Skip to content

Fock Space Exercises

These exercises practice Fock-space calculations in a fixed mode basis. They assume normalized number states. Bosonic creation and annihilation operators are denoted ai†,aia_i^\dagger,a_i, fermionic operators are denoted ci†,cic_i^\dagger,c_i, and di†,did_i^\dagger,d_i denotes either kind when the statistics do not affect the displayed formula.

For fermions, use canonical mode order 1,2,3,…1,2,3,\ldots. The sign in cic_i or ci†c_i^\dagger is (−1)Si(-1)^{S_i}, where

Si=∑k<inkS_i = \sum_{k<i} n_k

is the number of occupied modes before mode ii in the chosen order.

  1. For two bosonic modes, compute
a2†a1∣3,1⟩B.a_2^\dagger a_1\lvert3,1\rangle_B.
Solution

Operators act from right to left. First annihilate one boson in mode 11:

a1∣3,1⟩B=3 ∣2,1⟩B.a_1\lvert3,1\rangle_B = \sqrt3\,\lvert2,1\rangle_B.

Then create one boson in mode 22:

a2†∣2,1⟩B=2 ∣2,2⟩B.a_2^\dagger\lvert2,1\rangle_B = \sqrt2\,\lvert2,2\rangle_B.

Therefore

a2†a1∣3,1⟩B=6 ∣2,2⟩B.a_2^\dagger a_1\lvert3,1\rangle_B = \sqrt6\,\lvert2,2\rangle_B.

The factor 6\sqrt6 is the product of the removal factor from mode 11 and the creation factor into mode 22.

  1. Let
∣Ψ⟩B=12(∣2,0⟩B+∣0,2⟩B).\lvert\Psi\rangle_B = \frac{1}{\sqrt2} \bigl( \lvert2,0\rangle_B+\lvert0,2\rangle_B \bigr).

Compute N1∣Ψ⟩BN_1\lvert\Psi\rangle_B, Ntot∣Ψ⟩BN_{\mathrm{tot}}\lvert\Psi\rangle_B, and ⟨N1⟩Ψ\langle N_1\rangle_\Psi.

Solution

The mode number operator acts diagonally:

N1∣2,0⟩B=2∣2,0⟩B,N1∣0,2⟩B=0.N_1\lvert2,0\rangle_B = 2\lvert2,0\rangle_B, \qquad N_1\lvert0,2\rangle_B = 0.

Thus

N1∣Ψ⟩B=22∣2,0⟩B=2 ∣2,0⟩B.N_1\lvert\Psi\rangle_B = \frac{2}{\sqrt2}\lvert2,0\rangle_B = \sqrt2\,\lvert2,0\rangle_B.

This is not proportional to ∣Ψ⟩B\lvert\Psi\rangle_B, so the state is not an eigenstate of N1N_1. The total number operator is

Ntot=N1+N2,N_{\mathrm{tot}} = N_1+N_2,

and both components have two particles. Hence

Ntot∣Ψ⟩B=2∣Ψ⟩B.N_{\mathrm{tot}}\lvert\Psi\rangle_B = 2\lvert\Psi\rangle_B.

The expectation value of N1N_1 is

⟨N1⟩Ψ=12(2)+12(0)=1.\langle N_1\rangle_\Psi = \frac12(2)+\frac12(0) = 1.

This example separates “definite total particle number” from “definite occupation of each mode.”

  1. With fermionic mode order 1,2,3,41,2,3,4, compute
c2∣1,1,0,1⟩F,c3†∣1,1,0,1⟩F,c4†∣1,1,0,1⟩F.c_2\lvert1,1,0,1\rangle_F, \qquad c_3^\dagger\lvert1,1,0,1\rangle_F, \qquad c_4^\dagger\lvert1,1,0,1\rangle_F.
Solution

For c2c_2, the mode is occupied, and the number of occupied modes before mode 22 is

S2=n1=1.S_2=n_1=1.

