Second quantization

Reminder on Second Quantization

Second quantization
Second quantization and operators, two-body operator with
Morten Hjorth-Jensen1 examples
Wick’s theorem
National Superconducting Cyclotron Laboratory and Department of Physics and Thouless’ theorem and analysis of the Hartree-Fock equations
Astronomy, Michigan State University, East Lansing, MI 48824, USA1 using second quantization
Examples on how to use bit representations for Slater
Jun 27, 2017 determinants
c 2013-2017, Morten Hjorth-Jensen. Released under CC Attribution-NonCommercial 4.0 license

Second quantization Second quantization

In Eq. (1) the operator aᆠacts on the vacuum state |0i, which
We introduce the time-independent operators aα† and aα which does not contain any particles. Alternatively, we could define a
create and annihilate, respectively, a particle in the single-particle closed-shell nucleus or atom as our new vacuum, but then we need
state ϕα . We define the fermion creation operator aα† to introduce the particle-hole formalism, see the discussion to come.
In Eq. (2) aᆠacts on an antisymmetric n-particle state and creates
aα† |0i ≡ |αi, (1) an antisymmetric (n + 1)-particle state, where the one-body state
ϕα is occupied, under the condition that α 6= α1 , α2 , . . . , αn . It
and follows that we can express an antisymmetric state as the product
aα† |α1 . . . αn iAS ≡ |αα1 . . . αn iAS (2) of the creation operators acting on the vacuum state.

|α1 . . . αn iAS = aα† 1 aα† 2 . . . aα† n |0i (3)

Second quantization Second quantization

Using the Pauli principle

It is easy to derive the commutation and anticommutation rules for
the fermionic creation operators aα† . Using the antisymmetry of the |α1 . . . αi . . . αi . . . αn iAS = 0 (6)
states (3)
it follows that
|α1 . . . αi . . . αk . . . αn iAS = −|α1 . . . αk . . . αi . . . αn iAS (4) aα† i aα† i = 0. (7)
If we combine Eqs. (5) and (7), we obtain the well-known
we obtain
anti-commutation rule
aα† i aα† k = −aα† k aα† i (5)
aα† aβ† + aβ† aα† ≡ {aα† , aβ† } = 0 (8)
Second quantization Second quantization

What is the physical interpretation of the operator aα and what is

the effect of aα on a given state |α1 α2 . . . αn iAS ? Consider the
following matrix element
The hermitian conjugate of aᆠis
hα1 α2 . . . αn |aα |α10 α20 . . . αm
0
i (11)
aα = (aα† )† (9)
where both sides are antisymmetric. We distinguish between two
If we take the hermitian conjugate of Eq. (8), we arrive at cases. The first (1) is when α ∈ {αi }. Using the Pauli principle of
Eq. (6) it follows
{aα , aβ } = 0 (10)
hα1 α2 . . . αn |aα = 0 (12)
The second (2) case is when α ∈ / {αi }. It follows that an hermitian
conjugation
hα1 α2 . . . αn |aα = hαα1 α2 . . . αn | (13)

Second quantization Second quantization

For the last case, the minus and plus signs apply when the sequence
α, α1 , α2 , . . . , αn and α10 , α20 , . . . , αn+1
0 are related to each other via
Eq. (13) holds for case (1) since the lefthand side is zero due to the even and odd permutations. If we assume that α ∈ / {αi } we obtain
Pauli principle. We write Eq. (11) as
hα1 α2 . . . αn |aα |α10 α20 . . . αn+1
0
i=0 (15)
hα1 α2 . . . αn |aα |α10 α20 . . . αm
0
i = hαα1 α2 . . . αn |α10 α20 . . . αm
0
i (14)
when α ∈ {αi0 }. If α ∈
/ {αi0 }, we obtain
Here we must have m = n + 1 if Eq. (14) has to be trivially
different from zero. aα |α10 α20 . . . αn+1
0
i6=α = 0 (16)
| {z }

and in particular
aα |0i = 0 (17)

Second quantization Second quantization

If {ααi } = {αi0 }, performing the right permutations, the sequence

α, α1 , α2 , . . . , αn is identical with the sequence α10 , α20 , . . . , αn+1
0 . The action of the operator aα from the left on a state vector is to
This results in remove one particle in the state α. If the state vector does not
contain the single-particle state α, the outcome of the operation is
hα1 α2 . . . αn |aα |αα1 α2 . . . αn i = 1 (18) zero. The operator aα is normally called for a destruction or
annihilation operator.
and thus The next step is to establish the commutator algebra of aα† and aβ .
aα |αα1 α2 . . . αn i = |α1 α2 . . . αn i (19)
Second quantization Second quantization

The action of the anti-commutator {aα† ,aα } on a given n-particle If it is present we arrive at
state is
aα† aα |α1 α2 . . . αk ααk+1 . . . αn−1 i = aα† aα (−1)k |αα1 α2 . . . αn−1 i
aα† aα |α1 α2 . . . αn i = 0 = (−1)k |αα1 α2 . . . αn−1 i = |α1 α2 . . . αk ααk+1 . . . αn−1 i
| {z }
6=α
aα aα† |α1 α2 . . . αk ααk+1 . . . αn−1 i = 0 (21)
aα aα† |α1 α2 . . . αn i = aα |αα1 α2 . . . αn i = |α1 α2 . . . αn i (20)
| {z } | {z } | {z } From Eqs. (20) and (21) we arrive at
6=α 6=α 6=α

if the single-particle state α is not contained in the state. {aα† , aα } = aα† aα + aα aα† = 1 (22)

Second quantization Second quantization

The first case results in

aα† aβ |αβα1 α2 . . . αn−2 i = 0

aβ aα† |αβα1 α2 . . . αn−2 i = 0 (23)
n o
The action of aα† , aβ , with α 6= β on a given state yields three while the second case gives
possibilities. The first case is a state vector which contains both α
and β, then either α or β and finally none of them. aα† aβ |β α1 α2 . . . αn−1 i = |α α1 α2 . . . αn−1 i
| {z } | {z }
6=α 6=α

aβ aα† |β α1 α2 . . . αn−1 i = aβ |α βα1 α2 . . . αn−1 i

| {z } | {z }
6=α 6=α
= −|α α1 α2 . . . αn−1 i (24)
| {z }
6=α

Second quantization Second quantization

We can summarize our findings in Eqs. (22) and (26) as

Finally if the state vector does not contain α and β {aα† , aβ } = δαβ (27)

aα† aβ | α1 α2 . . . αn i = 0 with δαβ is the Kronecker δ-symbol.

| {z }
6=α,β The properties of the creation and annihilation operators can be
summarized as (for fermions)
aβ aα† | α1 α2 . . . αn i = aβ |α α1 α2 . . . αn i = 0 (25)
| {z } | {z }
6=α,β 6=α,β aα† |0i ≡ |αi,
For all three cases we have and
aα† |α1 . . . αn iAS ≡ |αα1 . . . αn iAS .
{aα† , aβ } = aα† aβ + aβ aα† = 0, α 6= β (26)
from which follows

|α1 . . . αn iAS = aα† 1 aα† 2 . . . aα† n |0i.

Second quantization Second quantization

The hermitian conjugate has the folowing properties

A very useful operator is the so-called number-operator. Most
aα = (aα† )† .
physics cases we will study in this text conserve the total number of
Finally we found particles. The number operator is therefore a useful quantity which
allows us to test that our many-body formalism conserves the
aα |α10 α20 . . . αn+1
0
i6=α = 0, in particular aα |0i = 0, number of particles. In for example (d, p) or (p, d) reactions it is
| {z } important to be able to describe quantum mechanical states where
particles get added or removed. A creation operator aᆠadds one
and
particle to the single-particle state α of a give many-body state
aα |αα1 α2 . . . αn i = |α1 α2 . . . αn i,
vector, while an annihilation operator aα removes a particle from a
and the corresponding commutator algebra single-particle state α.

