Fall 2010 Qualifier Solutions

Jack Wimberley
January 12, 2012

Part I

Problem 1.1.
Part 1.1.a. Drawing a parallel to the x-axis at the center of the circle, we can see that
the location of the mass is the sum of two vectors of magnitude a and at angles t and
t + . So, its x and y coordinates are:
x(t) = a cos t + a cos(t + (t))
y(t) = a sin t + a sin(t + (t))
Part 1.1.b. The kinetic energy of the mass (which is a function of time) is


1
T (t) = m x(t)
2 + y(t)
2
2

2
1
2

= ma
sin t ( + (t))
sin(t + (t))
2

2 

+ cos t + ( + (t)) cos(t + (t))



1
2

= ma2 2 + ( + (t))
+ 2( + (t))
cos (t)
2
Since there is no potential energy (we are ignoring gravity), the kinetic energy is the only
term in the Lagrangian:


1
2 + 2( + )
cos .
L = ma2 2 + ( + )
2
Part 1.1.c. The Euler-Lagrange equation for the coordinate is
h
i
d L L

= ma2 + 2 sin = 0.
dt

So, the equation of motion is

= 2 sin .
1

Part 1.1.d. In the small angle approximation,  1 and sin . The equation of
motion becomes = 2 , and with the initial conditions given the solution is
(t) = A cos t.
Evidently, the mass travels around the ring at the same frequency that the ring rotates
Part 1.1.e. The generalized momentum is
p =



L
= ma2 + + cos .

To calculate the Hamiltonian, we will need to rewrite in terms of p :

p
=
(1 + cos ).
ma2
Now we can calculate the Hamiltonian:
L=
H = p

p2
1
p (1 + cos ) ma2 2 sin2 .
2
2ma
2

The Hamiltonian is a constant of motion since it doesnt contain any explicit time dependence. It is a theorem of mechanics that the time derivative of a variable u(, p , t)
is
du
u
= {u, H} +
.
dt
t
So, since H/t = 0 and {H, H} = 0, H is a constant of motion.
Problem 1.2.
Part 1.2.a. Maxwells four equations in a vacuum / dielectric environment are
E
B =

 E
c2 t

B =0
E =

B
t

The third equation doesnt add anything new to our knowledge, since we have assumed
that By is the only nonzero component of the magnetic field and that it doesnt have any
y-variation. Similarly, the first equation doesnt tell us anything that the second does not.
The second equates vectors each with two nonzero components and the fourth equations

equates vectors with one nonzero component, leading to three total equations relating Ex ,
Ez , and By :
kc2
By

ic2 By
Ez =
 x

Ex =

iBy = ikEx

Ez
x

Part 1.2.b. Substituting the first two equations above into the third, its easy to show
that
 2



2
k + 2 By = 0.
x2
c
Part 1.2.c. To ensure that the fields are localized around the yz plane, we look for solutions
of the form By = Ae|x|/ , where > 0 is a penetration depth into the vacuum or metal.
For now x > 0 (the axes can be oriented in this way as long as we are considering each
region separately). Substituting this form into the differential equation in (b) for some
region where x > 0, we find that
=p

1
k2

 2 /c2

which is greater than 0 as demanded under the condition

k2 >

 2
.
c2

Part 1.2.d. At any chargless electromagnetic boundary, the tangential components of E

and H are continuous, so the boundary conditions here take the form
Ez+ = Ex
By+ = By
When  < 0, the penetration depth can be more conveniently rewritten
p
1
+ = + = k 2 + || (/c)2 .

In the vacuum, where  = 1, the penetration depth is instead

p
1
= = k 2 (/c)2 .