Therefore

c2∣1,1,0,1⟩F=−∣1,0,0,1⟩F.c_2\lvert1,1,0,1\rangle_F = -\lvert1,0,0,1\rangle_F.

For c3†c_3^\dagger, mode 33 is empty, and

S3=n1+n2=2.S_3=n_1+n_2=2.

Thus the sign is positive:

c3†∣1,1,0,1⟩F=∣1,1,1,1⟩F.c_3^\dagger\lvert1,1,0,1\rangle_F = \lvert1,1,1,1\rangle_F.

For c4†c_4^\dagger, mode 44 is already occupied. Pauli exclusion gives

c4†∣1,1,0,1⟩F=0.c_4^\dagger\lvert1,1,0,1\rangle_F = 0.
  1. With the same fermionic mode order, compute
c1†c3∣0,1,1,0⟩F.c_1^\dagger c_3\lvert0,1,1,0\rangle_F.

Explain why the sign should be derived from the operator order rather than guessed from the final bitstring.

Solution

Act first with c3c_3. Since mode 33 is occupied and one occupied mode lies before it,

S3=n1+n2=1.S_3=n_1+n_2=1.

Therefore

c3∣0,1,1,0⟩F=−∣0,1,0,0⟩F.c_3\lvert0,1,1,0\rangle_F = -\lvert0,1,0,0\rangle_F.

Now create in mode 11. There are no earlier modes, so no additional sign appears:

c1†(−∣0,1,0,0⟩F)=−∣1,1,0,0⟩F.c_1^\dagger \bigl( -\lvert0,1,0,0\rangle_F \bigr) = -\lvert1,1,0,0\rangle_F.

Thus

c1†c3∣0,1,1,0⟩F=−∣1,1,0,0⟩F.c_1^\dagger c_3\lvert0,1,1,0\rangle_F = -\lvert1,1,0,0\rangle_F.

The final bitstring alone does not remember how many fermionic operators had to pass through occupied modes. The sign belongs to the ordered operator calculation.

  1. Verify the fermionic identity
{c2,c2†}∣1,0,1⟩F=∣1,0,1⟩F\{c_2,c_2^\dagger\} \lvert1,0,1\rangle_F = \lvert1,0,1\rangle_F

by acting on the state explicitly.

Solution

The anticommutator is

{c2,c2†}=c2c2†+c2†c2.\{c_2,c_2^\dagger\} = c_2c_2^\dagger+c_2^\dagger c_2.

For the first term, mode 22 is initially empty. Since S2=n1=1S_2=n_1=1,

c2†∣1,0,1⟩F=−∣1,1,1⟩F.c_2^\dagger\lvert1,0,1\rangle_F = -\lvert1,1,1\rangle_F.

Acting again with c2c_2 on ∣1,1,1⟩F\lvert1,1,1\rangle_F also contributes a minus sign:

c2(−∣1,1,1⟩F)=−(−∣1,0,1⟩F)=∣1,0,1⟩F.c_2 \bigl( -\lvert1,1,1\rangle_F \bigr) = -\bigl( -\lvert1,0,1\rangle_F \bigr) = \lvert1,0,1\rangle_F.

For the second term,

c2∣1,0,1⟩F=0c_2\lvert1,0,1\rangle_F = 0

because mode 22 is empty. Hence

{c2,c2†}∣1,0,1⟩F=∣1,0,1⟩F.\{c_2,c_2^\dagger\} \lvert1,0,1\rangle_F = \lvert1,0,1\rangle_F.
  1. Let
H^0=∑iϵiNi.\widehat H_0 = \sum_i \epsilon_i N_i.

Find the energy of ∣2,0,1⟩B\lvert2,0,1\rangle_B and of ∣1,0,1⟩F\lvert1,0,1\rangle_F.

Solution

The number operators are diagonal on occupation-number states. For the bosonic state,

H^0∣2,0,1⟩B=(2ϵ1+ϵ3)∣2,0,1⟩B.\widehat H_0\lvert2,0,1\rangle_B = (2\epsilon_1+\epsilon_3) \lvert2,0,1\rangle_B.

