Solutions

Final Exam 2024

1. 50 points. Cartesian basis

Consider a lithium atom (neutral lithium has three electrons) in a [He] 2s electron configuration that is
illuminated by light with ϵ̂ = êx polarization that is resonant with a transition from [He] 2s to [He] 3p
(ignore electron spin). After a certain amount of time, the atom is in a 50/50 superposition of the
|Rn,L , L M ⟩ = |R2,0 , 0 0⟩ ≡ |ψg ⟩ state and some combination |ψe ⟩ of |R3,1 , L′ = 1, M ′ ⟩ states in the 3p
manifold. At some moment, the light is turned off and the atom’s state evolves according to
1
|ψ(t)⟩ = √ |ψg ⟩ + e−iωt |ψe ⟩ ,


2
where ℏω = ∆E is the energy difference between the 2s and 3p manifolds. (You should ignore the
possibility of spontaneous emission for this problem.)

(a) Using hydrogenic wavefunctions for the outer electron (use Z = 1 to account for shielding of the
nucleus), calculate the radial part of the E1 transition matrix element, reg ≡ ⟨R3,1 | r |R2,0 ⟩. You
to not need to simplify your answer after performing the integral and indicating the dependence
of your answer on constants (i.e. don’t spend time simplifying the pure-number part). For the
rest of this problem, you may use the symbol reg to denote this result. (This problem may
have been modified later to say that the integral does not need to be evaluated.)
(b) Write the operator d • ϵ̂ for this polarization in terms of spherical harmonics.
(c) Using this, calculate the transition dipole matrix elements for photon absorption,
⟨R3,1 , 1, M ′ | d • ϵ̂ |R2,0 , 0 0⟩, for the three allowed values of M ′ ∈ {−1, 0, 1} (in terms of reg ).
(d) Use this to construct a normalized state vector |ψe ⟩ = M ′ cM ′ |L′ = 1, M ′ ⟩ that maximizes the
P
magnitude of the transition dipole moment ⟨ψg | d • ϵ̂ |ψe ⟩; this is the state excited by the light.
Based on this, would you say that this light couples an s orbital to px , py , pz , or some combination
of these “chemistry-basis” orbitals?
(e) Use the form of the time-dependent state vector above to calculate the expectation value of the
electric dipole moment in |ψ(t)⟩ as a function of time.

(Hint: you may find it helpful to rewrite the dipole moment operator in terms of spherical harmonics.)

Solution
(a) Using the table at the back of the exam, we have
1
⟨r|R2,0 ⟩ = p 5 (r − 2a0 ) e−r/(2a0 )
2 2a0
s
2 2
⟨r|R3,1 ⟩ = r (r − 6a0 ) e−r/(3a0 )
81 3a70

1
 The radial integral we need to compute is
Z ∞ ! s !
1 2 2
⟨R3,1 | r |R2,0 ⟩ = dr r2 (r − 2a0 ) e−r/(2a0 ) r r (r − 6a0 ) e−r/(3a0 )
81 3a70
p
0 2 2a50
s Z ∞
1 2
= dr r4 (r − 2a0 )(r − 6a0 ) e−5r/(6a0 )
81 2a50 3a70 0
If the integral needs to be evaluated, if we identify
5
a≡ ,
6a0
we can use the integral in the back to the test to evaluate the three terms we need,
Z ∞
6! 6! · 67 7
dr r6 e−ar = 7 = a
0 a 57 0
Z ∞
8 · 5! 8 · 5! · 66 7
−8a0 dr r5 e−ar = − 6 a0 = − a0
0 a 56
Z ∞
12 · 4! 2 12 · 4! · 65 7
12a20 dr r4 e−ar = a = a0 ,
0 a5 0 55
which gives us

