Problem solutions, 9 May 2012

1
D. E. Soper
2
University of Oregon
16 May 2012
Here is a solution for problem 5.38, which seemed to cause the most diculty.
Please check the algebra: there could be errors.
Problem 5.38 With a potential V
0
cos(kz t), we write
V
0
cos(kz t) =
V
0
2
e
ikz
e
it
+
V
0
2
e
ikz
e
+it
. (1)
The rst term leads to transititions to states

with
f
=
1
+, where
1
is the energy of the ground state atom that we start with. The second term
leads to transititions to states

with
f
=
1
. Since the nal state of
interest has
f
>
1
, this term does not contribute.
The transition rate is (using eq. (58) from our notes)
dR = (2)(
f

1
) d p
f
V
2
0
4
|

p
f

e
ikz

1
_
|
2
(2)
We can write this using
f
= p
2
f
/(2m) and
d p
f
= mp
f
d
f
d
f
. (3)
This gives the transition rate per unit solid angle:
dR
d
f
=
mp
f
V
2
0
2
|

p
f

e
ikz

1
_
|
2
(4)
Lets look at the matrix element

p
f

e
ikz

1
_
. The nal state wave function
should really be an exact eigenstate for an electron in a Coulomb potential,
but we approximate this as a plane wave. Then we have factors exp(i p
f
r)
and exp(i

k r), where

k points along the z-axis. The product of these is
exp(iq r) where q = p
f

k.

p
f

e
ikz

1
_
=
1
(2)
3/2
_

0
r
2
dr
0
(r)
_
1
1
d cos e
iqr cos
_
2
0
d (5)
1
Copyright, 2012, D. E. Soper
2
soper@uoregon.edu
1
Here we have chosen the coordinate system for the position r of the electron
so that the z-axis is along q. We have denoted the radial part of the wave
function by
0
(r):

0
(r) =
1

_
Z
a
0
_
3/2
e
Zr/a
0
. (6)
Now we can perform the and integrals:

p
f

e
ikz

1
_
=
i
(2)
1/2
q
_

0
rdr
0
(r)
_
e
iqr
e
iqr

. (7)
We can perform the r integral using
_

0
rdre
r
=
1

2
. (8)
Thus

p
f

e
ikz

1
_
=
i
(2)
1/2
q
1

_
Z
a
0
_
3/2
_
1
[(Z/a
0
) iq]
2

1
[(Z/a
0
) + iq]
2
_
.
(9)
That is

p
f

e
ikz

1
_
=
i
(2)
1/2
q
1

_
Z
a
0
_
3/2
4iZq/a
0
[(Z
2
/a
2
0
) + q
2
]
2
. (10)
This simplies to

p
f

e
ikz

1
_
=
2
3/2

_
Z
a
0
_
5/2
1
[(Z
2
/a
2
0
) + q
2
]
2
. (11)
The dependence on angle is contained in
q
2
= p
2
f
+ k
2
2p
f
k cos , (12)
where is the angle between p
f
and the original z-axis.
With an incoming electromagnetic wave, there is an operator p, which
becomes p
f
and there is a polarization vector . The amplitude is propor-
tional to p
f
, so there is a strong angular dependence favoring emission in
the direction of the polarization, . In the present problem, the angular
dependence is weak if k is small compared to p
f
.
2