Solutions manual for Modern Quantum Chemistry

St Maxwell
st maxwell@outlook.com

1 Chapter 1
Exercise 1.1 a) Show that Oij = ~ei · O~ej . b) If O~a = ~b show that bi =
P
Oij aj .
j
Solution:
a) We already know:
n
X
O~ej = ~ek Okj
k=1

Hence:
n
X n
X n
X
~ei · O~ej = ~ei · ~ek Okj = ~ei~ek Okj = δik Okj = Oij
k=1 k=1 k=1

b) bi = ~ei · ~b = ~ei · O~a

Hence:
n
X n
X
bi = ~ei · Oaj ~ej = aj ~ei · O~ej
j=1 j=1

From the last problem, we know:

~ei · O~ej = Oij
Therefore:
n
X
bi = aj Oij
j=1

Exercise 1.2 Calculate [A, B] and {A, B} when

   
1 1 0 1 −1 1
   
A = 1 B = −1
   
2 2 0 0
   
0 2 −1 1 0 1

Solution:

[A, B] = AB − BA
       
1 1 0 1 −1 1 1 −1 1 1 1 0
       
= 1 2 2  · −1 0 − −1 0 · 1
       
0 0 2 2
       
0 2 −1 1 0 1 1 0 1 0 2 −1
 
0 −2 4
 
= 2
 
0 3
 
−4 −3 0

1
 {A, B} = AB + BA
       
1 1 0 1 −1 1 1 −1 1 1 1 0
       
= 1 2 2  · −1 0 + −1 0 · 1
       
0 0 2 2
       
0 2 −1 1 0 1 1 0 1 0 2 −1
 
0 0 −2
 
=  0 −2 3 
 
 
−2 3 −2

Exercise 1.3 If A is an N × M matrix and B is a M × K matrix show that (AB)† = B† A† .

Solution:
We set C = AB, hence: X
Cij = Aik Bkj
k

†
X
∗
Cij = Cji = A∗jk Bki
∗

Because A†kj = A∗kj †

Bik = ∗
Bki ,
† †
A†kj
X
Cij = Bik
k

C = B† A†
†

Exercise 1.4 Show that

a. trAB=trBA.
b. (AB)−1 = B−1 A−1 .
c. If U is unitary and B = U† AU, then A = UBU† .
d. If the product C = AB of two Hermitian matrices is also Hermitian, then A and B commute.
−1
e. If A is 
Hermitian then
 A , if it exists, is also Hermitian.
 
A11 A12 1 A −A12
f. If A =  , the A−1 =  22 .
A21 A22 (A11 A22 − A12 A21 ) −A21 A11
Solution:
a. We set C = AB D = BA, hence:
N
X
trC = Cnn
n=1
N X
X N
= Ank Bkn
n=1 k=1

N
X
trD = Dnn
n=1
N X
X N
= Bnk Akn
n=1 k=1

N P
P N
Replace n with k and k with n in trD = Bnk Akn respectively(n and k are dummy variables and have
n=1 k=1

2
same value range):
N X
X N
trD = Bkn Ank
k=1 n=1
N X
X N
= Ank Bkn
n=1 k=1

Thus:
trAB = trBA
b. Because AA−1 = A−1 A = 1, therefore:

(AB)−1 (AB) = 1

(AB)−1 AB = 1
(AB)−1 ABB−1 = B−1
(AB)−1 A = B−1
(AB)−1 AA−1 = B−1 A−1
(AB)−1 = B−1 A−1
c. From the definition of unitary matrix:
U† = U−1
Therefore:

UBU† = UU† AUU†

= UU−1 AUU−1
=A
c.
Aik Bkj = Cij
Bjk Aki = Cji
(Bjk Aki )∗ = Cji
∗

∗
Bjk A∗ki = Cji
∗

Because A, B and C are Hermitian matrices, hence:

Bij A∗ki = Cji

∗

d.
AB = C
(AB)† = C†
B† A† = C†
BA = C
Therefore:
AB = BA
[A, B] = 0
e. We already know that:
AA−1 = 1
(AA−1 )† = 1†
(A−1 )† A† = 1†

3
And therefore:
(A−1 )† A = 1
(A−1 )† = A−1
Thus A−1 is Hermitian.
 
x y
f. We suppose A−1 =  . Because AA−1 = 1, hence:
z w
    
A A12 x y 1 0
 11  = 
A21 A22 z w 0 1

And we have simultaneous equations: 

A11 x + A12 z = 1


A x + A z = 0
21 22
A11 y + A12 w = 0


A21 y + A22 w = 1


The solution is:  A22

 x=
11 22 − A12 A21


 A A
−A12



y =

A11 A22 − A12 A21
−A21

 z =
A 11 22 − A12 A21
A





w =
 A11
A11 A22 − A12 A21
At last, we have:  
1 A22 −A12
A−1 =  
(A11 A22 − A12 A21 ) −A21 A11

Exercise 1.5 Verify the above properties for 2 × 2 determinants.

Solution:  
 
0 a
1. Take the determinant   as example:

0 b
 
 
0 a
 =0×b−0×a=0
 
0 b
 
 
a 0 
2. For determinant  :

0 b 
 
 
a 0
  = ab − 0 = ab
 
0 b
   
   
a b c d
3. If A =  , and B = 
 
, then:

c d a b

det(B) = bc − ad = −(ad − bc) = −det(A)

4
    
 ∗
c∗ 
  
a b a
4. Suppose A =  , so A† =  :
 ∗
d∗ 
 
c d b
|A| = ad − bc
 †
A  = a∗ d∗ − b∗ c∗
 
∗
So it is obviously that |A| = A† 

5.  
 
1 2
A =  

3 4
 
 
4 3
B =  

2 1
 
 
8 5
AB =  

20 13
|A| = −2 |B| = −2
|AB| = 4
|A| |B| = |AB|

Exercise 1.6 Using properties (1)-(5) prove that in general

Solution:
6. Suppose the ith and jth columns in the determinant A are equal:
 
 a11 a12 ... a1i ... a1j ... a1n 
 
 
 a21 a22 ... a2i ... a2j . . . a2n 
 
A =  . .. .. .. ..  = x

 .. . . . . 

 
an1 an2 ... ani ... anj . . . ann 

Now exchange the ith and jth columns and have:

 
 a11 a12 ... a1j ... a1i ... a1n 
 
 
 a21 a22 ... a2j ... a2i . . . a2n 
 
0 0
A =  . .. .. .. ..  = x

 .. . . . . 


 
an1 an2 ... anj ... ani . . . ann 

Because A and A0 are same, x = x0 . And from the property (3), we know that x = −x0 . Finally, x = x0 = 0.

7.
AA−1 = 1
Hence:
AA−1  = |1|
 

|A| A−1  = 1
 

A  = (|A|)−1
 −1 

5
  
∗
8. Because |A| = A†  , therefore:
∗
(|A|) = A† 
 

Since AA† = 1,
|A| A†  = 1
 

So:
∗
|A| (|A|) = 1
9. From U† U = UU† = 1, we know
U† = U−1
And therefore:
U  = U  = (|U|)−1
 †   −1 

Because
U† OU = Ω
 †
U  |O| |U| = |Ω|
−1 −1
|U| U†  |O| |U| (|U|) = |U| |Ω| (|U|)
 

Hence:
|O| = |Ω|

Exercise 1.7 Using Eq.(1.39), note that the inverse of a 2 × 2 matrix A obtained in Exercise 1.4f can be
written as  
1 A 22 −A 12
A−1 =  
|A| −A21 A11

Show that the equation

Ac = 0
where A is an N × N matrix and c is a column matrix with elements ci , i = 1, 2, . . . , N can have a nontrivial
solution(c 6= 0) only when |A| = 0.

Solution:
From Exercise 1.4f we know
 
1 A −A12
A−1 =  22 
(A11 A22 − A12 A21 ) −A21 A11

Because:
|A| = A11 A22 − A12 A21
Thus  
1  A22 −A12
A−1 = 
|A| −A21 A11f

6 0, then A−1 exists.

If |A| =
A−1 Ac = 0
c=0

Exercise 1.8 Show that the trace of a matrix is invariant under a unitary transformation, i.e., if

Ω = U† OU

then show that trΩ = trO.

6
Solution:
Because trAB = trBA,
trΩ = trU† OU = trOUU†
Since
UU† = 1
trΩ = trO1 = trO

Exercise 1.9 Show that Eq.(1.90) contains Eq.(1.87) for all α = 1, 2, . . . , N .

Solution:
  
c11 c21 ... cN
1 ω1
  
 1
 c2 c22 . . . cN ω
 
2   2 
Uω =  .

. .
 
..

 .. .. ..  
  
. 
  
c1N c2N . . . cN N ωN
 
ω1 c1 ω2 c1 . . . ωN cN
1 2
1
 
 ω1 c12 ω2 c22 . . . ωN cN
 
2 
= .

..
 = OU
 ..

 . ... 
ω1 c1N ω2 c2N ... ωN cN
N

Since  
cα
1
 
 α
 c2 
cα = 
 .. 

 . 
 
cα
N

Uω = ω1 c1 , ω2 c2 , . . . , ωN cN

It is obviously that
Ocα = ωcα α = 1, 2, . . . , N

Exercise 1.9 & Exercise 1.10

Just have a try. And I don’t think approach (b) is friendly to human. It is convenient when works as a
computer program.
 
a1
 
a2
 
 or Acα = aα cα α = 1, 2, . . . , N . Show that
 
Exercise 1.12 Given that U† AU = a =  .. 
 . 
 
aN
a. det (An ) = an1 an2 · · · anN .
N
b. trAn = anα .
P
α=1
c. If G(ω) = (ω1 − A)−1 , then
N ∗ N
X Uiα Ujα X cα α∗
i cj
(G(ω))ij = =
α=1
ω − aα α=1
ω − aα

7
Show that using Dirac notation this can be rewritten as
X hi | αi hα | ji
(G(ω))ij ≡ hi | G (ω) | ji =
α
ω − aα

Solution:
a. Since An = Uan U† , thus
det (An ) = det (U) det (an ) det U†

Because U−1 = U† ,

det (An ) = det (U) det (an ) det U−1

n
= (det (a))
= an1 an2 · · · anN

b. Because tr U† An U = tr An UU† = tr (An ), hence

N
X
tr (An ) = tr (an ) = anα
α=1

c.
X = U† (ω1 − A)U
= ωU† 1U − U† AU = ω1 − a
 
ω − a1
 
ω − a2
 
 
= ..


 . 
 
ω − aN

It is diagonal matrix, therefore

 
ω − a1
 
ω − a2
 
  †
G(ω) = U 
 ..
U

 . 
 
ω − aN
XX
Uiα Xαβ U †


(G(ω))ij = βj
α β

Because Xαβ = δαβ (ω − aα )−1 ,

∗
X Uiα Ujα X cα α∗
i cj
(G(ω))ij = =
α
ω − aα α
ω − aα
Because
∗
Uiα = hi | αi Ujα = hα | ji
∗
X Uiα Ujα X hi | αi hα | ji
(G(ω))ij = =
α
ω − aα α
ω − aα

Exercise 1.13 If  
a b
A= 
b a

8
show that
f (a + b) + f (a − b) f (a + b) − f (a − b)
 

2
f (A) =  f (a + b) −
 2 
f (a − b) f (a + b) + f (a − b) 
2 2
Solution:
First of all, we should diagonalize A.  
a − ω
 
b 
 =0
a − ω
 
 b
We have
ω1 = a + b ω2 = a − b
So the diagonal matrix  
a+b 0
a= 
0 a−b
When ω = a + b,     
a b c1 c
    = (a + b)  1 
b a c2 c2
Solve the equations, and we have
   1 
c1 √
 12 
 1 =  
1
c2 √
2
Similarly, when ω = a − b, we get
   1 
c2 √
 12 
 1 =  
2
c2 −√
2
And therefore  1 1 
√ √
U =  12 2 

1 
√ −√
2 2
Hence
f (A) = Uf (a)U†
 1 1   1 1 
√ √ √ √
=  12
 2  f (a + b) 0
 12 2 
1  1 
√ −√ 0 f (a − b) √ −√
2 2 2 2
f (a + b) + f (a − b) f (a + b) − f (a − b)
 

2
=  f (a + b) −
 2 
f (a − b) f (a + b) + f (a − b) 
2 2

Exercise 1.14 Using the above representation of δ(x), show that

Z ∞
a(0) = dx a(x)δ(x)
−∞

9
Solution:
Z ∞ 
lim a(x) = lim dx a(x)δ(x)
ε→0 ε→0 −∞
Z −ε Z −ε Z ∞ 
= lim dx a(x)δ(x) + dx a(x)δ(x) + dx a(x)δ(x)
ε→0 −∞ +ε +ε
Z −ε
= lim dx a(x)δ(x)
ε→0 +ε
= a(0)

Exercise 1.15 As a further illustration of the consistency of our notation, consider the matrix representation
of an operator O in the basis {ψi (x)}. Starting with
X
Oψi (x) = ψj (x)Oji
j

Show that Z
Oji = dx ψj∗ (x)Oψi (x)

Then using Eqs.(1.127a) and (1.138) rewrite (1) in bra-ket notation and show that it is identical to Eq.(1.55).
Solution:
X
Oψi (x) = ψj (x)Oji
j
X
ψk∗ (x)Oψi (x) = ψk∗ (x)ψj (x)Oji
j
Z Z X
dx ψk∗ (x)Oψi (x) = dx ψk∗ (x)ψj (x)Oji
j
 
X Z X
= Oji  dx ψk∗ (x)ψj (x)
j j
X
= Oji δkj = Oki
j

Therefore Z
Oji = dx ψj∗ (x)Oψi (x)

Exercise 1.16

Solution:

Oφ(x) = ωφ(x)
∞
X ∞
X
Ocj ψj (x) = ωcj ψj (x)
j=1 j=1

Multiply ψi∗ (x) on both side:

∞
X ∞
X
ψi∗ (x)Ocj ψj (x) = ωcj ψi∗ (x)ψj (x)
j=1 j=1
∞
X ∞
X
Oij cj = ω cj δij
j=1 j=1

= ωci

10
Assume  
o1
 
 o2 
 
O=
 .. 
 on = (on1 , on2 , . . . , onj )
.
 
on
Therefore:
oi c = ωci
Generalize i from n to ∞:
Oij c = ωc

Exercise 1.17

Solution:
a. Z
dx hi | xi hx | ji = hi | ji = δij
Z
dx ψi∗ (x)ψj (x) = δij

b.
∞
X
hx0 | ii hi | xi = hx0 | xi = δ(x0 − x) = δ(x − x0 )
i=1
X
ψi∗ (x0 )ψi (x) = δ(x − x0 )
i
c. Z
dx hx0 | xi hx | ai = hx0 | ai

Because hx0 | xi = δ(x − x0 ), Z

dx δ(x − x0 )a(x) = a(x0 )

d.
O |ai = |bi
Multiply hx| on the left side and insert unity:

hx | bi = hx | O | ai
Z
= dx0 hx | O | x0 i hx0 | ai
Z
b(x) = dx0 O(x, x0 )a(x0 )
e.
O(x, x0 ) = hx | O | x0 i
XX
= hx | ii hi | O | ji hj | x0 i
i j
X
= ψi (x)Oij ψj∗ (x0 )
ij

Exercise 1.18

11
Solution:
Since |Φ̃i is normalized: D  E
Φ̃  Φ̃ = 1

Z ∞  ∗
2 2
N e−αx N e−αx dx = 1
−∞
Z ∞
2
N2 e−2αx dx = 1
−∞
r
π
N2 · =1
2α
r
2 2α
N =
π
Therefore,
∞
1 d2
D   E Z  
−αx2 2
Φ̃  H  Φ̃ = Ne − − δ(x) N e−αx dx
 
2 dx 2
−∞
Z ∞  
2 2 2 2
= N2 e−αx −2α2 x2 e−αx + αe−αx − δ(x)e−αx dx
−∞
Z ∞ 
2 2 2
= N2 −2α2 x2 e−2αx + αe−2αx − δ(x)e−2αx dx
−∞
 r r 
α π π
= N2 · − +α −1
2 2α 2α
r  r 
2α α π
= −1
π 2 2α
r
α 2α
= −
2 π
To minimize variation integral, r
∂E 1 1
= − =0
∂α 2 2πα
2
α=
π
So we get the result
1
E =−
π

Exercise 1.19

Solution:
D  E Z ∞
2
Φ̃  Φ̃ = N 2 e−2αr · r2 dr = 1

0

2 · π1/2
N2 · =1
8 · (2α)1/2
r
2 2α3
N =8
π

12
 D   E Z ∞ 2

1 1

2
E = Φ̃  H  Φ̃ = N e−αr − ∇2 − N e−αr r2 dr
 
0 2 r
Z ∞ 
2 −2αr 2 2 2
=N 2
3αr e − 2α2 r4 e−2αr − re−2αr dr
0
2 · π1/2 24 · π1/2
 