4! · 65 6 · 5 · 62
  
1 8·5·6
reg = a0 √ − + 12
81 3 55 52 5

which, frankly, is even more-simplified than is necessary. For the curious, this simplifies all the way to
√
27648 3
reg = a0 ≈ 3.06a0 .
15625
(b) Since we know that ê1 = − √12 (êx + iêy ) and ê−1 = √1 (êx
2
− iêy ), we have
2
ê1 − ê−1 = − √ êx ,
2
so
1
ϵ̂ = êx = − √ (ê1 − ê−1 ).
2
It is okay if anybody wrote this down by inspection. Next, it will be helpful to write the operator in
the form
d = − er
r
4π X
= − er Y (ω) ê∗q ,
3 q 1,q

so that
r !
4π
Y1,1 (ω) ê∗1 + Y1,−1 (ω) ê∗−1


d • ϵ̂ = r −e • êx
3
! r  
4π ∗ ∗

1 1
= r −e Y1,1 (ω) ê1 + Y1,−1 (ω) ê−1 • − √ ê1 + √ ê−1
3 2 2
r ! 
4π 1 1
= r −e − √ Y1,1 (ω) + √ Y1,−1 (ω)
3 2 2

2
(c) In terms of reg , we have
r ! 
′ 4π 1
⟨R3,1 , 1, M | d ϵ̂ |R2,0 , 0 0⟩ = reg
• −e −√ (⟨1, M ′ | Y1,1 (ω) |0 0)⟩ − ⟨1, M ′ | Y1,−1 (ω) |0 0)⟩) .
3 2

Why do calculus when you can do algebra instead? These integrals are over products of three spherical
harmonics,
Z
′ ∗
⟨1, M | Y1,q (ω) |0 0)⟩ = d2 ω Y1,M ′ (ω) Y1,q (ω) Y0,0 (ω)

which we know how to evaluate in terms of Wigner 3j symbols:

r   
3·3
Z
2 ∗ M′ 1 1 0 1 1 0
d ω Y1,M ′ (ω) Y1,q (ω) Y0,0 (ω) = (−1)
4π 0 0 0 −M ′ q 0
r   
′ 3·3 1 δ ′
= (−1)M −√ √ Mq
4π 2·1+1 2·1+1
δM ′ q
= √
4π
Putting all of this together gives us
r  
′ 2π δM ′ ,1 δM ′ ,−1
⟨R3,1 , 1, M | d ϵ̂ |R2,0 , 0 0⟩ = e reg
• √ − √
3 4π 4π
M′
= e reg √ .
6

(d) Since the M ′ = 0 component doesn’t couple to the ground state with this polarization, we can ignore
it. In terms of the cM ′ , we’d have the transition moment
1
⟨ψe | d • ϵ̂ |R2,0 , 0 0⟩ = e reg √ c∗1 − c∗−1 .


6
Since this is a normalized state vector, we require that
2 2
|c1 | + |c−1 | = 1.

One choice that maximizes the magnitude of this transition moment is

1
c1 = −c2 = √ ,
2

but this times any global phase eiϕ is just as correct. The main point is that this is a 50/50 superpo-
sition, and that the phase of the superposition is π .
Comparing this to the tesseral harmonics u1,1 and v1,1 , we see that the relative minus sign between the
Y1,1 (ω) and Y1,−1 (ω) parts of the spatial wavefunction indicates that we have something proportional
to u1,1 , which corresponds to px .

(e) We’re asked to calculate

1
⟨ψg | d |ψg ⟩ + ⟨ψe | d |ψe ⟩ + e−iωt ⟨ψg | d |ψe ⟩ + eiωt ⟨ψe | d |ψg ⟩ .


⟨ψ(t)| d |ψ(t)⟩ =
2

3
 We know that the first two terms are zero, as these are parity eigenstates. Since the q = 0 component
of a spherical vector can’t couple |ψg ⟩ to |ψe ⟩, we only need to consider the q = ±1 vector components
of d. Using the values for c1 and c−1 from above, we have
 
−iωt −iωt 1 1
e ⟨ψg | d |ψe ⟩ = e ⟨R2,0 , 0 0| d √ |R3,1 , 1, 1⟩ − √ |R3,1 , 1, −1⟩
2 2
r
4π −iωt 1 
= −e e reg √ ⟨0 0| Y1,−1 (ω) |1, 1⟩ ê∗−1 − ⟨0 0| Y1,1 (ω) |1, −1⟩ ê∗1
3 2

+ ⟨0 0| Y1,1 (ω) |1, 1⟩ ê∗1 − ⟨0 0| Y1,−1 (ω) |1, −1⟩ ê∗−1 .

