcional
No ratings yet
Atlas2000 Definicion Funcional
24 pages
Guard I/O Ethernet/Ip Safety Modules: 1791Es-Ib8Xobv4, 1791Es-Ib16
No ratings yet
Guard I/O Ethernet/Ip Safety Modules: 1791Es-Ib8Xobv4, 1791Es-Ib16
108 pages
Cobol
No ratings yet
Cobol
95 pages
AltaRica 3.0 Language Specification - V1.2
No ratings yet
AltaRica 3.0 Language Specification - V1.2
119 pages
Bitwig Studio User Guide English 3.0 PDF
No ratings yet
Bitwig Studio User Guide English 3.0 PDF
477 pages
Open Quake Model
No ratings yet
Open Quake Model
97 pages
Python GTK 3 Tutorial
No ratings yet
Python GTK 3 Tutorial
131 pages
Euler
No ratings yet
Euler
152 pages
Tutorial Python GTK 3
100% (1)
Tutorial Python GTK 3
131 pages
C++/CLI: The Visual C++ Language for .NET
From Everand
C++/CLI: The Visual C++ Language for .NET
julian Pino
No ratings yet
C++/CLI :The Visual C++ Language for NET
From Everand
C++/CLI :The Visual C++ Language for NET
abraham black
No ratings yet
Antenna Arraying Techniques in the Deep Space Network
From Everand
Antenna Arraying Techniques in the Deep Space Network
David H. Rogstad
No ratings yet
Introduction to Stochastic Search and Optimization: Estimation, Simulation, and Control
From Everand
Introduction to Stochastic Search and Optimization: Estimation, Simulation, and Control
James C. Spall
4/5 (2)
Prediction of Burnout: An Artificial Neural Network Approach
From Everand
Prediction of Burnout: An Artificial Neural Network Approach
Felix Ladstätter
No ratings yet
Open Data Structures: An Introduction
From Everand
Open Data Structures: An Introduction
Pat Morin
4/5 (4)
Designing cdma2000 Systems
From Everand
Designing cdma2000 Systems
Leonhard Korowajczuk
No ratings yet
Design and Analysis of Distributed Algorithms
From Everand
Design and Analysis of Distributed Algorithms
Nicola Santoro
No ratings yet
Quasi-Monte Carlo Methods in Finance: With Application to Optimal Asset Allocation
From Everand
Quasi-Monte Carlo Methods in Finance: With Application to Optimal Asset Allocation
Mario Rometsch
No ratings yet
Control Systems
From Everand
Control Systems
Francisco Luis Pagola y de las Heras
No ratings yet
Schaum's Outline of Programming with C++
From Everand
Schaum's Outline of Programming with C++
John R. Hubbard
No ratings yet
Is 7098
No ratings yet
Is 7098
17 pages
15276_133745925471693139_241119143420
No ratings yet
15276_133745925471693139_241119143420
2 pages
L700 Antenna Description
No ratings yet
L700 Antenna Description
4 pages
Bornay Aecon
100% (2)
Bornay Aecon
2 pages
My Seminar Report
No ratings yet
My Seminar Report
24 pages
WESM WESMMNM 033 - Revision34
No ratings yet
WESM WESMMNM 033 - Revision34
55 pages
Question Bank For Physics - Ii Regulation 2013
No ratings yet
Question Bank For Physics - Ii Regulation 2013
10 pages
104264AE
No ratings yet
104264AE
60 pages
Ball Valve Thread End
No ratings yet
Ball Valve Thread End
2 pages
LVAC Circuits Calculation Report（565）
No ratings yet
LVAC Circuits Calculation Report（565）
44 pages
Pow-R-Way III Bus Plugs - Eaton
No ratings yet
Pow-R-Way III Bus Plugs - Eaton
2 pages
Triacs BT136 Series E Sensitive Gate: General Description Quick Reference Data
No ratings yet
Triacs BT136 Series E Sensitive Gate: General Description Quick Reference Data
2 pages
PM2 Panasonic
No ratings yet
PM2 Panasonic
6 pages
12.smoothing Reactor & Reactive Power Control of HVDC System
50% (4)
12.smoothing Reactor & Reactive Power Control of HVDC System
32 pages
2023 Nec Changes Presentation
100% (2)
2023 Nec Changes Presentation
23 pages
Unit 8 Heating and Cooling of Electric Motors
100% (1)
Unit 8 Heating and Cooling of Electric Motors
20 pages
CRTD 084 Ne 6 N
No ratings yet
CRTD 084 Ne 6 N
9 pages
N700 Catalogo PDF
No ratings yet
N700 Catalogo PDF
40 pages
PD10055 Directional Panel: Home Page Antennas Omni Fiberglass PD10055
No ratings yet
PD10055 Directional Panel: Home Page Antennas Omni Fiberglass PD10055
2 pages
Amoled PDF
No ratings yet
Amoled PDF
17 pages
CHAPTER 3: Methodology
No ratings yet
CHAPTER 3: Methodology
3 pages
03 Hard Copy Format For Sep
No ratings yet
03 Hard Copy Format For Sep
18 pages
Types of Lightning Arresters - Circuit Globe
No ratings yet
Types of Lightning Arresters - Circuit Globe
9 pages
Diodes in Parallel
No ratings yet
Diodes in Parallel
4 pages
Assigment Week 2 - Rizky Fajar A. H - 04211841000016 PDF
No ratings yet
Assigment Week 2 - Rizky Fajar A. H - 04211841000016 PDF
7 pages
SR208C en 20120419 PDF
No ratings yet
SR208C en 20120419 PDF
35 pages