Vous êtes sur la page 1sur 5

Physics 141B, Spring 2010

Problem Set #7 Solutions


David Strubbe

1. (Kittel 10.4)
a) We have to solve the differential equation B 2 2 B = 0 with cylindrical
symmetry, i.e. B (, , z) = B (). In cylindrical coordinates, the Laplacian
is


1
2

1 2
2 =
(1)

+ 2 2+ 2


z
Given the cylindrical symmetry, only the first term is nonzero, so our equation becomes





1 B 2 B
1
+

= B 2
=0
(2)
B 2


2
or equivalently,
2 B 2

B
2B
2 2 2 = 0

(3)

Define R = / and A (R) = B (R). Then our equation after dividing by


2 is
R2 A R

2A
A
R2
=0
R
R2

(4)

At this point, to solve the differential equation and get properties of its
solutions, we will consult a reference: M Abramowitz and IA Stegun, Handbook of Mathematical Functions, NIST, 1972. Equations referred to in the
remainder of this problem are from this work. It is available online (via the
UC library) at http://www.knovel.com/web/portal/basic search/
display? EXT KNOVEL DISPLAY bookid=528.
1

Our equation has the form of the defining equation (9.6.1) for modified
Bessel functions (also known as hyperbolic Bessel functions)
z2


d2 w
dw
+z
z2 + 2 w = 0
2
dz
dz

(5)

with the parameter = 0. There are two kinds of solutions, those that are
zero at R = 0 and grow (the first kind, I ) and those that diverge at R = 0
and then go to zero at infinity (the second kind, K ). We need the second
kind since the magnetic field should diverge at the vortex core. As plotted
in Mathematica via Plot[BesselK[0, z], BesselI[0, z], z, 0, 3]:
5

0.5

1.0

1.5

2.0

2.5

3.0

Our solution then is B () = cK0 (/). To find the coefficient c, we use the
given condition
Z
B () d
(6)
0 = 2
Z
Z 0
= 2
cK0 (/) d = 2c2
RK0 (R) dR
(7)
0

We can do the integral in Mathematica, and find that it is 1 (or more


painfully use (9.6.28)).
0
22

(8)

0
K0 (/)
22

(9)

0 = 2c2 ; c =

B () =

b) Unfortunately, these limits cannot be derived by taking limits of the differential equation itself. Instead results for the limits of the modified Bessel
2

function K0 itself must be used. (9.6.8) tells us that for small arguments,
i.e. / 1,
K0 (/) ln (/) = ln (/)

(10)

The requested limit in the problem is because the London equation with constant is not correct inside the coherence length . Conversely,
for large arguments / 1, (9.7.2) gives
r
z
K0 (z) =
e + ...
(11)
2z
s
/
K0 (/)
e
(12)
2
Therefore the limits of the solution are
(
0
B () 2
2 ln (/)
q
B ()

0
22

/
2 e

(13)

2. (Kittel 10.6)

a) Let w be the dimension in the x-direction, the magnetic field be in the y~ = B y


direction, and the transport direction be z. Then the magnetic field B
can be derived from a vector potential A = Bx
z . A particle of charge q
tunneling in the z-direction will acquire a phase from this vector potential
of


 q

qA
(14)
exp i z = exp i Bxz
~c
~c
Specifically, for Cooper pairs (q = 2e) and a tunneling gap of T , the phase
becomes


2e
(15)
exp i BxT
~c
This origin of the phase can be seen from the substitution p p + qc A in
the presence of a vector potential.
Therefore the tunneling rate (called T in Kittels discussion) becomes
complex and position-dependent:


2e
= 0 exp i BxT
(16)
~c
3

Using this tunneling rate in equation 10.42, and taking the real part, we find



h
i
n2
2e
i
= Re 2ine
(17)
= Re 2i0 exp i BxT i
t
~c


2e
= 20 sin
BxT
(18)
~c
Thus we have an effective phase difference given by
(x) =

2e
BxT +
~c

(19)

where we are given = 2 .


To find the total current from this x-dependent expression, we average
x:


Z
Z
1 w/2

2e
1 w/2
dx
sin (x) dx =
sin
BxT +
w w/2
w w/2
~c
2


1 ~c
2e
w/2
=
cos
BxT +
w 2eBT
~c
2 w/2

 w/2

2e
1 ~c
sin
BxT
=
w 2eBT
~c
w/2
 



eBxT
eBxT
1 ~c
sin
sin
=
w 2eBT
~c
~c


1 ~c
eBxT
=
sin
w eBT
~c

over

(20)
(21)
(22)
(23)
(24)

We choose to integrate symmetrically around x = 0 so that the average


phase difference remains = 2 regardless of B.
Our final result is


eBwT
~c
sin
(25)
J = J0
eBwT
~c
3. (Kittel 16.2)
Inside the conducting sphere, the electric field is zero. It is the sum of the
applied field and the depolarization field of the sphere:
E = 0 = E0 + Esphere = E0

4
P
3

(26)

Therefore P = 3E0 /4. The dipole moment of the sphere is


= PV =

3E0 4 3
a = E0 a3 = E0
4 3

(27)

The result for the polarizability is = a3 .


4. (Kittel 16.5)
a) Inside the dielectric sphere, the electric field is
E = E0

4
4
P = E0
E
3
3

(28)

where = 1 + 4. Solving for E, which is the macroscopic (i.e. volumeaveraged) electric field, we get
E=

E0
3
E0
=
2
+

1 + 4

(29)

b)
P = E =

E0
1 + 4
3

(30)

Vous aimerez peut-être aussi