So, the magnetic field is

(
By =

A+ e x x > 0

A e x x < 0
3

The boundary condition on By implies that A+ = A = A. It follows that the electric

field component Ez is
(
+
ic2 By
iAc2
+ / || e x x > 0
Ez =
=

(x) x

e x
x<0
So, the boundary condition on Ez means that k must satisfy
+ = ||
k 2 + || (/c)2 = ||2 k 2 ||2 (/c)2
(||2 1)k 2 = || (|| + 1)(/c)2
s
|| (|| + 1)
k=
c
||2 1
Part 1.2.e. k when || 1, or  1. When  takes the form 1 02 / 2 , this
occurs when 02 / 2 = 2, or equivalent at the plasmon frequency
0
=
2
Problem 1.3.
Part 1.3.a. In a cubic lattice, there are c = 6 nearest neighbors.
Part 1.3.b. Since the atoms are randomly distributed, on average nA c of the nearest
neighbors are type A and nB c are type B. Assuming only nearest neighbor interaction,
the interaction energy for an atom of type A is thus




1
1
EA =
AA nA c +
AB nB c,
2
2
and the interaction energy for an atom of type B is




1
1
AB nA c +
BB nB c.
EB =
2
2
So, the average interaction energy for an arbitrary lattice point is


c
U = nA EA + nB EB =
AA n2A + 2AB nA nB + BB n2B .
2
Now since nA + nB = 1 and nA nB = x, we have nA = (1 + x)/2 and nB = (1 x)/2.
So, the average interaction energy per site becomes


c
U (x) =
AA (1 + x)2 + 2AB (1 x2 ) + BB (1 x)2 .
8
In the case where AA = BB = 0 and AB > 0 , this interaction energy becomes


c
U (x) =
(AB + 0 ) (AB 0 )x2 .
4
4

Part 1.3.c. The N atoms of type A and B can be distributed in N !/NA !NB ! ways. Using
Stirlings approximation, the computational entropy per atom can be found to be


kB
N!
S=
= kB (nA ln nA + nB ln nB ) .
ln
N
NA !NB !
In terms of the parameter x, the entropy is
S(x) = kB ln 2

kB
((1 + x) ln(1 + x) + (1 x) ln(1 x)) .
2

When x = 1, the entropy is equal to 0, which checks out since there is only one possible
configuration in these situations.
Part 1.3.d. The free energy is
F (x, T ) = C(T )

c 2 1
x kB T ((1 + x) ln(1 + x) + (1 x) ln(1 x)) ,
4
2

where C(T ) = c/4 + kB T ln 2. For convenience, we can rewrite this using the parameter
= c/2kB T as


1
F (x, T ) = C(T ) kB T (1 + x) ln(1 + x) + (1 x) ln(1 x) x2 .
2
Then the first and second derivatives with respect to x are
1
F 0 (x, T ) = kB T (ln(1 + x) ln(1 x) 2x) ,
2 

1
00
F (x, T ) = kB T
.
1 x2
The local extrema of F at constant temperature T are evidently the solutions to
ln(1 + x) ln(1 x) = 2x.
At high temperatures,  1, so to first order in the exponential exp(2x) = 1 + 2x
and so we have
1+x
1 + 2x =
.
1x
The only root of this equation in the range |x| 1 is x = 0, since the other is 1 1/  0.
The second derivative at x = 0 is obviously positive, so x = 0 is a local minimum. Since
there are no other extrema, this is also the global minimum in the range |x| 1.
In the other limit, when the temperature is very high, the parameter becomes very
large. So, to find a non-zero root x of the equation F 0 (x) = 0, we need an x so that either
or both of the terms ln(1 + x) and ln(1 x) are very large. ln z becomes arbitrarily large
5