For the fermionic state,

H^0∣1,0,1⟩F=(ϵ1+ϵ3)∣1,0,1⟩F.\widehat H_0\lvert1,0,1\rangle_F = (\epsilon_1+\epsilon_3) \lvert1,0,1\rangle_F.

The formula is the same for bosons and fermions; the allowed occupations differ.

  1. Consider two modes with
H^=ϵ1N1+ϵ2N2+t d1†d2+t∗d2†d1.\widehat H = \epsilon_1N_1+\epsilon_2N_2 +t\,d_1^\dagger d_2 +t^*d_2^\dagger d_1.

Compute H^∣1,0⟩\widehat H\lvert1,0\rangle and compare the action on ∣1,1⟩B\lvert1,1\rangle_B and ∣1,1⟩F\lvert1,1\rangle_F.

Solution

On the one-particle state ∣1,0⟩\lvert1,0\rangle, only mode 11 is occupied. The diagonal terms give

(ϵ1N1+ϵ2N2)∣1,0⟩=ϵ1∣1,0⟩.(\epsilon_1N_1+\epsilon_2N_2) \lvert1,0\rangle = \epsilon_1\lvert1,0\rangle.

The hopping term d1†d2d_1^\dagger d_2 gives zero because mode 22 is empty. The term d2†d1d_2^\dagger d_1 moves the particle from mode 11 to mode 22:

d2†d1∣1,0⟩=∣0,1⟩.d_2^\dagger d_1\lvert1,0\rangle = \lvert0,1\rangle.

Therefore

H^∣1,0⟩=ϵ1∣1,0⟩+t∗∣0,1⟩.\widehat H\lvert1,0\rangle = \epsilon_1\lvert1,0\rangle +t^*\lvert0,1\rangle.

For the two-boson state,

(ϵ1N1+ϵ2N2)∣1,1⟩B=(ϵ1+ϵ2)∣1,1⟩B,(\epsilon_1N_1+\epsilon_2N_2)\lvert1,1\rangle_B = (\epsilon_1+\epsilon_2)\lvert1,1\rangle_B,

and

d1†d2∣1,1⟩B=2 ∣2,0⟩B,d2†d1∣1,1⟩B=2 ∣0,2⟩B.d_1^\dagger d_2\lvert1,1\rangle_B = \sqrt2\,\lvert2,0\rangle_B, \qquad d_2^\dagger d_1\lvert1,1\rangle_B = \sqrt2\,\lvert0,2\rangle_B.

Thus

H^∣1,1⟩B=(ϵ1+ϵ2)∣1,1⟩B+2 t∣2,0⟩B+2 t∗∣0,2⟩B.\widehat H\lvert1,1\rangle_B = (\epsilon_1+\epsilon_2)\lvert1,1\rangle_B +\sqrt2\,t\lvert2,0\rangle_B +\sqrt2\,t^*\lvert0,2\rangle_B.

For two fermionic modes, both modes are already occupied in ∣1,1⟩F\lvert1,1\rangle_F. A hopping term would try to create a fermion in an already occupied target mode, so both hopping terms vanish:

H^∣1,1⟩F=(ϵ1+ϵ2)∣1,1⟩F.\widehat H\lvert1,1\rangle_F = (\epsilon_1+\epsilon_2)\lvert1,1\rangle_F.

This is Pauli blocking in a two-mode calculation.

  1. Show that the bilinear di†djd_i^\dagger d_j conserves total particle number by proving
[Ntot,di†dj]=0,Ntot=∑kNk.[N_{\mathrm{tot}},d_i^\dagger d_j]=0, \qquad N_{\mathrm{tot}}=\sum_k N_k.
Solution

For either bosons or fermions, the total number operator satisfies

[Ntot,di†]=di†,[Ntot,dj]=−dj.[N_{\mathrm{tot}},d_i^\dagger] = d_i^\dagger, \qquad [N_{\mathrm{tot}},d_j] = -d_j.

