Scribd
Upload
147 views8 pages

Final Solutions

The document is the solutions to a physics final exam. It contains answers to 8 multi-part questions covering topics in solid state physics including excitons, semiconductors, ferromagnetism, phase transitions, Fermi surfaces, dielectrics, superconductivity, and magnetism.

Uploaded by

Yenhao Philip Chen
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
147 views8 pages

Final Solutions

The document is the solutions to a physics final exam. It contains answers to 8 multi-part questions covering topics in solid state physics including excitons, semiconductors, ferromagnetism, phase transitions, Fermi surfaces, dielectrics, superconductivity, and magnetism.

Uploaded by

Yenhao Philip Chen
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 8

Physics 141B, Spring 2010

Final Exam Solutions


David Strubbe

1. (30) Short answer


a) (6) Wannier excitons are large in size compared with the lattice constant.
Frenkel excitons are localized on one unit cell. According to the hydrogenic
model, the binding energy of level n for a Wannier exciton depends on the
effective masses and dielectric constant via
e4
2n2 ~2 20

(1)

1
1
1
= +

me
mh

(2)

En =
where the reduced effective mass is

b) (6) To make Si n-type, you need an element with one more valence electron,
i.e. 5. To disrupt other properties as little as possible, it is best to pick the
neighboring element P, although N or As are also acceptable.
c) (6) A Bloch wall is a transition between two ferromagnetic domains, in
which spins rotate in the plane of the wall between the directions of the two
domains. The exchange interaction makes large changes in spin direction
costly in energy, so the change in direction is spread out over a domain wall.
A typical width is that of Fe, about 300 lattice constants.
d) (6) The Meissner effect is the expulsion of magnetic flux by a superconductor,
so that B = 0 inside. It is described by the London equation,
j=

c
A
42L

e) (6) See Kittel Fig. 12.20.


2. (20) Electronic and optical properties of a simple metal
1

(3)

a) (10) See Kittel Fig. 14.3. Below the plasmon frequency, there are no modes
to excite, and everything is reflected. Above the plasmon frequency, there
are propagating modes and transmission is possible. The plasmon frequency
is

1/2
4ne2
p =
= 1.8 1015 s1
(4)
m
which corresponds to 1.2 eV.
b) (10) The Fermi wavevector is
kF = 3 2 n

1/3

= 1.4 107 cm1

(5)

The extremal area of the Fermi surface is S = kF2 . The De Haas-van Alphen
period is given by
 
2e
1
=
= 1.5 107 gauss1
(6)

B
~cS
3. (25) Landau theory of phase transition
a) (6) The Landau free energy at E = 0 is
1
1
1
F (P, T, 0) = g0 + g2 P 2 + g4 P 4 + g6 P 6
2
4
6
Extrema are found where

F
= g2 P + g4 P 3 + g6 P 5 = P g2 + g4 P 2 + g6 P 4 = 0
P

There are two solutions: P = 0 or g2 + g4 P 2 + g6 P 4 = 0, i.e.


!1/2
p
g4 g42 4g2 g6
P =
2g6

(7)

(8)

(9)

b) (6) The condition


F (Ps , TC , 0) = F (0, TC , 0)

(10)

means
1
1
1
g0 + g2 Ps2 + g4 Ps4 + g6 Ps6 = g0
2
4
6
1
1
1
2
4
g2 Ps + g4 Ps + g6 Ps6 = 0
2
4
6

(11)
(12)
(13)

Using the equation for Ps above to eliminate g2 , we have


1
2
g4 Ps2 + g6 Ps4 = 0
2
3
1
2
g4 + g6 Ps2 = 0
2
3
3 |g4 |
3g
4
=
Ps2 (TC ) =
4g6
4g6

(14)
(15)
(16)

Note that unfortunately the desired answer was given in the exam without
the minus sign or absolute value.
c) (7) Now we use the given form of g2 (T ) and our two equations for Ps :
g2 (TC ) = (TC T0 ) = g4 Ps2 g6 Ps4 =

3g42
9g42
3g42
=

4g6
16g6
16g62
3g42
TC = T 0 +
16g6

(17)
(18)

d) (6) This is a first-order phase transition because at TC the polarization


changes discontinuously between 0 and Ps 6= 0.
4. (25) Fermi surfaces and carrier orbits
a) (6) The reciprocal lattice vectors are G1 =
2
2
A

= 3.1 108 cm1 .