2 2 1
= N 3α · − 2α · −
8 · (2α)3/2 32 · 2 · (2α)5/2 4α
r r
2α3 3
 
π 1
=8 −
π 16 2α 4α
r
3 2α
= α−2
2 π
To minimize variation integral, r
∂E 3 2
= − =0
∂α 2 πα
8
α=
9π
So the result is:
4
E =− = −0.4244
3π

Exercise 1.21

Solution:
a.
D  E XD  E D  E
Φ̃  Φ̃ = Φ̃0  Φα hΦα | Φβ i Φβ  Φ̃0
  
αβ
X D  E D  E
= Φ̃0  Φα Φβ  Φ̃0 δαβ
αβ
X D  E2
=  Φα  Φ̃0  = 1
α
D  E
Because Φ0  Φ̃0 = 0:

∞ D
X  E2
 Φα  Φ̃0  = 1
  
α=1
Therefore:
D   E XD  E D  E
Φ̃  H  Φ̃ = Φ̃0  Φα hΦα | H | Φβ i Φβ  Φ̃0
   
αβ
X D  E D  E
= Φ̃0  Φα Eα · δαβ Φβ  Φ̃0

αβ
∞
X D  E2
= Eα ·  Φα  Φ̃0 
  
α=1

Because α = 1, 2, . . ., Eα ≥ E1 : D   E
Φ̃  H  Φ̃ ≥ E1
 

b.
D  E  ∗  
Φ̃0  Φ̃0 = 1 = x |Φ̃0 i + y |Φ̃1 i x |Φ̃0 i + y |Φ̃1 i

 ∗  
= x∗ hΦ̃0 | + y ∗ hΦ̃1 | x |Φ̃0 i + y |Φ̃1 i
D  E D  E D  E D  E
= |x|2 Φ̃0  Φ̃0 + |y|2 Φ̃1  Φ̃1 + xy Φ̃0  Φ̃1 + Φ̃1  Φ̃0
   

13
Because |Φ̃α i is orthogonal, D  E
Φ̃0  Φ̃0 = x2 + y 2 = 1


c.
D   E  ∗  
Φ̃0  H  Φ̃0 = x |Φ̃0 i + y |Φ̃1 i H x |Φ̃0 i + y |Φ̃1 i
 
D   E D   E
= |x|2 Φ̃0  H  Φ̃0 + |y|2 Φ̃1  H  Φ̃1
   

= |x|2 E0 + 1 − |x|2 E1

= E1 − |x|2 (E1 − E0 )
Because E1 ≥ E0 , D   E
Φ̃0  H  Φ̃ = E1 − |x|2 (E1 − E0 ) ≥ E1
  0

E1 ≥ E 1

Exercise 1.22

Solution:
Firstly, we form the matrix representation of operator H in the basis:
(H)11 = h1s | H | 1si = h1s | H0 | 1si + h1s | F r cos θ | 1si
1
= − + h1s | F r cos θ | 1si
2
(H)22 = h2pz | H | 2pz i = h2pz | H0 | 2pz i + h2pz | F r cos θ | 2pz i
1
= − + h2pz | F r cos θ | 2pz i
8
(H)12 = h1s | H | 2pz i = h1s | H0 | 2pz i + h1s | F r cos θ | 2pz i
= h1s | F r cos θ | 2pz i
(H)21 = h2pz | H | 1si = h2pz | H0 | 1si + h2pz | F r cos θ | 1si
= h2pz | F r cos θ | 1si
(2pz is centrosymmetric, so h1s | H | 2pz i = 0.)
Z 2π Z π Z ∞
h1s | F r cos θ | 1si = π−1/2 e−r · F r cos θ · π−1/2 e−r · r2 sin θ dr dθ dφ
0 0 0
Z 2π Z π Z ∞
F
= sin θ cos θ · r3 e−2r dr dθ dφ
r 0 0 0
3F 2π π
Z Z
= sin θ cos θ dθ dφ
8r 0 0
=0
Z 2π Z π Z ∞
h2pz | F r cos θ | 2pz i = (32π)−1/2 re−r/2 cos θ · F r cos θ · (32π)−1/2 re−r/2 cos θ · r2 sin θ dr dθ dφ
0 0 0
Z 2π Z π Z ∞
F
= cos3 θ sin θ dθ · r5 e−r dr dφ = 0
32π 0 0 0
Z 2π Z π Z ∞
h1s | F r cos θ | 2pz i = π−1/2 e−r · F r cos θ · (32π)−1/2 re−r/2 cos θ · r2 sin θ dr dθ dφ
0 0 0
Z Z ∞ 2π Z π
F
=√ cos2 θ sin θ dθ · r4 e−3/2r dr dφ
32π 0 0 0
Z 2π Z ∞
F 2 4 −3/2r
=√ r e dr dφ
32π 0 0 3
256F
= √
243 2

14
Then we can solve the eigenvalue equation:
Hc = Ec
We have the determinant to be 0:  1 256F 
− − E √ 

 2 243 2  = 0
 256F 1
 √ − − E 
8

243 2
5 1 2562 F 2
E2 + E + − =0
8 6 2 · 2432
We solve the characteristic polynomial and the result is:
s !
2562 F 2

1 5 25 1
E= − ± −4 −
2 8 64 16 2 · 2432
r
5 1 9 2562 F 2
=− ± +
16 2 64 2 · 2432
r
5 3 128 · 2562 F 2
=− ± 1+
16 16 9 · 2432
We should talk about which sign be taken, but I’ll skip it and just show the result:
r
5 3 128 · 2562 F 2
E=− − 1+
16 16 9 · 2432
Using Taylor series expansion:

1 128 · 2562 F 2
 
5 3
E≈− − 1+
16 16 2 9 · 2432
1 1 3 128 · 2562 2
=− − · F
2 2 16 9 · 2432
So α = 2.96.

2 Chapter 2
Exercise 2.1

Solution:
Case 1. m = 2i − 1, n = 2j − 1
Z Z
α∗ ∗
hχm | χn i = dr ψm (r)ψnα (r) dω α (ω)α(ω) = α∗
dr ψm (r)ψnα (r) · 1

= δmn = 0

Case 2. m = 2i, n = 2j
Z Z
β∗ ∗
hχm | χn i = dr ψm (r)ψnβ (r) dω β (ω)β(ω) = β∗
dr ψm (r)ψnβ (r) · 1

= δmn = 0

Case 3. m = n = 2i − 1
Z Z
α∗ ∗
hχm | χn i = dr ψm (r)ψnα (r) dω α (ω)α(ω) = α∗
dr ψm (r)ψnα (r) · 1

= δmn = 1

15
Case 4. m = n = 2i
Z Z
hχm | χn i = β∗
dr ψm (r)ψnβ (r) dω β ∗ (ω)β(ω) = β∗
dr ψm (r)ψnβ (r) · 1

= δmn = 1

Case 5. m = 2i − 1, n = 2j
Z Z
hχm | χn i = α∗
dr ψm (r)ψnβ (r) dω α∗ (ω)β(ω) = α∗
dr ψm (r)ψnβ (r) · 0

=0

So we conclude that
hχm | χn i = δmn

Exercise 2.2

Solution:
N
X
h(i) (χi (x1 )χj (x2 ) . . . χk (xN )) = h(1) (χi (x1 )χj (x2 ) . . . χk (xN )) + h(2) (χi (x1 )χj (x2 ) . . . χk (xN )) + . . .
i=1
+ h(N ) (χi (x1 )χj (x2 ) . . . χk (xN ))
= ε1 (χi (x1 )χj (x2 ) . . . χk (xN )) + ε2 (χi (x1 )χj (x2 ) . . . χk (xN )) + . . .
+ εN (χi (x1 )χj (x2 ) . . . χk (xN ))
= (ε1 + ε2 + . . . + εN ) (χi (x1 )χj (x2 ) . . . χk (xN ))
= EΨHP
Where E = ε1 + ε2 + . . . + εN .

Exercise 2.3

Solution:
Z
1
1
√ χ∗i (x1 )χ∗j (x2 ) − χ∗j (x1 )χ∗i (x2 ) √ χi (x1 )χj (x2 ) − χj (x1 )χi (x2 ) dx1 dx2


hΨ | Ψi =
2 2
Z
1
= χi (x1 )χj (x2 )χi (x1 )χj (x2 ) − χi (x1 )χ∗j (x2 )χj (x1 )χi (x2 )
∗ ∗ ∗
2
− χ∗j (x1 )χ∗i (x2 )χi (x1 )χj (x2 ) − χ∗j (x1 )χ∗i (x2 )χj (x1 )χi (x2 ) dx1 dx2
1
= (1 − 0 − 0 + 1)
2
=1

Exercise 2.4

Solution:

H ΨHP
12 = (h(1) + h(2))χi (x1 )χj (x2 )
= εi χi (x1 )χj (x2 ) + εj χi (x1 )χj (x2 )
= (εi + εj )ΨHP
12

H ΨHP
21 = (h(1) + h(2))χi (x2 )χj (x1 )
= εj χi (x1 )χj (x2 ) + εi χi (x1 )χj (x2 )
= (εi + εj )ΨHP
21

16
 h
i
H Ψ(x1 , x2 ) = (h(1) + h(2)) 2−1/2 χi (x1 )χj (x2 ) − χj (x1 )χi (x2 )
= 2−1/2 h(1)χi (x1 )χj (x2 ) − h(1)χj (x1 )χi (x2 ) + h(2)χi (x1 )χj (x2 ) − h(2)χj (x1 )χi (x2 )

= 2−1/2 εi χi (x1 )χj (x2 ) − εj χj (x1 )χi (x2 ) + εj χi (x1 )χj (x2 ) − εi χj (x1 )χi (x2 )


h
i
= (εi + εj ) 2−1/2 χi (x1 )χj (x2 ) − χj (x1 )χi (x2 )
= (εi + εj )Ψ(x1 , x2 )

Exercise 2.5

Solution:
Z
1
1
√ χ∗i (x1 )χ∗j (x2 ) − χ∗j (x1 )χ∗i (x2 ) √ χk (x1 )χl (x2 ) − χl (x1 )χk (x2 ) dx1 dx2


hK | Li =
2 2
Z
1
= χ∗i (x1 )χ∗j (x2 )χk (x1 )χl (x2 ) − χ∗i (x1 )χ∗j (x2 )χl (x1 )χk (x2 )
2
− χ∗j (x1 )χ∗i (x2 )χk (x1 )χl (x2 ) + χ∗j (x1 )χ∗i (x2 )χl (x1 )χk (x2 ) dx1 dx2

1
= (δik δjl − δil δjk − δjk δil + δjl δik )
2
= δik δjl − δil δjk

Exercise 2.6

Solution:
Z
1
hψ1 | ψ1 i = (φ∗ + φ∗2 )(φ1 + φ2 ) dτ
2 (1 + S12 ) 1
Z
1
= (φ∗ φ1 + φ∗2 φ2 + φ∗1 φ2 + φ∗2 φ1 ) dτ
2 (1 + S12 ) 1
Z
1
= (1 + 1 + S12 + S12 ) dτ
2 (1 + S12 )
=1
Z
1
hψ2 | ψ2 i = (φ∗ − φ∗2 )(φ1 − φ2 ) dτ
2 (1 − S12 ) 1
Z
1
= (φ∗ φ1 + φ∗2 φ2 − φ∗1 φ2 − φ∗2 φ1 ) dτ
2 (1 − S12 ) 1
Z
1
= (1 + 1 − S12 − S12 ) dτ
2 (1 − S12 )
=1
Z
−1/2 −1/2
hψ1 | ψ2 i = [2 (1 + S12 )] [2 (1 − S12 )] (φ∗1 + φ∗2 )(φ1 − φ2 ) dτ
Z

−1/2 ∗
2
(φ1 φ1 − φ∗2 φ2 − φ∗1 φ2 + φ∗2 φ1 ) dτ

= 4 1 − S12
Z
 2

−1/2
= 4 1 − S12 (1 − 1 − S12 + S12 ) dτ

=0

Exercise 2.7

17
Solution:
The system(Benzene) has 42 electrons. So the number of possible determinants is
 
72
  = 72! = 1.643 × 1020
42 42!30!

The number of singly excited determinants is 42 × 30 = 1260. And there are ( 42 30

2 ) × ( 2 ) = 374535 doubly
excited determinants.
Exercise 2.8

Solution:
Z h i∗
Ψ34 dx1 dx2 2−1/2 (χ3 (x1 )χ4 (x2 ) − χ4 (x1 )χ3 (x2 ))

  34 
12
 h1  Ψ12 =
h i
× h(r1 ) 2−1/2 (χ3 (x1 )χ4 (x2 ) − χ4 (x1 )χ3 (x2 ))
Z
1
dx1 dx2 χ3 (x1 )∗ χ4 (x2 )∗ h(r1 )χ3 (x1 )χ4 (x2 ) + χ4 (x1 )∗ χ3 (x2 )∗ h(r1 )χ4 (x1 )χ3 (x2 )

=
2
− χ4 (x1 )∗ χ3 (x2 )∗ h(r1 )χ3 (x1 )χ4 (x2 ) − χ3 (x1 )∗ χ4 (x2 )∗ h(r1 )χ4 (x1 )χ3 (x2 )

Z
1
dx1 χ3 (x1 )∗ h(r1 )χ3 (x1 ) + χ4 (x1 )∗ h(r1 )χ4 (x1 )
 
=
2
1 1
= h3 | h(1) | 3i + h4 | h(1) | 4i
2 2

  34 
34   34 
By exactly the same procedure, one finds that Ψ34 12 h1 Ψ12 = Ψ12 h2 Ψ12 and thus
   

Ψ12  O1  Ψ34

34   
12 = h3 | h | 3i + h4 | h | 4i

Z h i∗
Ψ21  O1  Ψ34 dx1 dx2 2−1/2 (χ1 (x1 )χ2 (x2 ) − χ2 (x1 )χ1 (x2 ))

  
12 =
h i
× O1 2−1/2 (χ3 (x1 )χ4 (x2 ) − χ4 (x1 )χ3 (x2 ))
Z
1
dx1 dx2 χ1 (x1 )∗ χ2 (x2 )∗ O1 χ3 (x1 )χ4 (x2 ) + χ2 (x1 )∗ χ1 (x2 )∗ O1 χ4 (x1 )χ3 (x2 )

=
2
− χ1 (x1 )∗ χ2 (x2 )∗ O1 χ4 (x1 )χ3 (x2 ) − χ2 (x1 )∗ χ1 (x2 )∗ O1 χ3 (x1 )χ4 (x2 )


=0

12 O1 Ψ1 = 0
Ψ34

  2
 

Exercise 2.8

Solution:

hΨ0 | H | Ψ0 i = hΨ0 | O1 + O2 | Ψ0 i
= h1 | h | 1i + h2 | h | 2i + h12 | 12i − h12 | 21i
Ψ0  H  Ψ34
12 = Ψ0 O1 + O2 Ψ12

  
  34 
 
= h12 | 34i − h12 | 43i

12 H Ψ0 = Ψ12 O1 + O2 Ψ0
Ψ34

  
34   
   
= h34 | 12i − h34 | 21i
Ψ12  H  Ψ12 = Ψ12  O1 + O2  Ψ34

34   34 
34   
12
= h3 | h | 3i + h4 | h | 4i + h34 | 34i − h34 | 43i

18
Therefore:
 
h1 | h | 1i + h2 | h | 2i + h12 | 12i − h12 | 21i h12 | 34i − h12 | 43i
H = 
h34 | 12i − h34 | 21i h3 | h | 3i + h4 | h | 4i + h34 | 34i − h34 | 43i

Exercise 2.13

Solution:
Case 1. a 6= b, r 6= s:
|Ψra i = |χ1 . . . χr χb . . . χN i
|Ψsb i = |χ1 . . . χa χs . . . χN i
There are no two columns correspondingly to be equal. So

hΨra | O1 | Ψsb i = 0

Case 2. a = b, r 6= s:
|Ψra i = |χ1 . . . χr . . . χN i
|Ψsb i = |χ1 . . . χs . . . χN i
hΨra | O1 | Ψsb i = hr | h | si
Case 3. a 6= b, r = s:
|Ψra i = |χ1 . . . χr χb . . . χN i
|Ψsb i = |χ1 . . . χa χr . . . χN i = − |χ1 . . . χr χa . . . χN i
hΨra | O1 | Ψsb i = − hb | h | ai
Case 4. a = b, r = s:
|Ψra i = |χ1 . . . χr χb . . . χN i
|Ψsb i = |χ1 . . . χr χb . . . χN i
N
X
hΨra | O1 | Ψsb i = hc | h | ci − ha | h | ai + hr | h | ri
c

Exercise 2.14

Solution:

|N Ψ0 i = |χ1 . . . χa χb . . . χN i
|N −1 Ψa i = |χ1 . . . χa−1 χa+1 . . . χN i
So we have:
N
Ψ0  H  N Ψ0 = N Ψ0  O1 + O2  N Ψ0

