The dynamics of a Bose-Einstein condensate is usually described by the time-dependent Gross-Pitaevskii equation (TDGPE) $$ i\hbar\frac{\partial\psi(x,t)}{\partial t} = -\frac{\hbar^2\nabla^2}{2m}\psi(x,t)+V\psi + g |\psi(x,y)|^2\psi(x,t) $$ where $V$ is the external potential and $g$ the interaction strength.

To find the excitation spectrum for small perturbation, it is possible to consider the ansatz $$ \psi(x,t) = e^{-i \mu t/\hbar} ( \psi_0(x) + \delta\psi(x,t) ) \;, $$ where we can use a Fourier expansion $$ \delta\psi(x,t) = \sum_i [ u_i(x) e^{-i\omega_i t}+v_i^*(x) e^{i\omega_i t}] \;. $$ Inserting this ansatz in the TDGPE gives a system of equations: one time-independent GPE for $\psi_0$, and the two bogoliubov equations for $u_i$ and $v_i$.

Now my questions:

1) Is my above explanation of the procedure correct?

2) Sometimes it is mentioned that $\delta\psi(x,t)$ needs to be orthogonal to $\psi_0(x)$, to conserve the particle number. Therefore I would expect $\int dx (\psi(x)\delta\psi(x,t)) = 0$. If we ask for $\int dx \vert\psi(x,t)\vert^2 = 1$ and $\int dx \vert\psi_0(x,t)\vert^2 = 1$, to first order we find $\int dx (\psi(x)\delta\psi^\ast(x,t) + \psi^\ast(x)\delta\psi(x,t)) = 0$. How do I understand this as an orthogonality constraint? What will be the constraint on the $u_i$ and $v_i$?

3) Since excitations are expected to be bosonic quasiparticles, one wants bosonic commutation relations to be satisfied (this appears in second quantisation). It is known that this implies that $u_i$ and $v_i$ must obey $\int dx (u_i^2 - v_i^2) = 1$. Is there an argument to impose this in the approach I used above (without going to second quantisation)?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.