19Jan2011
Chemistry 21b – Spectroscopy
Lecture # 7 – TimeDependent Perturbation Theory & LightMatter Interactions
In Ch 21a you considered the effects of small, time independent perturbations to
various systems (see also the supplemental notes from Week #1). Spectroscopy, of course,
is concerned with the interaction of matter with electromagnetic radiation, which is
inherently a timedependent process. The general procedure we will follow is similar, but
here the question becomes “What is the probability that, some time
t
after the perturbation
is applied, the system has undergone a transition from an initial state
i
to a final state
f
?”
The problem we wish to solve has the total Hamiltonian
ˆ
H
(
r
,t
)
=
ˆ
H
0
(
r
) +
λ
ˆ
H
′
(
r
,t
)
,
where
λ
again is a parameter of smallness (perturbation theory notes).
Note that we have assumed the zerothorder Hamiltonian
ˆ
H
0
is timeindependent. In
that case, we know that the timedependent eigenfunctions of
ˆ
H
0
may be written as
ψ
n
(
r
,t
)
=
ϕ
n
(
r
)
e
−
iω
n
t
ˆ
H
0
ϕ
n
=
E
(0)
n
ϕ
n
≡
¯
hω
n
ϕ
n
.
Let us suppose that at time
t>
0 the system is in a state characterized by
ψ
(
r
,t
)
=
summationdisplay
n
c
n
(
t
)
ψ
n
(
r
,t
)
.
(7
.
1)
The superposition principle tells us that the probability that the system is in the state
ψ
n
at time
t
is simply given by

c
n
(
t
)

2
. Thus, to answer the question above we must find the
{
c
n
(
t
)
}
coefficients. To obtain them we proceed much as in the timeindependent case.
Clearly, the total wavefunction
ψ
(
r
,t
) must satisfy
ı
¯
h
∂ψ
∂t
=
(
ˆ
H
0
+
λ
ˆ
H
′
)
ψ .
If we insert Eq.
(7.1) into this expression, and as before operate from the left with
<ψ
f
(
r
,t
)

, we are left with
ı
¯
h
dc
f
dt
=
λ
summationdisplay
n
<f

H
′

n>c
n
.
(7
.
2)
In general, this is an infinite series of coupled equations for the coefficients
{
c
f
(
t
)
}
.
As
λ
→
0, the timedependent perturbation becomes negligible and the
c
f
coefficients remain
constant. Thus, as for the timedependent case let us seek a solution in a power series of
corrections to the initial constants, or
c
f
(
t
)
=
c
(0)
f
+
λc
(1)
f
(
t
) +
λ
2
c
(2)
f
(
t
) +
...
57
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Inserting this series into Eq. (7.2) and collecting like terms in
λ
results in the following
corrections to the timedependent coefficients [where the dot denotes a time derivative and
the
H
′
nf
are the matrix elements of
ˆ
H
′
(
r
,t
)]:
ı
¯
h
˙
c
(0)
f
=
0
ı
¯
h
˙
c
(1)
f
=
summationdisplay
n
H
′
fn
c
(0)
n
ı
¯
h
˙
c
(2)
f
=
summationdisplay
n
H
′
fn
c
(1)
n
.
.
.
ı
¯
h
˙
c
(
s
+1)
f
=
summationdisplay
n
H
′
fn
c
(
s
)
n
(7
.
3)
So, to get the corrections to order (
s
+ 1), we must know all of the coefficients to order
s
.
The first equation simply states that the initial coefficients are constant in time until the
perturbation is applied.
These are complex, long equations, and so at this point in the analysis it is common
to make a simplifying assumption, namely that the system starts in a definite eigenstate
of
ˆ
H
0
, call it
ψ
i
(
r
,t
). If we use the initial time as
t
=
−∞
, we have
ψ
(
r
,t
)
∼
ψ
i
(
r
,t
)
=
summationdisplay
n
δ
ni
ψ
n
(
r
,t
)
c
(0)
n
=
δ
ni
Substituting these expressions into the firstorder part of Eq. (7.3) gives
This is the end of the preview.
Sign up
to
access the rest of the document.
 Fall '10
 list
 Chemistry, matrix elements, Axial multipole moments, quadrupole, Multipole expansion

Click to edit the document details