Outline

Semiconductor device modeling • Basic concepts and physics model

with SCAPS-1D • Overview of capabilities
• Input/Output
• Brief literature survey of problems/results
Lecture 4 • Hands on: running standard models,
creating a simple model
Special T opics:
Device Modeling

SCAPS-1D SCAPS-1D
• SCAPS (a Solar Cell Capacitance Simulator) is a • Originally developed for polycrystalline cell structures
one-dimensional solar cell simulation program of the CuInSe2 and the CdT e family
• Developed at the Department of Electronics and – Basis reference paper published in 2000
Information Systems (ELIS) of the University of
• Designed to accommodate thin films, multiple
Gent, Belgium
interfaces, large band gaps (Eg=1.12eV for Si, but
• T he program is freely available to the research 2.4eV for CdS used as window layer)
community (reference it in your publications)
• T he package evolved over the years to include
• Continuously supported and further developed additional mechanisms, e.g., Auger recombination,
(http://scaps.elis.ugent.be); runs under Windows tunneling, multiple enhancement to user interface, etc.

Device model Device model: energy space

metal • Device is modeled in one • Band diagram of a realistic
p-doped semiconductor dimension across layers, V2 p-n heterojunction with back
V1
n-doped semiconductor
and formed junctions in p-type barrier and defect states at
physical space thermodynamic equilibrium
Metal/TCO
• Depth is x coordinate
EF • Compared to AMP S: better
n-type
description of recombination
1-D direction, • Spatial distributions of Defect sta tes
processes; several tunneling
x coordinate fields, charge carriers, semiconductors metal
mechanisms are included in
defects, etc. SCAP S

1
 Physics model: governing Physics model: solving
equations governing equations
• T he Poisson equation, and the continuity equations • Constitutive relations
for electrons and holes (same as in AMPS-1D)  n  E Fn  p  E Fp
jn   ; jp  
    q    1  x x
  ( x)       n  p  N A  N D   def ( n , p )  q q
x  x  0  q 
• Solve for potential and quasi-Fermi levels
jn n p,n – free carrier concentrations
  G  U n (n, p )  N±D,A – charged dopants • Boundary conditions at interfaces and contacts
x t def (n,p) – defect distributions • Structure is discretized, meshing refined around interfaces
j p p jn , jp – the electron and hole current
  G  U p (n, p )  densities • Newton-Raphson method with algorithm modifications
x t Un,p - the net recombination rates;
G - the generation rate

Physics model: interfaces Physics model: grading

• T he quasi-fermi levels are allowed to be • Almost all parameters can be graded; the principles
discontinuous at the interfaces of the algorithms used to simulate graded solar cell
• Recombination at the interface states is handled structures
• All parameters are consistently derived from the
Example: recombination composition grading of a layer
between electrons of CdS – Each layer is assumed to have composition A 1-y By
and holes in the CdTe on the – Define values of a parameter P for pure compounds A, B,
right side of the interface and the composition grading y(x) over the layer thickness
– Specify some grading law for P(y)

Physics model: grading Physics model: generation

• Grading laws: • From internal calculation under illumination
uniform, linear, – Dark or light, power level (~ND filter), choice of the
exponential, etc. illuminated side, choice of the spectrum
• Alternatively, • From user supplied generation g(x) file, at the x-
the composition coordinates of the nodes used by SCAPS
grading profile – allows for modeling of radiation detectors, EBIC
can be loaded measurements
from a file – Solar cell efficiency and QE cannot be calculated; may
use collection efficiency, based on “ ideal” device current

2
 Physics model: recombination Physics model: tunneling
• Direct band-to-band • The following tunneling mechanisms are treated: band to
– Between the occupied states in the CB and the vacant band tunneling, intraband tunneling, tunneling to interface
states in the VB defects and tunneling to contacts
• Indirect, or Shockley-Read-Hall
– Through a defect state in the gap
– Also through interface states
• Auger recombination
– Involves three carriers: after recombination, the energy is
given an electron in the conduction band (as opposed to
emission as a photon or phonon)

