Sunday, October 26, 2008

This week in cond-mat

While there are several interesting recent papers, one in particular touches on a nice piece of physics that I'd like to describe.

arxiv:0810.4384 - Bluhm et al., Persistent currents in normal metal rings
One of the remarkable properties that makes superconductors "super" is their capability to sustain currents that flow in closed loops indefinitely, without dissipation. We exploit this all the time in, for example, MRI magnets. What many people do not realize, however, is that normal metals (e.g., gold, silver) can also sustain persistent currents at very low temperatures, at least over length scales comparable to the coherence length. Think of electrons as waves for a minute. The coherence length is the distance that electrons can propagate in a metal and still have a well-defined phase (that is, the crests and troughs of the electron waves have a reproducible location relative to some initial point). At non-zero temperatures, inelastic interactions between the electrons and other degrees of freedom (including other electrons) fuzz out this phase relationship, suppressing quantum interference effects over a characteristic distance scale (the coherence length). Anyway, suppose you have a metal loop smaller than the coherence length. The phase of the electronic wave when you do one complete lap around the loop must increase (or decrease) by an integer multiple of 2pi (that is, there must be an integer multiple of wavelengths going around the loop) for the wave picture to make sense. The gradient of that phase is related to the current. It turns out that as T goes to zero, the allowed electronic states of such a loop have to have nonzero currents so that this phase winding picture holds. These currents also lead to a magnetic response - tiny current loops are magnetic dipoles that can be detected, and trying to thread external magnetic flux through these loops changes in the persistent currents (via the Aharonov-Bohm effect). These persistent currents have been measured before (see here, for a great example). However, there has been ongoing controversy concerning the magnitude and sign of these currents. In this experiment, Kam Moler's group at Stanford has used the incredibly sensitive scanning SQUID microscope to look at this phenomenon, one ring at a time, as a function of temperature and external magnetic field. This is a very pretty experiment probing some extremely finicky physics.

2 comments:

Anonymous said...

There is the heating/dephasing issue due to the SQUID noise. Although there are heat sinks to cool down the electrons, there still could be direct dephasing effect due to noise. The coherence length quoted in the paper is measured for wires, not for small rings. But anyway, a very interesting experiment and innovative data analysis.

Anonymous said...

Forgot another possible issue: when the size of the ring is comparable to coherence length, the heat sink need to be considered in the calculation of coherence since the coherent electrons may dwell inside the heat sink.