CNSE, Qualifying Examination, August 30/31, 2004

Problem 1.

Consider edge dislocation lines free to move y x

laterally (glide) along the x-direction in parallel b
planes (see illustration). (a) Assume the
Burger’s vectors of the dislocation lines point
along the same direction (parallel). Determine
the lowest stable energy configuration of the
dislocation lines assuming only dislocations in
neighboring planes interact. (b) Assume the
Burger’s vectors of the dislocation lines
alternate in direction, i.e. the Burger’s vector of
the top plane is anti-parallel to the Burger’s vector of the plane below it. Determine the lowest
stable energy configuration of the dislocation lines assuming only dislocations in neighboring
planes interact. For both parts (a) and (b) rigorously justify your answers.

You may find the following relations useful:

Gb 2
Force between parallel screw dislocations separated by a distance r: Fs = ± where upper
2πr
(lower) sign refers to parallel (anti-parallel) burgers vectors.

Force between parallel edge dislocations separated by a distance r:

Gb
Fx = ±bτ xy τ xy = cos θ cos 2θ
, where 2π (1 − ν )r .
F y = m bσ x − Gb
σx = sin θ (2 + cos 2θ )
2π (1 − ν )r

The upper (lower) sign refers to edge dislocations with parallel (anti-parallel) burgers vectors. The
angle, θ, is the angle made between a line connecting the centers of the dislocations with the x-axis.
 CNSE, Qualifying Examination, August 30/31, 2004

Problem 2.

For the ground state of the hydrogen atom with the wave function given below

Ψ (r ) = N exp[− r / a o ]

a.) find N.
b.) find the expectation (average) value of the position.
c.) find the most probable outcome of a measurement of r.

Where ao is the Bohr radius and its numerical value is not needed.
 CNSE, Qualifying Examination, August 30/31, 2004

Problem 3.

8
A - Prove that the ideal c-to-a ratio for an hcp structure is .
3

B - Derive the atomic packing factor for ruthenium (Ru) which has an hcp structure.
C - Estimate the density of hcp ruthenium (Mmass = 101 g/mol, radius = 133 pm).
D - Derive the atomic packing factor for rhodium (Rh) which has an fcc structure.
 CNSE, Qualifying Examination, August 30/31, 2004

Problem 4.

Given the following effects model for a designed experiment with one variable observed at several
different levels:
y ij = µ + τ i + ε ij where levels i ~ 1 to a, replicates j ~ 1 to n;
and given model errors are normally, independently distributed random variables with mean of zero
and constant variance for all levels;

a – Set out the null and alternative hypotheses associated with this analysis of variance (ANOVA).
b – Define type I and type II errors and indicate the symbol used to represent their probability.
c - Show how the experimental total variance can be separated into its component parts (MStreatments
1 a n
& MSerror). (V(yij) = ∑∑
an − 1 i j
( y ij − y•• ) 2 ).

Additionally, given the expectation value for MStreatments & MSerror (E(MStreatments) =
a
n
σ2 + ∑
(a − 1) i
τ i2 & E(MSerror) = σ2) and given SStotal/σ2, SStreatments/σ2, SSerror/σ2 are all χ2
distributed quantities with (an-1), (a-1), (an-a) degree of freedom, respectively;

d – Explain how to obtain an appropriate test statistic to evaluate the null hypothesis set out in 1-a.
e – Explain how to verify the assumptions associated with this type of analysis of variance
(ANOVA).
 CNSE, Qualifying Examination, August 30/31, 2004

Problem 5.

1. (a) You are asked to build a 600 µm square, 25 µm thick diaphragm in an oxidized (100) silicon
wafer which is 300 µm thick. Assume that you will be using an anisotropic etching solution of
KOH/water which attacks the (111) crystal planes 100 times slower than the (100) planes of Si,
and results in an etch rate of (100) silicon of 10 µm/hr.
(i) Find the etching time
(ii) Find the dimensions of the mask opening you would use. Assume you can neglect etching of
the (111) planes.

