70

SCATTERING

Now consider B; k , λ A(x, t) · p I I A(x , t ) · p A; k, λ (2.83)

The state I can contain either 0 or 2 photons because A(x , t ) changes the number of photons by ±1 depending upon which one of the two terms in A(x , t ) is selected. To get an intermediate state with zero photons, we † select from Equation (2.83) aλ (k )ε ∗ λ (k ) exp(+iω t) from the first matrix element and aλ (k)ε λ (k) exp(−iωt ) from the second matrix element. To get an intermediate state with two photons, we select aλ (k)ε λ (k) exp(−iωt) from the first matrix element and a† λ (k )ε λ (k ) exp(+iω t ) from the second matrix element. We use again the dipole (or long wavelength) approximation and set exp[(k − k ) · x] = 1 and Equation (2.82) becomes
(2) ˙ f (t) = (−i)2 − λ2 c t t1

e m

2

4π √ 2V ωω

B ε λ (k) · p e−iωt I
I

×

+iω t dt I ε∗ A ei(EB −EI )t ei(EI −EA )t λ (k ) · p e

+ (−i)2 − ×
t t1

e m

2

4π √ 2V ωω

+iω t I B ε∗ λ (k ) · p e I

dt I ελ (k) · p e−iωt A ei(EB −EI )t ei(EI −EA )t

(2.84)

Note the structure of the two terms, it is easy to spot errors when comparing them. We want to integrate this result over t and it is tempting to set all integration limits to −∞ and +∞ as we did in Section 1.3 to get Equation (1.99). This would be incorrect here because the upper integration limit t of the integral over t is itself an integration variable when integrating c ˙2 (t). When doing the integral over t , the value of t should be kept fixed f at an arbitrary finite value and should not be set to infinity before the integration of t is completed. So consider the integral over t in the first term of Equation (2.84)
t t1

dt ei(ω +EI −EA )t = = =

t t1

dt eλt ei(ω +EI −EA )t
t t1

eλt ei(ω +EI −EA )t λ + i(ω + EI − EA ) eλt+i(ω +EI −EA )t

− λ + i(ω + EI − EA )

eλt1 +i(ω +EI −EA )t1

(2.85)