763312A QUANTUM MECHANICS I - solution set 13 - autumn 2012

1. One-dimensional anharmonic oscillator: perturbation

matrix elements
Consider a one-dimensional anharmonic oscillator. Let the perturbation be of
the form H1 = σ~ωx3 . Calculate the matrix elements
r
3 (n + 3)(n + 2)(n + 1)
hψn+3 |x |ψn i = ,
8
 3
n+1 2
hψn+1 |x3 |ψn i = 3 ,
2
 n  32
hψn−1 |x3 |ψn i = 3 ,
r 2
n(n − 1)(n − 2)
hψn−3 |x3 |ψn i = .
8

Solution
In the lecture notes section 17.2 we derived a recurrence relation for the har-
monic oscillator energy eigenfunctions. Let us repeat it here:
r r
n+1 n
xψn (x) = ψn+1 (x) + ψn−1 (x). (1.1)
2 2

Employing Eq. (1.1), we obtain

 
2 1q 1 1q
x ψn (x) = (n + 2)(n + 1)ψn+2 (x) + n+ ψn (x) + n(n − 1)ψn−2 (x). (1.2)
2 2 2

Employing Eqs. (1.2) and (1.1), we obtain

3 2
x ψn (x) = xx ψn (x)
 
1q 1 1q
= (n + 2)(n + 1)xψn+2 (x) + n + xψn (x) + n(n − 1)xψn−2 (x)
2 2 2
s
 3/2
(n + 3)(n + 2)(n + 1) n+1
= ψn+3 (x) + 3 ψn+1 (x)
8 2
s
 3/2
n n(n − 1)(n − 2)
+3 ψn−1 (x) + ψn−3 (x). (1.3)
2 8

The energy eigenfunctions ψn (x) are orthonormal. Therefore, by multiplying

both sides by ψn+3 (x) and integrating from minus infinity to infinity, we obtain
r
3 (n + 3)(n + 2)(n + 1)
hψn+3 |x |ψn i = . (1.4)
8

1
Similarly we obtain
 3/2
3 n+1
hψn+1 |x |ψn i = 3 ,
2
 n 3/2
hψn−1 |x3 |ψn i = 3 ,
r 2
n(n − 1)(n − 2)
hψn−3 |x3 |ψn i = . (1.5)
8

2
2. One-dimensional anharmonic oscillator: perturbative
treatment
Let us consider an anharmonic oscillator in which the perturbation is of the
form H1 = σ~ωx3 . First, show that the energy levels are (to the second order
in σ)  
1 2 15 1 2 7
En = (n + )~ω − σ ~ω (n + ) + .
2 4 2 16
Then, show that the energy difference
 
15
En − En−1 = 1 − σ 2 n ~ω.
2
Finally, determine the energy eigenfunctions to the first order in σ.
This topic is discussed in the lecture notes section 17.2.3.

Solution
The energy levels are, to the second order in σ, [cf. Eq. (144)]
(0) (0)
X |hψm |H1 |ψn i|2
En = En(0) + hψn(0) |H1 |ψn(0) i + (0) (0)
. (2.1)
m6=n En − Em
In the previous problem, we calculated the values of all nonzero matrix elements.
(0)
Moreover, we know that the harmonic oscillator energy levels are En = (n +
1/2)~ω, where n ∈ {0, 1, · · · }. Using these results, we obtain
(0) (0) (0) (0) (0) (0)
(0) |hψn−3 |H1 |ψn i|2 |hψn−1 |H1 |ψn i|2 |hψn+1 |H1 |ψn i|2
En = En + (0) (0)
+ (0) (0)
+ (0) (0)
En − En−3 En − En−1 En − En+1
(0) (0)
|hψn+3 |H1 |ψn i|2
+ (0) (0)
− En En+3
9n3 9(n + 1)3
  