Input: action panel

Physics model: tunneling
• Tunneling is only
taken into account
in the solution of
the dc-problem
• Only indirect tunnel
influences on the
admittance (through
the setting of the dc-
state of the sample)

Output Device definition

• In each calculation the running parameter (V, f, or l)

is varied in the specified range
• Plot all calculated parameters, such as I/V, C/V, C/f,
Q(l), band diagrams, concentrations, and currents
• All calculations can be saved in ASCII format
• Device is represented as a stack of layers, up to 7
• When divergence occurs, the points calculated so far semiconductor layers with specified properties
are shown on the corresponding graphs • Separate entries for interface parameters
• Batch calculations possible; presentation of results
• T wo additional layers for contacts, front and back
and settings as a function of batch parameters

3
 Quantum efficiency Quantum efficiency
• T he Q.E. is the ratio of the number of carriers
collected by the solar cell to the number of photons
of a given energy incident on the solar cell
• If all photons of a certain wavelength are absorbed
and the resulting minority carriers are collected,
then the quantum efficiency at that particular
wavelength is unity
• T he quantum efficiency for photons with energy
below the band gap is zero Image from ASU PV education website
http://www.pveducation.org/pvcdrom/solar-cell-operation/quantum-efficiency

Admittance spectroscopy Admittance spectroscopy

• Small signal analysis, C(V) of C(f) • The C–V profiling tests the spatial charge distribution (1/C2 vs V gives
total defect + dopants concentration)
• T he cell structure is analyzed as if it were a • Frequency dependent admittance Y() is generally attributed to defects
parallel connection of a (frequency dependent)
• In response to the small testing ac
capacitance and a (frequency dependent) electric potential, defects change
conductance their occupation numbers
depending on their relaxation
~ times
J
Y ( )  ~  i  C ( )  G ( ) • Changing  (and E=kTln(t0))
u ext – scan energies of resonant states
T. Walter, R. Herberholz, C. Müller, H.W. Schock, Determination of defect distributions from admittance measurements and • Changing bias and bend bending
application to Cu(In,Ga)Se based heterojunctions, Journal of Applied Physics, 80 (1996) 4411-4420.
2
– scan along distance from CB
K. Decock , S. Khelifi, S. Buecheler, F. Pianezzi, A.N. Tiwari, M. Burgelman, Defect distributions in thin film solar cells
deduced from admittance measurements under different bias voltages, Jou rnal o f App lied Physics, 1 10 (20 11) 06 372 2
Contribute to admittance

Output Output

Fewer levels at Ew

From standard I-V curves at different levels of illumination To temperature-dependent admittance spectroscopy (C/f)

4
 Data analysis features
• Data analysis is supported within the package: any
ASCII-text file can be read as a measurement file SCAPS-1D
– The file extension indicates which kind of measurement
it contains: ‘ .iv’, ‘ .cv’, ‘ .cf’ or ‘ .qe’ BRIEF LITERATURE SURVEY
• A built-in curve fitting facility
OF PROBLEMS AND RESULTS
• Quantum efficiency panel
• A panel for the interpretation of admittance
measurements (C/f and C/V)

Analysis of graded band gap solar cells with SCAPS

M. B urgelman and J. Marleinthin, 23rd European PV Analysis of graded band gap solar
Conf erence, Valencia, Spain, 2008
cells with SCAPS
• In Cu(In,Ga)(Se,S) 2
devices the absorber
materials are
engineered to have an
optimized band gap
energy Eg (trade-off
between high current
for low Eg and high
voltage for high Eg) Ga- content should be low over most of the CIGS bulk, but high in a
narrow region at the back contact and at the interface; a characteristic
width < 50 nm seems to be sufficient to combine high Jsc with high Voc

Advance d e lectrical simulation of thin film solar cells Advanced electrical simulation of thin film solar cells
M. B urgelman, K. Decock, S. Khelif i and A. Abass, Thin Solid
Films, 535 (2013) 296-301 EC