2
4
A

= 1.6 108 cm1 and G2 =

b) (6) In two dimensions, we have


A
(2)2 kF2

(19)

kF = (2n)1/2

(20)

N =2

2
The concentration n = 1/8
A = 1.25 1015 cm2 . Therefore kF = 8.9 107
cm1 , which is a little more than 12 G1 .

c) (7) The Fermi surfaces in the first and second Brillouin zones, with shaded
areas occupied:

d) (6) The magnetic force on a charged particle is


q
e
F = v B = k B
c
~c

(21)

Both orbits are electron-like since k points outward. The first is an open
orbit and the second is a closed orbit.
5. (25) Low-frequency dielectric function of a polar insulator
a) (8) From the Kramers-Kronig relations,
Z
2
s2 (s)
1 () = + P
ds
2 2

s
0

(22)

where is the electronic contribution, since 2 as given describes only the


ionic contribution.
Z
L2 2
sA ( T )
2
ds
(23)
2
= + P

s2 2
T 2
0

2 2 + 2A
T
2
T
T
= + A 2
=
(24)
2
2
2
T
T
 2A

2
2 +
L2 2 = T
T
(25)

The solution is

A=

2
L2 T

2
T

(26)

b) (9) The optical parameters n and k (both of which are real) are related to
the dielectric function by
1 = n2 k 2 , 2 = 2nk
4

(27)

Since 2 = 0 for 6= T , either n = 0 or k = 0 for 6= T . 1 > 0 for < T

and > L , so it must be k = 0 and n = 1 . 1 < 0 for T < < L , so

it must be n = 0 and k = 1 .
The reflectivity at normal incidence is


n + ik 1 2

(28)
R =
n + ik + 1

2


1 1
< T , > L
1 1 2
1 +1




=
=
(29)
i1 1 2 < <
1 + 1
T
L
i +1
1

Below is plotted 1 and R for T = 1, L = 1. Note that R = 1 in the


forbidden region T < < L .

c) (8) The coefficient A must be positive for 2 to represent absorption. For


this to be the case, assuming > 0, then L > T . Alternately, the
longitudinal optical phonons take more energy to create because they induce
long-range electric fields, whereas the transverse phonons do not. You can

also appeal to the Lydane-Sachs-Teller relation


L2
0
2 =
T

(30)

or
2
L2 = T
+

4nQ2
M

(31)

6. (25) Superconductivity
a) (8) The stabilization energy density of a superconductor, from thermodynamic arguments with an applied field, is HC2 /8. However, the stabilization
energy of the electrons upon forming Cooper pairs is also 12 D (F ) 2 because
electrons within of the Fermi energy lower their energy by . Therefore
HC2 /8 21 D (F ) 2 .

b) (8) The Josephson effects are caused by coherent tunneling of Cooper pairs
between two superconductors. In the DC Josephson effect, for zero applied
voltage, there is a DC current IC . In the AC Josephson effect, a DC voltage
above VC induces an oscillating current. For coherence to be maintained
across the tunneling barrier, the thickness must not be large compared to
the coherence length .
c) (9) The velocity of a particle is
v=
The current density is

1 
q 
q 
1 
p A =
i~ A
m
c
m
c
j = q v =

nq 
q 
~ A
m
c

(32)

(33)

Due to the Meissner effect, in the interior of a ring, j = 0, so ~c = qA.


The flux through the ring is
Z
I
I
~c
~c
= B ds = A dl =
(34)
dl = 2s
q
q
for some integer s. Thus the quantum of flux, since q = 2e for Cooper
pairs, is 0 = hc/2e.
7. (25) Magnetic response of a metal

a) (8) In the bcc structure, there are two atoms per unit cell. Li is monovalent,
so the electron concentration is
2
22
3
n=
(35)
 = 4.7 10 cm
2
3.5A
The Fermi energy is
F =

2/3
~2 kF2
~2
3 2 n
= 7.6 1012 erg = 4.7 eV
=
2m
2m

(36)

The Pauli susceptibility is


=

3n2B
3N 2
=
= 8.0 107
2kB TF V
2F

(37)

Note that the magnetic susceptibility is dimensionless.


b) (9) The magnetic moment is
e~
= 1.65 1023 erg/gauss
(38)
2M c
The Langevin paramagnetic susceptibility of classical spins at 10 K is
= 3.26N = 3.26

n2
= 3.1 109
3kB T

(39)

The exact result needs to use a quantum-mechanical treatment with the


Brillouin function, but is of this order of magnitude.


B
N gJB BJ KB T
M
=
=
(40)
B
B
In the limit B/kB T 1, this becomes

nJ (J + 1) 2
= 1.2 108
3kB T

(41)

c) (8) We need to find the gyromagnetic ratio.

= ~J

(42)

3.26N
= 3
= 1.04 104 s1 gauss1
~J
~
2

(43)

The NMR frequency is related to the magnetic field by the gyromagnetic


ratio:
= B = 1.04 108 s1
7

(44)

8. (25) Magnetism of a crystal of atoms with local spins


a) (8) The thermodynamic average of the magnetization is
M =n

eB/kB T + eB/kB T
B
=
n
tanh
kB T
eB/kB T + eB/kB T

(45)

b) (8) Taking the first term of the Taylor series when B kB T ,


M = n tanh

n2 B
C
B

=
kB T
kB T
T

(46)

c) (9) The effective field is due to the exchange interaction. At 0 K all the
spins are aligned. The energy of one spin due to its Z nearest neighbors is
BE = ZJS (S + 1) = 34 JZ. Therefore
BE =

3JZ
4

(47)

Another form, using 3kB TC = 2ZJS (S + 1), is


BE =

3kB TC
2

(48)

You might also like