in magnitude as z 0, so we can expect to find a root for x = (1 ) for some small

positive constant . Taking x = 1 , we get
ln(2 ) ln = 2(1 ) ln 2 ln 2.
So = 2 exp(2). This makes the three local extrema x = 0 and x = (1 2 exp(2)).
The second derivative at x = 0 is negative, making this a local maximum, and the other
two are local minima.
Part 1.3.e. As stated in the problem, instability onsets when the second derivative of F
is positive, i.e. when the free energy lies in a well. Such a system cannot lower its free
energy by segrating into two distinct phases. Consider what would happen if it tried to
split into two phases with mixtures x + and x , for small . Then the total free energy
would be
F (x , T ) + F (x + , T )
1
F (x, T ) + 2 F 00 (x, T ) > F (x, T ).
2
2
At high temperatures, we showed in (d) that F (x, T ) reaches a global minimum at x = 0
and F 00 (x, T ) is positive everywhere, and consquently all mixed states for any x are stable.
However, at lower temperatures, we found that the x = 0 is a local maximum and there
are local minima on either side of x. Evidently, at some critical temperature the mixed
state becomes unstable (the sign of F 00 must switch), and there is a phase change in which
the segregated regions have elements A and B in the ratio
nA = 1 ec/kB T , nB = ec/kB T
or vice versa (these were derived from the values of x at the minima). The critical temperature and admixture at which instability first occurs is clearly located at x = 0, where
the free energy switches from a local minima to a local maxima when


c
00
F (0, T ) = kB T 1
= 0,
2kB T
i.e. when the temperature is
T =

c
.
2kB

Part 1.3.f . Above, we found that for T  T , x1 (T ) and x2 (T ) are given approximately
by x1 (T ) = 1 2 exp(c/kB T ) and x2 (T ) = x1 (T ). Outside of this limit, they are the
positive and negative solutions to


1 + xi (T )
c
=
xi (T ).
ln
1 xi (T )
kB T

If x < x1 (T ), the state of the alloy will be mixed (mostly of type B), and similarly if
x2 (T ) < x then the state will be mixed (mostly of type A). When x1 (T ) < x < x2 (T ),
the allow will segregate in sections that are predominantly type A or type B, and will be
in an unmixed state. A complication: should there not be a small range inside of each of
x1 and x2 in which the state is stable and mixed? How close are the inflection points to
the minima? Plotting some figures, this seems to be a big problem with my analysis. The
regions of stability and instability are determined by the second derivative of F , not the
first. The free energy can be minizied by a phase change into the segragated state, but
it would have to make a big jump between states of higher free energy. Perhaps the fact
that in this regions one of the alloy constituents is so outnumbered makes this jumping
easier, since a relatively small number of atoms need to clump together.
As T 0, clearly x1 (T ) 1 and x2 (T ) 1, looking at the approximations we found
for T  1. In the ground state of a binary alloy at T = 0, the alloy is segrated into regions
that are purely type A or type B, respectively. Since the self interaction 0 is less than
the cross interaction energy AB , this ensures that the interaction energy in the alloy is
minimized (especially when the surface area of the segrated areas is minimized).
Part 1.3.g. Based on the above analysis, Tc (x) must be temperature at which F 00 (x, Tc ) =
0. So,
c
Tc (x) =
(1 x2 ) = T (1 x2 ).
2kB
Clearly T is the maximal value of Tc (x) (at x = 0). We can rewrite the second derivative
of F using Tc as
kB (T Tc (x))
F 00 (x, T ) =
.
1 x2
When T > Tc (x), the second derivative F 00 (x, T ) is positive and the free energy of the
system lies in a well, and around this value of x the system is in a stable mixed state.
When T < Tc (x), the system is unstable and can lower its free energy by segrating into
unmixed states.
Problem 1.4.
Part 1.4.a. The minimum energy condition occurs when there is no relative motion between the outgoing particles, and they can be treated as one particle. Letting E be the
energy of the photon and letting the proton and pion be labeled by p and , conservation
of momentum and energy in this case necesitate that
Et = pt c = pp c + p c
q
Et + mp c2 = (pp c + p c)2 + (mp + m )2 c4
So, we have
Et2 + 2mp c2 Et + m2p c4 = Et2 + (mp + m )2 c4
7

and therefore



m
Et = 1 +
m c2 .
2mp

Part 1.4.b. Set up the system to that the photon travels to the right and the proton to
the left. Let this new intertial frame be found by an intertial boost with velocity to the
right, so that the protons velocity is to the left. The photons four-momentum (the first
two components, at least) is given by