N   
  
E0 =
N
X N X
X N
= hχm | h | χm i + hχm χn k χm χn i
m m n>m

N −1

N −1
Ψa  H  N −1 Ψa = N −1 Ψa  O1 + O2  N −1 Ψa
  
  
Ea =
N
X N
X N
X
= hχx | h | χx i + hχx χy k χx χy i
x(x6=a) x(x6=a) y>x(y6=a)

Therefore
a−1
X N
X
N
E 0 − N −1 E a = hχa | h | χa i + hχm χa k χm χa i + hχa χn k χa χn i
m=1 n=a+1

19
Because hχa χn k χa χn i = hχn χa k χn χa i and hχa χa k χa χa i = 0.
N
X
N N −1
E0 − E a = hχa | h | χa i + hχm χa k χm χa i
m=1

Exercise 2.15

Solution:
N!
1 X
H |Ψi = √ (−1)pi H Pi {χi (1)χj (2) . . . χk (N )}
N! i
Because H and P commute with each other.
N!
1 X
H |Ψi = √ (−1)pi Pi H {χi (1)χj (2) . . . χk (N )}
N! i

Since
N
X
H {χi (1)χj (2) . . . χk (N )} = h(i) {χi (1)χj (2) . . . χk (N )}
i
N
X
= {χi (1)χj (2) . . . h(i)χs (i) . . . χk (N )}
i
N
X
= εs (i) {χi (1)χj (2) . . . χk (N )}
i
= (ε1 + ε2 + . . . + εN ) {χi (1)χj (2) . . . χk (N )}
N!
1 X
H |Ψi = √ (−1)pi Pi H {χi (1)χj (2) . . . χk (N )}
N! i
N!
1 X
=√ (−1)pi Pi (ε1 + ε2 + . . . + εN ) {χi (1)χj (2) . . . χk (N )}
N! i
N!
ε1 + ε2 + . . . + εN X
= √ (−1)pi Pi {χi (1)χj (2) . . . χk (N )}
N! i
= (ε1 + ε2 + . . . + εN ) |Ψi

Exercise 2.17

Solution:

Ψ12  H  Ψ34

34   
12 = h3 | h | 3i + h4 | h | 4i + h34 | 34i − h34 | 43i
Z
h3 | h | 3i = dr1 dω1 ψ2∗ (r1 )α∗ (ω1 )hψ2 (r1 )α(ω1 )
Z
= dr1 ψ2∗ (r1 )hψ2 (r1 )

= (2|h|2)
Z
h4 | h | 4i = dr2 dω2 ψ2∗ (r2 )β ∗ (ω2 )hψ2 (r2 )β(ω2 )
Z
= dr2 ψ2∗ (r2 )hψ2 (r2 )

= (2|h|2)

20
 Z
h34 | 34i = dr1 dω1 dr2 dω2 ψ2∗ (r1 )α∗ (ω1 )ψ2∗ (r2 )β ∗ (ω2 )
−1
× r12 ψ2 (r1 )α(ω1 )ψ2 (r2 )β(ω2 )
Z
−1
= dr1 dr2 ψ2∗ (r1 )ψ2∗ (r2 )r12 ψ2 (r1 )ψ2 (r2 )

= (22|22)
Z
h34 | 43i = dr1 dω1 dr2 dω2 ψ2∗ (r1 )α∗ (ω1 )ψ2∗ (r2 )β ∗ (ω2 )
−1
× r12 ψ2 (r1 )β(ω1 )ψ2 (r2 )α(ω2 )
=0
Therefore
12 H Ψ12 = 2(2|h|2) + (22|22)
Ψ34

  34 
 

Ψ0  H  Ψ34

  
12 = h12 | 34i
Z
= dr1 dω1 dr2 dω2 ψ1∗ (r1 )α∗ (ω1 )ψ1∗ (r2 )β ∗ (ω2 )
−1
× r12 ψ2 (r1 )α(ω1 )ψ2 (r2 )β(ω2 )
Z
−1
= dr1 dr2 ψ1∗ (r1 )ψ1∗ (r2 )r12 ψ2 (r1 )ψ2 (r2 )

= (12|12)

12 H Ψ0 = h34 | 12i
Ψ0 Ψ34

  
 
Z
= dr1 dω1 dr2 dω2 ψ2∗ (r1 )α∗ (ω1 )ψ2∗ (r2 )β ∗ (ω2 )
−1
× r12 ψ1 (r1 )α(ω1 )ψ1 (r2 )β(ω2 )
Z
−1
= dr1 dr2 ψ2∗ (r1 )ψ2∗ (r2 )r12 ψ1 (r1 )ψ1 (r2 )

= (21|21)

Exercise 2.18

Solution:

 hab k rsi 2 =  hab | rsi − hab | sri 2

X  X 

abrs abrs
X

= hab | rsi hrs | abi + hab | sri hsr | abi − hab | rsi hsr | abi − hab | sri hrs | abi
abrs

21
We let I1 , I2 , I3 , I4 be equal to each part.
X
I1 = hab | rsi hrs | abi
abrs
N X
X N 2K
X 2K
X
 
 
= hab | rsi hrs | abi + hāb | rsi hrs | ābi + ab̄  rs rs  ab̄
a=1 b=1 r=N +1 s=N +1

 
 
+ hab | r̄si hr̄s | abi + hab | rs̄i hrs̄ | abi + āb̄  rs rs  āb̄

 
 
+ hāb | r̄si hr̄s | ābi + hāb | rs̄i hrs̄ | ābi + ab̄  r̄s r̄s  ab̄

 
 
 
 
+ ab̄  rs̄ rs̄  ab̄ + hab | r̄s̄i hr̄s̄ | abi + āb̄  r̄s r̄s  āb̄

 
 
 
 
+ āb̄  rs̄ rs̄  āb̄ + hāb | r̄s̄i hr̄s̄ | ābi + ab̄  r̄s̄ r̄s̄  ab̄

 
 

+ āb̄  r̄s̄ r̄s̄  āb̄
N X
X N 2K
X 2K
X
 
 
 
 

= hab | rsi hrs | abi + hāb | r̄si hr̄s | ābi + ab̄  rs̄ rs̄  ab̄ + āb̄  r̄s̄ r̄s̄  āb̄
a=1 b=1 r=N +1 s=N +1
N/2 N/2 K K
XX X X
=4 hab | rsi hrs | abi
a=1 b=1 r=N/2+1 s=N/2+1

Similarly, we can get the second term after cancelling the 0 term from the summation:
N/2 N/2 K K
XX X X
I2 = 4 hab | sri hsr | abi
a=1 b=1 r=N/2+1 s=N/2+1

By interchange the spartial orbitals s, r, I1 and I2 are found to be equal.

I1 = I2

The I3 part has some differences with either I1 or I2 :

N X
X N 2K
X 2K
X
 
 

I3 = hab | rsi hsr | abi + āb̄  r̄s̄ s̄r̄  āb̄
a=1 b=1 r=N +1 s=N +1
N/2 N/2 K K
XX X X
=2 hab | rsi hsr | abi
a=1 b=1 r=N/2+1 s=N/2+1

And I4 is the same:

N X
X N 2K
X 2K
X
 
 

I4 = hab | sri hrs | abi + āb̄  s̄r̄ r̄s̄  āb̄
a=1 b=1 r=N +1 s=N +1
N/2 N/2 K K
XX X X
=2 hab | sri hrs | abi
a=1 b=1 r=N/2+1 s=N/2+1

I3 = I4
Because εi = εī , the denominators in each term are all equal to εa + εb − εr − εs . Thus
N/2 N/2 K K
(2) 1 XX X X 8 hab | rsi hrs | abi + 4 hab | rsi hsr | abi
E0 =
4 a=1 εa + εb − εr − εs
b=1 r=N/2+1 s=N/2+1
N/2 N/2 K K

XX X X hab | rsi 2 hrs | abi + hsr | abi
=
a=1 b=1 r=N/2+1 s=N/2+1
εa + εb − εr − εs

22
Exercise 2.19

Solution:
Z
−1 ∗
Jii = (ii|ii) = dr1 dr2 ψi∗ (r1 )ψi (r1 )r12 ψi (r2 )ψi (r2 )
Z
−1 ∗
Kii = (ii|ii) = dr1 dr2 ψi∗ (r1 )ψi (r1 )r12 ψi (r2 )ψi (r2 )

It is obviously that
Jii = Kii
Z
−1 ∗
Jij = (ii|jj) = dr1 dr2 ψi∗ (r1 )ψi (r1 )r12 ψj (r2 )ψj (r2 )
Z
∗ −1
Jij = dr1 dr2 ψi (r1 )ψi∗ (r1 )r12 ψj (r2 )ψj∗ (r2 )
∗
Jij = Jij
Z
−1 ∗
Kij = (ij|ji) = dr1 dr2 ψi∗ (r1 )ψj (r1 )r12 ψj (r2 )ψi (r2 )
Z
∗ −1
Kij = dr1 dr2 ψi (r1 )ψj∗ (r1 )r12 ψj (r2 )ψi∗ (r2 )

If we exchange the label of electron 1 and electron 2, we find that

∗
Kij = Kij
Z
−1 ∗
Jji = (jj|ii) = dr1 dr2 ψj∗ (r1 )ψj (r1 )r12 ψi (r2 )ψi (r2 )
Z
−1 ∗
Kji = (ji|ij) = dr1 dr2 ψj∗ (r1 )ψi (r1 )r12 ψi (r2 )ψj (r2 )

Thus
Jij = Jji
Kij = Kji

Exercise 2.19

Solution:

Exercise 2.21

Solution:
 
2(1|h|1) + (11|11) (12|12)
H= 
(21|21) 2(2|h|2) + (22|22)
Because the spatial molecular orbitals are real, thus

(21|21) = (12|12) = K12

Therefore  
2h11 + J11 K12
H= 
K12 2h22 + J22

Exercise 2.22

23
Solution:
Z
 H  ΨHP −1
ΨHP dr1 dr2 ψ1∗ (r1 )ψ2∗ (r2 )(h1 + h2 + r12

  
↑↓ ↑↓ = )ψ1 (r1 )ψ2 (r2 )
Z
× dω1 dω2 α∗ (ω1 )β ∗ (ω2 )α(ω1 )β(ω2 )

= h11 + h22 + J12

With exactly the same procedure, the result of parallel spin is
Z
−1
Ψ↓↓ H Ψ↑↓ = dr1 dr2 ψ1∗ (r1 )ψ2∗ (r2 )(h1 + h2 + r12

HP   HP 
  )ψ1 (r1 )ψ2 (r2 )
Z
× dω1 dω2 β ∗ (ω1 )β ∗ (ω2 )β(ω1 )β(ω2 )

= h11 + h22 + J12

Exercise 2.24

Solution:

(a†1 a†2 + a†2 a†1 ) |χ1 χ2 i = a†1 a†2 |χ1 χ2 i + a†2 a†1 |χ1 χ2 i
=0+0=0

(a†1 a†2 + a†2 a†1 ) |χ1 χ3 i = a†1 a†2 |χ1 χ3 i + a†2 a†1 |χ1 χ3 i
=0+0=0

(a†1 a†2 + a†2 a†1 ) |χ1 χ4 i = a†1 a†2 |χ1 χ4 i + a†2 a†1 |χ1 χ4 i
=0+0=0

(a†1 a†2 + a†2 a†1 ) |χ2 χ3 i = a†1 a†2 |χ2 χ3 i + a†2 a†1 |χ2 χ3 i
=0+0=0

(a†1 a†2 + a†2 a†1 ) |χ1 χ2 i = a†1 a†2 |χ2 χ4 i + a†2 a†1 |χ2 χ4 i
=0+0=0

(a†1 a†2 + a†2 a†1 ) |χ1 χ2 i = a†1 a†2 |χ3 χ4 i + a†2 a†1 |χ3 χ4 i
= |χ1 χ2 χ3 χ4 i + |χ2 χ1 χ3 χ4 i
= |χ1 χ2 χ3 χ4 i − |χ1 χ2 χ3 χ4 i = 0

Exercise 2.25

Solution:

(a1 a†2 + a†2 a1 ) |χ1 χ2 i = a1 a†2 |χ1 χ2 i + a†2 a1 |χ1 χ2 i

= a†2 |χ2 i
=0

(a1 a†1 + a†1 a1 ) |χ1 χ2 i = a1 a†1 |χ1 χ2 i + a†1 a1 |χ1 χ2 i

= a†1 |χ2 i
= |χ1 χ2 i

24
Exercise 2.26

Solution:
D E D E
† †
 ai aj  =  δij − aj ai 
  

= δij h | i
= δij

Exercise 2.27

Solution:
D   E
χ1 χ2 . . . χN  a†i aj  χ1 χ2 . . . χN
 

i and j must be in {1, 2, . . . , N }, otherwise the integral is zero.

D   E D   E
χ1 χ2 . . . χN  a†i aj  χ1 χ2 . . . χN = χ1 χ2 . . . χN  δij − aj a†i  χ1 χ2 . . . χN
   
D   E
= δij hχ1 χ2 . . . χN | χ1 χ2 . . . χN i − χ1 χ2 . . . χN  aj a†i  χ1 χ2 . . . χN
 

= δij

Exercise 2.28

Solution:
(a)
χr is a virtual orbital, which is not involved in HF wave function.

ar |Ψ0 i = 0 = hΨ0 | a†r

(b)
χa is already in HF wave function.
a†a |Ψ0 i = 0 = hΨ0 | aa
(c)

a†r aa |χ1 . . . χa . . . χN i = −a†r aa |χa . . . χ1 . . . χN i

= −a†r |. . . χ1 . . . χN i
= − |χr . . . χ1 . . . χN i
= |χ1 . . . χr . . . χN i

(d)

†
†
|Ψ0 i = a†r aa |Ψ0 i
hΨ0 | = hΨ0 | a†a ar

Exercise 2.29

25
Solution:
D E
hi | h | ji  a2 a1 a†i aj a†1 a†2 
X
hΨ0 | O1 | Ψ0 i =
 
ij
D E
hi | h | ji  a2 a1 (δij − aj a†i )a†1 a†2 
X
=
 
ij
 D E D E
† † † † †
X
= hi | h | ji δij  a2 a1 a1 a2  −  a2 a1 aj ai a1 a2 
 
ij
 D   E
†
X
= hi | h | ji δij hΨ0 | Ψ0 i − Ψ0  aj ai  Ψ0

ij
Because i and j fall within 1 and 2. The second term has creation operators acting on the existing spin orbitals
and is zero as a result. X
hΨ0 | O1 | Ψ0 i = hi | h | ji δij = h1 | h | 1i + h2 | h | 2i
ij

Exercise 2.30

Solution:
D   E
hi | h | ji Ψ0  a†a ar a†i aj  Ψ0
X
hΨra | O1 | Ψ0 i =
 
ij
  E
X
 †   D  † †
= hi | h | ji δir Ψ0 aa aj Ψ0 − Ψ0  aa a ar aj  Ψ0
  
i
ij
  E
X
   D  † †
† 
= hi | h | ji δir Ψ0 δaj − aj aa Ψ0 − Ψ0  aa ai ar aj  Ψ0
 
ij

χr is not in |Ψ0 i. Creation operator a†r acting on it makes a result of zero.