Now we can evaluate all four of these with our identity,

r   
3· 0 1 1 0 1 1
⟨0 0| Y1,q (ω) |1, M ′ ⟩ ê∗q = ê∗q
4π 0 0 0 0 q M′
r   
3· 1 δ−M ′ ,q
= −√ √ ê∗q
4π 3 3
δ−M ′ ,q ∗
= − √ êq ,
4π
which shows us that the last two terms are zero. We are left with
r  
−iωt 2π −iωt 1 1
e ⟨ψg | d |ψe ⟩ = − e e reg − √ ê∗−1 + √ ê∗1
3 4π 4π
e −iωt 1
reg √ ê∗−1 − ê∗1


= √ e
3 2
 
e −iωt 1
= √ e reg − √ (ê1 − ê−1 )
3 2
e reg
= √ e−iωt êx .
3
Since matrix elements are just numbers, the other one is the complex conjugate of this,
e reg
eiωt ⟨ψe | d |ψg ⟩ = √ eiωt êx ,
3
giving us
 
1 e reg −iωt e reg
⟨ψ(t)| d |ψ(t)⟩ = √ e êx + √ eiωt êx
2 3 3
e reg
= √ cos(ωt) êx .
3

Broadly speaking, the point here is that x-polarized light coupled an s orbital to px , because the electric
dipole interaction induces an electric dipole moment that oscillates in the x direction.

2. 20 points. Equivalent electrons

In this problem, you will show that the terms that arise from an np3 electron configuration are
n o
4 o
S3/2 , 2 Do5/2 , 2 Do3/2 , 2 Po3/2 , 2 Po1/2 .

(a) Calculate the number of eigenstates that this electron configuration will have.

4
 (b) Construct a table B with the (MS , ML , MJ ) triples of these allowed states.
(c) Identify a term from your table and cross off entries you can assign to that term, then repeat
until you have no more entries.
You do not need to rewrite your table a bunch of times for part (c); just indicate clearly which term
you’re associating with each entry. (Hint: make sure your table in part (b) has the number of entries
you calculated in part (a).)

Solution
Perhaps surprisingly, this is actually easier than the nd2 configuration you did on Problem Set 06. So if you
did that problem, this should be fairly straightforward (while still tedious).
(a) Here we need the statistical weight of an np3 subshell,

gp !
g(np3 ) =
3!((gp − 3)!)

We know that a p orbital has gp = 2(2 · 1 + 1) = 6 berths in it, so we have

6! 6·5·4
g(np3 ) = = = 5 · 4 = 20
3! (3!) 3!

states in this electron configuration.

(b) The order here is unimportant (I just number the lines to help me keep track). I’ll start by enumerating
all of the states for which MJ > 0, and then the rest will just be these same 10 rows multiplied by a
minus sign.

ML 1 ML2 ML 3 MS1 MS2 MS3 (ML , MS , MJ )

1. 1 1 0 1/2 −1/2 1/2 (2, 1/2, 5/2)
2. 1 1 0 1/2 −1/2 −1/2 (2, −1/2, 3/2)
3. 1 0 0 1/2 1/2 −1/2 (1, 1/2, 3/2)
4. 1 0 0 −1/2 1/2 −1/2 (1, −1/2, 1/2)
5. 1 1 −1 1/2 −1/2 1/2 (1, 1/2, 3/2)
6. 1 1 −1 1/2 −1/2 −1/2 (1, −1/2, 1/2)
7. 1 0 −1 1/2 1/2 1/2 (0, 3/2, 3/2)
8. 1 0 −1 1/2 1/2 −1/2 (0, 1/2, 1/2)
9. 1 0 −1 1/2 −1/2 1/2 (0, 1/2, 1/2)
10. 1 0 −1 −1/2 1/2 1/2 (0, 1/2, 1/2)

Adding to this the same entries with minus signs leaves us with the table B:

5
 (ML , MS , MJ )
1. (2, 1/2, 5/2)
2. (2, −1/2, 3/2)
3. (1, 1/2, 3/2)
4. (1, −1/2, 1/2)
5. (1, 1/2, 3/2)
6. (1, −1/2, 1/2)
7. (0, 3/2, 3/2)
8. (0, 1/2, 1/2)
9. (0, 1/2, 1/2)
10. (0, 1/2, 1/2)
11. (−2, −1/2, −5/2)
12. (−2, 1/2, −3/2)
13. (−1, −1/2, −3/2)
14. (−1, 1/2, −1/2)
15. (−1, −1/2, −3/2)
16. (−1, 1/2, −1/2)
17. (0, −3/2, −3/2)
18. (0, −1/2, −1/2)
19. (0, −1/2, −1/2)
20. (0, −1/2, −1/2)

(c) We start by identifying the entry(ies) with maximum ML , which is lines 1. and 2. Of these, the
maximum MS is line 1., so this is our first one to identify: this is an L = 2, S = 1/2 state, which is a
2
DJ term. We will have two allowed values of J ∈ {5/2, 3/2}. For MJ = 5/2, we only have one and
we cross of line 1. For MJ = 3/2, lines 2. and 3. work (because they are allowed and distinct from
one another). MJ = 1/2 will also have two, and line 4. works, but we can’t use line 6. because it’s not
distinct from line 4., so we cross off line 8. instead. Adding 10 to each line number we’ve used (to get
the negative MJ versions of these) allows P us to cross off lines 11., 12., 13., 14., and 18. So we have
identified two allowed terms, and since i Li = 3, all of these will have odd parity:
2
Do5/2 and 2 Do3/2

Our table B now looks like

(ML , MS , MJ )
5. (1, 1/2, 3/2)
6. (1, −1/2, 1/2)
7. (0, 3/2, 3/2)
9. (0, 1/2, 1/2)
10. (0, 1/2, 1/2)
15. (−1, −1/2, −3/2)
16. (−1, 1/2, −1/2)
17. (0, −3/2, −3/2)
19. (0, −1/2, −1/2)
20. (0, −1/2, −1/2)

Next, we again look for a stretched state and we see that line 5. works, and corresponds to a 2 PJ term.
This has J ∈ {3/2, 1/2}. We can now cross off lines 5., 6., 9., 15., 16., and 19. to add
2
Po3/2 and 2 Po1/2 .

Our table, then, is left with

6
 (ML , MS , MJ )
7. (0, 3/2, 3/2)
10. (0, 1/2, 1/2)
17. (0, −3/2, −3/2)
20. (0, −1/2, −1/2)

Here, we see that the stretched state is L = 0 and S = 3/2, giving a

4 o
S3/2
term. The four states in the table therefore complete our analysis.
3. 30 points. M 1 transitions
Consider a magnetic dipole transition in an atom in free space between the ground 2 Po3/2 and excited
P1/2 manifolds with the same electron configuration. There are two σ − transitions in this system
2 o

(here, since we are considering an M 1 transition, σ − refers to the projection of the magnetic field of
the light on the quantization axis) that connect M in the ground state to M ′ = M − 1 in the excited
state. You may assume gs ≈ 2 and I = 0.
E
(a) Write the J- and MJ -doubly-stretched state ( 2 Po3/2 , M = 3/2 ) in the uncoupled basis |L ML , S MS ⟩.
E
(b) Write the state 2 Po1/2 , M ′ = 1/2 in the uncoupled basis |L ML′ , S MS′ ⟩.
(c) Use these to calculate the M 1 transition dipole moment for absorption, µ(3/2), defined via
D E
µ(M ) ≡ 2 Po1/2 , M − 1 µ • ê−1 2 Po3/2 , M .

(Hint: rewrite the operator in terms of projection-raising and -lowering operators.)

(d) Calculate the other M 1 transition dipole moment for σ − polarization: µ(1/2). For this, you can
either expand the states again in the uncoupled basis or just use the Wigner-Eckart theorem to
relate µ(1/2) to µ(3/2).

