Abstract
We show that, except for the correction in the longitudinal superfluid density =ρ[1-(3/2)(ħ/δ) ‖(l^⋅∇)l^‖], all the nonanalytic contributions to the supercurrent which have been studied previously vanish. It is shown that, if one does not use a gauge transformation, one has to go beyond the linear approximation for the spatial variations of the gap in order to treat consistently the superfluid velocity.
- Received 25 January 1985
DOI:https://doi.org/10.1103/PhysRevB.32.2960
©1985 American Physical Society