1 n(n − 1)(n − 2)
= n+ ~ω + σ 2 ~ω + −
2 24 8 8

(n + 3)(n + 2)(n + 1)
−
24
  
1 1 9 9
= n+ ~ω + σ 2 ~ω (n3 − 3n2 + 2n) + n3 − (n3 + 3n2 + n + 1)
2 24 8 8

1 3
− (n + 6n2 + 11n + 6)
24
   
1 15 1 7
= n+ ~ω − σ 2 ~ω (n + )2 + . (2.2)
2 4 2 16
The energy difference
 
1 15 1 7
En − En−1 = (n + )~ω − σ 2 ~ω (n + )2 +
2 4 2 16
 
1 2 15 1 7
− (n − 1 + )~ω + σ ~ω (n − 1 + )2 +
2 4 2 16
 
15 2
= 1− σ n ~ω. (2.3)
2

3
The nth energy eigenfunction is, to the first order in σ, [cf. Eq. (144)]
(0)
X hψm (0)
(0) |H1 |ψn i (0)
ψn = ψn + (0) (0)
ψm
m6=n En − Em
(0) σ (0) 3 (0) (0) (0) (0) (0)
= ψn + hψ |x |ψn iψn−3 + σhψn−1 |x3 |ψn iψn−1
3 n−3
(0) (0) (0) σ (0) (0) (0)
+ σhψn+1 |x3 |ψn iψn+1 − hψn+3 |x3 |ψn iψn+3
3
 r
(0) 1 n(n − 1)(n − 2) (0)  n 3/2
(0)
= ψn + σ ψn−3 + 3 ψn−1
3 8 2
n + 1 3/2 (0)
  r 
1 (n + 3)(n + 2)(n + 1) (0)
−3 ψn+1 − ψn+3 . (2.4)
2 3 8

4
3. van der Waals constant
Calculate the van der Waals constant
e4
C = h1s; 1s|(xA xB + yA yB − 2zA zB )2 |1s; 1si
2E1s
2
e4  2

rA 
= − 6 h1s| |1si
2E1s 3
= 12a60 × Ry.

Solution
The van der Waals force between two hydrogen atoms is discussed in the lecture
notes section 17.2.4. Here we begin from the result

e4
C=− h1s; 1s|(xA xB + yA yB − 2zA zB )2 |1s; 1si. (3.1)
2E1s
Since the position operators commute with each other,
(xA xB + yA yB − 2zA zB )2 = x2A x2B + yA
2 2 2 2
yB + 4zA zB
+ 2xA xB yA yB − 4xA xB zA zB − 4yA yB zA zB . (3.2)
Substitution Eq. (3.2) into Eq. (3.1) yields
e4

C=− h1s; 1s|x2A x2B |1s; 1si + h1s; 1s|yA
2 2 2 2
yB |1s; 1si + 4h1s; 1s|zA zB |1s; 1si
2E1s
+2h1s; 1s|xA xB yA yB |1s; 1si − 4h1s; 1s|xA xB zA zB |1s; 1si

−4h1s; 1s|yA yB zA zB |1s; 1si . (3.3)

The wave function of the state 1s is (cf. p. 148)

1
ψ100 (r) = √ e−r . (3.4)
π

The Cartesian and spherical coordinates of electron A are related by


 xA = rA sin θA cos ϕA ,
yA = rA sin θA sin ϕA , (3.5)
zA = rA cos θA .


Naturally, the Cartesian and spherical coordinates of electron B are related in

a similar manner.
Since we know that the state 1s is spherically symmetric, we have the intu-
ition that the last three matrix elements at the right-hand side of Eq. (3.3) are
zero. However, let us also calculate the third last matrix element. Employing

5
Eqs. (3.4) and (3.5), we obtain
Z Z
∗ ∗
h1s; 1s|xA xB yA yB |1s; 1si = ψ100 (rA ) ψ100 (rB ) xA xB yA yB
R3 R3
×ψ100 (rA )ψ100 (rB ) dVA dVB
1
Z ∞ Z π Z 2π
−2rA 4 3
= e rA drA sin θA dθA cos ϕA sin ϕA dϕ
π2 0 0 0
Z ∞ Z π Z 2π
−2rB 4 3
× e rB drB sin θB dθB cos ϕB sin ϕB dϕ
0 0 0
| {z }
=0

= 0. (3.6)

Indeed, the third last matrix element vanishes as we expected. Similarly we can
show that both the second last and the last matrix element indeed vanish. Thus
we now have the result
e4 
h1s; 1s|x2A x2B |1s; 1si + h1s; 1s|yA
2 2 2 2