X  
hΨra | O1 | Ψ0 i = hi | h | ji δir Ψ0  δaj − aj a†a  Ψ0

  

ij
X  
hi | h | ji δir δaj hΨ0 | Ψ0 i − δir Ψ0  aj a†a  Ψ0

  
=
ij
X
= hi | h | ji δir δaj
ij

= hr | h | ai

Exercise 2.31

Solution:
1X D   E
hΨra | O2 | Ψ0 i = hij | kli Ψ0  a†a ar a†i a†j al ak  Ψ0
 
2
ijkl
 D   E
1X  E D 
= hij | kli δir Ψ0  a†a a†j al ak  Ψ0 − Ψ0  a†a a†i ar a†j al ak  Ψ0
   
2
ijkl
D   E D   E
δir Ψ0  a†a a†j al ak  Ψ0 = −δir Ψ0  a†j a†a al ak  Ψ0
   
D   E D   E
= −δir δal Ψ0  a†j ak  Ψ0 + δir Ψ0  a†j al a†a ak  Ψ0
   
D   E D   E D   E
= −δir δal Ψ0  a†j ak  Ψ0 + δir δak Ψ0  a†j al  Ψ0 − δir Ψ0  a†j al ak a†a  Ψ0
     
D   E D   E
= −δir δal Ψ0  a†j ak  Ψ0 + δir δak Ψ0  a†j al  Ψ0
   

26
 D   E D   E D   E
− Ψ0  a†a a†i ar a†j al ak  Ψ0 = −δrj Ψ0  a†a a†i al ak  Ψ0 + Ψ0  a†a a†i a†j ar al ak  Ψ0
     
D   E
= δrj Ψ0  a†i a†a al ak  Ψ0
 
D   E D   E
= δrj δal Ψ0  a†i ak  Ψ0 − δrj Ψ0  a†i al a†a ak  Ψ0
   
D   E D   E D   E
= δrj δal Ψ0  a†i ak  Ψ0 − δrj δak Ψ0  a†i al  Ψ0 + δrj Ψ0  a†i al ak a†a  Ψ0
     
D   E D   E
= δrj δal Ψ0  a†i ak  Ψ0 − δrj δak Ψ0  a†i al  Ψ0
   


1X D   E D   E
hΨra | O2 | Ψ0 i = hij | kli δrj δal Ψ0  a†i ak  Ψ0 − δrj δak Ψ0  a†i al  Ψ0
   
2
ijkl
D   E D   E
− δir δal Ψ0  a†j ak  Ψ0 + δir δak Ψ0  a†j al  Ψ0
   
 
1 X X X X
= hir | iai − hir | aii − hrj | jai + hrj | aji
2 i i j j
X 
1 X X X
= hri | aii − hri | iai − hri | iai + hri | aii
2 i i i i
X

= hrb | abi − hrb | bai
b
X
= hrb k abi
b

Exercise 2.32

Solution:
a)

s+ |αi = (sx + isy ) |αi

1 i
= |βi + i · |βi
2 2
=0

s+ |βi = (sx + isy ) |βi

1 i
= |αi + i · − |αi
2 2
= |αi
s− |αi = (sx − isy ) |αi
1 i
= |βi − i · |βi
2 2
= |βi
s− |βi = (sx − isy ) |βi
1 i
= |αi − i · − |αi
2 2
=0
b)
s2 = s2x + s2y + s2z

27
 s+ s− = (sx + isy )(sx − isy ) = s2x + s2y − i(sx sy − sy sx )
= s2x + s2y − i[sx , sy ]
= s2x + s2y + sz
s− s+ = (sx − isy )(sx + isy ) = s2x + s2y + i(sx sy − sy sx )
= s2x + s2y + i[sx , sy ]
= s2x + s2y − sz
Therefore
s2 = s+ s− − sz + s2z
s2 = s− s+ + sz + s2z

Exercise 2.33

Solution:

  
    
α  s2  α α  s2  β 3/4 0
s2 = 
  
   = 
  
β s α
 2
β s β
 2 0 3/4
   
hα | sz | αi hα | sz | βi 1/2 0
sz =  =
   
hβ | sz | αi hβ | sz | βi 0 −1/2
   
hα | s+ | αi hα | s+ | βi
 0 1
s+ =  =

hβ | s+ | αi hβ | s+ | βi 0 0
   
hα | s− | αi hα | s− | βi
 0 0
s− =  =

hβ | s− | αi hβ | s− | βi 1 0

And  
1 0
s+ s− =  
0 0
 
0 0
s− s+ =  
0 1
It is obvious that
s2 = s+ s− − sz + s2z
s2 = s− s+ + sz + s2z

Exercise 2.34

Solution:
 2 
s , sz = s2 sz − sz s2
= s2x + s2y + s2z sz − sz s2x + s2y + s2z

= s2x sz + s2y sz − sz s2x − sz s2y

Because
sx sy − sy sx = isz
sy sz − sz sy = isx

28
 sz sx − sx sz = isy
Therefore
s2 , sz = sx (sz sx − isy ) + sy (sz sy + isx ) − (sx sz + isy )sx − (sy sz − isx )sy
 

= sx sz sx − isx sy + sy sz sy + isy sx − sx sz sx − isy sx − sy sz sy + isx sy

=0

Exercise 2.35

Solution:
Because operator A commutes with H ,

H (A |Φi) = H A |Φi = A H |Φi = A E |Φi = E(A |Φi)

Thus A |Φi is the eigenfunction of Hamiltonian operator with eigenvalue E.

Both |Φi and A |Φi have the same eigenvalue E. If |Φi is nondegenerate, |Φi and A |Φi must describe the
same state(based on superposition principle). Therefore

A |Φi = a |Φi

where a is a constant.
Eigenfunctions |Φi i of Hermitian operator H form a complete basis set. Suppose eigenfunction |ψi with
eigenvalue k can be expanded as X
|ψi = ci |Φi i
i

The coefficient ci can be determined as following:

X
hΦi | ψi = cj hΦi | Φj i = ci
j

Because |ψi and |Φi i are eigenfunctions of Hermitian operator H , |φi and |Ψi i will be orthognal if they have
different eigenvalues and the corresponding coeffcient will vanish.
n
X
|ψi = ci |Φi i
i=1

where |Φi i are n degeneracy and all have the eigenvalue k.

With the same process, we know that A |ψi is also the eigenfunction of H with eigenvalue k. We can conclude
that
A |ψi = a |ψi
And therefore
n
X
A |ψi = a ci |Φi i
i=1

Exercise 2.36

Solution:
A is a Hermitian operator, so eigenfunctions |Ψ1 i and |Ψ2 i with different eigenvalues are orthognal. In
addition, we have already know |Ψ1 i and |Ψ2 i are eigenfunctions of H as well:

H |Ψ1 i = k1 |Ψ1 i

H |Ψ2 i = k2 |Ψ2 i
And
hΨ1 | H | Ψ2 i = k2 hΨ1 | Ψ2 i = 0

29
If A is S 2 , while |Ψ1 i and |Ψ2 i being singlet and triplet spin-adapted configurations wavefinctions:

S 2 |Ψ1 i = 0 |Ψ1 i

S 2 |Ψ2 i = 2 |Ψ2 i
Thus we can affirm that the element of the Hamiltonian between singlet and triplet spin-adapted configurations
is zero.
Exercise 2.37

Solution:
N!
1 X
Sz |χi χj . . . χk i = Sz √ (−1)pi Pi {χi χj . . . χk }
N! i
N!
1 X
=√ (−1)pi Sz Pi {χi χj . . . χk }
N! i
Because Sz is invariant to permutation and therefore commutes with Pi :
N!
1 X
Sz |χi χj . . . χk i = √ (−1)pi Pi Sz {χi χj . . . χk }
N! i

1
Sz {χi χj . . . χk } = Ms {χi χj . . . χk } =
(Nα − Nβ ){χi χj . . . χk }
2
No matter how we permute the labels of spin orbtals, the total spin’s z component will never change.
N!
1 1 X
Sz |χi χj . . . χk i = (Nα − Nβ ) √ (−1)pi Pi {χi χj . . . χk }
2 N! i
1
= (Nα − Nβ ) |χi χj . . . χk i
2

Exercise 2.38

Solution:

Sz |ψi ψ̄i ψj ψ̄j . . .i = 0

And therefore
Sz2 |ψi ψ̄i ψj ψ̄j . . .i = 0
As for operator S+ :
N!
1 X
S+ |ψi ψ̄i ψj ψ̄j . . .i = √ (−1)pi S+ Pi {ψi ψ̄i ψj ψ̄j . . .}
N! i
N!
1 X
=√ (−1)pi Pi S+ {ψi ψ̄i ψj ψ̄j . . .}
N! i

Because S+ commutes with Pi .

2K
X
S+ {ψi ψ̄i ψj ψ̄j . . .} = s+ (j){ψi ψ̄i ψj ψ̄j . . .}
j

= {ψi ψi ψj ψ̄j . . .} + {ψi ψ̄i ψj ψj . . .} + . . .

K
X
= {ψi ψi ψj ψ̄j . . .}
j

30
 N! K
1 X X
S+ |ψi ψ̄i ψj ψ̄j . . .i = √ (−1)pi Pi {ψi ψi ψj ψ̄j . . .}
N! i j
K N!
X 1 X
= √ (−1)pi Pi {ψi ψi ψj ψ̄j . . .}
j
N! i
=0
Because there are same columns in the determinant.

S− S+ |ψi ψ̄i ψj ψ̄j . . .i

In the end:
S 2 |ψi ψ̄i ψj ψ̄j . . .i = 0

Exercise 2.39

Solution:
a)

S 2 (α(1)β(2) − β(1)α(2)) = (S− S+ + Sz + Sz2 ) (α(1)β(2) − β(1)α(2))

= S− S+ (α(1)β(2)) − S− S+ (β(1)α(2)) + Sz (α(1)β(2))
− Sz (β(1)α(2)) + Sz2 (α(1)β(2)) − Sz2 (β(1)α(2))

2
X
S+ (α(1)β(2)) = s+ (i)(α(1)β(2))
i=1
= α(1)α(2)
2
X
S+ (β(1)α(2)) = s+ (i)(β(1)α(2))
i=1
= α(1)α(2)
Therefore the sum of first two terms diminishes.

Sz (α(1)β(2)) = Sz (β(1)α(2)) = 0

And
Sz2 (α(1)β(2)) = Sz2 (β(1)α(2)) = 0
Thus
S 2 (α(1)β(2) − β(1)α(2)) = 0
So |1 Ψ21 i is a singlet.

b)

S 2 (α(1)β(2) + β(1)α(2)) = (S− S+ + Sz + Sz2 ) (α(1)β(2) + β(1)α(2))

= S− S+ (α(1)β(2)) + S− S+ (β(1)α(2)) + Sz (α(1)β(2))
+ Sz (β(1)α(2)) + Sz2 (α(1)β(2)) + Sz2 (β(1)α(2))

S− S+ (α(1)β(2)) = β(1)α(2) + α(1)β(2)

S− S+ (β(1)α(2)) = β(1)α(2) + α(1)β(2)
We have the result that
S 2 (α(1)β(2) + β(1)α(2)) = 2 (α(1)β(2) + β(1)α(2))
Therefore |3 Ψ21 i is a triplet.

31
Exercise 2.40

Solution:

Ψ21  H  1 Ψ21 = 1 Ψ21  O1  1 Ψ21 + 1 Ψ21  O2  1 Ψ21

1
  
  
  

Ψ1  O1  1 Ψ21 = 1 Ψ21  h1  1 Ψ21 + 1 Ψ21  h2  1 Ψ21

1 2   
  
  
Z  ∗

1 2   1 2  1
Ψ1 h1 Ψ1 =
  dr1 dr2 dω1 dω2 (ψ1 (1)ψ2 (2) + ψ1 (2)ψ2 (1)) (α(1)β(2) − β(1)α(2)) h1
2
 
1
(ψ1 (1)ψ2 (2) + ψ1 (2)ψ2 (1)) (α(1)β(2) − β(1)α(2))
2
Z
1 n
= dr1 dr2 dω1 dω2 (α∗ (1)β ∗ (2) − β ∗ (1)α∗ (2)) (α(1)β(2) − β(1)α(2)) ψ1∗ (1)ψ2∗ (2)h1 ψ1 (1)ψ2 (2)
4
o
+ ψ1∗ (1)ψ2∗ (2)h1 ψ1 (2)ψ2 (1) + ψ1∗ (2)ψ2∗ (1)h1 ψ1 (1)ψ2 (2) + ψ1∗ (2)ψ2∗ (1)h1 ψ1 (2)ψ2 (1)
1
(h11 + h22 )
=
2
With the same procedure

1  1
Ψ21  h2  1 Ψ21 = (h11 + h22 )
 
2
So
Ψ21  O1  1 Ψ21 = h11 + h22

1   
 −1  1
Ψ21  O2  1 Ψ21 = 1 Ψ21  r12

1   
2

 Ψ1
Z
1 h
−1 −1
= dr1 dr2 ψ1∗ (1)ψ2∗ (2)r12 ψ1 (1)ψ2 (2) + ψ1∗ (1)ψ2∗ (2)r12 ψ1 (2)ψ2 (1)
2
i
−1 −1
+ ψ1∗ (2)ψ2∗ (1)r12 ψ1 (1)ψ2 (2) + ψ1∗ (2)ψ2∗ (1)r12 ψ1 (2)ψ2 (1)
1
= (J12 + K12 + K12 + J12 )
2
= J12 + K12
Finally the result is
Ψ21  H  1 Ψ21 = h11 + h22 + J12 + K12

1   

Exercise 2.41

Solution:
a)
S 2 = S− S+ + Sz + Sz2
¯
Sz |ψ1α ψ1β i = 0
¯
Sz2 |ψ1α ψ1β i = 0
¯ ¯
S 2 |ψ1α ψ1β i = S− S+ |ψ1α ψ1β i
1  ¯ ¯

= S− S+ √ ψ1α (1)ψ1β (2) − ψ1α (2)ψ1β (1)
2
1  α 
= √ ψ1 (1)ψ1β (2)S− S+ α(1)β(2) − ψ1α (2)ψ1β (1)S− S+ α(2)β(1)
2
1  α 
= √ ψ1 (1)ψ1β (2)S− α(1)α(2) − ψ1α (2)ψ1β (1)S− α(2)α(1)
2
1  α 
= √ ψ1 (1)ψ1β (2)(β(1)α(2) + α(1)β(2)) − ψ1α (2)ψ1β (1)(α(2)β(1) + β(2)α(1))
2
1  α 
= √ ψ1 (1)ψ1β (2) − ψ1α (2)ψ1β (1) (α(2)β(1) + β(2)α(1))
2

32
 ¯
If ψ1α = ψ1β = ψ1 , two terms is equal and the last formula diminishes. Therefore |ψ1α ψ1β i is the eigenfunction
¯
of S 2 and is pure singlet. Otherwise, |ψ1α ψ1β i can’t be the eigenfunction of S 2 .
b)

¯ 1  
S 2 |ψ1α ψ1β i = √ ψ1α (1)ψ1β (2) − ψ1α (2)ψ1β (1) (α(2)β(1) + β(2)α(1))
2
1  α 
= √ ψ1 (1)α(1)ψ1β (2)β(2) − ψ1α (2)α(2)ψ1β (1)β(1)
2
1  
− √ ψ1β (1)α(1)ψ1α (2)β(2) − ψ1α (1)β(1)ψ1β (2)α(2)
2
=K −J

Therefore
K  S 2  K = hK | K − Ji = hK | Ki − hK | Ji

  

|Ki is normalized, so hK | Ki = 1. And

1
Z  ∗
hK | Ji = dr1 dr2 dω1 dω2 ψ1α (1)α(1)ψ1β (2)β(2) − ψ1α (2)α(2)ψ1β (1)β(1)
2
 
ψ1β (1)α(1)ψ1α (2)β(2) − ψ1α (1)β(1)ψ1β (2)α(2)
Z
1 ∗
= dr1 dr2 dω1 dω2 ψ1α ∗ (1)ψ1β (2)ψ1β (1)ψ1α (2)α∗ (1)β ∗ (2)α(1)β(2)
2
∗
+ ψ1α ∗ (2)ψ1β (1)ψ1α (1)ψ1β (2)α∗ (2)β ∗ (1)β(1)α(2)
 αβ 2
 
= S11 

In conclusion,
 αβ 2
 
K  S  K = 1 − S11

 2 


3 Chapter 3
Exercise 3.1

Solution:

hχi | f | χj i = hχi | h | χj i + χi  v HF  χj

  
Z

 HF   X −1
χi v
  χj = dx1 dx2 χ∗i (1)χ∗b (2)r12 (1 − P12 ){χb (2)χj (1)}
b
XZ
−1
= dx1 dx2 χ∗i (1)χ∗b (2)r12 χb (2)χj (1)
b
Z 
−1
− dx1 dx2 χ∗i (1)χ∗b (2)r12 χj (2)χb (1)
X
= ([ij|bb] − [ib|bj])
b
X
= hib k jbi
b

Therefore X
hχi | f | χj i = hi | h | ji + hib k jbi
b

Exercise 3.2

33
Solution:
XX
L [{χa }] = E0 [{χa }] − εba ([a|b] − δab )
a b
XX
L ∗ [{χa }] = E0∗ [{χa }] − ε∗ba ([a|b]∗ − δab
∗
)
a b

Because L and E0 are real, and [a|b]∗ = [b|a], δab

∗
= δba .
XX
L [{χa }] = E0 [{χa }] − ε∗ba ([b|a] − δba )
a b

Because a, b are dummy variables, we can just exchange them.

XX
L [{χb }] = E0 [{χb }] − εab ([b|a] − δba )
a b

By comparing the last two equations, we conclude that ε∗ba must be equal to εab .

Exercise 3.3

Solution:
h is Hermitian operator, so
[χa |h|δχa ] = [δχa |h|χa ]∗
[δχa χa |χb χb ] = [χa δχa |χb χb ]
The complex conjugate of first two terms is the last two terms(we can just exchange the subscripts).

