Phase fluctuations in a strongly disordered s-wave superconductor close to the metal-insulator transition

Understanding the role of interaction and disorder is an extremely important problem in the context of many-body quantum systems such as high-temperature superconductors, quantum Hall effect and superfluid helium. In recent years, it has been understood that the ground state of conventional superconductors, previously thought to be unaffected by disorder, gets destabilized at a critical value of disorder. A number of recent experiments reveal the formation of unusual metallic and insulating states close to critical disorder, where superconducting correlations persist in the resistive state. 

In our lab we have been trying to understand the effects of disorder on the fundamental superconducting properties using a combination of transport, Hall effect, electronic tunneling, and penetration depth studies. Our samples consist of 3-dimensional epitaxial NbN films with disorder ranging from the moderately clean limit (kFl~10.12) to the very dirty limit (kFl~1.24, very close to the disorder driven metal insulator transition). The Tc decreases from ~17K to less than 350mK with increasing disorder. The resistivity and Hall effect studies show that with increasing disorder the transport properties of these films evolve from conventional metallic behavior (  RH independent of T) for our most ordered film to an unusual metallic state with governed by strong e-e interactions. The Tc and conductivity at the lowest temperature (0) both asymptotically approach zero in the limit kFl1, indicating that the superconductivity gets destroyed at the metal-insulator transition (MIT). 

We show that close to critical disorder, the essential physics of the superconducting state is governed by strong phase fluctuations. A spectroscopic gap-map obtained via scanning tunneling spectroscopy shows a spatial inhomogeneity in the magnitude of the superconducting energy gap. The gap feature is also seen to persist above Tc giving rise to a “pseudogapped” state at T>Tc. We also observe that the zero temperature superfluid density is suppressed and that it varies approximately linearly over a large range of temperature in contrast with BCS predictions. The coherence peaks in the density of states are suppressed and we also see an appearance of mid-gap states in the conductance spectra. I will highlight the interpretation of these results on the basis of a dissipative model for quantum and classical phase fluctuations.