 




ECM B /c

Et /c
Et /c
=
= (1 )
.
ECM B /c

Et /c
Et /c
Letting x = ECM B /Et , we obviously have
s
(1 ) =

1
= x,
1+

and therefore

1 x2
= 1 2x2 + O(x4 ).
1 + x2
The corresponding Lorentz factor is
=

1
1
1
1 + x2
=
+ x + O(x2 ).
=p
=
2
2x
2x
2
1
The protons four-momentum is thus

 


 
mp c
mp c
Ep /c

,
=
=
mp c
0
pp

and so
Ep = mp c2


1
2 1
= mp c
+ x + O(x2 )
2
x


Et
ECM B
1
+
mp c2
2
ECM B
Et
!



1
2
1
m
m
c
m
E

CM
B
= mp c2
1+
+ 1+
2
2mp ECM B
2mp
m c2
m2p
m2 + 2mp m 4
=
c + 2
ECM B .
4ECM B
m + 2mp m

Part 1.4.c. Since ECM B is so small, the second term in Ep is almost entirely negligible.
Using the numbers provided, we find that
Ep 67.9 TeV.
This is an extremely high energy proton, an order of magnitude more energetic than can
be produced in the LHC at CERN.
Part 1.4.d. I have no idea why it is appropriate to use the higher CMB photon energy in
calculating the estimate. I should ask someone. Perhaps there is enough variation in the
CMB that these higher energy photons are not too uncommon over long enough distances?
Part 1.4.e. The protons could not have come from very far away, because over long
enough distances the probability of a scattering event with a CMB photon to lose energy
and produce a pion becomes nearly 100%. Since the rate of this reaction is limited by the
electromagnetic force, it will not especially unlikely and this distance will not have to be
extremely far.
Problem 1.5.
Part 1.5.a. This is a simple matter of applying Kirchoffs Circuit Laws. The drop in
potential across the inductor is
dI
V = L0 x,
dt
assuming that x is small enough that the current is in phase over the width, and so the
potential at the next capacitor is
V (x + x) = V (x) V = V (x) (L0 x)

I(x)
.
t

In the limit x 0, the derivative of the voltage at x is

V
V (x + x) V (x)
I
= lim
= L0 .
x0
x
x
t
Similarly, the current that flows through the capacitor at x + x is
I = (C0 x)

V (x + x)
V (x)
(C0 x)
,
t
t

since the charge stored on the capicitor is Q = C0 V . Since the voltage is measured relative
to the grounded bottom line, this charge flows from the top line to the bottom, since this
is direction positive charges flow. So, the current flow out of the junction is
I(x + x) = I(x) I = I(x) (C0 x)
9

V (x)
,
t

and so in the limit x 0 the change we have

I
I(x + x) I(x)
V
= lim
= C0
.
x x0
x
t
These are the Telegraphers Equations.
Part 1.5.b. The second derivative of the potential is
2V
2I
2I
2V
=
L
=
L
=
L
C
,
0
0
0
0
x2
xt
tx
x2
which is the wave equation with velocity
v=

1
.
L0 C0

Part 1.5.c. Inserting a solution of the given form into the wave equation, we have
k 2 V (x, t) =

2
V (x, t),
v2

and thus (taking the positive root)

= vk =

k
.
L0 C0

Inserting these forms into the first of the Telegraphers Equations yields
iC0 V0+ = ikI0+ ,
or
V0+
k
Z0 =
=
=
I0+
C0

L0
.
C0

For a wave travelling to the left, the dispersion relation between and k is unchanged and
the impedence Z0 picks up a negative sign. By convention, the exponential displacement
is i(t + kx).
Part 1.5.d. The AC voltage generator generateds a voltage
V (0, t) = V0 eit .
Extending this solution to the entire seminfinite axis, we see that the voltage is a right
travelling wave:
V (x, t) = V0 ei(tkx) .