(b) Assuming you use the same KOH/water etch as for part (a), indicate whether and why you
would you choose (100)-, (110)- or (111)-oriented Si for fabricating
(i) Diaphragm-based pressure sensors
(ii) High aspect ratio comb actuators

(c) Define each of the following terms.

(i) Stress
(ii) Strain
(iii) Young’s modulus
(iv) Poisson’s ratio
(v) Yield Stress
(vi) Thermal Expansion Coefficient
 CNSE, Qualifying Examination, August 30/31, 2004

Problem 6.

The figure shows a model of a rectifying metal-n-type semiconductor barrier where an insulating
layer of thickness d is prepared over the semiconductor before the metal contact is applied. Assume
that donors (Nd=n) are uniformly distributed in the semiconductor material

The following assumptions apply to this model:

a) 0 ≤ x ≤ d, insulating layer where no mobile charges or ionized impurities are present.

b) d ≤ x ≤ w, the usual space-charge region at a metal-semiconductor interface. (Assume no
carrier generation or recombination occurs in this region).
c) x > w, bulk semiconductor conditions apply as usual.

Given the above assumptions:

I. Sketch the electric field and electrostatic potential for this system as a function of x;
II. Calculate an expression for the semiconductor barrier capacitance per unit area (junction
capacitance) in terms of d, VB, V (applied potential across the barrier) and material parameters
of the semiconductor.

Metal

VB

EF
d

W
X=0
 CNSE, Qualifying Examination, August 30/31, 2004

Problem 7

Consider a crystal with an equilibrium lattice spacing, ao, of 3.6 Å. A uniaxial tensile stress is
applied as shown in the schematic below. Define x as the relative displacement of the atomic rows
from their equilibrium positions in the direction of applied stress. The ideal stress vs. displacement
σ
20

σ (GPa)
a = x + ao

0
0 1 2
σ x (Angstroms)

curve for this situation is also shown below. Assume this plot can be approximated by a half-period
of a sine curve. The maximum of that curve represents the stress required to fracture the crystal
along an atomic plane (~25 GPa). This occurs at a relative atomic displacement of 0.9 Å.

(a) Calculate the Young’s modulus of the crystal.

(b) Complete separation (fracture) the crystal is represented by the formation of two new
planar surfaces. From the information given above estimate the energy per unit area
associated with these two new surfaces.
 CNSE, Qualifying Examination, August 30/31, 2004

Problem 8.

Find the transmission and reflection coefficients for a particle of mass m incident from the left with
E = 2 Vo for the square barrier potential as shown below.

Vo

x
0 a
 CNSE, Qualifying Examination, August 30/31, 2004

Problem 9.

A - Solid material is obtained for a rod by cooling a melt of elements A & B to just below its
eutectic point. Calculate the relative masses of α & β phases present.

0 10 60 70 100
Composition B (Mass %)

B- The stress (σ) relaxation of a rod may be modeled by first order kinetics, where the
relaxation time @ 25 oC is 60 days. If the rod is initially stressed to 20 MPa, how much
time will elapse before the rod is stressed to only 1 MPa
C - If the relaxation time @ 45 oC is found to be only 1/8 that @ 25 oC, how much less time
will elapse before the rod is stressed under only 1 MPa @ 35 oC.

0%
 CNSE, Qualifying Examination, August 30/31, 2004

Problem 10 (+7 pages attachment).

A process engineer is interested in the effect of spin speed on the non-uniformity of photoresist
deposited on wafers of 200mm diameter. He is interested in minimizing the within wafer non-
uniformity as well as the wafer-to-wafer non-uniformity. He found the following data that he
deemed appropriate for his study and, subsequently, he completed the attached statistical analysis.