{aα† , aβ† } = {aα , aβ } = 0 {aα† , aβ } = δαβ .

Second quantization Second quantization

Let us consider an operator proportional with aα† aβ and α = β. It

acts on an n-particle state resulting in The operator X
 N̂ = aα† aα (30)
0
 α∈/ {αi } α
†
aα aα |α1 α2 . . . αn i = (28) is called the number operator since it counts the number of


|α1 α2 . . . αn i α ∈ {αi } particles in a given state vector when it acts on the different
single-particle states. It acts on one single-particle state at the time
Summing over all possible one-particle states we arrive at and falls therefore under category one-body operators. Next we
! look at another important one-body operator, namely Ĥ0 and study
X its operator form in the occupation number representation.
aα† aα |α1 α2 . . . αn i = n|α1 α2 . . . αn i (29)
α

Second quantization Second quantization

Defining X
ĥ0 (xi )ψαi (xi ) = ψα0k (xi )hαk0 |ĥ0 |αk i (32)
We want to obtain an expression for a one-body operator which α0k
conserves the number of particles. Here we study the one-body
operator for the kinetic energy plus an eventual external one-body we can easily evaluate the action of Ĥ0 on each product of
potential. The action of this operator on a particular n-body state one-particle functions in Slater determinant. From Eq. (32) we
with its pertinent expectation value has already been studied in obtain the following result without permuting any particle pair
coordinate space. In coordinate space the operator reads !
X
X ĥ0 (xi ) ψα1 (x1 )ψα2 (x2 ) . . . ψαn (xn )
Ĥ0 = ĥ0 (xi ) (31)
i
i X
= hα10 |ĥ0 |α1 iψα01 (x1 )ψα2 (x2 ) . . . ψαn (xn )
and the anti-symmetric n-particle Slater determinant is defined as α01
X
1 X + hα20 |ĥ0 |α2 iψα1 (x1 )ψα02 (x2 ) . . . ψαn (xn )
Φ(x1 , x2 , . . . , xn , α1 , α2 , . . . , αn ) = √ (−1)p P̂ψα1 (x1 )ψα2 (x2 ) . . . ψαn (xn ).
n! p α02
+ ...
X
+ hαn0 |ĥ0 |αn iψα1 (x1 )ψα2 (x2 ) . . . ψα0n (xn ) (33)
α0n
Second quantization Second quantization

If we interchange particles 1 and 2 we obtain We can continue by computing all possible permutations. We
! rewrite also our Slater determinant in its second quantized form
X and skip the dependence on the quantum numbers xi . Summing up
ĥ0 (xi ) ψα1 (x1 )ψα1 (x2 ) . . . ψαn (xn )
all contributions and taking care of all phases (−1)p we arrive at
i
X X
= hα20 |ĥ0 |α2 iψα1 (x2 )ψα02 (x1 ) . . . ψαn (xn ) Ĥ0 |α1 , α2 , . . . , αn i = hα10 |ĥ0 |α1 i|α10 α2 . . . αn i
α02 α01
X X
+ hα10 |ĥ0 |α1 iψα01 (x2 )ψα2 (x1 ) . . . ψαn (xn ) + hα20 |ĥ0 |α2 i|α1 α20 . . . αn i
α01 α02
+ ... + ...
X X
+ hαn0 |ĥ0 |αn iψα1 (x2 )ψα1 (x2 ) . . . ψα0n (xn ) (34) + hαn0 |ĥ0 |αn i|α1 α2 . . . αn0 i (35)
α0n α0n

Second quantization Second quantization

In Eq. (35) we have expressed the action of the one-body operator
of Eq. (31) on the n-body state in its second quantized form. This
equation can be further manipulated if we use the properties of the In the number occupation representation or second quantization we
creation and annihilation operator on each primed quantum get the following expression for a one-body operator which
number, that is conserves the number of particles
X
|α1 α2 . . . αk0 . . . αn i = aα† 0 aαk |α1 α2 . . . αk . . . αn i (36) Ĥ0 = hα|ĥ0 |βiaα† aβ (38)
k
αβ

Inserting this in the right-hand side of Eq. (35) results in

Obviously, Ĥ0 can be replaced by any other one-body operator
X
Ĥ0 |α1 α2 . . . αn i = hα10 |ĥ0 |α1 iaα† 0 aα1 |α1 α2 . . . αn i which preserved the number of particles. The stucture of the
α01
1 operator is therefore not limited to say the kinetic or single-particle
X energy only.
+ hα20 |ĥ0 |α2 iaα† 0 aα2 |α1 α2 . . . αn i The opearator Ĥ0 takes a particle from the single-particle state β
2
α02 to the single-particle state α with a probability for the transition
+ ... given by the expectation value hα|ĥ0 |βi.
X
+ hαn0 |ĥ0 |αn iaα† 0 aαn |α1 α2 . . . αn i
n
α0n
X
= hα|ĥ0 |βiaα† aβ |α1 α2 . . . αn i (37)
α,β

Second quantization Second quantization

Using the commutation relations for the creation and annihilation

operators we have
It is instructive to verify Eq. (38) by computing the expectation aα1 aα† aβ aα† 2 = (δαα1 − aα† aα1 )(δβα2 − aα† 2 aβ ), (40)
value of Ĥ0 between two single-particle states
X which results in
hα1 |ĥ0 |α2 i = hα|ĥ0 |βih0|aα1 aα† aβ aα† 2 |0i (39)
αβ h0|aα1 aα† aβ aα† 2 |0i = δαα1 δβα2 (41)

and X
hα1 |ĥ0 |α2 i = hα|ĥ0 |βiδαα1 δβα2 = hα1 |ĥ0 |α2 i (42)
αβ
Two-body operators in second quantization Operators in second quantization
We can now let ĤI act on all terms in the linear combination for
Let us now derive the expression for our two-body interaction part,
|α1 α2 . . . αn i. Without any permutations we have
which also conserves the number of particles. We can proceed in
exactly the same way as for the one-body operator. In the  
X
coordinate representation our two-body interaction part takes the  V (xi , xj ) ψα1 (x1 )ψα2 (x2 ) . . . ψαn (xn )
following expression X i<j
ĤI = V (xi , xj ) (43) X
= hα10 α20 |v̂ |α1 α2 iψα0 1 (x1 )ψα0 2 (x2 ) . . . ψαn (xn )
i<j
α01 α02
where the summation runs over distinct pairs. The term V can be + ...
an interaction model for the nucleon-nucleon interaction or the X
+ hα10 αn0 |v̂ |α1 αn iψα0 1 (x1 )ψα2 (x2 ) . . . ψα0 n (xn )
interaction between two electrons. It can also include additional
α01 α0n
two-body interaction terms.
+ ...
The action of this operator on a product of two single-particle X
functions is defined as + hα20 αn0 |v̂ |α2 αn iψα1 (x1 )ψα0 2 (x2 ) . . . ψα0 n (xn )
X α02 α0n
V (xi , xj )ψαk (xi )ψαl (xj ) = ψα0 k (xi )ψα0 l (xj )hαk0 αl0 |v̂ |αk αl i (44) + ... (45)
α0k α0l
where on the rhs we have a term for each distinct pairs.

