Nina Markovic gave an interesting talk about work she had been
involved in at Harvard as a postdoc. They fabricated silicon/silicon
nitride wafers which they hydroflourically etched channels into. They
placed carbon nanowires (5-25 nm in width) across the gaps, and then
sputtered the surface of the wafers and the carbon wires with a thin
film (4-10 nm) of moly-germanium. They placed their samples in a
cryostat at brought them below the critical temp. They monitored the
voltage drop across the wafer superconducting film, and across the
nanowire film bridging the gap.
They observed that as the width of the film on the nanowire approached
the coherence length, the film turned resistive. Additional cooling
did not effect the result, and the bulk film across the surface of
the wafer remained superconducting. Using modeling techniques,
Markovic and her team surmised that the loss of superconductivity in
the nanowires was predominantly due to quantum tunneling phase slips,
leading to decoherence--thus the coherence length of a material places
a fundamental limit on how small a superconductor can be made.
One of the supercon gurus in the audience argued that a more likely
explanation was a reduction in the cooper pair/phonon population with
decreasing dimensions.
What do you think? Apparently the interpretation of the data is
somewhat controversial at present.
Interesting stuff.
DAH 1/6/04
*-----------------------*
Posted at:
www.GroupSrv.com
*-----------------------*
.
|
