Wednesday, October 31, 2007

A Spin Triplet supercurrent though the half-metallic ferromagnet CrO2

by R. S. Keizer, S. T. B. Goennenwein, T. M. Klapwijk, G. Miao, G. Xiao and A. Gupta

Journal club talk by Patrick Morales
Summary by Fazel Fallah Tafti



Superconductivity and ferromagnetism are competing ordered states
of matter but it appears that they can coexist. In this article we
learned about another example of SC in a FM material which exists in
the half-metallic material CrO2.

The experiment is realized by growing a single crystal thin film of
CrO2 on TiO2 substrate (epitaxial growth) and patterning two
superconducting islands made of NbTiN on top it using organic resist
mask with conventional electron beam lithography.

NbTiN is a conventional S-wave BCS superconductor and CrO2 is a
spin polarized half metallic ferromagnet i.e. it has a finite DOS at
the Fermi level for the up spin polarization but it shows a gap of ~
2eV in the down spin polarization.

The supercurrent tunnels through the Josephson junction
NTN-CrO2-NTN which is shown in IV characteristic curve. Fraunhofer
pattern is observed for Ic as a function of induced magnetic field
which is regarded as another evidence of supercurrent tunneling
through the FM material.

The distance between the two NTN leads is about 300nm which is much
higher than the coherence length expected for a FM metal. This is even
higher than the coherence length in most normal metals (~100 nm) which
is surprising.

Such persistence of a supercurrent through a FM half-metal can be
justified by the idea of spin triplet cooper pairs. In this picture we
can still have the main features of the BCS theory but the pairing
mechanism in unconventional such that instead of singlet spin pairs we
have a pair made of two electrons with parallel spins.

The question to be answered is that what are is the mechanism
behind singlet-triplet conversion at the junction between the normal
s-wave SC and CrO2. In fact there are a number of proposed mechanisms
for such phenomenon but they all agree on the fact that such process
has to be caused by an interplay between the exchange field of the FM
at the FM-SC junction and a non-homogeneous magnetic field across the
junction. In order to make a correct theory to describe such process
one has to include various important factors such as the effect of
domain walls at the interface, spin-orbit interactions in the SC,
Double exchange interaction in CrO2 and perhaps the effect of local
magnetic impurities.

No comments: