Brigitte starts her presentation
by presenting the general motivation of the presented work: understanding the
effect of enhanced inhomogeneity on the superconducting (SC) properties of thin
NbN films.
The principal probes are
transport measurement (with or without magnetic field), TEM and STS.
First she makes the distinction
between granular and homogeneously disordered systems. Granular materials are
coherent with Coupled Josephson-junction. The disappearance of SC properties
into granular films follows a phase fluctuation model (described by the XY
model). For inhomogeneous materials, there is two major models for the SC :
it’s either fermionic (localized fermion into the insolator) or bosonic
(localized cooper pairs)
About the experiment: The films
are grown on Al2O3 substrate and then measured ex-situ. Tc are from 9 to 2.4K
and is evolving other the time, as the sheet resistance.
TEM measurement: have shown a
crystallization of the sample next to the substrate on few layers and a
passivation layer probably made of Nb2O5.
STS measurement: have shown gap
inhomogeneity. Cross correlation map between the spectroscopic and topographic
pictures show nothing. Auto-Correlation of the spectroscopic picture show a
typical range of inhomogeneity distribution of 100nm. This distribution is stable between 300mK and 4K
Electronic-transport
measurements have shown a non-monotonous behavior of Tc as a function of
thickness . It is not following finkelstein’s theory.
By fitting the influence of
fluctuations, a simple 2d Aslamasov-Larkin fluctuation model (AL) is enough to
explain SC fluctuation for the cleanest samples. It seems in accordance with an
amplitude fluctuation model for the destruction of SC, supposed to be in
accordance with a fermionic model of the suppression of SC. There is no need to
take into account Maki-Thomson or DOS correction. For more disordered samples,
fluctuations fit with a 0d-AL fluctuations. There is a 0d-2d crossover near to
Tc (defined as R=0). These 0d-AL fluctuation supports the picture of coupled
grains which size is corresponding to the (electronic) inhomogeneity
distribution.
She finished the talk by a
renormalization analysis coherent with nu=2/3 and z=1, obtained from
electric-field and magnetic field variation. These values are compatible with
a XY model.
To summarize, Brigitte concluded
on the presence of both bosonic and fermionic features into the sample.
The question of 0d fluctuations
is open, as the competition between a fermionic and bosonic picture of the
destruction of the SC in disordered thin films, is open.
Blogged by Vincent Humbert
No comments:
Post a Comment