Operators in second quantization Operators in second quantization

For the other terms on the rhs we obtain similar expressions and
summing over all terms we obtain We introduce second quantization via the relation
X
HI |α1 α2 . . . αn i = hα10 α20 |v̂ |α1 α2 i|α10 α20 . . . αn i aα† 0 aα† 0 aαl aαk |α1 α2 . . . αk . . . αl . . . αn i
k l
α01 ,α02
+ ... = (−1)k−1 (−1)l−2 aα† 0 aα† 0 aαl aαk |αk αl α1 α2 . . . αn i
X k l | {z }
+ hα10 αn0 |v̂ |α1 αn i|α10 α2 . . . αn0 i 6=αk ,αl

α01 ,α0n = (−1)k−1 (−1)l−2 |αk0 αl0 α1 α2 . . . αn i

| {z }
+ ... 6=α0k ,α0l
X
+ hα20 αn0 |v̂ |α2 αn i|α1 α20 . . . αn0 i = |α1 α2 . . . αk0 . . . αl0 . . . αn i (47)
α02 ,α0n
+ ... (46)

Operators in second quantization Operators in second quantization

P
Inserting this in (46) gives Here we let 0 indicate that the sums running over α and β run
X over all single-particle states, while the summations γ and δ run
HI |α1 α2 . . . αn i = hα10 α20 |v̂ |α1 α2 iaα† 0 aα† 0 aα2 aα1 |α1 α2 . . . αn i over all pairs of single-particle states. We wish to remove this
1 2
α01 ,α02 restriction and since
+ ...
X
= hα10 αn0 |v̂ |α1 αn iaα† 0 aα† 0 aαn aα1 |α1 α2 . . . αn i hαβ|v̂ |γδi = hβα|v̂ |δγi (49)
1 n
α01 ,α0n
we get
+ ...
X X X
= hα20 αn0 |v̂ |α2 αn iaα† 0 aα† 0 aαn aα2 |α1 α2 . . . αn i hαβ|v̂ |γδiaα† aβ† aδ aγ = hβα|v̂ |δγiaα† aβ† aδ aγ (50)
2 n
α02 ,α0n αβ αβ
X
+ ... = hβα|v̂ |δγiaβ† aα† aγ aδ (51)
X 0
αβ
= hαβ|v̂ |γδiaα† aβ† aδ aγ |α1 α2 . . . αn i (48)
α,β,γ,δ where we have used the anti-commutation rules.
Operators in second quantization Operators in second quantization

Changing the summation indices α and β in (51) we obtain

X X With this expression we can now verify that the second
hαβ|v̂ |γδiaα† aβ† aδ aγ = hαβ|v̂ |δγiaα† aβ† aγ aδ (52)
quantization form of ĤI in Eq. (53) results in the same matrix
αβ αβ
between two anti-symmetrized two-particle states as its
From this it follows that the restriction on the summation over γ corresponding coordinate space representation. We have
and δ can be removed if we multiply with a factor 12 , resulting in 1 X
hα1 α2 |ĤI |β1 β2 i = hαβ|v̂ |γδih0|aα2 aα1 aα† aβ† aδ aγ aβ† 1 aβ† 2 |0i.
1 X 2
ĤI = hαβ|v̂ |γδiaα† aβ† aδ aγ (53) αβγδ
2 (54)
αβγδ

where we sum freely over all single-particle states α, β, γ og δ.

Operators in second quantization Operators in second quantization

Using the commutation relations we get

The vacuum expectation value of this product of operators becomes
aα2 aα1 aα† aβ† aδ aγ aβ† 1 aβ† 2
= aα2 aα1 aα† aβ† (aδ δγβ1 aβ† 2 − aδ aβ† 1 aγ aβ† 2 ) h0|aα2 aα1 aα† aβ† aδ aγ aβ† 1 aβ† 2 |0i

= aα2 aα1 aα† aβ† (δγβ1 δδβ2 − δγβ1 aβ† 2 aδ − aδ aβ† 1 δγβ2 + aδ aβ† 1 aβ† 2 aγ ) = (δγβ1 δδβ2 − δδβ1 δγβ2 )h0|aα2 aα1 aα† aβ† |0i

= aα2 aα1 aα† aβ† (δγβ1 δδβ2 − δγβ1 aβ† 2 aδ = (δγβ1 δδβ2 − δδβ1 δγβ2 )(δαα1 δβα2 − δβα1 δαα2 ) (56)

−δδβ1 δγβ2 + δγβ2 aβ† 1 aδ + aδ aβ† 1 aβ† 2 aγ ) (55)

Operators in second quantization Operators in second quantization

The two-body operator can also be expressed in terms of the

Insertion of Eq. (56) in Eq. (54) results in anti-symmetrized matrix elements we discussed previously as

1 1 X
hα1 α2 |ĤI |β1 β2 i = hα1 α2 |v̂ |β1 β2 i − hα1 α2 |v̂ |β2 β1 i ĤI = hαβ|v̂ |γδiaα† aβ† aδ aγ
2 2
 αβγδ
−hα2 α1 |v̂ |β1 β2 i + hα2 α1 |v̂ |β2 β1 i 1 X
= [hαβ|v̂ |γδi − hαβ|v̂ |δγi] aα† aβ† aδ aγ
= hα1 α2 |v̂ |β1 β2 i − hα1 α2 |v̂ |β2 β1 i 4
αβγδ
= hα1 α2 |v̂ |β1 β2 iAS . (57) 1 X
= hαβ|v̂ |γδiAS aα† aβ† aδ aγ (58)
4
αβγδ
Operators in second quantization Particle-hole formalism
Second quantization is a useful and elegant formalism for
constructing many-body states and quantum mechanical operators.
One can express and translate many physical processes into simple
The factors in front of the operator, either 14 or 12 tells whether we pictures such as Feynman diagrams. Expecation values of
use antisymmetrized matrix elements or not. many-body states are also easily calculated. However, although the
We can now express the Hamiltonian operator for a many-fermion equations are seemingly easy to set up, from a practical point of
system in the occupation basis representation as view, that is the solution of Schroedinger’s equation, there is no
particular gain. The many-body equation is equally hard to solve,
X 1 X irrespective of representation. The cliche that there is no free lunch
H= hα|t̂ + ûext |βiaα† aβ + hαβ|v̂ |γδiaα† aβ† aδ aγ . (59)
4 brings us down to earth again. Note however that a transformation
α,β αβγδ
to a particular basis, for cases where the interaction obeys specific
This is the form we will use in the rest of these lectures, assuming symmetries, can ease the solution of Schroedinger’s equation.
that we work with anti-symmetrized two-body matrix elements. But there is at least one important case where second quantization
comes to our rescue. It is namely easy to introduce another
reference state than the pure vacuum |0i, where all single-particle
states are active. With many particles present it is often useful to
introduce another reference state than the vacuum state|0i. We
will label this state |ci (c for core) and as we will see it can reduce
considerably the complexity and thereby the dimensionality of the
many-body problem. It allows us to sum up to infinite order specific
many-body correlations. The particle-hole representation is one of
these handy representations.
Particle-hole formalism Particle-hole formalism

If we use Eq. (60) as our new reference state, we can simplify

In the original particle representation these states are products of considerably the representation of this state
the creation operators aᆠi acting on the true vacuum |0i. Following
Eq. (3) we have |ci ≡ |α1 α2 . . . αn−1 αn i = aα† 1 aα† 2 . . . aα† n−1 aα† n |0i (63)

|α1 α2 . . . αn−1 αn i = aα† 1 aα† 2 . . . aα† n−1 aα† n |0i (60) The new reference states for the n + 1 and n − 1 states can then be
|α1 α2 . . . αn−1 αn αn+1 i = aα† 1 aα† 2 . . . aα† n−1 aα† n aα† n+1 |0i (61) written as

|α1 α2 . . . αn−1 i = aα† 1 aα† 2 . . . aα† n−1 |0i (62) |α1 α2 . . . αn−1 αn αn+1 i = (−1)n aα† n+1 |ci ≡ (−1)n |αn+1 ic (64)
|α1 α2 . . . αn−1 i = (−1)n−1 aαn |ci ≡ (−1)n−1 |αn−1(65)
ic

