Decay width calculation1 D. E.

Soper2 University of Oregon 21 May 2012

Abstract I oer here a solution for exercise 12.1 in the notes on time dependent perturbation theory.

Problem setup
=

We need to calculate dk 2 (E2 + E1 ) 1, 0, 0; k, V 2, 1, m; 0

2

(1)

Here the atom states are denoted by n, l, m and there is an additional photon with polarization and momentum k that accompanies the nal 1, 0, 0 state. The photon energy is = |k |. The atom states have energies E1 = E0 , E0 E2 = , 4 where I use the shorthand notation E0 = Thus = We have 1, 0, 0; k, V 2, 1, m; 0 = e (k, ) 1, 0, 0 eikx p 2, 1, m 2m . (5)
1 2

(2)

e2 13.6 eV . 2a0 3 E0 . 4

(3)

(4)

Copyright, 2012, D. E. Soper soper@uoregon.edu

Furthermore, we can use the dipole approximation, 1, 0, 0 eikx p 2, 1, m im 1, 0, 0 x 2, 1, m With this approximation, ie 1, 0, 0; k, V 2, 1, m; 0 = (k, ) 1, 0, 0 x 2, 1, m 2 Thus =

(6)

. (7)

dk 2 (E2 + E1 )

e2 (k, ) v (m) 4 2

(8)

where v (m) = 1, 0, 0 x 2, 1, m . (9)

Integrating over the emitted photon

e2 (k, ) v (m) 4 2
2

Write this as (using |k | = ) =

2 d

d 2 (E2 + E1 )
2

e2 3 d (k, ) v (m) 2
2

(10) ,

Now (k, ) v (m)

= |v (m)|2 |k v (m)|2 /k 2 .

(11)

Also we can use the fact that the tensor k i k j averaged over the angles of k is invariant under rotations, so must be proportional to ij . Thus d |k v (m)|2 = d [v (m) ]i v (m)j k i k j 1 ij 2 k 3 (12)

= 4 [v (m) ]i v (m)j = Also d = 4 . 4 |v (m)|2 k 2 . 3

(13)

Thus d

(k, ) v (m)

8 |v (m)|2 . 3

(14)

These results give = 4e2 3 |v (m)|2 , 3 (15)

The matrix element

v (m) = 1, 0, 0 x 2, 1, m . (16)

Now we need This is easiest if we take the m = 0 case. We note that v (0) is invariant under rotations about the z -axis, so it must point in the z -direction. Thus, letting u(0) be a unit vector in the z -direction, we have v (0) = u(0) 1, 0, 0 r cos() 2, 1, 0 The matrix element is a straightforward integral,

(17)

1, 0, 0 r cos() 2, 1, 0 =
0

r2 dr

d r cos()

R10 (r)Y00 (, )R21 (r)Y10 (, )

1

=
0

r2 dr 2
1

d cos r cos() 1 (4 )1/2

1
3/2 a0

2 er/a0

1/2 1 r r/(2a0 ) 3 e cos (2a0 )3/2 31/2 a0 (4 )1/2 1 r4 dr e3r/(2a0 ) = 3/2 4 2 a0 0 1

d cos cos2 1

= =

25 a5 0 3 2 2 3 4 5 3 / 2 3 3 2 a0 215/2 a0 35 (18)

215/2 a0 u(0) . (19) 35 For m = 1, we get almost the same result, but with a unit vector 1 u(1) = (1, i, 0) . (20) 2 Thus in each case, we get v (0) = v (m) = where |u(m)|2 = 1. 215/2 a0 u(m) . 35 (21)

Thus

Result
= 4e2 3 215 2 217 2 2 3 a = e a0 , 3 310 0 311 = 3 E0 4 e2 . 2a0 (22)

Putting this together, we have

We recall that with E0 = Thus 3 = This gives = With

(23)

(24)

33 e 4 E0 28 a2 0

(25)

29 6 e E0 , 38

(26)

E0 13.6 eV , 1 , e2 = 137 this is

(27)

29 1 (28) 13.6 eV 4.13 107 eV . 38 1373 See G. W. F. Drake, J. Kwela, and A. van Wijngaarden, Phys. Rev. A 46, 113 (1992) for theory and experimental results.