[δχa χa |χb χb ]∗ = [χb χb |δχa χa ] = [χa χa |δχb χb ]

The second summation is therefore

1 XX  
[δχa χa |χb χb ] + [δχa χa |χb χb ] + [δχa χa |χb χb ]∗ + [δχa χa |χb χb ]∗
2 a
b
XX
= [δχa χa |χb χb ] + complex conjugate
a b

Exercise 3.4

Solution:
From the previous result, X
hχi | f | χj i = hi | h | ji + hib k jbi
b

And it is obviously that X

hχj | f | χi i = hj | h | ii + hjb k ibi
b
∗
h is a Hermitian operator, so hi | h | ji = hj | h | ii . And
X XZ
−1
hib k jbi = dx1 dx2 χ∗i (1)χ∗b (2)r12 χb (2)χj (1)
b b
Z 
−1
− dx1 dx2 χ∗i (1)χ∗b (2)r12 χj (2)χb (1)

X XZ
−1
hjb k ibi = dx1 dx2 χ∗j (1)χ∗b (2)r12 χb (2)χi (1)
b b
Z 
−1
− dx1 dx2 χ∗j (1)χ∗b (2)r12 χi (2)χb (1)

34
 P P ∗ ∗
It is easy to find that hib k jbi = hjb k ibi . So hχi | f | χj i = hχj | f | χi i , and Fock operator is a Hermitian
b b
operator.
Exercise 3.5

Solution:
N −2
Ψcd  H  N −2 Ψcd − N Ψ0  H  N Ψ0

N −2   
  
Ecd =
N N
 X 1X
Ψ0  H  N Ψ0 =

N  
ha | h | ai + hab k abi
a
2
a,b

N N
X 1 X
Ψcd  H  N −2 Ψcd =

N −2   
ha | h | ai + hab k abi
2
a6=c,d a,b6=c,d

Therefore
 N N N N
N −2 1 X X X X 
Ecd = − hc | h | ci − hd | h | di − hac k aci + had k adi + hca k cai + hda k dai
2 a a a a

− hcd k cdi − hdc k dci

 N
X   N
X 
= − hc | h | ci + hca k cai − hd | h | di + hda k dai + hcd k cdi
a a
= −εc − εd + hcd | cdi − hcd | dci

Exercise 3.6

Solution:
N N   NX
+1 N +1
N
X 1X 1 X 
E0 −N +1 E r = ha | h | ai + hab k abi − ha | h | ai + hab k abi
a
2 a
2
a,b a,b
N N
1 X X 
= − hr | h | ri − har k ari + hrb k rbi
2
a(b≡r) b(a≡r)
N
X
= − hr | h | ri − hrb k rbi
b
= −εr

Exercise 3.7

Solution:
N!
1 X
H |Ψ0 i = H √ (−1)pi Pi {χ1 χ2 . . . χN }
N! i
N!
1 X
=√ (−1)pi H Pi {χ1 χ2 . . . χN }
N! i
N!
1 X
=√ (−1)pi Pi H {χ1 χ2 . . . χN }
N! i

35
Because
N
X
H0 {χ1 χ2 . . . χN } = f (a) {χ1 χ2 . . . χN }
a=1
N
X
= εa {χ1 χ2 . . . χN }
a=1

N N!
X 1 X
H0 |Ψ0 i = εa √ (−1)pi Pi {χ1 χ2 . . . χN }
a=1 N ! i

Slater determinant is the eigenfunction of Hartree-Fock Hamiltonian H0 with eigenvalue

P
a εa .
Suppose
H0 = f (1) + f (2) + . . . + f (i) + f (j) + . . . + f (N )
And after permutation operator Pi j being applied on it, it becomes

Pij H0 = f (1) + f (2) + . . . + f (j) + f (i) + . . . + f (N )

It is obvious that thay are just identical.

Pij H0 = H0 Pij
So H0 commutes with Pij .
Exercise 3.9

Solution:
Suppose χi has α spin function.
Z
hχi | h | χi i = dr1 dω1 ψj∗ (r1 )α∗ (ω1 )h(1)ψj (r1 )α(ω1 )
Z
= dr1 ψj∗ (r1 )h(1)ψj (r1 )

= (ψj |h|ψj )

N N/2 N/2
X X X

hχi χb k χi χb i = hψj ψb k ψj χb i + ψj ψ̄b ψj ψ̄b
b b b

The first term:

N/2 N/2  Z
X X
−1
hψj ψb k ψj ψb i = dr1 dr2 dω1 dω2 ψj∗ (r1 )α∗ (ω1 )ψb∗ (r2 )α∗ (ω2 )r12 ψj (r1 )α(ω1 )ψb (r2 )α(ω2 )
b b
Z 
−1
− dr1 dr2 dω1 dω2 ψj∗ (r1 )α∗ (ω1 )ψb∗ (r2 )α∗ (ω2 )r12 ψb (r1 )α(ω1 )ψj (r2 )α(ω2 )
N/2  Z
X
−1
= dr1 dr2 ψj∗ (r1 )ψb∗ (r2 )r12 ψj (r1 )ψb (r2 )
b
Z 
−1
− dr1 dr2 ψj∗ (r1 )ψb∗ (r2 )r12 ψb (r1 )ψj (r2 )
N/2
X 
= (jj|bb) − (jb|bj)
b

36
The last term:
N/2 N/2  Z
X
 X −1
dr1 dr2 dω1 dω2 ψj∗ (r1 )α∗ (ω1 )ψb∗ (r2 )β ∗ (ω2 )r12

ψj ψ̄b ψj ψ̄b =
ψj (r1 )α(ω1 )ψb (r2 )β(ω2 )
b b
Z 
−1
− dr1 dr2 dω1 dω2 ψj∗ (r1 )α∗ (ω1 )ψb∗ (r2 )β ∗ (ω2 )r12 ψb (r1 )β(ω1 )ψj (r2 )α(ω2 )
N/2  Z 
X
−1
= dr1 dr2 ψj∗ (r1 )ψb∗ (r2 )r12 ψj (r1 )ψb (r2 )
b
N/2
X
= (jj|bb)
b

We have
N N/2
X X 
hχi χb k χi χb i = 2(jj|bb) − (jb|bj)
b b

The orbital energies in closed-shell expression is

N/2
X 
εj = (ψj |h|ψj ) + 2(jj|bb) − (jb|bj)
b
N/2
X
= hjj + (2Jjb − Kjb )
b

Exercise 3.10

Solution:
K
X
|ψi i = |φµ i Cµi
µ=1

K
X
∗
hψi | = Ciµ hφµ |
µ=1

Because molecular orbitals are orthonormal,

δij = hψi | ψj i
K
X K
X
∗
= Ciµ hφµ | · |φν i Cνj
µ=1 ν=1
K X
X K
∗
= Ciµ hφµ | φν i Cνj
µ=1 ν=1
K X
X K
∗
= Ciµ Sµν Cνj
µ=1 ν=1

It is equivalent with the following expression:

I = C† SC

Exercise 3.11

37
Solution:

ρ(r) = hΨ0 | ρ̂(r) | Ψ0 i

N N! N! Z
X 1 XX ∗
= pi
(−1) (−1) pj
dx1 . . . dxN Pi {χ1 (1)χ2 (2) . . . χN (N )}
s
N ! i j

δ(rs − r)Pj {χ1 (1)χ2 (2) . . . χN (N )}

Because spin orbitals are orthogonal, the intergral will be zero if permutaions i and j are different.
N N! Z
X 1 X ∗
ρ(r) = dx1 . . . dxN Pi {χ1 (1)χ2 (2) . . . χN (N )}
s
N ! i
δ(rs − r)Pi {χ1 (1)χ2 (2) . . . χN (N )}
The electron labeled s occupies spin orbitals {χi |i = 1, 2, . . . N } in turn, and other N −1 electrons have (N −1)!
kinds of arrangement.
N N Z
X (N − 1)! X
ρ(r) = dxs χi (xs )∗ δ(rs − r)χi (xs )
s
N ! i
N
X 1 X N Z
= dxs χi (xs )∗ δ(rs − r)χi (xs )
s
N i
N N/2 Z
X 2 X
= drs ψi (rs )∗ δ(rs − r)ψi (rs )
s
N i
N N/2
X 2 X
= ψi (r)∗ ψi (r)
s
N i
N/2
X
=2 ψi (r)∗ ψi (r)
i

Exercise 3.12

Solution:
X
(PSP)µδ = Pµν Sνω Pωδ
νω
 
N/2 N/2
X X X
∗ ∗ 
= 2 Cµa Cνa · Sνω · 2 Cωb Cδb
νω a b
N/2 N/2
" X  #
XX
∗ ∗
=4 Cµa Cνa Sνω Cωb Cδb
a b νω
N/2 N/2
XX
∗
=4 Cµa δab Cδb
a b
N/2
X
∗
=4 Cµa Cδa
a
= 2Pµδ
i.e.
PSP = 2P

Exercise 3.13

38
Solution:
N/2 Z 
X
−1
f (r1 ) = h(r1 ) + dr2 ψb∗ (r2 )(2 − P12 )r12 ψb (r2 )
b
N/2
" #
X Z X X
−1
= h(r1 ) + dr2 φ∗σ (r2 )Cσb
∗
(2 − P12 )r12 φ∗λ (r2 )Cλb
∗

b σ λ
N/2 Z 
XX
−1
= h(r1 ) + ∗
Cσb Cλb dr2 φ∗σ (r2 )(2 − P12 )r12 φλ (r2 )
σλ b
Z 
1X −1
= h(r1 ) + Pσλ dr2 φ∗σ (r2 )(2 − P12 )r12 φλ (r2 )
2
σλ

Exercise 3.15

Solution:

U† SU = s
SU = Us
X
Sµν ciν = si ciµ
ν

Multiply by ci∗
µ on both side and sum
X X
ci∗ i
µ Sµν cν = ci∗ i
µ si cµ
µν µν
Z
X X  2
ci∗
µ dr φ∗µ (r)φν (r) ciν = si ciµ 
µν µ
Z X  2
dr φ0∗ 0
i (r)φi (r) = si ciµ 
µ
Z X  2
dr φ0∗ 0
i (r)φi (r) = si
ciµ 
µ
Z X  2
2
dr |φ0i (r)| = si ciµ 
µ

Because intergral and summation are positive, the eigenvalues si must be positive.

Exercise 3.16

Solution:
K
X
0
ψi = Cµi φ0µ
µ=1

The expansion by original basis set is

K
X
ψi = Cνi φν
ν=1

The transformation within two basis set is

K
X
φ0µ = Xνµ φν
ν

39
 K
X K
X K
X
0
ψi = Cµi φ0µ = 0
Cµi Xνµ φν
µ=1 µ=1 ν

It is obviously that
K
X
0
Cνi = Cµi Xνµ
µ=1

i.e.
C = XC0

X
X† FX X†



µν
= µi
Fij Xjν
ij
X Z
∗
= Xiµ dr φ∗i (1)f (1)φj (1) Xjν
ij
Z
= dr φ0∗ 0
µ (1)f (1)φν (1)

0
= Fµν

Exercise 3.17

Solution:
N/2
X
E0 = (haa + faa )
a
N/2 N/2
!
X X
= 2 hψa | h | ψa i + (2Jab − Kab )
a b
N/2 K Z
X X
∗
= 2 Cµa Cνa dr φ∗µ (1)h(1)φν (1)+
a µν
N/2  K Z K Z !
X X X
∗ −1 ∗ −1 ∗
2 Cµa Cνa dr φ∗µ (1)φν (1)r12 ψb (2)ψb (2) − ∗
Cµa Cνa dr φ∗µ (1)ψb (1)r12 ψb (2)φν (2)
b µν µν
N/2 K  N/2 
XX X
∗ core


= Cµa Cνa 2Hµν + 2(µν|bb) − (µb|bν)
a µν b
K
1 X
core
= Pνµ (Hµν + Fµν )
2 µν

Exercise 3.19

Solution:
 3/4
4αβ  
φGF
1s (α, r − RA )φGF
1s (β, r − RB ) = exp − α|r − RA |2 − β|r − RB |2
π2
The exponent in the result can be changed as following:
 
−α|r − RA |2 − β|r − RB |2 = −(α + β)|r|2 + (2αRA + 2βRB ) · r − α|RA |2 + β|RB |2

40
We set p = α + β,

 αRA + βRB 2
 
αRA + βRB
−α|r − RA |2 − β|r − RB |2 = −p|r|2 + 2pr · − p  
α+β α+β 
 2 
 αRA + βRB  
+ p   − α|RA |2 + β|RB |2
α+β 

Set RP = (αRA + βRB )/(α + β):

  αβ  
−α|r − RA |2 − β|r − RB |2 = − p|r|2 + 2pr · RP − p|RP |2 − |RA |2 + |RB |2 − 2RA · RB
α+β
αβ
= −p|r − RP |2 − |RA − RB |2
α+β
On the other hand,
 3/4  3/4
4αβ 2αβ 2(α + β)
=
π2 (α + β)π π
 3/4  3/4
2αβ 2p
=
(α + β)π π

Therefore, the result is

φGF GF GF
1s (α, r − RA )φ1s (β, r − RB ) = KAB φ1s (p, r − RP )

Where  3/4
2αβ  
KAB = exp αβ/(α + β)|RA − RB |2
(α + β)π
 3/4
2p  
φGF
1s (p, r − RP ) = exp − p|r − RP |2
π

Exercise 3.21

Solution:

φCGF GF
1s (ζ = 1.0, STO-1G) = φ1s (0.270950)

Because α0 = α(ζ = 1.0) × ζ 2

φCGF GF
1s (ζ = 1.24, STO-1G) = φ1s (0.41661272)

φGF
1s is already normalized.

Z  3/4  3/4
2α −α|r−RA |2 2α 2
S12 = e · e−α|r−RB | dr
π π
 3/2 Z
2α 2 2
= e−α|r−RA | · e−α|r−RB | dr
π
 3/2 Z
2α e e−p|r−RP |2 dr
= ·K
π
 3/2 Z ∞
2α 2
= 4π ·K e r2 e−pr dr
π 0
 3/2  1/2
2α e·1 π
= 4π ·K
π 4 p3

41
Because  
αβ h α i
K = exp −
e · |RA − RB | = exp − · |RA − RB |2
2
α+β 2
p = α + β = 2α
Therefore
 3/2  1/2
2α −α/2·|RA −RB |2 1 π
S12 = 4π ·e ·
π 4 8α3
2
= e−α/2·|RA −RB |
= 0.6648

Exercise 3.22

Solution:
ψ1 is σg symmetry and ψ2 is σu symmetry. So they have the corresponding formation:

ψ1 = C1 (φ1 + φ2 ), ψ2 = C1 (φ1 − φ2 )

Both ψ1 and ψ2 are normalized.

hψ1 | ψ1 i = 1
C12


hψ1 | ψ1 i + hψ2 | ψ2 i + hψ1 | ψ2 i + hψ2 | ψ1 i = 1
1
C12 =
2 + 2S12
C1 = ±[2(1 + S12 )]−1/2
There is no matter whether C1 is positive or negative. We just set it positive.

C1 = [2(1 + S12 )]−1/2

With the same procedure

C2 = [2(1 − S12 )]−1/2

Exercise 3.23

Solution:

Hcore C = SCε
Use matrix presentation
      
H core core
H12 C C2 S S12 C C2 ε1 0
 11  1  =  11  1  
core core
H21 H22 C1 −C2 S21 S22 C1 −C2 0 ε2
(
core core
H11 C1 + H12 C1 = ε1 (S11 C1 + S12 C1 )
core core
H11 C2 − H12 C2 = ε2 (S11 C2 − S12 C2 )
Eliminate C1 and C2 on both side, and S11 = 1
core core
ε1 = (H11 + H12 )/(1 + S12 )
= (−1.1204 − 0.9584)/(1 + 0.6593)
= −1.2528 a.u.
core core
ε2 = (H11 − H12 )/(1 − S12 )
= (−1.1204 + 0.9584)/(1 − 0.6593)
= −0.4755 a.u.