Particle-hole formalism Particle-hole formalism

The first state has one additional particle with respect to the new The physical interpretation of these new operators is that of
vacuum state |ci and is normally referred to as a one-particle state so-called quasiparticle states. This means that a state defined by
or one particle added to the many-body reference state. The the addition of one extra particle to a reference state |ci may not
second state has one particle less than the reference vacuum state necesseraly be interpreted as one particle coupled to a core. We
|ci and is referred to as a one-hole state. When dealing with a new define now new creation operators that act on a state α creating a
reference state it is often convenient to introduce new creation and new quasiparticle state
annihilation operators since we have from Eq. (65)
(
aα† |ci = |αi, α>F
6 0
aα |ci = (66) bα† |ci = (70)
aα |ci = |α−1 i, α ≤ F
since α is contained in |ci, while for the true vacuum we have
aα |0i = 0 for all α. where F is the Fermi level representing the last occupied
The new reference state leads to the definition of new creation and single-particle orbit of the new reference state |ci.
annihilation operators which satisfy the following relations The annihilation is the hermitian conjugate of the creation operator

bα |ci = 0 (67) bα = (bα† )† ,

{bα† , bβ† } = {bα , bβ } = 0
resulting in
{bα† , bβ } = δαβ (68)
( ( a α>F
aα† α > F α
We assume also that the new reference state is properly normalized bα† = bα = (71)
aα α ≤ F aα† α ≤ F
hc|ci = 1 (69)
Particle-hole formalism Particle-hole formalism
With the new creation and annihilation operator we can now
construct many-body quasiparticle states, with
one-particle-one-hole states, two-particle-two-hole states etc in the We express the one-body operator Ĥ0 in terms of the quasi-particle
same fashion as we previously constructed many-particle states. We creation and annihilation operators, resulting in
can write a general particle-hole state as X X h i
Ĥ0 = hα|ĥ0 |βibα† bβ + hα|ĥ0 |βibα† bβ† + hβ|ĥ0 |αibβ bα
|β1 β2 . . . βnp γ1−1 γ2−1 . . . γn−1
h
i ≡ bβ† 1 bβ† 2 . . . bβ† n bγ†1 bγ†2 . . . bγ†n |ci αβ>F α>F
p
| {z }| {z }h
>F ≤F β≤F
(72) X X
+ hα|ĥ0 |αi − hβ|ĥ0 |αibα† bβ (77)
We can now rewrite our one-body and two-body operators in terms
α≤F αβ≤F
of the new creation and annihilation operators. The number
operator becomes The first term gives contribution only for particle states, while the
X X X last one contributes only for holestates. The second term can
N̂ = aα† aα = bα† bα + nc − bα† bα (73) create or destroy a set of quasi-particles and the third term is the
α α>F α≤F
contribution from the vacuum state |ci.
where nc is the number of particle in the new vacuum state |ci.
The action of N̂ on a many-body state results in

N|β1 β2 . . . βnp γ1−1 γ2−1 . . . γn−1

h
i = (np +nc −nh )|β1 β2 . . . βnp γ1−1 γ2−1 . . . γn−1
h
i
(74)
Here n = np + nc − nh is the total number of particles in the
quasi-particle state of Eq. (72). Note that N̂ counts the total
Particle-hole formalism
number of particles present Particle-hole formalism
X
Nqp = bα† bα , (75)
α
Before we continue with the expressions for the two-body operator, The two-particle operator in the particle-hole formalism is more
gives us the number
we introduce of quasi-particles
a nomenclature as for
we will use canthebe rest
seen of
bythis
computing
text. It complicated since we have to translate four indices αβγδ to the
is inspired by the notation
−1 −1used in−1quantum chemistry. We reserve possible combinations of particle and hole states. When performing
Nqp = |β1 β2 . . . βnp γ1 γ2 . . . γnh i = (np +nh )|β1 β2 . . . βnp γ1−1 γ2−1 . . . γn−1 i
the labels i, j, k, . . . for hole states and a, b, c, . . . for states above h the commutator algebra we can regroup the operator in five
(76)
F , viz. particle states. This means also that we will skip the different terms
where nqp = np + nh is the total number of quasi-particles.
constraint ≤ F or > F in the summation symbols. Our operator Ĥ0
(a) (b) (c) (d) (e)
reads now ĤI = ĤI + ĤI + ĤI + ĤI + ĤI (79)
X Xh i
Ĥ0 = ha|ĥ|biba† bb + ha|ĥ|iiba† bi† + hi|ĥ|aibi ba Using anti-symmetrized matrix elements, bthe term ĤI
(a)
is
ab ai
X X 1X
+ hi|ĥ|ii − hj|ĥ|iibi† bj (78) ĤI
(a)
= hab|V̂ |cdiba† bb† bd bc (80)
i ij
4
abcd

Particle-hole formalism Particle-hole formalism

The first line stands for the creation of a two-particle-two-hole
state, while the second line represents the creation to two
The next term ĤI
(b)
reads one-particle-one-hole pairs while the last term represents a
contribution to the particle single-particle energy from the hole
(b) 1 X  states, that is an interaction between the particle states and the
ĤI = hab|V̂ |ciiba† bb† bi† bc + hai|V̂ |cbiba† bi bb bc (81)
4 hole states within the new vacuum state. The fourth term reads
abci
(d) 1 X 
This term conserves the number of quasiparticles but creates or ĤI = hai|V̂ |jkiba† bk† bj† bi + hji|V̂ |akibk† bj bi ba +
(c) 4
removes a three-particle-one-hole state. For ĤI we have aijk
1 X 
1 X  hai|V̂ |jiiba† bj† + hji|V̂ |aii − hji|V̂ |iaibj ba (. 83)
ĤI
(c)
= hab|V̂ |ijiba† bb† bj† bi† + hij|V̂ |abiba bb bj bi + 4
aij
4
abij
1X 1X The terms in the first line stand for the creation of a particle-hole
hai|V̂ |bjiba† bj† bb bi + hai|V̂ |biiba† bb . (82) state interacting with hole states, we will label this as a
2 2
abij abi two-hole-one-particle contribution. The remaining terms are a
particle-hole state interacting with the holes in the vacuum state.
Finally we have

(e) 1X 1X 1X
ĤI = hkl|V̂ |ijibi† bj† bl bk + hij|V̂ |kjibk† bi + hij|V̂ |iji
4 2 2
ijkl ijk ij
(84)
Summarizing and defining a normal-ordered Hamiltonian Summarizing and defining a normal-ordered Hamiltonian

A
1 X Y n o
ΦAS (α1 , . . . , αA ; x1 , . . . xA ) = √ (−1)P P̂ ψαi (xi ), ap† , aq = δpq , p, q ≤ αF
A i=1
P̂ n o
ap , aq† = δpq , p, q > αF
which is equivalent with |α1 . . . αA i = aα† 1 . . . aα† A |0i. We have also
with i, j, . . . ≤ αF , a, b, . . . > αF , p, q, . . . − any
ap† |0i = |pi, ap |qi = δpq |0i
n o ai |Φ0 i = |Φi i, aa† |Φ0 i = |Φa i
δpq = ap , aq† ,
and
and n o n o ai† |Φ0 i = 0 aa |Φ0 i = 0
0 = ap† , aq = {ap , aq } = ap† , aq†

|Φ0 i = |α1 . . . αA i, α 1 , . . . , α A ≤ αF

Summarizing and defining a normal-ordered Hamiltonian Summarizing and defining a normal-ordered Hamiltonian

One- and two-body operators

The one-body operator is defined as
X We can also define a three-body operator
F̂ = hp|fˆ|qiap† aq
1 X
pq V̂3 = hpqr |v̂3 |stuiAS ap† aq† ar† au at as
36 pqrstu
while the two-body opreator is defined as
with the antisymmetrized matrix element
1X
V̂ = hpq|v̂ |rsiAS ap† aq† as ar
4 pqrs hpqr |v̂3 |stuiAS = hpqr |v̂3 |stui + hpqr |v̂3 |tusi + hpqr |v̂3 |usti − hpqr |v̂3 |suti − hpqr |v̂3 |tsui − hpqr |v̂3 |utsi.
(85)
where we have defined the antisymmetric matrix elements

hpq|v̂ |rsiAS = hpq|v̂ |rsi − hpq|v̂ |sr i.