10

Obviously V0+ = V0 and so by part c we know that I0+ = V0+ /Z0 , therefore
r
C0 i(tkx)
I(x, t) = V0
e
.
L0
The power leaving the generator and entering the line is
P (0, t) = < [V (0, t)] < [I(0, t)] =

V02

C0
cos2 t.
L0

The time averaged power is

V2
P (0) = 0
2

C0
.
L0

Since the impedence is completely real, power is transmitted to the line very efficiently. If
there had been an imaginary component to the impedence, some of the current would have
been reflected backwards. For example, if the impedence had been completely imaginary,
there would have been no real current at the origin and the power would have been zero.
Part 1.5.e. The voltage and current are the sums of a right travelling wave and a left
travelling wave. The voltage is
V (x, t) = V0+ ei(tkx) + V0 ei(t+kx) ,
and this must match the ideal generator at x = 0, i.e.
< [V (0, t)] = (V0+ + V0 ) cos t V0 cos t
and so V0+ + V0 = V0 . Since the impedence is positive for the right travelling wave and
negative for the left travelling wave, the current is
r

C0 
I(x, t) =
V0+ ei(tkx) V0 ei(t+kx) .
L0
The right-going amplitude is
r
I0+ = V0+

C0
L0

and the left-going amplitude is similarly

r
I0 = V0

C0
.
L0

The total real current must be 0 at the termination, i.e.

< [I(`, t)] = I0+ cos (t k`) + I0 cos (t + k`) 0.
11

Trig identity time! This implies that

(I0+ + I0 ) cos2 t + (I0+ I0 ) cos2 k` 0.
For this current to be identically zero, we must have k = n/2` for some odd integer n (an
expected condition) and also I0+ + I0 = 0, or
I0 = I0+ .
In terms of the voltage amplitudes, this condition says that V0 = V0+ , and these must
add to V0 we have V0+ = V0 = V0 /2. The real voltage and current are given by
V (x, t) = V0 cos(kx) cos(t)
r
C0
sin(kx) sin(t)
I(x, t) = V0
L0

Part II

Problem 2.1. This is very similar to a problem from Spring 2011.

Part 2.1.a. Since P (P (x)) = (x), the only possible eigenvalues of P are 1 and 1,
the square roots of unity. The corresponding eigenspaces are the spaces of even and odd
functions, respectively.
Part 2.1.b. At x 6= 0, Schrodingers equation is
~2 2
(x) = E(x),
2m
and ( x) and + (x) obviously satisfy this equation with appropriate choice of k . In
particular,
2
~2 k
E =
.
2m
The function (x) is odd and so has parity 1, while + (x) is even (note the absolute
value) and has parity 1.
Part 2.1.c. The wave function must satisfy (a) = (a) = 0. The condition (4) is
easily seen by integrating Schrodingers equation over the range (, ) and taking the
limit  0.
Part 2.1.d. The boundary condition on (x) is
(a) = sin(k a) = 0

12

(the condition at a is superfluous and the condition (4) is satisfied since (x) is continuous and (0) = 0). So, the wave vector is constrained to values
n
k =
| n = 1, 2, 3, . . .
a
for integer values n. The corresponding eigenenergy E is
2 ~2 n2
.
2ma2
Obviously, E is not affected by the presence/absence of the spike potential since the odd
wave function has no value at the spike.
E =

Part 2.1.e. For the even wave function + (x), the boundary conditions at a are trivially
satisfied and so we must focus on condition (4). The discontinuity in the derivative of + (x)
at x = 0 is
0
0
+
(+0) +
(0) = 2k+ cos(k+ a),
and so condition (4) requires that
k+ cos(k+ a) =
or that
tan(k+ a) =

m
sin(k+ a),
~2

k+ a
ma
, where = 2 .