42
Exercise 3.24

Solution:
(
C11 = C1 = [2(1 + S12 )]−1/2
C21 = C1 = [2(1 + S12 )]−1/2
(
C12 = C2 = [2(1 − S12 )]−1/2
C22 = −C2 = −[2(1 − S12 )]−1/2
P1 ∗
Because Pµν = 2 a=1 Cµa Cνa
∗
= (1 + S12 )−1


 P11 = 2C11 C11
∗
= (1 + S12 )−1

P
12 = 2C11 C21
∗


 P 21 = 2C21 C11 = (1 + S12 )−1
∗
P22 = 2C21 C21 = (1 + S12 )−1

 
1 1
P = (1 + S12 )−1  
1 1

The corresponding density matrix for H+

2 is
 
01 −1 
1 1
P = P = [2(1 + S12 )] 
2 1 1

Exercise 3.25

Solution:

core
X h 1 i
F11 = H11 + Pλσ (φ1 φ1 |φσ φλ ) − (φ1 φλ |φσ φ1 )
2
λσ

core
h 1 i
= H11 + P11 (φ1 φ1 |φ1 φ1 ) − (φ1 φ1 |φ1 φ1 )
2
h 1 i
+ P12 (φ1 φ1 |φ2 φ1 ) − (φ1 φ1 |φ2 φ1 )
2
h 1 i
+ P21 (φ1 φ1 |φ1 φ2 ) − (φ1 φ2 |φ1 φ1 )
2
h 1 i
+ P22 (φ1 φ1 |φ2 φ2 ) − (φ1 φ2 |φ2 φ1 )
2
Because
(φ1 φ1 |φ2 φ1 ) = (φ1 φ1 |φ2 φ1 ) = (φ1 φ1 |φ1 φ2 ) = (φ1 φ2 |φ1 φ1 )
h1 1 i
F11 core
= H11 + (1 + S12 )−1 (φ1 φ1 |φ2 φ1 ) + (φ1 φ1 |φ1 φ2 ) + (φ1 φ1 |φ2 φ2 ) − (φ1 φ2 |φ2 φ1 )
2 2

Exercise 3.33

Solution:
Z
f α (r1 ) = dω1 α∗ (ω1 )f (r1 , ω1 )α(ω1 )
XZ
−1
f (r1 , ω1 ) = h(1) + dx2 χ∗a (2)r12 (1 − P12 )χa (2)
a

43
Therefore
Z Z " #
XZ
−1
α
f (r1 ) = ∗
dω1 α (ω1 )h(1)α(ω1 ) + ∗
dω1 α (ω1 ) dx2 χ∗a (2)r12 (1 − P12 )χa (2) α(ω1 )
a

The first term, where core-Hamiltonian has nothing to do with spin:

Z
dω1 α∗ (ω1 )h(1)α(ω1 ) = h(1)

The next term can be divided into two parts based on spin:
α α
N Z
X N Z
X
−1 −1
dω1 x2 α ∗
(ω1 )χ∗a (x2 )r12 (1 − P12 )χa (x2 )α(ω1 ) = dω1 x2 α∗ (ω1 )χ∗a (x2 )r12 χa (x2 )α(ω1 )
a a
α
N Z
X
−1
− dω1 x2 α∗ (ω1 )χ∗a (x2 )r12 χa (x1 )α(ω2 )
a
Nα Z
X
−1
= dω1 r2 ω2 α∗ (ω1 )α∗ (ω2 )ψa∗α (r2 )r12 α(ω2 )ψaα (r2 )α(ω1 )
a
α
N Z
X
−1
− dω1 r2 ω2 α∗ (ω1 )α∗ (ω2 )ψa∗α (r2 )r12 α(ω1 )ψaα (r1 )α(ω2 )
a
α
N Z
X
−1 α
= dr2 ψa∗α (r2 )r12 ψa (r2 )
a
α
N Z
X
−1 α
− dr2 ψa∗α (r2 )r12 ψa (r1 )
a
Nα
X  α
Ja (1) − Kaα (1)

=
a

Similarly
β β
N Z
X N Z
X
−1 −1
dω1 x2 α ∗
(ω1 )χ∗a (x2 )r12 (1 − P12 )χa (x2 )α(ω1 ) = dω1 r2 ω2 α∗ (ω1 )β ∗ (ω2 )ψa∗β (r2 )r12 β(ω2 )ψaβ (r2 )α(ω1 )
a a
β
N Z
X
−1
− dω1 r2 ω2 α∗ (ω1 )β ∗ (ω2 )ψa∗β (r2 )r12 β(ω1 )ψaβ (r1 )α(ω2 )
a
Nβ Z
X
−1 β
= dr2 ψa∗β (r2 )r12 ψa (r2 )
a
Nβ
X
= Jaβ (1)
a

So the result is
α β
N
X N
α
 α  X
f (r1 ) = h(1) + Ja (1) − Kaα (1) + Jaβ (1)
a a

Exercise 3.35

Solution:
Z
dr1 ψi∗α hψiα = hα
ii

44
 Z
−1 α
dr1 dr2 ψi∗α (1)ψa∗α (2)r12 ψa (2)ψiα (1) = Jia
αα

Z
−1 α
dr1 dr2 ψi∗α (1)ψa∗α (2)r12 ψa (1)ψiα (2) = Kia
αα

Z
−1 β αβ
dr1 dr2 ψi∗α (1)ψa∗β (2)r12 ψa (2)ψiα (1) = Jia

Therefore
εα α α α
i = (ψi |f |ψi )
α β
N
X N
X αβ
= hα
ii + αα
(Jia − αα
Kia ) + Jia
a a

εβi = (ψiβ |f β |ψiβ )

β α
N
X N
X
= hβii + ββ
(Jia ββ
− Kia )+ βα
Jia
a a
α β α α
N N N N
X X 1 X X αα
E0 = εα
i + εβi − αα
(J − Kia )
i i
2 i a ia
Nβ Nβ Nα Nβ
1 X X ββ ββ
X X αβ
− (Jia − Kia )− Jia
2 i a i a

Exercise 3.36

Solution:
Z Z
ρS (r) dr = ρα (r) − ρβ (r) dr

= Nα − Nβ
Because
1 α
Sz |Ψi = N − N β |Ψi


2
So the eigenvalue(expectation value) of Sz is 21


N − Nβ .
α

Z
ρS (r) dr = 2hSz i

Exercise 3.37

Solution:
α α
N
X N
X
α 2
ρ (r) = |ψaα (r)| = ψaα∗ (r) · ψaα (r)
a a
X
ψaα = α
Cµa φµ
µ
α !
N
X X X
α α∗ ∗ α
ρ (r) = Cνa φν (r) · Cµa φµ (r)
a ν µ
Nα
XX
= α
Cµa α∗
Cνa · φµ (r)φ∗ν (r)
µν a

45
Define density matrix for α electrons
α
N
X
α α α∗
Pµν = Cµa Cνa
a
X
ρα (r) = α
Pµν φµ (r)φ∗ν (r)
µν

Exercise 3.38

Solution:
N
X N
X
hO1 i = hΨ | h(i) | Ψi = hχi | h(i) | χi i
i i
α
N
X N
X Nβ D
X   E
hχi | h(i) | χi i = hψiα | h(i) | ψiα i + ψiβ  h(i)  ψiβ
 
i i i
α α
N
X N XX
X
∗
hψiα | h(i) | ψiα i = Cµi Cνi (ν|h|µ)
i i µ ν
XX
α
= Pµν (ν|h|µ)
µ ν

With the same procedure

β
N D
X   E XX
ψiβ  h(i)  ψiβ = β
Pµν (ν|h|µ)
 
i µ ν
XX XX
α β
hO1 i = Pµν (ν|h|µ) + Pµν (ν|h|µ)
µ ν µ ν
XX
T
= Pµν (ν|h|µ)
µ ν

Exercise 3.39

Solution:

hρ̂S i = Ψ0  ρ̂S  Ψ0

  

N! N Z
2 XX
= (−1)pi (−1)pj dx1 . . . xN Pi {χ1 (1) . . . χk (N )}
N ! ij m
δ(rm − R)sz (m)Pj {χ1 (1) . . . χk (N )}
The permutation Pi and Pj are required to be the same. Otherwise there must be electrons occupying

46
different spin orbitals(not the electron m), and the corresponding term is equal to zero.
N! N Z
2 XX
hρ̂S i = dx1 . . . xN Pi {χ1 (1) . . . χk (N )}
N! i m
δ(rm − R)sz (m)Pi {χ1 (1) . . . χk (N )}
N X
N Z
2 X
= dxm χi (m)δ(rm − R)sz (m)χi (m)
N i m
α
N  N Z
2 X X
= drm dωm ψiα (m)α(ωm )δ(rm − R)sz (m)ψiα (m)α(ωm )
N m i
β
N Z
X 
+ drm dωm ψiβ (m)β(ωm )δ(rm − R)sz (m)ψiβ (m)β(ωm )
i
N  Nα Z
2 X 1X
= drm ψiα (m)δ(rm − R)ψiα (m)
N m 2 i
Nβ Z 
1X
− drm ψiβ (m)δ(rm − R)ψiβ (m)
2 i
α β
N Z
X N Z
X
= dR ψiα (R)ψiα (R) − dR ψiβ (R)ψiβ (R)
i i
= ρα (R) − ρβ (R)
= ρS (R)

Exercise 3.40

Solution:
α β
N N Nα Nα
X X 1 X X αα
hα hβaa αα


E0 = aa + + Jab − Kab
a a
2 a
b
Nβ X
Nβ N X
N α β
1 X ββ ββ

X αβ
+ Jab − Kab + Jab
2 a a
b b
X
hα α α
aa = (ψa |h|ψa ) =
α∗
Cµa α
(φµ |h|φν )Cνa
µν

Nα
X Nα X
X X
hα
aa =
α∗
Cµa α
(φµ |h|φν )Cνa = α
Pνµ core
Hµν
a a µν µν

In the same way,

β β
N
X N X
X X
hβaa = β∗
Cµa β
(φµ |h|φν )Cνa = β
Pνµ core
Hµν
a a µν µν
X
αα
= ψaα ψaα |ψbα ψb = α α∗ α α∗ α


Jab Cµa Cνa Cλb Cσb (µν|λσ)
µνλσ
X
αα
= ψaα ψbα |ψbα ψa = α α∗ α α∗ α


Kab Cµa Cνa Cλb Cσb (µλ|σν)
µνλσ

ββ
X
Jab = ψaβ ψaβ |ψbβ ψb = β
β∗ β
Cµa β∗ β
Cνa Cλb Cσb (µν|λσ)
µνλσ

47
 ββ
X
= ψaβ ψbβ |ψbβ ψaβ = β∗ β β∗ β


Kab Cµa Cνa Cλb Cσb (µλ|σν)
µνλσ

αβ
X
Jab = ψaα ψaα |ψbβ ψb = β
α∗ α
Cµa β∗ β
Cνa Cλb Cσb (µν|λσ)
µνλσ

α α α α
N N N N
1 X X αα αα

1 XX α∗ α
X
α∗ α
 
Jab − Kab = Cµa Cνa Cλb Cσb (µν|λσ) − (µλ|σν)
2 a 2 a
b µνλσ b
1 X α α 
= Pνµ Pσλ (µν|λσ) − (µλ|σν)
2
µνλσ

β β β β
N N N N
1 X X ββ ββ
1 X X β∗ β X β∗ β  
Jab − Kab = Cµa Cνa Cλb Cσb (µν|λσ) − (µλ|σν)
2 a 2 a
b µνλσ b
1 X β β  
= Pνµ Pσλ (µν|λσ) − (µλ|σν)
2
µνλσ

α β α β
X N
N X N
XX N
X
αβ α∗ α β∗ β
Jab = Cµa Cνa Cλb Cσb (µν|λσ)
a b µνλσ a b
X β
α
= Pνµ Pσλ (µν|λσ)
µνλσ

Therefore, the total energy is

X
α core
X
β core 1 X α α 
E0 = Pνµ Hµν + Pνµ Hµν + Pνµ Pσλ (µν|λσ) − (µλ|σν)
µν µν
2
µνλσ
1 X β β   X α β
+ Pνµ Pσλ (µν|λσ) − (µλ|σν) + Pνµ Pσλ (µν|λσ)
2
µνλσ µνλσ
1 X T core 1 X α core 1 X β core
= P H + P H + P H
2 µν νµ µν 2 µν νµ µν 2 µν νµ µν
1 X αh α β
α
i
+ Pνµ Pσλ + Pσλ (µν|λσ) − Pσλ (µλ|σν)
2
µνλσ
1 X β h α β

β
i
+ Pσλ Pνµ + Pνµ (µν|λσ) − Pνµ (µλ|σν)
2
µνλσ

Because µ, ν, λ, σ are dumb variables, and (µν|λσ) = (λσ|µν). The last term can be expressed as
1 X βh α β
β
i
Pνµ Pλσ + Pλσ (µν|λσ) − Pλσ (µλ|σν)
2
µνλσ

48
 1 X T core 1 X α core 1 X β core
E0 = P H + P H + P H
2 µν νµ µν 2 µν νµ µν 2 µν νµ µν
1 X αh α β
α
i
+ Pνµ Pσλ + Pσλ (µν|λσ) − Pσλ (µλ|σν)
2
µνλσ
1 X βh α β
β
i
+ Pνµ Pλσ + Pλσ (µν|λσ) − Pλσ (µλ|σν)
2
µνλσ
 i
1 X T core 1 X α core
Xh
T α
= P H + P Hµν + Pσλ (µν|λσ) − Pσλ (µλ|σν)
2 µν νµ µν 2 µν νµ µν
 i
1X β core
Xh
T β
+ P Hµν + Pσλ (µν|λσ) − Pσλ (µλ|σν)
2 µν νµ µν
1 X h T core α α β β
i
= Pνµ Hµν + Pνµ Fµν + Pνµ Fµν
2 µν

4 Chapter 4
Singly-excited singlet spin-adapted configurations:
 
|1 Ψra i = 2−1/2 |Ψr̄ā i + |Ψra i

Doubly-excited singlet spin-adapted configurations:

|1 Ψrr rr̄
aa i = |Ψaā i
 
−1/2
|1 Ψrs
aa i = 2 |Ψ rs̄
aā i + |Ψ sr̄
aā i
 
−1/2
|1 Ψrr
ab i = 2 |Ψ r̄r
āb i + |Ψ rr̄
ab̄ i
 
−1/2
|A Ψrs
ab i = (12) 2 |Ψrs r̄s̄ s̄r r̄s rs̄ sr̄
ab i + 2 |Ψāb̄ i − |Ψāb i + |Ψāb i + |Ψab̄ i − |Ψab̄ i

1  s̄r 
|B Ψrs
ab i = |Ψāb i + |Ψr̄s
āb i + |Ψ rs̄
ab̄ i + |Ψ sr̄
ab̄ i
2

 1 r
Ψra  H − E0  1 Ψsb = hΨa | H − E0 | Ψsb i + Ψra  H − E0  Ψs̄b̄

1  
  
2
 
+ Ψr̄ā  H − E0  Ψsb + Ψr̄ā  H − E0  Ψs̄b̄

  


hΨra | O1 | Ψsb i hΨra | O2 | Ψsb i

a 6= b, r 6= s 0 hrb k asi
N
P
a = b, r 6= s hr | h | si hrn k sni − hra k sai
n
PN
a 6= b, r = s − hb | h | ai − hbn k ani − hbr k ari
n
N N
1 X X
N hmn k mni − han k ani
2 m,n
X
hm | h | mi − ha | h | ai n
a = b, r = s m N
X
+ hr | h | ri + hrn k rni − hra k rai
n

49
 Ψra  O1  Ψs̄b̄ Ψra  O2  Ψs̄b̄

  
  


a 6= b, r 6= s 0 rb̄ as̄
a = b, r 6= s 0 hrā k as̄i


a 6= b, r = s 0 rb̄ ar̄
a = b, r = s 0 hrā k ar̄i

hΨr̄ā | O1 | Ψsb i hΨr̄ā | O2 | Ψsb i

a 6= b, r 6= s 0 hr̄b k āsi
a = b, r 6= s 0 har̄ k sāi
a 6= b, r = s 0 hr̄b k āri
a = b, r = s 0 har̄ k rāi

Ψā  O1  Ψs̄b̄ Ψr̄ā  O2  Ψs̄b̄

r̄   
  


a 6= b, r 6= s 0 r̄b̄ ās̄
N
P
a = b, r 6= s hr̄ | h | s̄i hr̄n k s̄ni − hr̄ā k s̄āi
n

   N

P 

a 6= b, r = s − b̄  h  ā − b̄n ān − b̄r̄ ār̄
n
N N
1X X
N hmn k mni − hān k āni
2 m,n
X
hm | h | mi − hā | h | āi n
a = b, r = s m N
X
+ hr̄ | h | r̄i + hr̄n k r̄ni − hr̄ā k r̄āi
n

In addition, only when a = b and r = s, the following terms are not zero:

hΨra | E0 | Ψsb i = Ψr̄ā  E0  Ψs̄b̄ = E0

  

With the above results, we can evaluate the matrix elements.

a 6= b, r 6= s
 1

Ψa  H − E0  1 Ψsb =

1 r  

hrb k asi + rb̄ as̄ + hr̄b k āsi + r̄b̄ ās̄
2
= 2(ra|bs) − (rs|ba)

a = b, r 6= s
N
 1 X
Ψra  H − E0  1 Ψsb =

1  
hr | h | si + hrn k sni − hra k sai + hrā k as̄i
2 n
N
X 
+ har̄ k sāi + hr̄ | h | s̄i + hr̄n k s̄ni − hr̄ā k s̄āi
n

N
X
hr | h | si + hrn k sni = hr | f | si
n

50
 N
X
hr̄ | h | s̄i + hr̄n k s̄ni = hr̄ | f | s̄i
n

These two terms are the non-diagonal of Fock matrix, which are zero. The remaining is
 1 
Ψa  H − E0  1 Ψsb =

1 r  
4(ra|as) − 2(rs|aa)
2
Because a = b, we can be free to substitute a with b.

