The thinnest superconducting wires ever made, only 10 nm wide, have
been used in an experiment showing how the superconducting state gets
extinguished as the wire narrows.
Just as traffic becomes more problematic as you reduce flow on an interstate
from four lanes down to three and then down to two and finally to one
lane, so electron pairs (or Cooper pairs, which constitute the supercurrent)
moving through very thin passages are sensitive to quantum effects not
noticeable in larger wires.
A quantum phase slip (QPS) is one such effect. It is a quantum fluctuation
in which the superconducting wavefunction spontaneously tunnels from
one state into another, a process which results in a momentary voltage,
and therefore a nonzero electrical resistance, even if the temperature
could somehow be reduced to absolute zero.
Armed with thin wires (10-20 nm) consisting of molybdenum-germanium
deposited onto carbon nanotubes, Michael Tinkham (Tinkham@RSJ.Harvard.edu)
and his colleagues at Harvard have conducted the most thorough study
yet made of this phenomenon and have definitely shown that resistance
goes up as the wire gets thinner.
The quantum resistance effect only becomes noticeable for wires below
about 30 nm in size, far smaller than most wires used in today's computers,
so there is no bottleneck yet. Future advanced superconducting computers,
however, might have trouble; by going to lower temperatures you can
eliminate resistivity arising from thermal fluctuations, but not from
quantum fluctuations. (Lau
et al., Physical Review Letters, 19 November 2001.)