Since k+ a is dimensionless, so must be , and therefore has same units as ~2 /ma, or

(Js)2 /(kgm) = J m.
Part 2.1.f . In the limit 0, we also have
(k+ a) cot(k+ a) = 0.
The solutions to z cot z = 0 are z = 0 and z = n/2 for odd integers n, and so the permitted
nontrivial values of k+ are
n
k+ =
| n = 1, 3, 5, . . .
2a
and the energy levels are
2 ~2 n2
E+ =
| n = 1, 3, 5, . . .
8ma2
The corresponding wavefunctions take the form


 nx 
n |x| n
+ (x) = sin

= cos
.
2a
2
2a
Together with the odd parity solutions (x), this provides the complete spectrum for the
wavefunction of a particle in a box ( = 0) centered at x = 0. If the coordinates are shifted
so that a 0, then the wavefunctions (x) and + (x) take the forms sin(nx/2a) for
any integer n.
13

Part 2.1.g. In the limit , the wave number k+ is restricted to values such that
tan(k+ a) =

k+ a
0.

Since the roots of tan z = 0 are the same as the roots of sin z = 0, i.e. z = n, the allowed
wave numbers are
n
k+ =
| n = 1, 2, 3, . . .
a
and the energy levels are
2 ~2 n2
| n = 1, 2, 3, . . .
E+ =
2ma2
These wavenumbers are identical to those of the odd parity eigenstates. Moreover, their
corresponding eigenstates are


  nx 
n |x|


+ (x) = sin
n = sin
.
a
a
These are just the absolute values of the eigenstates (x).
Part 2.1.h. The four lowest energy states can be characterized by a certain parameter n
equal 1/2, 1, 3/2, 2, respectively, with the energy of each states proportional to n2 . Their
parities are 1, 1, 1, 1 respectively. As increases from 0 to , the energies of the 1/2
and 3/2 states increase and asymptotically approach the energies of the 1 and 2 states. So,
once becomes sufficiently large the system develops parity degeneracy, wherein each
energy level has degenerate odd and even eigenstates.
Problem 2.2.
Part 2.2.a. Since the population of particles in state |ai has a mean lifetime of a , the
time dependent population is
pa (t) = pa (t0 )e(tt0 )/a .
The multistep reaction is limited by the slowest step. If |bi |gi at least as fas |ai |bi,
i.e. if b a , then the population pb (t) will always be zero (this is an expected value of
course, and in reality there will be random fluctuations). In this case the fluoresence is
limited by the decay rate |ai |bi, and so
F (t) =

1
pa (t0 ) (tt0 )/a
pa =
e
.
a
a

Otherwise when b > a , then the differential equation determining pb (t) is

dpb
1
dpa
1
1
= pb
= pb + pa (t0 )e(tt0 )/a .
dt
b
dt
b
a
14

The solution to this equation with pb (t0 ) = 0 is



b
pb (t) =
pa (t0 ) e(tt0 )/b e(tt0 )/a .
b a
The fluoresence signal is the number of decays |bi |gi per unit time:
F (t) =


1
pa (t0 )  (tt0 )/b
e(tt0 )/a .
e
pb =
b
b a

Part 2.2.b. When a  b , the first form of the fluoresence is in effect (the reaction is
limited by a ) and so
pa (t0 ) (tt0 )/a
F (t)
e
.
a
When a  b , the reaction is limited by b and
F (t)

pa (t0 ) (tt0 )/b

e
b

When a b , the reaction is still limited by a since particles leave |bi as fast as they enter
and so
pa (t0 ) (tt0 )/ab
F (t)
e
.
a,b
Part 2.2.c. Since a > b , the fluoresence is limited by the decay |ai |bi. Use the graph
to find times t1 and t2 (after F reaches its maximum) between which F decreases by an
order of magnitude. This is sufficient to calculate that
a 282 ns.
Problem 2.3.
Part 2.3.a. In one dimension (z 0 ), Schrodingers equation can be rearranged and we


2
1
0
=

2m
E

V
(x,
y,
z
)
.
z 02
~2
Consider the incoming particle. In the high momentum limit, the semi-classical approximation is appropriate, and so
(x, y, z) = exp (iS(z)/~)
where

S(z) =

p
2m [E V (x, y, z 0 )] dz 0 .

