We extend to finite temperature the time-dependent effective theory for the Goldstone field (the phase of the pair field) $\ensuremath{\theta}$ which is appropriate for a superfluid containing one species of fermions with s-wave interactions, described by the BCS Lagrangian. We show that, when Landau damping is neglected, the effective theory can be written as a local time-dependent nonlinear Schr\"odinger Lagrangian which preserves the Galilean invariance of the zero-temperature effective theory and is identified with the superfluid component. We then calculate the relevant Landau terms that are nonlocal and that destroy the Galilean invariance. We show that the retarded $\ensuremath{\theta}$ propagator (in momentum space) can be well represented by two poles in the lower-half frequency plane, describing damping with a predicted temperature, frequency, and momentum dependence. It is argued that the real parts of the Landau terms can be approximately interpreted as contributing to the normal fluid component.
The effective propagator for the Goldstone mode (phase of the order parameter) is calculated for a neutral BCS system in the long-wavelength/low-frequency limit, with inclusion of Landau damping terms, for temperatures between
T=0 and
T=0.6Tc. The Landau terms are first evaluated numerically, and then accurate closed-form expressions are found for them. The resulting propagator is shown to be well approximated by the product of two simple poles at complex energy, corresponding to a damped mode with linear (and
T-dependent) dispersion for both the real and imaginary parts. Damping is only significant for
T> 0.4Tc. By considering the Fourier transform of the inverse of this pole-dominated propagator, an effective local equation of motion for the phase degree of freedom is obtained, which includes a specific damping term. The damping may be phenomenologically included in the equivalent time-dependent nonlinear Schr\"odinger equation by giving the pair mass a small temperature-dependent positive imaginary part.
A standard perturbative expansion around the mean-field solution is used to derive the low-energy effective action for superconductors at
T=0. Taking into account the density fluctuations at the outset we get the effective action where the density $\ensuremath{\rho}$ is the conjugated momentum to the phase $\ensuremath{\theta}$ of the order parameter. In the hydrodynamic regime, the dynamics of the superconductor is described by a time-dependent nonlinear Schr\"odinger (TDNLS) equation for the field $\ensuremath{\Psi}(x)=\sqrt{\ensuremath{\rho}/2}{e}^{i\ensuremath{\theta}}.$ The evolution of the density fluctuations in the crossover from weak-coupling (BCS) to strong-coupling (Bose condensation of localized pairs) superconductivity is discussed for the attractive Hubbard model. In the bosonic limit, the TDNLS equation reduces to the Gross-Pitaevskii equation for the order parameter, as in the standard description of superfluidity. The conditions under which a phase-only action can be derived in the presence of a long-range interaction to describe the physics of the superconductivity of ``bad metals'' are discussed.
Using a functional approach to the Keldysh formalism we develop a time-dependent Ginzburg-Landau equation for the long-wavelength and low-frequency dynamics of a weak-coupling superconductor, which is also valid in situations that deviate significantly from thermal equilibrium. The theory takes the interactions between quasiparticles and Cooper pairs into account exactly and is explicitly Galilean-invariant at zero temperature. However, the temporal and spatial dependence of the order parameter is assumed to be such that the effects of Landau damping can be neglected. We also consider the screening properties of the superconductor and show that the inclusion of the polarizability of the electron gas is necessary for a complete description.
We analyze the evolution of the dynamics of a neutral
s-wave superconductor between the BCS and BEC regimes. We consider 2d case when a BCS-BEC crossover occurs already at weak coupling as a function of the ratio of the two scales---the Fermi energy
EF and the bound state energy for two fermions in a vacuum,
E0. The BCS and BEC limits correspond to ${E}_{F}\ensuremath{\gg}{E}_{0}$ and ${E}_{F}\ensuremath{\ll}{E}_{0}$, respectively. The chemical potential $\ensuremath{\mu}={E}_{F}\ensuremath{-}{E}_{0}$ changes the sign between the two regimes. We use the effective action approach, derive the leading terms in the expansion of the effective action in the spatial and time derivative of the slowly varying superconducting order parameter $\mathrm{\ensuremath{\Delta}}(r,\ensuremath{\tau})$, and express the action in terms of a derivatives of the phase $\ensuremath{\phi}(r,\ensuremath{\tau})$ of $\mathrm{\ensuremath{\Delta}}(r,\ensuremath{\tau})=\mathrm{\ensuremath{\Delta}}{e}^{i\ensuremath{\phi}(r,\ensuremath{\tau})}$. The action contains ${(\ensuremath{\nabla}\ensuremath{\phi})}^{2}$ and ${\stackrel{\ifmmode \dot{}\else \.{}\fi{}}{\ensuremath{\phi}}}^{2}$ terms, which determine the dispersion of collective phase fluctuations, and $i\ensuremath{\pi}A\stackrel{\ifmmode \dot{}\else \.{}\fi{}}{\ensuremath{\phi}}$ term. For continuous $\ensuremath{\phi}(r,\ensuremath{\tau})$, the latter reduces to the contribution from the boundary and does not affect the dynamics. We show that this long-wavelength action does not change through the BCS-BEC crossover. We apply our approach to a moving vortex, for which $\ensuremath{\phi}$ is singular at the center of the vortex core, and $i\ensuremath{\pi}{A}_{\mathrm{vort}}\stackrel{\ifmmode \dot{}\else \.{}\fi{}}{\ensuremath{\phi}}$ term affects vortex dynamics. We find that this term has two contributions. One comes from the states away from the vortex core and has
Avort,1=n/2, where
n is the fermion density. The other comes from electronic states inside the vortex core and has ${A}_{\mathrm{vort},2}=\ensuremath{-}{n}_{0}/2$, where
n0 is the fermion density at the vortex core. This last term comes from the continuous part of the electronic spectrum and has no contribution from discrete levels inside the core; it also does not change if we add impurities. We interpret this term as the contribution to vortex dynamics in the continuum limit, when the spacing between energy levels $\ensuremath{\omega}$ is set to zero, while fermionic lifetime $\ensuremath{\tau}$ can be arbitrary. The total ${A}_{\mathrm{vort}}=(n\ensuremath{-}{n}_{0})/2$ determines the transversal force acting on the vortex core, $\ensuremath{\pi}{A}_{\mathrm{vort}}\stackrel{\ifmmode \dot{}\else \.{}\fi{}}{\mathbf{R}}\ifmmode\times\else\texttimes\fi{}\stackrel{\ifmmode \hat{}\else \^{}\fi{}}{z}$, where $\stackrel{\ifmmode \dot{}\else \.{}\fi{}}{\mathbf{R}}$ is the velocity of the vortex core and $\stackrel{\ifmmode \hat{}\else \^{}\fi{}}{z}$ a unit vector perpendicular to the 2d sample. The difference $(n\ensuremath{-}{n}_{0})/2$ changes through the BEC-BCS crossover as
n0 nearly compensates
n in the BCS regime, but vanishes in the BEC regime.