Solution
(a) The only way to make MJ = 3/2 if L = 1 and S = 1/2 is if ML = 1 and MS = 1/2:
E
2 o
P3/2 , M = 3/2 = |L 1, S 1/2⟩

(b) For this, we need the expansion

√
 
L S J
E
L−S+M ′
X
′
2 o
P1/2 , M = 1/2 = (−) 2J + 1 ′ |L ML′ , S MS′ ⟩.
′ ′
M L MS′ −M ′
ML ,MS

The only two ways to make MJ′ = 1/2 from (ML′ , MS′ ) are (1, −1/2) and (0, 1/2), so we have
E √  
√
 
1 1/2 1/2 1 1/2 1/2
P1/2 , M ′ = 1/2 = 2
2 o
|L 1, S, −1/2⟩ + 2 |L 0, S, 1/2⟩
1 −1/2 −1/2 0 1/2 −1/2
Using the table in the back of the test, we see that
 
1 1/2 1/2 1
= −√
1 −1/2 −1/2 3
 
1 1/2 1/2 1
= √ ,
0 1/2 −1/2 6

7
 so we get
r
E 2 1
2
Po1/2 , M ′ = 1/2 = − |L 1, S, −1/2⟩ + √ |L 0, S, 1/2⟩ .
3 3

(c) The magnetic moment operator is

µ = −µB (L + 2S).

For this, it will be helpful to write this in terms of projection-raising and -lowering operators (L± and
S± ),
 
(1) (1)
µ • ê−1 = − µB T−1 [L] + 2 T−1 [S]
 
L− S−
= − µB √ + 2 √
2 2
 
µB
= −√ (L− + 2S− ).
2
We can work out the action of these operators on our basis states,
p √
L− |L 1, S 1/2⟩ = 1(1 + 1) − 1(1 − 1) |L 0, S 1/2⟩ = 2 |L 0, S 1/2⟩

and
s    
1 1 1 1
S− |L 1, S 1/2⟩ = +1 − − 1 |L 1, S, −1/2⟩ = |L 1, S, −1/2⟩.
2 2 2 2

We can now calculate the desired matrix element,

D E
µ(3/2) = 2 Po1/2 , 1/2 µ • ê−1 2 Po3/2 , 3/2
 
µB D2 o E
= −√ P1/2 , 1/2 (L− + 2S− ) 2 Po3/2 , 3/2
2
  r !
µB 2 1
= −√ − ⟨L 1, S, −1/2| + √ ⟨L 0, S 1/2| (L− + 2S− ) |L 1, S 1/2⟩
2 3 3
!

µB
 r
2 1 √ 
= −√ − ⟨L 1, S, −1/2| + √ ⟨L 0, S 1/2| 2 |L 0, S 1/2⟩ + 2 |L 1, S, −1/2⟩
2 3 3
  r r !
µB 2 2
= −√ −2 +
2 3 3
µB
= √ .
3

(d) The fast way to do this is to use the Wigner-Eckart theorem. We’re taking the matrix element of the

8
 q = −1 component of a rank 1 irreducible tensor operator,

J′
 
′ ′ (1) J ′ −M ′ 1 J
⟨J M | T−1 [µ] |J M ⟩ = (−1) f (J ′ , |µ|, J)
−M ′ −1 M
 
1/2 1 3/2
= (−1) 1/2−M +1
f (J ′ , |µ|, J)
1 − M −1 M
 
3/2 1/2 1
= (−1) 3/2−M
f (J ′ , |µ|, J)
M 1 − M −1
 
3/2 1/2 1
= (−1) 3/2−M +3
f (J ′ , |µ|, J)
−M M − 1 1
s
(M + 12 )( 32 + M )
= (−1)1/2−M (−1)1/2+M −1 f (J ′ , |µ|, J)
24
s
1 (M + 12 )( 32 + M )
= f (J ′ , |µ|, J)
2 6

Since the reduced matrix element is the same for both transitions, the ratio of the matrix elements is
the ratio of the front factors,
s
µ(1/2) ( 12 + 12 )( 23 + 12 )
=
µ(3/2) ( 32 + 12 )( 23 + 32 )
1
= √
3
and we get
1
µ(1/2) = √ µ(3/2)
3
µB
=
3

3. 2 bonus points. Professor Campbell’s midterm mistake

On the midterm exam, Professor Campbell wrote that the electron-orbital contribution to the magnetic
moment was

µ = µB L.

What is wrong with this?

Solution
It is off by a minus sign ,

µ = −µB L.