Obviously the wavefunction is given by (1).

15

Part 2.3.b. When E  V , then

r
p

V
2m(E V ) = 2mE 1
E


V
2mE 1
2E
r

m
= 2mE
V
2E
1
= ~k V.
v
So, the wave function is given by
 Z
(x, y, z) = exp ik

dz



Z z
i
0
0
V (x, y, z ) dz
exp
v~

and since kz = k r (we ignore the phase at z 0 = ), this is



Z z
i
ikr
0
0
k (r) = e
exp
V (x, y, z ) dz .
v~
Part 2.3.c. Substituting (2) into (3), and noting that qz 0, we have
Z

Z
Rz
i
m
v~
V (b,z 0 ) dz 0
0
2 iqb

f (k, k ) =
d be
dz V (b, z)e
.
2~2

Now since
i

V (b, z)e v~

Rz

V (b,z 0 ) dz 0

= iv~

o
z
d n i R
V (b,z 0 ) dz 0
e v~
,
dz

we have (recalling mv = ~k)

Z
h i Rz
i
0
0
imv
d2 beiqb e v~ V (b,z ) dz
f (k, k ) =
2~

Z h
i
R
i
0 ) dz 0
k
v~
V
(b,z

e
=
1 eiqb d2 b.
2i
0

Part 2.3.d. The integration over z 0 is very straightforward, yielding


Z
2V0 a2 b2 b < a
0
0
V (b, z ) dz =
0
ba

16

To see this, just imagine a sphere being pierced a distance b = |b| from its axis. Now the
integral over b will be zero when b a, since e0 1 = 0. So, we have
 Z

h 2V0 2 2
i
2
tot = =
ei v~ a b 1 d2 b
i

 Z b<ah
i
2V
2
i v~0 a2 b2
d b
=< 2
1e
b<a


Z a
2V0 p 2
2
r dr d
1 cos
a r
=2
v~
0


Z a
2V0
1 cos
= 4
x
x dx
v~
0
Z
4 a
(x x cos x) dx
= 2
0
where is the constant

2V0
2mV0
= 2 .
v~
~ k
What this means is that the scattering cross section displays a quasi-periodic function of
a, with nodes at a = 2n or, following through a bunch of algebra, at wavelengths
where
V0 a
=
.
E() n
=

This agrees with our naive expectation that quantum interference effects should arise when
incoming particles have a wavelength that is a fraction of the radius a of the sphere
(adjusted by the factor V0 /E(), which might not have been guessed but seems sensical).
Part 2.3.e. To first order in V , equation (4) becomes
Z
m
0
f (k, k ) =
V (r)eiqr d3 r.
2~2
This is simply equation (3) taking k (r) to be the plane wave exp(ik r), which is none
other than the Born approximation. For this limit to be applicable, we need
Z
1
2V0 p 2
V (b, z 0 ) dz 0 =
a b2  1
v~
v~
for all b. The maximum of this expression occurs when b = 0, so the condition is equivalent
to
v~  V0 a.

Since mv = ~k = 2mE, we can also see that this is also equivalent to the more revealing
expression
E  V0 a.
17

Problem 2.4.
Part 2.4.a. The spherical harmonic Y`m (, ) is proportional to eim , so the spherically
symmetric state with spin j is
|j, 0i .
Part 2.4.b. All states of the form
{|0, 0i , |1, 0i , . . . , |2j, 0i}
are spherically symmetric elements of the composite systems eigenbasis with spin in the
range |j j| jtot j + j. So, the spherically symmetric state of this system is a linear
combination of these eigenbasis elements.
Part 2.4.c.
Part 2.4.d.
Problem 2.5. This is the next problem. This text will show up on the right after a short
delay.
Part 2.5.a.
Part 2.5.b.
Part 2.5.c.
Part 2.5.d.

18