C=− yB |1s; 1si + 4h1s; 1s|zA zB |1s; 1si . (3.7)
2E1s
The first matrix element inside the brackets,
Z Z
2 2 ∗ ∗ 2 2
h1s; 1s|xA xB |1s; 1si = ψ100 (rA ) ψ100 (rB ) xA xB ψ100 (rA )ψ100 (rB ) dVA dVB
R3 R3
Z Z
∗ 2 ∗ 2
= ψ100 (rA ) xA ψ100 (rA ) dVA ψ100 (rB ) xB ψ100 (rB )dVB
R3 R3
| {z }| {z }
=h1s|x2 |1si =h1s|x2 |1si
2 2
= h1s|x |1si . (3.8)

For the second last and the last matrix element we obtain similar results. Thus
we have the result
e4 
h1s|x2 |1si2 + h1s|y 2 |1si2 + 4h1s|z 2 |1si2 .

C=− (3.9)
2E1s
Since the state 1s is spherically symmetric, we have the intuition that the matrix
elements at the right-hand side of Eq. (3.9) are equal (we could verify this by a
simple calculation). Application of this fact yields

e4
C = − 6h1s|r2 /3|1si2
2E1s
e4
= − h1s|r2 |1si2 , (3.10)
3E1s
where r2 = x2 + y 2 + z 2 . On the other hand,
Z
h1s|r2 |1si = ψ100 (r)∗ r2 ψ100 (r) dV
R3
Z ∞ Z π Z 2π
1
= r4 e−2r dr sin θ dθ dϕ
π 0
|0 {z 0
}
=4π
Z ∞
= 4 r4 e−2r dr
0
= 3, (3.11)

6
where the last equality follows from integration by parts. Writing the Bohr
radius a0 explicitly, we obtain

h1s|r2 |1si = 3a20 . (3.12)

The shorthand e2 is defined on the lecture notes p. 162. We wish to express it

in terms of Rydbergs and the Bohr radius. We obtain (cf. pp. 144-145)
2
q2

4
e =
4π0
2
1 q2

= a20
4π0 a0
| {z }
=2 Ry

= 4a20 × (Ry)2 . (3.13)

Rydberg is defined as the ionization energy of hydrogen. Thus

E1s = −1 Ry. (3.14)

As we substitute Eqs. (3.12) - (3.14) into Eq. (3.10), we obtain

C = 12a60 × Ry. (3.15)

7
4. A fermion system
Consider a system composed of noninteracting spin– 12 fermions in a three-
dimensional harmonic oscillator potential. First, describe the shell structure of
the system. Then, suppose that there are a total of 100 particles in the system
and that all the energy states of the system are filled up to energy EF = 1 Ry.
Estimate the strength of the harmonic potential, V0 = ~ω.

Solution

EN degeneracy total
E0 2 2
E1 6 8
E2 12 20
E3 20 40
E4 30 70
E5 42 112
E6 56 168
.. .. ..
. . .

Table 1: Shell structure

Since we assume that the fermions do not interact with each other, the harmonic
oscillator potential
1
V (r) = mω 2 r2 (4.1)
2
is the potential each fermion experiences. Therefore the energy levels of each
fermion are (cf. p. 130)
3
EN = (N + )~ω. (4.2)
2
For spinless particles, the degeneracy of level EN is (cf. p. 130)
1
fN = (N + 1)(N + 2). (4.3)
2
Since we are now, however, dealing with spin– 21 particles, the degeneracy of
level EN is 1

gN = 2fN
= (N + 1)(N + 2). (4.4)
1 The state of a fermion is not specified by the spatial wave function alone, but by the spin

also. In the case of spin– 21 particles, the spin can obtain two different values.

8
In the second column of Table 1, the degeneracies of the first seven energy levels
are given. In the third column, the total number of states, up to the level in
question, is given.
Let us suppose that there are a total of 100 particles in the system and that
all the energy states of the system are filled up to the energy EF = 1 Ry. Now,
according to Table 1, the sixth shell is the last shell having nonzero occupation
number. Thus we conclude that

EF = E5
 
3
= 5+ ~ω. (4.5)
2

Consequently,
2
~ω = Ry. (4.6)
13

9
5. Binding energies in hydrogen, helium, and lithium
Consider a hydrogen, a helium and a lithium atom and assume that there is no
Coulomb interaction between the electrons. Calculate the binding energies of
the outmost electrons. Does this model give the correct shell structure for these
atoms, or for heavier atoms?
Since we assume that there is no Coulomb interaction between the electrons, the
shell structures of all the three atoms can be solved in a similar manner.

Solution
It is well-known that when the hydrogen atom is in its ground state, the binding
energy of the electron is 13.6 eV. What comes to the heavier elements, we begin
from the assumption that there is no Coulomb interaction between the electrons.
Consequently, on part of each electron, the system is like the hydrogen atom
with the difference that the charge of the nucleus is Ze, where Z is the number
of protons in the nucleus. Thus the time-independent Schrödinger equation for
each electron reads (cf. p. 144)
~2 2 1 Ze2
− ∇r ψ(r) − ψ(r) = Eψ(r). (5.1)
2m 4π0 r
Let us express the quantities in natural units with the exceptions that we now
measure length in multiples of a0 /Z and energy in multiples of Z 2 ~2 /(2ma20 )
(cf. pp. 144-145). That is, let us make the substitutions
r
ψ(r) = u(r0 ), where r0 = ,
a0 /Z
a0 0
r = r,
Z
Z 2 ~2 0
E = E,
2ma20
1 Ze2 Z~2
= 2 . (5.2)
4π0 a0 2ma20
We obtain
~2 2 0 Z 2 ~2 1 Z 2 ~2 0 0
− ∇r u(r ) − 2 2 u(r0 ) = E u(r ). (5.3)
2m 2ma0 r 0 2ma20
As we cancel out common factors and rearrange, we obtain
 2
Z2 0

Z 1
∇r u(r ) + 2 2 0 + 2 E u(r0 ) = 0.
2 0
(5.4)
a0 r a0
Let us evaluate the first term at the left-hand side. Firstly, as we use the
notation r0 = (x0 , y 0 , z 0 ) and apply the chain rule, we obtain
∂u ∂u ∂x0 ∂x0 1
= || =
∂x ∂x0 ∂x ∂x a0 /Z
Z ∂u
= . (5.5)
a0 ∂x0

10
As we employ the previous equation and the fact that the order of partial
differentiations can be swapped, we obtain
∂2u Z ∂ ∂u
=
∂x2 a0 ∂x0 ∂x
Z 2 ∂2u
= . (5.6)
a20 ∂x0 2
This leads us to see that
Z2 2 0
∇2r u(r0 ) = ∇ u(r ), (5.7)
a20
2 2 2
∂ ∂ ∂
where ∇2 = ∂x 0 2 + ∂y 0 2 + ∂z 0 2 . Indeed, we have now evaluated the first term at

the left-hand side of Eq. (5.4). Substituting and cancelling out common factors,
we obtain
 
0 2
2
∇ u(r ) + + E u(r0 ) = 0.
0
(5.8)
r0
We note that this is the lecture notes Eq. (132) with the exceptions that ψ has
been denoted by u, r by r0 and E by E 0 . Like in the lecture notes, we obtain
the result that the energy levels are
1
En0 = − , (5.9)
n2
where n ∈ {1, 2, 3, · · · }. However, due to our choice of units, energy is now
measured in multiples of Z 2 Rydbergs (cf. Eq. (5.2) and pp. 144-145). For
example, the energies of the first and the second level are −Z 2 Rydbergs and
−Z 2 /4 Rydbergs, respectively.
When we do not take the spin into account, we obtain the result that the
energy level En0 is n2 -fold degenerate (cf. p. 147). Taking the spin into account
doubles the degeneracy of each level. Thus, when a helium atom is in its ground
state, its electrons occupy the lowest energy level. Their binding energy
EB = −E1 · 22 Rydbergs
= 4 Rydbergs. (5.10)
When a lithium atom is in its ground state, the outmost of its electrons occupies
the second energy level (lithium has three electrons and up to two electrons can
occupy the lowest energy level simultaneously). Thus the binding energy of the
outmost electron is
EB = −E2 · 32 Rydbergs
9
= Rydbergs. (5.11)
4
This model does not give the correct shell structure for helium or any element
heavier than it. This is due to the fact that neglecting the Coulomb interaction
between the electrons is a crude approximation.

11