Hartree-Fock in second quantization and stability of HF Hartree-Fock in second quantization and Thouless’ theorem
solution
We wish now to derive the Hartree-Fock equations using our
second-quantized formalism and study the stability of the Let us give a simple proof of Thouless’ theorem. The theorem
equations. Our ansatz for the ground state of the system is states that we can make a linear combination av particle-hole
approximated as (this is our representation of a Slater determinant excitations with respect to a given reference state |ci. With this
in second quantization) linear combination, we can make a new Slater determinant |c 0 i
which is not orthogonal to |ci, that is
|Φ0 i = |ci = ai† aj† . . . al† |0i.
hc|c 0 i =
6 0.
We wish to determine û HF so that E0HF = hc|Ĥ|ci becomes a local
minimum. To show this we need some intermediate steps. The exponential
In our analysis here we will need Thouless’ theorem, which states product of two operators exp Â × exp B̂ is equal to exp ( + B̂)
that an arbitrary Slater determinant |c 0 i which is not orthogonal to only if the two operators commute, that is
Yn
a determinant |ci = aα† i |0i, can be written as [Â, B̂] = 0.
i=1
 
X X 
|c 0 i = exp Cai aa† ai |ci
 
a>F i≤F
Thouless’ theorem Thouless’ theorem
If the operators do not commute, we need to resort to the We note that
Baker-Campbell-Hauersdorf. This relation states that YX X
Cai aa† ai Cbi ab† ai |ci = 0,
exp Ĉ = exp  exp B̂, i a>F b>F

with and all higher-order powers of these combinations of creation and

annihilation operators disappear due to the fact that (ai )n |ci = 0
1 1 1 when n > 1. This allows us to rewrite the expression for |c 0 i as
Ĉ = Â + B̂ + [Â, B̂] + [[Â, B̂], B̂] − [[Â, B̂], Â] + . . .
2 12 12 ( )
Y X
From these relations, we note that in our expression for |c 0 i we |c 0 i = 1+ Cai aa† ai |ci,
have commutators of the type i a>F

[aa† ai , ab† aj ], which we can rewrite as

( )
and it is easy to convince oneself that these commutators, or higher Y X
|c 0 i = 1+ Cai aa† ai |ai†1 ai†2 . . . ai†n |0i.
powers thereof, are all zero. This means that we can write out our
i a>F
new representation of a Slater determinant as
   !2  The last equation can be written as
X X  Y X X  ( ) !
|c 0 i = exp Cai aa† ai |ci = 1+ Cai aa† ai + Cai aa† ai + . . . |ci Y X X
    |c 0 i = 1+ Cai aa† ai |ai†1 ai†2 . . . ai†n |0i = 1+ Cai1 aa† ai1 ai†1
a>F i≤F i a>F a>F
i a>F a>F
(86)
! !
0
New operators Showing thatX
× 1+
|c̃iC =a†|c i a† Y
ai†
X
Cai aa†
a
ai2 a i2 i2 . . . |0i = + |0i.
If we define a new creation operator a>F that |c̃i = |c 0 i. We need
We need to show i also toa>F assume that the
X new state is not orthogonal to |ci, that is hc|c̃i = 6 0. From this(87) it
bi† = ai† + Cai aa† , (88) follows that
a>F   
X in in
X
we have hc|c̃i = h0|ain . . . ai1  fi1 p ap†   fin t at†  |0i,
! p=i1 t=i1
Y Y X
|c 0 i = bi† |0i = ai† + Cai aa† |0i,
which is nothing but the determinant det(fip ) which we can, using
i i a>F
the intermediate normalization condition, normalize to one, that is
meaning that the new representation of the Slater determinant in
second quantization, |c 0 i, looks like our previous ones. However, det(fip ) = 1,
this representation is not general enough since we have a restriction
meaning that f has an inverse defined as (since we are dealing with
on the sum over single-particle states in Eq. (88). The
orthogonal, and in our case unitary as well, transformations)
single-particle states have all to be above the Fermi level. The
question then is whether we can construct a general representation X
fik fkj−1 = δij ,
of a Slater determinant with a creation operator
k
X
b̃i† = fip ap† , and X
p fij−1 fjk = δik .
j
where fip is a matrix element of a unitary matrix which transforms
our creation and annihilation operators a† and a to b̃ † and b̃.
These newitoperators
Wrapping up define a new representation of a Slater Thouless’ theorem
determinant as Y †
Using these relations we can
|c̃i then
= define
b̃i |0i.the linear combination of This means that we can actually write an ansatz for the ground
creation (and annihilation as well)i operators as state of the system as a linear combination of terms which contain
the ansatz itself |ci with an admixture from an infinity of
X X ∞
X ∞
X X one-particle-one-hole states. The latter has important consequences
fki−1 b̃i† = fki−1 fip ap† = ak† + fki−1 fip ap† .
when we wish to interpret the Hartree-Fock equations and their
i i p=i1 i p=in+1
stability. We can rewrite the new representation as
Defining X
ckp = fki−1 fip , |c 0 i = |ci + |δci,
i≤F
where |δci can now be interpreted as a small variation. If we
we can redefine approximate this term with contributions from one-particle-one-hole
∞
X X ∞
X (1p-1h) states only, we arrive at
ak† + fki−1 fip ap† = ak† + ckp ap† = bk† , !
i p=in+1 p=in+1
X
|c 0 i = 1 + δCai aa† ai |ci.
our starting point. We have shown that our general representation ai

of a Slater determinant In our derivation of the Hartree-Fock equations we have shown that
Y † Y †
|c̃i = b̃i |0i = |c 0 i = bi |0i, hδc|Ĥ|ci = 0,
i i
which means that we have to satisfy
with
∞
X n o
† † †
X
†
Hartree-Fock in second quantization and stability of HF Hartree-Fock in second quantization and stability of HF
solution solution
The variational condition for deriving the Hartree-Fock equations
guarantees only that the expectation value hc|Ĥ|ci has an extreme The norm of |c 0 i is given by (using the intermediate normalization
value, not necessarily a minimum. To figure out whether the condition hc 0 |ci = 1)
extreme value we have found is a minimum, we can use second
XX
quantization to analyze our results and find a criterion for the hc 0 |c 0 i = 1 + |δCai |2 + O(δCai3 ).
above expectation value to a local minimum. We will use Thouless’ a>F i≤F
theorem and show that
The expectation value for the energy is now given by (using the
hc 0 |Ĥ|c 0 i Hartree-Fock condition)
≥ hc|Ĥ|ci = E0 ,
hc 0 |c 0 i X X
hc 0 |Ĥ|c 0 i = hc|Ĥ|ci + δCai∗ δCbj hc|ai† aa Ĥab† aj |ci+
with ab>F ij≤F
|c 0 i = |ci + |δci.
1 X X 1 X X ∗ ∗
Using Thouless’ theorem we can write out as |c 0 i δCai δCbj hc|Ĥaa† ai ab† aj |ci+ δCai δCbj hc|aj† ab ai† aa Ĥ|ci+. . .
2! 2!
  ab>F ij≤F ab>F ij≤F
X X 
|c 0 i = exp δCai aa† ai |ci (89)
 
a>F i≤F
 
 XX 1 X X 
† † †
= 1+ δCai aa ai + δCai δCbj aa ai ab aj + . . .
 2! 
a>F i≤F ab>F ij≤F
Hartree-Fock in second quantization and stability of HF
(90) Hartree-Fock in second quantization and stability of HF
solution
where the amplitudes δC are small.
solution