Spin Speed (rpm)

1000 1250 1500 1750 2000
2.7 (5) 4.6 (12) 5.4 (18) 7.4 (10) 2.7 (11)
Non-uniformity
across wafer

2.7 (3) 6.6 (21) 7.0 (15) 9.7 (23) 3.9 (19)
(% 1σ)

5.8 (6) 4.6 (22) 7.0 (1) 8.5 (9) 4.3 (14)
4.3 (24) 7.0 (16) 7.4 (4) 7.4 (17) 5.8 (7)
3.5 (20) 7.0 (13) 7.4 (25) 8.9 (8) 4.3 (2)
* The order each wafer of this data set was processed is set out in parentheses following
its respective across wafer non-uniformity measurement.

a – List the stages of a well designed experiment.

b – Indicated all of the stages set out in question 4-a that the process engineer addressed and
explain how he addressed each of them (Include the basic principles of experimental design
where appropriate).
c - The statistical analysis provided includes analysis of variance, Dunnett’s, Tukey,
Bonferroni/Bartlett’s/Levene’s tests; list the null and alternative hypotheses associated with
each of these tests.
d - Provide your own conclusions and recommendations concerning both the minimization of
within wafer non-uniformity as well as wafer-to-wafer non-uniformity, relative to the current
process of record which uses a spin speed of 1250 rpm.
 CNSE, Qualifying Examination, August 30/31, 2004

Problem 11.
An exposure is performed using a step-and-repeat printing system. The light source has a
wavelength of 365 nm (I-line of a Hg arc lamp). The mask pattern is a grating with a line-to-line
spacing of 2 µm. The projection optics provide a 4x demagnification onto the image plane. This
lithography is performed with standard process parameters k1=0.5 and k2=0.5.

a) Calculate the minimum Numerical Aperture to print a 0.5 µm line-to-line spacing in resist in the
image plane.

b) What is the Depth of Focus for this imaging process?

c) Given that the intensity in the image plane from a point source uniformly illuminating an
aberration-free lens with a circular aperture is:
I=I0[2J1(x)/x]2,
Where J1(x) is a Bessel function, x≈(2πrR)/(fλ), where r is the distance in the image plane from the
peak intensity, R is the radius of the lens, f is the distance from the lens to the image plane, and λ is
the wavelength of the illumination.
The Rayleigh Resolution is defined as the distance between two incoherently illuminated point
sources such that the maximum intensity of the image of one point source is located at the first
minimum in intensity of the other point source. Assume that the lens is sufficiently far from the
point sources that it is uniformly illuminated. Derive the Rayleigh Resolution in terms of
wavelength and numerical aperture of the lens.
You may find the graph below helpful.

d) In the image plane there are two peak intensities corresponding to the image of each point
source. Halfway between these two peaks is a local minimum in intensity. How much lower is
the intensity at this local minimum compared to the intensity at the peaks when imaging at the
Rayleigh Resolution?

e) Using the process parameters from above (k1=k2=0.5), would you expect to be able to print
images of both pinholes? Why or why not?

0.5
0.45
0.4
0.35
0.3
0.25
J1(x)/x

0.2
0.15
0.1
0.05
0
-0.05 0 1 2 3 4 5
-0.1
x
 CNSE, Qualifying Examination, August 30/31, 2004

Problem 12.

A silicon sample maintained at 300K under thermal equilibrium has a non-uniform doping
concentration profile, such that the electron concentration, n, varies linearly from 1x1012 cm-3 to
5x1017 cm-3 while going from point x1 to point x2 (see figure below). Assuming that the mobility is
constant at 1000 cm2/Vs throughout the sample, answer the following. (Eg=1.1 eV and
ni=1.5x1010cm -3)

n
5x1017 cm-3

Linearly varying
12 -3
1x10 cm

X1 X2 x
10µm

(a) Calculate the diffusion coefficient, Dn (in cm2/s) for the electrons.
(b) Would the electrons diffuse everywhere such that the concentration is uniform throughout?
Explain.
(c) Plot the diffusion current density (A/cm2) for the electrons as a function of x. Mark the
numerical value on the graph.
(d) Plot the drift current density for electrons as a function of x.
(e) Sketch the energy band diagram as a function of x.
(f) What is the potential difference (give a numerical value) between the two ends of the
sample?