Use the product rule for commutators:

[A,BC]=[A,B]C+B[A,C].[A,BC] = [A,B]C+B[A,C].

Then

[Ntot,di†dj]=[Ntot,di†]dj+di†[Ntot,dj]=di†dj−di†dj=0.\begin{aligned} [N_{\mathrm{tot}},d_i^\dagger d_j] &= [N_{\mathrm{tot}},d_i^\dagger]d_j +d_i^\dagger[N_{\mathrm{tot}},d_j] \\ &= d_i^\dagger d_j -d_i^\dagger d_j \\ &= 0. \end{aligned}

The physical content is simple: di†djd_i^\dagger d_j creates one particle and annihilates one particle, so the net particle number is unchanged.

  1. A one-particle Hamiltonian in a two-mode basis has matrix elements
h11=ϵ1,h22=ϵ2,h12=t,h21=t∗.h_{11}=\epsilon_1, \qquad h_{22}=\epsilon_2, \qquad h_{12}=t, \qquad h_{21}=t^*.

Write its second-quantized one-body operator. Then add a density-density interaction that assigns an extra energy UU when both modes are occupied.

Solution

The one-body rule is

H^1=∑ijhijdi†dj.\widehat H_1 = \sum_{ij}h_{ij}d_i^\dagger d_j.

For the given two-mode matrix,

H^1=ϵ1d1†d1+ϵ2d2†d2+t d1†d2+t∗d2†d1.\widehat H_1 = \epsilon_1d_1^\dagger d_1 +\epsilon_2d_2^\dagger d_2 +t\,d_1^\dagger d_2 +t^*d_2^\dagger d_1.

Using Ni=di†diN_i=d_i^\dagger d_i, this is

H^1=ϵ1N1+ϵ2N2+t d1†d2+t∗d2†d1.\widehat H_1 = \epsilon_1N_1+\epsilon_2N_2 +t\,d_1^\dagger d_2 +t^*d_2^\dagger d_1.

A density-density interaction that adds energy UU when both modes are occupied is

H^int=UN1N2.\widehat H_{\mathrm{int}} = U N_1N_2.

The full model is therefore

H^=ϵ1N1+ϵ2N2+t d1†d2+t∗d2†d1+UN1N2.\widehat H = \epsilon_1N_1+\epsilon_2N_2 +t\,d_1^\dagger d_2 +t^*d_2^\dagger d_1 +U N_1N_2.

For fermions, N1N2N_1N_2 is a projector onto the state with both modes occupied. For bosons, it counts cross-mode pairs, so the same expression gives an interaction energy Un1n2U n_1n_2 on ∣n1,n2⟩B\lvert n_1,n_2\rangle_B.

  • Dropping the n\sqrt n and n+1\sqrt{n+1} factors in normalized bosonic number states.
  • Reading fermionic bitstrings from left to right without tracking the signs from the operator order.
  • Treating di†djd_i^\dagger d_j as particle-number changing because it contains a creation operator.
  • Confusing a mode occupation with a labeled particle.
  • Writing only diagonal number-operator terms for a Hamiltonian whose one-particle matrix has off-diagonal hopping or mixing.
  • Forgetting that Pauli blocking can make a fermionic hopping term vanish even when the analogous bosonic term is nonzero.
  • P. A. M. Dirac, The Principles of Quantum Mechanics, 4th ed., Oxford University Press, 1958.
  • A. L. Fetter and J. D. Walecka, Quantum Theory of Many-Particle Systems, McGraw-Hill, 1971.
  • J. W. Negele and H. Orland, Quantum Many-Particle Systems, Addison-Wesley, 1988.
  • R. Shankar, Principles of Quantum Mechanics, 2nd ed., Springer, 1994.
  • A. Altland and B. Simons, Condensed Matter Field Theory, 2nd ed., Cambridge University Press, 2010.