1
$\begingroup$

I am trying to find the mean number of particles in the BCS ground state $|\psi>$, but I am stuck on a step.

$$|\psi> = \Pi_{k}(u_{k} + v_{k}c^{{\dagger}}_{{k}{\uparrow}}c^{\dagger}_{{-k}{\downarrow}})$$

where $u_{k}$, $v_{k}$ are complex numbers. So to begin, we start with a specific spin state $$<\psi|N_{{p}{\uparrow}}|\psi> =<\psi|c^{{\dagger}}_{{p}{\uparrow}}c_{{p}{\uparrow}}|\psi>$$

This is what I get

$$\Pi_{k} |v_{k}|^{2} (c_{{k}{\uparrow}}c_{{-k}{\downarrow}}c^{{\dagger}}_{{p}{\uparrow}}c_{{p}{\uparrow}}c^{{\dagger}}_{{k}{\uparrow}}c^{\dagger}_{{-k}{\downarrow}})$$

I don't know how to simplify this or if I am even on right track. Any hints?

$\endgroup$
1
  • $\begingroup$ put the orders in normal order using the anti-commutation rule for fermion operators. Search Wick's theorem. $\endgroup$
    – wcc
    Commented Apr 30, 2019 at 13:32

1 Answer 1

2
$\begingroup$

Here is a calculation based on a general Bogoluibov transformation $$ a_i= u_{i\beta}b_\beta +v^*_{i\beta}b^\dagger_\beta\\ a^\dagger_i= v_{i\alpha} b_\alpha +u^*_{i\alpha}b^\dagger_\alpha. $$ With $\vert 0 \rangle_b$ being the vacuum state for the $b_\alpha$'s we have
$$ N=\sum_i { _b\langle{0}\vert{ a^\dagger_i a_i}\vert{0}\rangle _b} \\ ={\sum_{i,\alpha,\beta} { _b\langle{0}\vert{( v_{i\alpha} b_\alpha +u^*_{i\alpha}b^\dagger_\alpha)(u_{i\beta}b_\beta +v^*_{i\beta}b^\dagger_\beta)}\vert{0}\rangle_b}}\\ = \sum_{i,\alpha,\beta} (v_{i\alpha} v^*_{i\beta}) {_b\langle{0}\vert{ b_\alpha b^\dagger_\beta }\vert{0}\rangle_b}\\ =\sum_{\alpha,\beta} v_{i\alpha} v^*_{i\beta} \delta_{\alpha\beta}\\ = \sum_\alpha |v_\alpha|^2. $$ We can also write $$ N= \frac 12 \sum_{\alpha=1}^N\left((|v_\alpha|^2-|u_\beta|^2) + (|u_\alpha|^2+|v_\alpha|^2)\right)\\ =\frac 12 \sum_{\alpha=1}^N\left((|v_\alpha|^2-|u_\beta|^2) + 1\right), $$ which looks like the contribution of the Dirac sea negative-energy Bogoliubov-de Gennes equation eigenstates $(v^*_\alpha, u^*_\alpha)$, but corrected by an infinite normal-ordering counterterm (the sum of all the ``+1''s).

As an aside, note I'm not sure that your BCS ground state is normalized. I like to write $$ \vert{0}\rangle _b ={\mathcal N} \exp\left\{\frac 12 a^\dagger_i a^\dagger_jS_{ij}\right\} \vert{0}\rangle_a $$ where $$ S_{ij}= v^*_{i\alpha}(u^*)^{-1}_{\alpha j} $$ is skew symmetric and $$ {\mathcal N}= {\rm det}^{-1/4}(I+S^\dagger S) $$ is derived by using the fermionic version of MacMahon's master theorem. This way makes clear that the BCS ground state is a coherent state superposition of Cooper pairs with pair "wavefunction" $S_{ij}$.

Note added: Yes your state is normalized. It's just not the way I like to write it.

$\endgroup$
7
  • $\begingroup$ May I ask how is the last equation about the normalization constant derived (in detail)? or if it has been shown at any reference? Thanks a lot! $\endgroup$
    – Ogawa Chen
    Commented Nov 11, 2020 at 11:04
  • $\begingroup$ It's a rather lengthy and indiect proof using canonical forms. The bose version is in Schwinger's essay On Angular Momentum which is reprinted in Biedenharn and Van Dam, Quantum Theory of Angular Momentum, Academic Press, 1965. The only derivation I have seen in print of the fermionic version is in Pressley and Segal's book Loop Groups which is in my inaccessible office. $\endgroup$
    – mike stone
    Commented Nov 11, 2020 at 12:51
  • $\begingroup$ Can you tell me which section? Thanks! $\endgroup$
    – Ogawa Chen
    Commented Nov 11, 2020 at 13:32
  • $\begingroup$ another related question is that, let's say, by tuning the parameter in the BdG Hamiltonian, we obtain a new pairing ground state proportional to $\Omega_1 = \exp(\frac{1}{2}\tilde{S}_{i,j}c^\dagger_i c_j)|0\rangle$ (the normalization factor can be calculated just use the formula you give above), then what would be the overlap between $|\Omega_1\rangle$ and $|\Omega_0 \rangle = \exp(\frac{1}{2}S_{i,j}c^\dagger_i c_j)|0\rangle$ (which is proportional to the ground state of initial Hamiltonian)? Is there also a simple formula for it? Thanks! $\endgroup$
    – Ogawa Chen
    Commented Nov 11, 2020 at 14:23
  • $\begingroup$ I just noticed that the overlap between different pairing states has been studied before. There is a so-called Onishi formula, but with a "sign problem", which is solved in this beautiful paper: journals.aps.org/prc/abstract/10.1103/PhysRevC.50.2874 $\endgroup$
    – Ogawa Chen
    Commented Nov 17, 2020 at 18:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.