We have already calculated the second term on the right-hand side 1   1  †

of the previous equation hc| {aj† ab }{ai† aa }V̂N |ci = hc| V̂N {aa† ai }{ab† aj } |ci
2! 2!
  XX  
hc| {ai† aa }Ĥ{ab† aj } |ci = δCai∗ δCbj hp|ĥ0 |qihc| {ai† aa }{ap† aq }{ab† aj } |ci which is nothing but
pq ijab
1   1X
(91) hc| V̂N {aa† ai }{ab† aj } |ci∗ = (hij|v̂ |abi)∗ δCai∗ δCbj∗
  2! 2
1 XX ∗ ijab
+ δCai δCbj hpq|v̂ |rsihc| {ai† aa }{ap† aq† as ar }{ab† aj } |ci,
4 pqrs or
ijab 1X
(92) (hab|v̂ |iji)δCai∗ δCbj∗
2
ijab
resulting in where we have used the relation
X X X
E0 |δCai |2 + |δCai |2 (εa − εi ) − haj|v̂ |biiδCai∗ δCbj . ha|Â|bi = (hb|† |ai)∗
ai ai ijab
due to the hermiticity of Ĥ and V̂ .

Hartree-Fock in second quantization and stability of HF Hartree-Fock in second quantization and stability of HF
solution solution
With these definitions we write out the energy as
!
X X X
hc 0 |H|c 0 i = 1 + |δCai |2 hc|H|ci + |δCai |2 (εHF HF
a − εi ) + Aai,bj δCai∗ δCbj +
ai ai ijab
We define two matrix elements
(93)
Aai,bj = −haj|v̂ bii 1X ∗ 1X
Bai,bj δCai δCbj + Bai,bj δCai∗ δCbj∗ + O(δCai3 ),
2 2
ijab ijab
and
(94)
Bai,bj = hab|v̂ |iji
both being anti-symmetrized. which can be rewritten as
!
X
hc 0 |H|c 0 i = 1+ |δCai |2 hc|H|ci + ∆E + O(δCai3 ),
ai

and skipping higher-order terms we arrived

hc 0 |Ĥ|c 0 i ∆E
= E0 + P .
hc 0 |c 0 i (1 + ai |δCai |2 )
Hartree-Fock in second quantization and stability of HF Hartree-Fock in second quantization and stability of HF
solution solution

The condition
1
We have defined ∆E = hχ|M̂|χi ≥ 0
1 2
∆E = hχ|M̂|χi for an arbitrary vector
2
with the vectors
χ = [δC δC ∗ ]T
χ = [δC δC ∗ ]T
and the matrix means that all eigenvalues of the matrix have to be larger than or
  equal zero. A necessary (but no sufficient) condition is that the
∆+A B matrix elements (for all ai )
M̂ = ,
B∗ ∆ + A∗
(εa − εi )δab δij + Aai,bj ≥ 0.
with ∆ai,bj = (εa − εi )δab δij .
This equation can be used as a first test of the stability of the
Hartree-Fock equation.

Operators in second quantization Operators in second quantization

Assume that we have at our disposal n different single-particle
orbits α0 , α2 , . . . , αn−1 and that we can distribute among these
In the build-up of a shell-model or FCI code that is meant to tackle orbits N ≤ n particles.
large dimensionalities is the action of the Hamiltonian Ĥ on a A Slater determinant can then be coded as an integer of n bits. As
Slater determinant represented in second quantization as an example, if we have n = 16 single-particle states α0 , α1 , . . . , α15
and N = 4 fermions occupying the states α3 , α6 , α10 and α13 we
|α1 . . . αn i = aα† 1 aα† 2 . . . aα† n |0i.
could write this Slater determinant as
The time consuming part stems from the action of the Hamiltonian
ΦΛ = aα† 3 aα† 6 aα† 10 aα† 13 |0i.
on the above determinant,
  The unoccupied single-particle states have bit value 0 while the
X X
 hα|t + u|βiaα† aβ + 1 † †
hαβ|v̂ |γδiaα aβ aδ aγ  aα† 1 aα† 2 . . . aα† n |0i. occupied ones are represented by bit state 1. In the binary notation
4 we would write this 16 bits long integer as
αβ αβγδ

A practically useful way to implement this action is to encode a α0 α1 α2 α3 α4 α5 α6 α7 α8 α9 α10 α11 α12 α13 α14 α15
Slater determinant as a bit pattern. 0 0 0 1 0 0 1 0 0 0 1 0 0 1 0 0

which translates into the decimal number

23 + 26 + 210 + 213 = 9288.

We can thus encode a Slater determinant as a bit pattern.

Operators in second quantization Operators in second quantization

We assume again that we have at our disposal n different
single-particle orbits α0 , α2 , . . . , αn−1 and that we can distribute
among these orbits N ≤ n particles. The ordering among these
states is important as it defines the order of the creation operators.
With N particles that can be distributed over n single-particle We will write the determinant
states, the total number of Slater determinats (and defining thereby
the dimensionality of the system) is ΦΛ = aα† 3 aα† 6 aα† 10 aα† 13 |0i,
  in a more compact way as
n
dim(H) = .
N
Φ3,6,10,13 = |0001001000100100i.
The total number of bit patterns is 2n .
The action of a creation operator is thus

aα† 4 Φ3,6,10,13 = aα† 4 |0001001000100100i = aα† 4 aα† 3 aα† 6 aα† 10 aα† 13 |0i,

which becomes

−aα† 3 aα† 4 aα† 6 aα† 10 aα† 13 |0i = −|0001101000100100i.

Operators in second quantization Operators in second quantization

Consider the action of aᆠ2 on various slater determinants:

Similarly
aα† 2 Φ00111 = aα† 2 |00111i = 0 × |00111i
aα† 6 Φ3,6,10,13 = aα† 6 |0001001000100100i = aα† 6 aα† 3 aα† 6 aα† 10 aα† 13 |0i, aα† 2 Φ01011 = aα† 2 |01011i = (−1) × |01111i
aα† 2 Φ01101 = aα† 2 |01101i = 0 × |01101i
which becomes aα† 2 Φ01110 = aα† 2 |01110i = 0 × |01110i
−aα† 4 (aα† 6 )2 aα† 10 aα† 13 |0i = 0! aα† 2 Φ10011 = aα† 2 |10011i = (−1) × |10111i
This gives a simple recipe: aα† 2 Φ10101 = aα† 2 |10101i = 0 × |10101i
If one of the bits bj is 1 and we act with a creation operator aα† 2 Φ10110 = aα† 2 |10110i = 0 × |10110i
on this bit, we return a null vector aα† 2 Φ11001 = aα† 2 |11001i = (+1) × |11101i
If bj = 0, we set it to 1 and return a sign factor (−1)l , where l aα† 2 Φ11010 = aα† 2 |11010i = (+1) × |11110i
is the number of bits set before bit j.
What is the simplest way to obtain the phase when we act with one
annihilation(creation) operator on the given Slater determinant
representation?

