Tuesday 28 July 2015- Uwe Pracht - Pairing enhancement versus phase disordering in coupled nanograins of a conventional superconductor

Uwe Pracht takes us back to the longest-standing problem discussed in this workshop. Granular aluminum exhibits dome-like superconductivity, where the critical temperature increases with resistivity, contrary to what has been seen on much of the materials discussed in this workshop. The experimental facts are known already for a long time, but have never been satisfyingly explained.

Actually, as Uwe shows, many of the exotic phenomena that are discussed this week, such as a superconducting dome in the phase diagram, but also evidence of a pseudogap, have already been shown in this system. In his talk, Uwe tries to answer two main questions: how do the relevant energy scales (Delta, superfluid stiffness, Tc) evolve in the phase diagram, and what is the origin of the superconducting dome.

Uwe has studied these materials by looking at the electromagnetic response of various granular aluminum films. From the frequency-dependent response functions \sigma_1 and \sigma_2, he extracts the energy gap \Delta and the superfluid stiffness, respectively, using a conventional Mattis-Bardeen fit of the response.

In this way, he has been able to identify different regions in the phase diagram of granular aluminum. At low disorder, Tc increases, together with \Delta. This is attributed to a decoupling of the different superconducting grains, which enhances the shell effect in the individual grains. At the same time, the superfluid density is continuously decreasing, which has a detrimental effect on the superconductor. After a cross-over regime, the phase fluctuations in the system take over. Here, the spectral gap stays more or less constant, whereas Tc is going down.

In this region exactly, it seems that a pseudo-gap like feature seems to appear in the data. The analysis of this feature is based on the fact that individual grains that remain superconducting, but loose long-range phase coherence. However, the data in this regime is noisy, and the errorbars do not exclude a gapless state above Tc. It is definitely a very interesting system that deserves being studied in much more detail.

Blogged by Eduard Driessen