• T he result of measurements performed on CIGS

based solar cells depends on history of the sample
• T he model includes band gap grading, multivalent
EV
defects and metastable transitions between defects
– The occupation of metastable defects is set during initial
conditions at higher temperature, and then frozen in during
cell operation at lower temperature
– Metastable states of the double vacancy type (VSe –VCu)
were introduced
A band gap grading profile was set that results in a small but
influential hump or energy barrier in the conduction band

5
 Modeling metastabilities in chalcopyrite-based thin f ilm solar cells
Advance d e lectrical simulation of thin film solar cells K. Decock, P. Zabierowski, and M. B urgelman, J. Appl. Phys.
111, 043703 (2012)

• T he effect of voltage induced metastabilities on

the capacitance-voltage characteristics of CIGS
• Defect, (VSe-VCu ) complex, transitioning from
charged donor to charged acceptor configuration
• T he agreement between the simulation and
measurement results has been obtained using a
simple model, and optimizing only 5 parameters
• A barrier in the CB impedes the collection of electrons that are – E TR, the shallow doping density N A, the (VSe -VCu)
generated in the absorber to the left of the barrier complex density N M, the density N t of the additional
• Results in light current reduction at FB thus affecting FF acceptor defect and its energy E t .

Modeling metastabilities in De sign and optimization of large area thin-film CdTe

chalcopyrite-based thin film solar cells de te ctor for radiation therapy imaging applications
Sy mbols represent measurement results E. I. Parsai, D. Shvydka, J. Kang, Med. Phys. 37 (2010) 3980

• Investigate performance of thin-film CdTe

material in detecting 6MV x-rays
• The utilization of this material has become technologically
feasible only in recent years due to significant development
in large area photovoltaic applications
• Extensive MCNP simulations to evaluate geometrical
parameters, detective quantum efficiency, scatter
• Results of absorbed energy simulation were used in device
operation modeling (with SCAPS) to predict the detector
output signal, and compared with measured signal

Design and optimization of large area Design and optimization of large area
thin-film CdTe thin-film CdTe
14
2

6x10
Current density, nA/cm
-3 -1

75 CdTe thickness
e-h pairs, cm s

14
10 m
5x10 30 m
50 100 m
14
4x10 CdTe thickness
25
10 m
14
30 m
3x10 100 m 0
0.0 0.1 0.2 0.3
0 20 40 60 80 100 Voltage, V
Depth, m -25

• MC modeled electron-hole generation profiles were used as input for SCAPS

• CdTe/CdS solar cell with a built-in junction field of ~104 V/cm; the resulting
readout voltages are >100 mV for detector layers working in PV mode • Better quality of material results in higher readout signals
• Defect density study was only possible for the standard solar cell device
thickness range (4-10m)

6
Modeling the ef f ect of 1 MeV electron irradiation on the
perf ormance of n+–p–p+ silicon space solar cells Modeling the effect of 1 MeV
A. Hamache, N. Sengouga, A. Meftah, M. Henini, Radiation Physics
and Chemistry 123 (2016) 103–108 electron irradiation
• P erformance of cells used
in space degrades after ~50m
irradiation; anomalous
increase in performance
right before failure
• Analyzed several defects
(experimentally
established) which act as
recombination centers
and/or traps of free
carriers It was concluded that a shallower donor trap is responsible for the phenomenon
while the deeper donor trap enhances this phenomenon

Summary References
• SCAPS-1D is a versatile package for • SCAP S manual, M. Burgelman, K. Decock, A. Niemegeers,
semiconductor device modeling J. Verschraegen, S. Degrave, Version: 17 february 2016
• Output for I/V, C/V, C/f, Q(l), band diagrams, • SCAP S 3.0, An introduction, K. Decock and M. Burgelman
concentrations, and currents • M. Burgelman, P . Nollet, S. Degrave, Thin Solid Films 361
(2000) 527
• Data analysis for I-V, C-V, C-f
• Additional references are provided within slides
• A number of standard models available with the
distribution package
• Well-developed user interface, convenient
scripting facilities