Ψa  H − E0  1 Ψsb = 2(ra|bs) − (rs|ba)

1 r   

With the same procedure, when a 6= b and r = s, the result is

Ψa  H − E0  1 Ψsb = 2(ra|bs) − (rs|ba)

1 r   

a = b, r = s
N N
 1 X 1X
Ψra  H − E0  1 Ψsb =

1  
hm | h | mi − ha | h | ai + hr | h | ri + hmn k mni
2 m 2 m,n
N
X N
X
− han k ani + hrn k rni − hra k rai + hrā k ar̄i
n n
N
X
+ har̄ k rāi + hm | h | mi − hā | h | āi + hr̄ | h | r̄i
m
N N N
1X X X
+ hmn k mni − hān k āni + hr̄n k r̄ni
2 m,n n n

− hr̄ā k r̄āi − 2E0
1 
= − 2εa + 2εr + 4(ra|ar) − 2(rr|aa)
2
After selective substitutions, the result is

Ψra  H − E0  1 Ψsb = −εa + εr + 2(ra|bs) − (rs|ba)

1   

And we can conclude that

Ψra  H − E0  1 Ψsb = (−εa + εr )δab δrs + 2(ra|bs) − (rs|ba)

1   

Ψ0  H  1 Ψrr
aa = Ψ0 O1 Ψaā + Ψ0 O2 Ψaā

  
  rr̄ 
  rr̄ 
   
= haā k rr̄i
= [ar|ār̄] − [ar̄|ār]
= Kar

1 
  sr̄  
Ψ0  H  1 Ψrs Ψ0 H  Ψrs̄ aā + Ψ0 H Ψaā

    

aa = √
 
2
1 
  sr̄  
Ψ0 O2  Ψrs̄
aā + Ψ0 O2 Ψaā
 

=√  
2
1  
= √ [ar|ās̄] − [as̄|ār] + [as|ār̄] − [ar̄|ās]
2
1  
= √ (ar|as) + (as|ar)
2
√
= 2(sa|ra)

51
Similarly,

1 
  rr̄  
Ψ0  H  1 Ψrr O O

    r̄r 

ab = √ Ψ 0 2
 Ψ āa + Ψ 0
 2
Ψ
ab̄
2
1  
= √ [ār̄|br] − [ār|br̄] + [ar|b̄r̄] − [ar̄|b̄r]
2
1  
= √ (ar|ar) + (ar|br)
2
√
= 2(rb|ra)

1 
Ψ0  H  A Ψrs 2 hΨ0 | H | Ψrs ab i + 2 Ψ0 H Ψāb̄ − Ψ0 H Ψāb

  
  r̄s̄ 
  s̄r 
ab = √
   
12
 sr̄  
+ Ψ0  H  Ψr̄s āb + Ψ0 H Ψab̄ − Ψ0 H Ψab̄

  
  rs̄ 

   
 
1   
=√ 2 [ar|bs] − [as|br] + 2 [ār̄|b̄s̄] − [ās̄|b̄r̄] − [ās̄|br] + [ār|bs̄]
12

+ [ār̄|bs] − [ās|br̄] + [ar|b̄s̄] − [as̄|b̄r] − [as|b̄r̄] + [ar̄|b̄s]
1 h
=√ 2(ar|bs) − 2(as|br) + 2(ar|bs) − 2(as|br)
12
i
− (as|br) + (ar|bs) + (ar|bs) − (as|br)
√ h i
= 3 (ra|sb) − (rb|sa)

aa H − E0
 Ψaa = Ψaā  H  Ψrr̄

1
Ψrr
  1 rr 
rr̄   
aā − E0

N
X
= hm | h | mi − ha | h | ai − hā | h | āi + hrh | ri + hr̄ | h | r̄i
m
N N N
1X X X
+ hmn k mni − han k ani − hān k āni
2 m,n n n
N
X N
X
+ hāa k āai + hrn k rni + hr̄n k r̄ni + hr̄r k r̄ri
n n
− har k ari − har̄ k ar̄i − hār k āri
N
X N 
1X
− hār̄ k ār̄i − hm | h | mi + hmn k mni
m
2 m,n
= −2εa + 2εr + Jaa + Jrr − 4Jra + 2Kra

 1 rs  1 
rs̄ 
aa H − E0 Ψaā  H  Ψrs̄ aā + Ψaā H Ψaā

1
Ψrs
  
sr̄   sr̄ 
  Ψaa =  
2 
+ Ψrs̄ H H

  sr̄ 
sr̄   rs̄ 
aā
  Ψ aā + Ψ aā
  Ψaā − 2E0

52
For each part,
N
 rs̄  X
Ψrs̄ H


aā
  Ψaā = hm | h | mi − ha | h | ai − hā | h | āi + hr | h | ri + hs̄ | h | s̄i
m
N N
1X X X
+ hmn k mni − han k ani − hān k āni + hāa k āai
2 m,n n n
N
X N
X
+ hrn k rni + hs̄n k s̄ni + hrs̄ k rs̄i
n n
− har k ari − has̄ k as̄i − hār k āri − hās̄ k ās̄i

N
 sr̄  X
Ψsr̄ H


aā
  Ψ aā = hm | h | mi − ha | h | ai − hā | h | āi + hs | h | si + hr̄ | h | r̄i
m
N N
1X X X
+ hmn k mni − han k ani − hān k āni + hāa k āai
2 m,n n n
N
X N
X
+ hsn k sni + hr̄n k r̄ni + hr̄s k r̄si
n n
− has k asi − har̄ k ar̄i − hās k āsi − hār̄ k ār̄i

Ψaā  H  Ψsr̄

rs̄   
aā = hrs̄ k sr̄i = Krs

aā H Ψaā = hsr̄ k rs̄i = Krs

Ψsr̄

  rs̄ 
 
Therefore
 1 rs  1 
H

1
Ψrs

aa
 − E0
 Ψaa = − 4εa + 2εr + 2εs + 2Jaa + 2Jrs
2 
− 4Jra − 4Jsa + 2Kra + 2Ksa + 2Krs
= εr + εs − 2εa + Jaa + Jrs + Krs
− 2Jsa − 2Jra + Ksa + Kra
Exercise 4.1

Solution:
X
cde Ψa H Ψcde
ctuv

r  tuv 
 
c<d<e
t<u<v

If there is no one in c, d, e equal to a, and no one in t, u, v equal to r, the integral will be zero. This requires
at least one term in c, d, e to be equal to a. And the similar requirment is applied to t, u, v. For example, we
let c be a, and t be r, after which we change the dumb variables. The result is:
X
acd Ψa H Ψacd
crtu

r  rtu 
 
c<d
t<u

Exercise 4.2

Solution:
    
0 K12 c1 c
    = E  1
K12 2∆ c2 c2

53
   
−E K12 c
   1 = 0
K12 2∆ − E c2
 
 −E
 
K12 
 =0
K12 2∆ − E 
 

E 2 − 2∆E − K12
2
=0
p
2∆ ± 4∆2 + 4K122
E=
2
2 2 1/2


= ∆ ± ∆ + K12
Ecorr is the lowest eigenvalue.

1/2
Ecorr = ∆ − ∆2 + K12
2

Exercise 4.3

Solution:

Ecorr = −0.020562
K12 = 0.1813
c = Ecorr /K12 = −0.1134
As R → ∞, two-electron integral tends to 12 (φ1 φ1 |φ1 φ1 ), and ε1 , ε2 → E(H). That is
1
lim K12 = (φ1 φ1 |φ1 φ1 )
R→∞ 2
lim ∆ = 0
R→∞
Therefore
limR→∞ K12
lim c = 
1/2 
R→∞ 2
limR→∞ ∆ − ∆2 + K12 − 2∆
1
2 (φ1 φ1 |φ1 φ1 )
=
− 12 (φ1 φ1 |φ1 φ1 )
= −1
 
 
1  ψ 1 (1) ψ 1 (2)
|Ψ0 i = √  
2 ψ̄1 (1) ψ̄1 (2)
 
 
22̄ 1  ψ 2 (1) ψ 2 (2)
|Ψ11̄ i = √   
2 ψ̄2 (1) ψ̄2 (2)
Because ψ1 and ψ2 are the linear combination of φ1 and φ2 .
ψ1 = c1 (φ1 + φ2 )
ψ2 = c2 (φ1 − φ2 )
Substitute the expands into determinants
c2 h
|Ψ0 i = √1 φ1 (1)φ̄1 (2) + φ1 (1)φ̄2 (2) + φ2 (1)φ̄1 (2) + φ2 (1)φ̄2 (2)


2

i
− φ1 (1)φ̄1 (2) + φ1 (1)φ̄2 (2) + φ2 (1)φ̄1 (2) + φ2 (1)φ̄2 (2)
1

= |φ1 φ̄1 i + |φ1 φ̄2 i + |φ2 φ̄1 i + |φ2 φ̄2 i
2 + 2S12

54
 c2 h
|Ψ212̄1̄ i = √2 φ1 (1)φ̄1 (2) − φ1 (1)φ̄2 (2) − φ2 (1)φ̄1 (2) + φ2 (1)φ̄2 (2)


2

i
− φ1 (1)φ̄1 (2) − φ1 (1)φ̄2 (2) − φ2 (1)φ̄1 (2) + φ2 (1)φ̄2 (2)
1

= |φ1 φ̄1 i − |φ1 φ̄2 i − |φ2 φ̄1 i + |φ2 φ̄2 i
2 − 2S12
As R → ∞, c → −1, and S12 → 0:

|Φ0 i = |Ψ0 i − |Ψ212̄1̄ i

1

= 2 |φ1 φ̄2 i + 2 |φ2 φ̄1 i
2
= |φ1 φ̄2 i + |φ2 φ̄1 i

After normalization:
1

|Φ0 i = √ |φ1 φ̄2 i + |φ2 φ̄1 i
2

Exercise 4.4

Solution:
Z
γij = dx1 dx01 χ∗i (x1 )γ(x1 , x01 )χj (x01 )
hZ i∗
∗
γji = dx1 dx01 χ∗j (x1 )γ(x1 , x01 )χi (x01 )
Z
∗
dx1 dx01 χ∗i (x01 ) γ(x1 , x01 ) χj (x1 )

=

Substitute γ(x1 , x01 ) with its definition:

Z Z
0 ∗ 0
γij = N dx1 dx1 χi (x1 )χj (x1 ) dx2 . . . dxN Φ(x1 , . . . , xN )Φ∗ (x01 , . . . , xN )
Z Z
∗
γji =N dx1 dx01 χ∗i (x01 )χj (x1 ) dx2 . . . dxN Φ(x01 , . . . , xN )Φ∗ (x1 , . . . , xN )

Because x1 and x01 are dumb variables, γij = γji

∗
. And that matrix γ is a Hermitian matrix is proved.

Exercise 4.5

Solution:
N
X N Z
X
tr γ = γii = dx1 dx01 χ∗i (x1 )γ(x1 , x01 )χ∗i (x01 )
i i
N Z
X Z
= dx1 dx01 χ∗i (x1 )χ∗i (x01 ) · N dx2 . . . dxN Φ(x1 , . . . , xN )Φ∗ (x01 , . . . , xN )
i

Make x1 = x01 . Then the result of integral is obvious:

tr γ = N

Exercise 4.6

55
Solution:
(a)
N
X
hΦ | O1 | Φi = hΦ | h(x1 ) | Φi
i
Z
=N dx1 dx2 . . . dxN Φ∗ (x1 , . . . , xN )h(x1 )Φ(x1 , . . . , xN )
Z Z
=N dx1 h(x1 ) · dx2 . . . dxN Φ∗ (x1 , . . . , xN )Φ(x1 , . . . , xN )
Z
1 h i
=N· dx1 h(x1 )γ(x1 , x01 ) 0
N x1 =x1
Z h i
= dx1 h(x1 )γ(x1 , x01 ) 0
x1 =x1

(b)
Z X
hΦ | O1 | Φi = dx1 h(x1 ) χi (x1 )γij χ∗j (x1 )
ij
XZ
= dx1 χ∗j (x1 )h(x1 )χi (x1 ) · γij
ij
X
= hji γij
ij
X
= (hγ)jj
j

= tr hγ

Exercise 4.8

Solution:
(a)
K
X h i 1X K X K h i
|1 Φ0 i = c0 |ψ1 ψ̄1 i + cr1 2−1/2 (|ψ1 ψ̄r i + |ψr ψ̄1 i) + crs 2 −1/2
(|ψr ψ̄ s i + |ψ s ψ̄ r i)
r=2
2 r=2 s=2 11
K K K
 1 X j  
1 X X ij

= c0 |ψi ψ̄j i i=1 + √ c1 |ψi ψ̄j ii=1 + |ψi ψ̄j ii=1 + √ c11 |ψi ψ̄j i + |ψj ψ̄i i
j=1 2 j=2 2 2 i=2 j=2
K K K K K K
1 X i1 1 X 1j 1 X X ij 1 X X ji
= c11
  
0 |ψi ψ̄j i i=1 + √ c1 |ψi ψ̄j ij=1 + √ c1 |ψi ψ̄j ii=1 + √ c11 |ψi ψ̄j i + √ c11 |ψi ψ̄j i

j=1 2 i=2 2 j=2 2 2 i=2 j=2 2 2 i=2 j=2

Clearly CI expansion can be expressed as

K X
X K
|1 Φ0 i = Cij |ψi ψ̄j i
i=1 j=1

For configurations |ψr ψ̄s i and |ψs ψ̄r i, they are the same in some sense. Thus crs sr
11 and c11 are equal.
And it can be concluded in general that
Cij = Cji
C is a symmetric matrix.
(b)
1 X
|1 Φ0 i = √


Cij i(1)j̄(2) − i(2)j̄(1)
2 ij

56
 Z    
0 1 X
1 X ∗ ∗ 0 ¯∗ ∗ ¯∗ 0


γ(1, 1 ) = 2 dx2 √ Cij i(1)j̄(2) − i(2)j̄(1) · √ Ckl k (1 )l (2) − k (2)l (1 )
2 ij 2 kl
XXZ
∗
i(1)j̄(2)k ∗ (10 )¯l∗ (2) − i(1)j̄(2)k ∗ (2)¯l∗ (10 )

= dx2 Cij Ckl
ij kl

− i(2)j̄(1)k ∗ (10 )¯l∗ (2) + i(2)j̄(1)k ∗ (2)¯l∗ (10 )


XX
∗
i(1)k ∗ (10 )δjl + j̄(1)¯l∗ (10 )δik
 
= Cij Ckl
ij kl

For the first part

XX
∗
I1 = Cij Ckl i(1)k ∗ (10 )δjl
ij kl
XX
∗
= Cij Ckj i(1)k ∗ (10 )
ij k
XX
= Cij (C † )jk i(1)k ∗ (10 )
ik j
X
= (CC † )ik i(1)k ∗ (10 )
ik
X
= (CC † )ij i(1)j ∗ (10 )
ij

The second part

XX
I2 = ∗
Cij Ckl j̄(1)¯l∗ (10 )δik
ij kl
XX
= Cij Cil∗ j̄(1)¯l∗ (10 )
ij l
XX
= Cij (C † )li j̄(1)¯l∗ (10 )
ij l
XX
= Cji (C † )il j̄(1)¯l∗ (10 )
jl i
X
= (CC † )jl j̄(1)¯l∗ (10 )
jl
X
= (CC † )ij ī(1)j̄ ∗ (10 )
ij
X
γ(1, 10 ) = (CC † )ij i(1)j ∗ (10 ) + ī(1)j̄ ∗ (10 )
 
ij

(c)
U† CU = d
UC† U = d† = d
Because d is diagonal.
U† C = dU†
C† U = U† d

U† CC† U = dU† U† d = dU† U−1 d = d2

(d) X
ζi = ψk Uki
k

57
 X
ψi = (U † )ik ζk
k

X X X X X 
γ(1, 10 ) = (CC † )ij (U † )ik ζk (1) · (U † )∗lj ζl∗ (10 ) + (U † )ik ζ̄k (1) · (U † )∗lj ζ̄l∗ (10 )
ij k l k l
XX h i
= (U † )ik (CC † )ij (U † )∗lj ζk (1)ζl∗ (10 ) + ζ̄k (1)ζ̄l∗ (10 )
ij kl
XX h i
= (U † )ki (CC † )ij Ujl ζk (1)ζl∗ (10 ) + ζ̄k (1)ζ̄l∗ (10 )
ij kl
X h i
= d2k δkl ζk (1)ζl∗ (10 ) + ζ̄k (1)ζ̄l∗ (10 )
kl
X h i
= d2i ζi (1)ζi∗ (10 ) + ζ̄i (1)ζ̄i∗ (10 )
i

(e)
X
|1 Φ0 i = 2−1/2 Cij i(1)j̄(2) − i(2)j̄(1)
 
ij
X X X X X 
= 2−1/2 Cij (U † )ik ζk (1) · (U † )jl ζ̄l (2) − (U † )ik ζk (2) · (U † )jl ζ̄l (1)
ij k l k l
XX  
= (U † )ik Cij (U † )jl 2−1/2 ζk (1)ζ̄l (2) − ζk (2)ζ̄l (1)
ij kl
XX  
= (U † )ki Cij Ujl 2−1/2 ζk (1)ζ̄l (2) − ζk (2)ζ̄l (1)
ij kl
X
= dk δkl |ζk ζ̄l i
kl
X
= di |ζi ζ̄i i
i

Exercise 4.9

Solution:
(a)
hu | ui = K 2 (a2 + b2 ) = 1
a2 − b2
hu | vi = K 2 (a2 − b2 ) =
a2 + b2
(b) It’s just obvious and skip it.
Exercise 4.10

Solution:

h11 1̄1 12 1̄2 | H | 11 1̄1 21 2̄1 i = h11 1̄1 12 1̄2 | O1 | 11 1̄1 21 2̄1 i + h11 1̄1 12 1̄2 | O2 | 11 1̄1 21 2̄1 i
= h11 1̄1 12 1̄2 | O2 | 11 1̄1 21 2̄1 i
= [12 21 |1̄2 2̄1 ] − [12 2̄1 |1̄2 21 ]
= (12 21 |12 21 )
=0

Exercise 4.9

58
Solution:
(a) D   E
Ψ0  H − E  Ψ2111 2̄1̄11 = K12
 

D   E
Ψ0  H − E  Ψ1222 1̄2̄22 = K12
 

D   E
Ψ2111 2̄1̄11  H − E  Ψ1222 1̄2̄22 = 0
 

D   E
Ψ2111 2̄1̄11  H − E  Ψ2111 2̄1̄11 2122 2̄1̄22 = K12
 

D   E
Ψ2122 2̄1̄22  H − E  Ψ2111 2̄1̄11 2122 2̄1̄22 = K12
 

D   E
Ψ2111 2̄1̄11 2122 2̄1̄22  H  Ψ2111 2̄1̄11 2122 2̄1̄22 = 4h22 + 2J22
 

The integrals, such as (21 (1)21 (1)|22 (2)22 (2)) = J22 and (21 (1)22 (1)|22 (2)21 (2)) = K22 , are zero.

