In this talk, Matthias Eschrig discussed the modification of the pairing
state of a superconductor (SC) due to proximity with a magnetic material.
He began by reviewing the case of a ferromagnet (F) to SC junction, studied
by Buzdin in 1982. Due to the Fermi energy mismatch in the F, characterized
by the parameter J, one expects a split Fermi surface. This furthermore implies
pairing at a nonzero wavevector q, i.e., oscillations in the pairing phi(z)
as a function of the position z in the F region:
phi(z) ~ Exp[-z/xi1]*Exp[i z/xi2]
Where xi1 and xi2 are parameters that can be determined theoretically. xi1
decreases with increasing T, while xi2 increases with increasing T.
Matthias discussed two experiments verifying this picture, each with
a SC-F-SC function. The first (by Kontos et al) showed a transtion between 0 and Pi phases
of the junction (signalled by a vanishing of the critical Josephson current Ic)
with increasing width df of the F region and the second (by Ryazonov
et al) showed a Pi-to-0 transition with increasing T
At this point Dirk Morr pointed out that one can have a zero of Ic without
a transition between Pi and 0 states. But the location of the transition
agreed with theory.
Turning to the case of an interface between a SC and a half metal Ferromagnet,
the main topic, naively one expects no proximity effect. However, recent experiments
by Keizer et al, Nature 2006, on Josephson junctions with NbTiN SC
linked by Cr02 half metal, showed a large-distance Josephson effect.
An initial clue was that the Josephson effect was observed to be
very sensitive to surface properties. The experimentalists
observed hysteresis of the Fraunhofer diffraction pattern. After
subtracting the hysteresis, the pattern was shifted by Pi from
the usual case.
Matthias's work on this problem was published in 2003 in PRL and in 2006 on
cond-mat, and is based on the notion that the important physics occurs
at the interface.
The first effect to consider is spin mixing. Thus, one expects different
phase shifts of spin-up and spin-down fermions scattering at such an
interface, characterized by an angle theta. By itself, this leads to
singlet (S) - triplet (T) mixing, the magnitude of which is proportional
to Sin theta.
To understand the experiments, however, additional scattering properties
must be included. The additional properties included were surface scattering
at the interfaces that were assumed to have a local interface magnetization
m. The two relevant interface magnetizations, m1 and m2, can be labelled by
their angles alpha_i with respect to the magnetization M of the FM regime
and also by the angle between them. The resulting critical Josephson current
is sensitive to the angles alpha, and theta, while the shift in the
Fraunhofer diffraction pattern depends on the difference between the
local interface magnetizations m1 and m2. Future work will focus
on determining the precise physical mechanism behind the interface
magnetizations m1 and m2.
Friday, August 31, 2007
Posted by DSheehy at 9:39 AM