Operators in second quantization Operators in second quantization

We have an SD representation The action

ΦΛ = aα† 0 aα† 3 aα† 6 aα† 10 aα† 13 |0i, aα0 Φ0,3,6,10,13 = |0001001000100100i,

in a more compact way as can be obtained by subtracting the logical sum (AND operation) of
Φ0,3,6,10,13 and a word which represents only α0 , that is
Φ0,3,6,10,13 = |1001001000100100i.
|1000000000000000i,
The action of
from Φ0,3,6,10,13 = |1001001000100100i.
aα† 4 aα0 Φ0,3,6,10,13 = aα† 4 |0001001000100100i = aα† 4 aα† 3 aα† 6 aα† 10 aα† 13 |0i, This operation gives |0001001000100100i.
Similarly, we can form aα† 4 aα0 Φ0,3,6,10,13 , say, by adding
which becomes |0000100000000000i to aα0 Φ0,3,6,10,13 , first checking that their
logical sum is zero in order to make sure that orbital α4 is not
−aα† 3 aα† 4 aα† 6 aα† 10 aα† 13 |0i = −|0001101000100100i.
already occupied.

Operators in second quantization Exercises

It is trickier however to get the phase (−1)l . One possibility is as Exercise 1
follows This exercise serves to convince you about the relation between two
Let S1 be a word that represents the 1−bit to be removed and different single-particle bases, where one could be our new
all others set to zero. Hartree-Fock basis and the other a harmonic oscillator basis.
Consider a Slater determinant built up of single-particle orbitals ψλ ,
In the previous example S1 = |1000000000000000i
with λ = 1, 2, . . . , A. The unitary transformation
Define S2 as the similar word that represents the bit to be X
added, that is in our case ψa = Caλ φλ ,
S2 = |0000100000000000i. λ

Compute then S = S1 − S2 , which here becomes brings us into the new basis. The new basis has quantum numbers
S = |0111000000000000i a = 1, 2, . . . , A. Show that the new basis is orthonormal. Show that
the new Slater determinant constructed from the new single-particle
wave functions can be written as the determinant based on the
Perform then the logical AND operation of S with the word previous basis and the determinant of the matrix C . Show that the
containing old and the new Slater determinants are equal up to a complex
Φ0,3,6,10,13 = |1001001000100100i, constant with absolute value unity. (Hint, C is a unitary matrix).
Starting with the second quantization representation of the Slater
which results in |0001000000000000i. Counting the number of
determinant
1−bits gives the phase. Here you need however an algorithm for Yn
bitcounting. Several efficient ones available. Φ0 = aα† i |0i,
i=1
Exercises Exercises
Exercise 2
Calculate the matrix elements
Exercise 3
hα1 α2 |F̂ |α1 α2 i Show that the onebody part of the Hamiltonian
and X
Ĥ0 = hp|ĥ0 |qiap† aq ,
hα1 α2 |Ĝ |α1 α2 i pq
with
can be written, using standard annihilation and creation operators,
|α1 α2 i = aα† 1 aα† 2 |0i,
X in normal-ordered form as
F̂ = hα|fˆ|βiaα† aβ , X n o X
αβ
Ĥ0 = hp|ĥ0 |qi ap† aq + hi|ĥ0 |ii.
Z pq i

hα|fˆ|βi = ψα∗ (x)f (x)ψβ (x)dx,

Explain the meaning of the various symbols. Which reference
1 X vacuum has been used?
Ĝ = hαβ|ĝ |γδiaα† aβ† aδ aγ ,
2
αβγδ

and
Z Z
hαβ|ĝ |γδi = ψα∗ (x1 )ψβ∗ (x2 )g (x1 , x2 )ψγ (x1 )ψδ (x2 )dx1 dx2

Compare these results with those from exercise 3c).

Exercises Exercises
Exercise 4 Exercise 5
Show that the twobody part of the Hamiltonian The aim of this exercise is to set up specific matrix elements that
will turn useful when we start our discussions of the nuclear shell
1X
ĤI = hpq|v̂ |rsiap† aq† as ar , model. In particular you will notice, depending on the character of
4 pqrs the operator, that many matrix elements will actually be zero.
Consider three N-particle Slater determinants |Φ0 , |Φai i and |Φab ij i,
can be written, using standard annihilation and creation operators, where the notation means that Slater determinant |Φai i differs from
in normal-ordered form as |Φ0 i by one single-particle state, that is a single-particle state ψi is
1X n o X n o 1X replaced by a single-particle state ψa . It is often interpreted as a
ĤI = hpq|v̂ |rsi ap† aq† as ar + hpi|v̂ |qii ap† aq + hij|v̂ |iji. so-called one-particle-one-hole excitation. Similarly, the Slater
4 pqrs 2
pqi ij
determinant |Φabij i differs by two single-particle states from |Φ0 i
Explain again the meaning of the various symbols. and is normally thought of as a two-particle-two-hole excitation.
This exercise is optional: Derive the normal-ordered form of the We assume also that |Φ0 i represents our new vacuum reference
threebody part of the Hamiltonian. state and the labels ijk . . . represent single-particle states below the
Fermi level and abc . . . represent states above the Fermi level,
1 X so-called particle states. We define thereafter a general onebody
Ĥ3 = hpqr |v̂3 |stuiap† aq† ar† au at as ,
36 pqr normal-ordered (with respect to the new vacuum state) operator as
stu
X n o
and specify the contributions to the twobody, onebody and the F̂N = hp|f |βi ap† aq ,
pq
scalar part.
with Z
∗
Exercises Exercises: Using hp|f |qi = ψp (x)f (x)ψq (x)dx,
sympy to compute matrix elements
Exercise 6 and a general normal-ordered two-body operator
Write a program which sets up all possible Slater determinants n o
1X
given N = 4 eletrons which can occupy the atomic single-particle ĜN = hpq|g |rsiAS ap† aq† as ar ,
4 pqrs
states defined by the 1s, 2s2p and 3s3p3d shells. How many
single-particle states n are there in total? Include the spin degrees Exercise 7
with for example the direct matrix element given as
of freedom as well. We include here a python program which may
Compute the matrix
Z Z element
aid in this direction. It uses bit manipulation functions from
http://wiki.python.org/moin/BitManipulation. hpq|g |rsi = ψp∗ (x1 )ψq∗ (x2 )g 0(x10 , x02 )ψr (x1 )ψs (x2 )dx1 dx2
hα1 α2 α3 |Ĝ |α1 α2 α3 i,
import math
""" with being theorem
usinggWick’s invariant and
under the interchange
express the two-bodyof the coordinates
operator of
G in the
A simple Python class for Slater determinant manipulation two particles.
occupation The single-particle
number states ψi representation.
(second quantization) are not necessarily
Bit-manipulation stolen from:
eigenstates of fˆ. The curly brackets mean that the operators are
http://wiki.python.org/moin/BitManipulation normal-ordered with respect to the new vacuum reference state.
"""
How would you write the above Slater determinants in a second
# bitCount() counts the number of bits set (not an optimal function) quantization formalism, utilizing the fact that we have defined a
def bitCount(int_type): new reference state?
""" Count bits set in integer """ Use thereafter Wick’s theorem to find the expectation values of
count = 0
while(int_type):
int_type &= int_type - 1 hΦ0 |F̂N |Φ0 i,
count += 1
return(count)
and
hΦ0 ĜN |Φ0 i.
Exercises: Using sympy to compute matrix elements Exercises: Using sympy to compute matrix elements
The last exercise can be solved using the symbolic Python package
called SymPy. SymPy is a Python package for general purpose We can expand the above Python code by defining one-body and
symbolic algebra. There is a physics module with several interesting two-body operators using the following SymPy code
submodules. Among these, the submodule called secondquant, # This code sets up a two-body Hamiltonian for fermions
from sympy import symbols, latex, WildFunction, collect, Rational
contains several functionalities that allow us to test our algebraic from sympy.physics.secondquant import F, Fd, wicks, AntiSymmetricTensor, substitute_dummies, NO
manipulations using Wick’s theorem and operators for second # setup hamiltonian
quantization. p,q,r,s = symbols(’p q r s’,dummy=True)
from sympy import * f = AntiSymmetricTensor(’f’,(p,),(q,))
from sympy.physics.secondquant import * pr = NO((Fd(p)*F(q)))
v = AntiSymmetricTensor(’v’,(p,q),(r,s))
i, j = symbols(’i,j’, below_fermi=True) pqsr = NO(Fd(p)*Fd(q)*F(s)*F(r))
a, b = symbols(’a,b’, above_fermi=True) Hamiltonian=f*pr + Rational(1)/Rational(4)*v*pqsr
p, q = symbols(’p,q’) print "Hamiltonian defined as:", latex(Hamiltonian)
print simplify(wicks(Fd(i)*F(a)*Fd(p)*F(q)*Fd(b)*F(j), keep_only_fully_contracted=True))
Here we have used the AntiSymmetricTensor functionality, together
The code defines single-particle states above and below the Fermi with normal-ordering defined by the NO function. Using the latex
level, in addition to the genereal symbols pq which can refer to any option, this program produces the following output
type of state below or above the Fermi level. Wick’s theorem is n o 1 n o
implemented between the creation and annihilation operators Fd fqp ap† aq − vsrqp ap† aq† ar as
and F, respectively. Using the simplify option, one can lump 4
together several Kronecker-δ functions.