D   E
Ψ2111 2̄1̄11 2122 2̄1̄22  H − E  Ψ1211 1̄2̄11 1222 2̄1̄22 = 4h22 + 2J22 − E0
 

= 4ε2 − 8J12 + 4K12 + 2J22 − 4ε1 + 2J11

= 4(ε2 − ε1 ) + 2J11 + 2J22 − 8J12 + 4K12
= 4∆

Thus we could construct the full CI equation:

    
0 K12 K12 0 1 1
    
    
K12 2∆ 0 K12  c1  2 c 
    = Ecorr  1 
    
K12 0 2∆ K12  c2  c2 
    
0 K12 K12 4∆ c3 c3

5 Chapter 5
Exercise 5.1

Solution:
(a)
XX | hab k rsi |2 | h11̄ k 22̄i |2 |(12|21)|2 2
K12
Ecorr (FO) = = = =
εa + εb − εr − εs ε1 + ε1 − ε2 − ε2 2(ε1 − ε2 ) 2(ε1 − ε2 )
a<b r<s

(b)

∆ − (∆2 + K12
2 1/2
 
1 )
Ecorr = ∆ − (∆2 + K122 1/2
) =∆
∆
2 1/2
 
 K
= ∆ 1 − 1 + 12
∆2
2 
 
 K12
≈∆ 1− 1+
2∆2
2 2
K K12
= − 12 = −
2∆ 2(ε2 − ε1 )
= Ecorr (FO)

59
Exercise 5.2

Solution:
X
1i 1̄i Ψ0 H Ψ1i 1̄i
ctu

  tu 
e1i 1̄i =  
t<u
D   E
= c21ii 2̄1̄ii Ψ0  H  Ψ21ii 2̄1̄ii
 

= c21ii 2̄1̄ii K12

D   E XD   E
e1i 1̄i c21ii 2̄1̄ii = Ψ21ii 2̄1̄ii  H  Ψ0 + Ψ21ii 2̄1̄ii  H − E0  Ψtu tu
1i 1̄i c1i 1̄i
   
t<u
XD   E
 H − E0  Ψ1ki 1̄il c1ki 1̄il
 2 2̄ 2 2̄
= K12 + Ψ21ii 2̄1̄ii

kl
D   E
For Ψ21ii 2̄1̄ii  H − E0  Ψ21ki 1̄2̄il ,
 

case 1: i = k = l D   E D   E
Ψ21ii 2̄1̄ii  H − E0  Ψ21ki 1̄2̄il = Ψ21ii 2̄1̄ii  H − E0  Ψ21ii 2̄1̄ii = 2∆
   

case 2: i 6= k = l = j
D   E D   E
 2 2̄
Ψ21ii 2̄1̄ii  H − E0  Ψ21ki 1̄2̄il = Ψ21ii 2̄1̄ii  H − E0  Ψ1ij1̄ij = h2i 2̄i k 2j 2̄j i = 0
  

case 3: i = k 6= l
D   E D   E X
Ψ21ii 2̄1̄ii  H − E0  Ψ21ki 1̄2̄il = Ψ21ii 2̄1̄ii  H − E0  Ψ21ii 2̄1̄li = h2̄i a k 2̄l ai = 0
   
a

e1i 1̄i c21ii 2̄1̄ii = K12 + 2∆c21ii 2̄1̄ii

Exercise 5.4

Solution:

|aābb̄i = (2−1/2 )4 |(11 + 12 )(1̄1 + 1̄2 )(11 − 12 )(1̄1 − 1̄2 )i

1 
= |11 (1̄1 + 1̄2 )(11 − 12 )(1̄1 − 1̄2 )i + |12 (1̄1 + 1̄2 )(11 − 12 )(1̄1 − 1̄2 )i
4
1 
= − |11 (1̄1 + 1̄2 )12 (1̄1 − 1̄2 )i + |12 (1̄1 + 1̄2 )11 (1̄1 − 1̄2 )i
4
1


= − |11 1̄1 12 (1̄1 − 1̄2 )i + |11 1̄2 12 (1̄1 − 1̄2 )i + |12 1̄1 11 (1̄1 − 1̄2 )i + |12 1̄2 11 (1̄1 − 1̄2 )i
4
1


= − − |11 1̄1 12 1̄2 i + |11 1̄2 12 1̄1 i + − |12 1̄1 11 1̄2 i + |12 1̄2 11 1̄1 i
4
1 
= |11 1̄1 12 1̄2 i − |11 1̄2 12 1̄1 i − |12 1̄1 11 1̄2 i + |12 1̄2 11 1̄1 i
4
= |11 1̄1 12 1̄2 i

Exercise 5.5

Solution:

  ss̄  
hΨ0 | H | Ψ∗∗ −1/2
Ψ0  H  Ψrr̄ aā + Ψ0 H Ψaā
 

aā i = 2
 
 
= 2−1/2 haā k rr̄i + haā k ss̄i
= 2−1/2 K12

60
 1 
rr̄ 
aā | H − E0 | Ψaā i =
hΨ∗∗ Ψaā H − E0  Ψrr̄
aā + Ψaā H − E0 Ψaā
∗∗
 
rr̄   ss̄ 
 
2
 ss̄  
+ Ψss̄ aā H − E0 Ψaā + Ψaā H − E0 Ψaā

  rr̄ 
ss̄ 
   

1 1
Ψaā  H  Ψrr̄

rr̄   
aā = 2hbb + 2hrr + J11 + J22 + 2J12 − K12
2 2
E0 = 2haa + 2hbb + 2J11

εa = haa + J11
εb = hbb + J11

3 1
aā H Ψaā = 2(εb + εr ) − J11 + J22 − 2J12 + K12
Ψrr̄

  rr̄ 
 
2 2
E0 = εa + εb − 2J11

1 1
aā H − E0 Ψaā = 2(ε2 − ε1 ) + J11 + J22 − 2J12 + K12
Ψrr̄

  rr̄ 
 
2 2
1
Ψaā  H − E0  Ψss̄

rr̄   
aā = hrr̄ k ss̄i = J22
2
 ss̄  1
aā H − E0 Ψaā = Ψaā H − E0 Ψaā = J22
Ψss̄

  rr̄ 
rr̄ 
   
2
1 1
aā H − E0 Ψaā = Ψaā H − E0 Ψaā = 2(ε2 − ε1 ) + J11 + J22 − 2J12 + K12
Ψss̄

  ss̄ 
rr̄   rr̄ 
   
2 2
1
aā | H − E0 | Ψaā i = 2(ε2 − ε1 ) + J11 + J22 − 2J12 + K12 = 2∆
hΨ∗∗ ∗∗ 0
2

6 Chapter 6
Exercise 6.3

Solution:

(2)
X0 hΨ0 | V | ni hn | V | Ψ0 i
E0 = (0) (0)
n E0 − En
X0 hΨ0 | P v(i) | ni hn | P v(i) | Ψ0 i
i i
= (0) (0)
n E0 − En
Because perturbation operator is the sum of one-particle operator, |ni and |Ψ0 i must differ with no more than
two spin orbitals. And since n can’t be 0, |ni must be single-excited determinant, which can be noted as |Ψra i.

(2)
X hΨ0 | P v(i) | Ψr i hΨr | P v(i) | Ψ0 i
i a a i
E0 =
ar
hΨ0 | H 0 | Ψ 0 i − hΨ r | H | Ψr i
a 0 a

Based on the rule of the element of one-particle operator matrix

*   +
X 
Ψ0  v(i)  Ψra = ha | v | ri
 
 
i

And the eigenvalue of H0 is the sum of spin orbital energy

(0)
X
hΨ0 | H0 | Ψ0 i = εb
b

61
 (0)
X
hΨra | H0 | Ψra i = εb + ε(0)
r
b6=a

(2)
X ha | v | ri hr | v | ai X var vra
E0 = (0) (0)
= (0) (0)
ar εa − εr ar εa − εr

Exercise 6.4

Solution:
a.

(3)
X0 | hΨ0 | V | ni |2
(1)
B0 = −E0  2
(0) (0)
n E0 − En
X X var vra
=− vaa  2
(0) (0)
a ar εa − εr
X vaa vbr vrb
=−  2
(0) (0)
abr εb − εr

b. With the same discussion stated in last exercise, |ni and |mi are single-excited determinant. We note them
as |Ψra i and |Ψsb i correspondingly.

(3)
X var vsb hΨra | H | Ψsb i
A0 =   
(0) (0) (0) (0)
abrs εa − εr εb − εs

c. Just follow the rule of evaluating element of one-particle operator matrix.

d.
(3) (3) (3)
X var vsb hΨra | V | Ψsb i X vaa vbr vrb
E0 = A0 + B 0 =   − 2
(0) (0) (0) (0)

εa − εr εb − εs (0) (0)
abrs abr ε − εr b

First, seperate the first term based on whether s is equal to r.

var vsb hΨra | V | Ψsb i v v hΨra | V | Ψrb i v v hΨra | V | Ψsb i

 ar rb   ar sb 
X X X
  = + 
(0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0)
abrs εa − εr εb − εs abr εa − εr εb − εr abrs εa − εr εb − εs
s6=r

Then for two situations b = a and b 6= a, each term can be divided into two parts.

v v hΨra | V | Ψrb i X var vra (P vcc − vaa + vrr ) X var vrb vba
 ar rb 
X
c
= 2 −   
(0) (0) (0) (0) (0) (0) (0) (0)

(0) (0)
abr εa − εr εb − εr ab εa − εr abr εa − εr εb − εr
b6=a
X vra var vcc X var vra vaa
=  2 −  2
(0) (0) (0) (0)
abc εa − εr ab εa − εr
X var vra vrr X var vrb vba
+  2 − 
(0) (0)

(0) (0)

(0) (0) ε − ε ε − ε
ab εa − εr abr a r b r
b6=a

X var vsb hΨra | V | Ψsb i X var vsb vrs

  =   
(0) (0) (0) (0) (0) (0) (0) (0)
abrs εa − εr εb − εs ars εa − εr εa − εs
s6=r s6=r

62
 " #
(3)
X var vsb vrs X var vra vrr
E0 =   + 2
(0) (0) (0) (0)

εa − εr εa − εs (0) (0)
ars
s6=r
ab εa − εr
" #
X var vrb vba X var vra vaa
−   +   
(0) (0) (0) (0) (0) (0) (0) (0)
abr εa − εr εb − εr ab εa − εr εb − εr
b6=a
" #
X var vra vcc X vaa vbr vrb
+  2 −  2
(0) (0) (0) (0)
abc εa − εr abr εb − εr
X var vsb vrs X var vrb vba
=   −   
(0) (0) (0) (0) (0) (0) (0) (0)
ars εa − εr εa − εs abr εa − εr εb − εr
X var vsb vrs X vbr var vab
=   −   
(0) (0) (0) (0) (0) (0) (0) (0)
ars εa − εr εa − εs abr εa − εr εb − εr

e. No need to say more.

Exercise 6.8

Solution:

(2) 1X | hab k rsi |2

E0 =
4 εa + εb − εr − εs
abrs
1 X (hab | rsi − hab | sri)(hrs | abi − hsr | abi)
=
4 εa + εb − εr − εs
abrs
1 X hab | rsi hrs | abi − hab | rsi hsr | abi − hab | sri hrs | abi + hab | sri hsr | abi
=
4 εa + εb − εr − εs
abrs

The second term:

X hab | rsi hsr | abi X hab | rsi hrs | bai
I2 = =
εa + εb − εr − εs εa + εb − εr − εs
abrs abrs

The third term

X hab | sri hrs | abi X hba | sri hrs | bai
I3 = =
εa + εb − εr − εs εa + εb − εr − εs
abrs abrs

Because a and b can exchange(the summation is symmetric in a and b)

X hba | sri hrs | bai X hab | rsi hrs | bai
I3 = =
εa + εb − εr − εs εa + εb − εr − εs
abrs abrs

The last term

X hab | sri hsr | abi X hab | rsi hrs | abi
I4 = =
εa + εb − εr − εs εa + εb − εr − εs
abrs abrs

Therefore we got the result

(2) 1 X hab | rsi hrs | abi 1 X hab | rsi hrs | bai

E0 = −
2 εa + εb − εr − εs 2 εa + εb − εr − εs
abrs abrs

Exercise 6.9

Solution:

63
Set D = ε2 − ε1 and X = J11 + J22 − 4J12 + 2K12
 1/2
 1  1 2 2
Ecorr = D + X − D + X + K12
2 2

Introduce perturbation on each matrix element.

  2 1/2
 1 1
Ecorr = D + λX − D + λX + λ2 K12
2
2 2
" #1/2
 1   1  λ2 K12
2
= D + λX − D + λX 1 +
2
2 2 D + 1 λX 2

Use (1 + x)1/2 = 1 + 12 x + . . .
" #1/2
 1   1  λ2 K12
2
Ecorr = D + λX − D + λX 1 +
2
2 2 D + 12 λX
" #
 1   1  1 λ2 K122
= D + λX − D + λX 1 +
2 2 2 D + 1 λX 2


2
1 λ2 K12 2
=−
2 D + 21 λX
λ2 K12
2
1
=−
2D 1 + λX
2D

Use (1 − x)−1 = 1 + x + . . .

λ2 K12
2
 
λX
Ecorr =− 1−
2D 2D
λ2 K12
2
λ3 K12
2
X
=− + 2
2D 4D
Second-order energy:
2 2
(2) K12 K12
E0 =− =
2D 2(ε1 − ε2 )
Third-order energy:
2 2
(3) K12 X K12 (J11 + J22 − 4J12 + 2K12 )
E0 = 2
=
4D 4(ε2 − ε1 )2

Exercise 6.10

Solution:
2N
(1) 1X
E0 = hΨ0 | V | Ψ0 i = − hab k abi
2
ab
N
X N 
1 X
=− h1i 1̄i k 1i 1̄i i + h1̄i 1i k 1̄i 1i i
2 i i
= −N J11
D   E D   E D   E
Ψ21ii 2̄1̄ii  H − H0  Ψ21ii 2̄1̄ii = Ψ21ii 2̄1̄ii  H  Ψ21ii 2̄1̄ii − Ψ21ii 2̄1̄ii  H0  Ψ21ii 2̄1̄ii
     

D   E
Ψ21ii 2̄1̄ii  H  Ψ21ii 2̄1̄ii = (2N − 2)h11 + 2h22 + (N − 1)J11 + J22
 

64
 D   E
Ψ21ii 2̄1̄ii  H0  Ψ21ii 2̄1̄ii = (2N − 2)ε1 + 2ε2
 

Because ε1 = h11 + J11 and ε2 = h22 + 2J12 − K12

D   E
Ψ21ii 2̄1̄ii  H − H0  Ψ21ii 2̄1̄ii = (N − 1)J11 + J22 − (2N − 2)J11 − 4J12 + 2K12 = −N J11 + J11 + J22 − 4J12 + 2K12
 

65