Claire Marrache Kikuchi gave a detailed presentation on disordered amorphous NbxSi1-x focusing on 3 different ways to control disorder: (i) By tuning the thickness of the sample, (ii) by changing the composition and (iii) through controlled annealing. She started by reminding the audience that different kinds of disorder can have fundamentally different effects. For example, with decreasing thickness surface phonon softening can play an important role and has often been argued to be the reason for enhancement of T_c in observed in Al and Sn. In addition, she reminded the audience, the difference between homogeneous and granular disorder has been highlighted by many authors.
Amorphous thin films of NbxSi1-x were grown through co-deposition of Nb ans Si using electron beam evaporation. For thick films (thickness>50nm) samples with composition 11-18% Nb showed superconductivity. The samples have electronic density of states at Fermi energy in the range 1041 states/J-cm3 which is comparable to that of Au. TEM images of the samples do not show any granular structure.
The composition tuned samples have a metal-insulator transition at 9.9% of Nb, with samples below this composition range exhibiting Efrot-Schlovsky behavior in their transport properties. Interestingly, in this system superconductivity gets apparently destroyed well inside the metallic regime unlike amorphous InOx and TiN where superconductivity persists in the disorder driven insulating regime. This point towards a very different mechanism, arising maybe from the increase in e-e interactions resulting from a loss of effective screening, which drives the superconducting transition temperature in this material. However, further measurements down to lower temperatures would be needed to confirm this.
In the composition range where the samples are superconducting (e.g. 14%, 15% and 18%), Tc monotonically decreases with decresing thickness. The dependence of Tc on the sheet resistance is qualitatively consistent with an increase in e-e repulsive interaction resulting from the loss of effective screening. However, there are quantitative discrepancies with theory.
The most intriguing aspect of these samples was the effect of annealing. While there is no discernible change in the morphology with annealing up to 5000C, contrary to usual expectation the sheet resistance gradually increases and the Tc decreases with increasing annealing temperatures. There was considerable discussion on the microscopic changes that could be responsible for this behavior. It was felt that this issue needs to be looked into further using other microscopic tools. However, one important point was that Tc evolves in the same way with sheet resistance for both composition driven and annealing driven disordered sample, whereas the thickness tuned samples did not fall on the same curve.
In summary, the well characterized samples of NbxSi1-x would definitely provide us with another system where various scenarios proposed for the superconductor-insulator transition could be compared with experiments. It would be interesting to study these sample with various experimental probes such as STM and penetration depth measurements and compare the commonalities and differences with TiN, NbN and InOx. Prof. Claude Chapelier mention during the discussion that STM measurements on this system are underway. Hopefully, as more systems get studied, a common picture for the disorder driven SIT in various systems will gradually emerge.
blogged by Pratap Raychaudhuri
No comments:
Post a Comment