Exercises: Using sympy to compute matrix elements Exercises: Using sympy to compute matrix elements
We can continue along these lines and define a normal-ordered
We can now use this code to compute the matrix elements between Hamiltonian with respect to a given reference state. In our first
two two-body Slater determinants using Wick’s theorem. step we just define the Hamiltonian
from sympy import symbols, latex, WildFunction, collect, Rational, simplify from sympy import symbols, latex, WildFunction, collect, Rational, simplify
from sympy.physics.secondquant import F, Fd, wicks, AntiSymmetricTensor, substitute_dummies, NO, evaluate_deltas
from sympy.physics.secondquant import F, Fd, wicks, AntiSymmetricTensor, substitute_dummies, NO, evaluate_deltas
# setup hamiltonian # setup hamiltonian
p,q,r,s = symbols(’p q r s’,dummy=True) p,q,r,s = symbols(’p q r s’,dummy=True)
f = AntiSymmetricTensor(’f’,(p,),(q,)) f = AntiSymmetricTensor(’f’,(p,),(q,))
pr = NO((Fd(p)*F(q))) pr = Fd(p)*F(q)
v = AntiSymmetricTensor(’v’,(p,q),(r,s)) v = AntiSymmetricTensor(’v’,(p,q),(r,s))
pqsr = NO(Fd(p)*Fd(q)*F(s)*F(r)) pqsr = Fd(p)*Fd(q)*F(s)*F(r)
Hamiltonian=f*pr + Rational(1)/Rational(4)*v*pqsr #define the Hamiltonian
c,d = symbols(’c, d’,above_fermi=True) Hamiltonian = f*pr + Rational(1)/Rational(4)*v*pqsr
a,b = symbols(’a, b’,above_fermi=True) #define indices for states above and below the Fermi level
index_rule = {
expression = wicks(F(b)*F(a)*Hamiltonian*Fd(c)*Fd(d),keep_only_fully_contracted=True, simplify_kronecker_deltas=True)
’below’: ’kl’,
expression = evaluate_deltas(expression) ’above’: ’cd’,
expression = simplify(expression) ’general’: ’pqrs’
print "Hamiltonian defined as:", latex(expression) }
Hnormal = substitute_dummies(Hamiltonian,new_indices=True, pretty_indices=index_rule)
The result is as expected, print "Hamiltonian defined as:", latex(Hnormal)

δac fdb − δad fcb − δbc fda + δbd fca + vcd

ab
. which results in
1 sr † †
fpq aq† ap + vqp as ar ap aq
4

Exercises: Using sympy to compute matrix elements Exercises: Using sympy to compute matrix elements
In our next step we define the reference energy E0 and redefine the
Hamiltonian by subtracting the reference energy and collecting the We can now go back to exercise 7 and define the Hamiltonian and
coefficients for all normal-ordered products (given by the NO the second-quantized representation of a three-body Slater
function). determinant.
from sympy import symbols, latex, WildFunction, collect, Rational, simplify from sympy import symbols, latex, WildFunction, collect, Rational, simplify
from sympy.physics.secondquant
from sympy.physics.secondquant import F, Fd, wicks, AntiSymmetricTensor, substitute_dummies, NO, evaluate_deltas import F, Fd, wicks, AntiSymmetricTensor, substitute_dummies, NO, evaluate_deltas
# setup hamiltonian # setup hamiltonian
p,q,r,s = symbols(’p q r s’,dummy=True) p,q,r,s = symbols(’p q r s’,dummy=True)
f = AntiSymmetricTensor(’f’,(p,),(q,))
pr = Fd(p)*F(q) v = AntiSymmetricTensor(’v’,(p,q),(r,s))
v = AntiSymmetricTensor(’v’,(p,q),(r,s)) pqsr = NO(Fd(p)*Fd(q)*F(s)*F(r))
pqsr = Fd(p)*Fd(q)*F(s)*F(r) Hamiltonian=Rational(1)/Rational(4)*v*pqsr
#define the Hamiltonian a,b,c,d,e,f = symbols(’a,b, c, d, e, f’,above_fermi=True)
Hamiltonian=f*pr + Rational(1)/Rational(4)*v*pqsr
#define indices for states above and below the Fermi level expression = wicks(F(c)*F(b)*F(a)*Hamiltonian*Fd(d)*Fd(e)*Fd(f),keep_only_fully_contracted=True, simplify_kronecker_deltas=True)
index_rule = { expression = evaluate_deltas(expression)
’below’: ’kl’, expression = simplify(expression)
’above’: ’cd’, print latex(expression)
’general’: ’pqrs’
} resulting in nine terms (as expected),
Hnormal = substitute_dummies(Hamiltonian,new_indices=True, pretty_indices=index_rule)
E0 = wicks(Hnormal,keep_only_fully_contracted=True)
Hnormal = Hnormal-E0 −δad vefcb −δae vfd
cb cb
+δaf ved −δbd vefac −δbe vfd
ac ac
+δbf ved +δcd vefab +δce vfd
ab ab
−δcf ved
w = WildFunction(’w’)
Hnormal = collect(Hnormal, NO(w))
Hnormal = evaluate_deltas(Hnormal)
print latex(Hnormal)
which gives us
Exercises: Derivation of Hartree-Fock equations Exercises: Derivation of Hartree-Fock equations
Exercise 9
Consider the ground state |Φi of a bound many-particle system of
fermions. Assume that we remove one particle from the
single-particle state λ and that our system ends in a new state
Exercise 8 |Φn i. Define the energy needed to remove this particle as
What is the diagrammatic representation of the HF equation? X
Eλ = |hΦn |aλ |Φi|2 (E0 − En ),
n
X n
−hαk |u HF |αi i + [hαk αj |v̂ |αi αj i − hαk αj |v |αj αi i] = 0
j=1 where E0 and En are the ground state energies of the states |Φi
and |Φn i, respectively.
(Represent (−u HF ) by the symbol − − −X .) Show that
Eλ = hΦ|aλ† [aλ , H] |Φi,
where H is the Hamiltonian of this system.
If we assume that Φ is the Hartree-Fock result, find the
relation between Eλ and the single-particle energy ελ for states
λ ≤ F and λ > F , with

ελ = hλ|t̂ + û|λi,

and X
hλ|û|λi = hλβ|v̂ |λβi.
β≤F

We have assumed an antisymmetrized matrix element here. Discuss

the result.
The Hamiltonian operator is defined as
X 1 X
H= hα|t̂|βiaα† aβ + hαβ|v̂ |γδiaα† aβ† aδ aγ .
2
αβ αβγδ