A hint of negative electrical resistance
emerges from a new
experiment in which microwaves of two different frequencies are directed at a
2-dimensional electron gas. The electrons, moving at the interface
between two semiconductor crystals, are subjected to an electric
field in the forward (longitudinal) direction and a faint magnetic
field in the direction perpendicular to the plane. In such
conditions the electrons execute closed-loop trajectories which
will, in addition, drift forward depending on the strength of the
applied voltage.
A few years ago, two experimental groups observed that
when, furthermore, the electrons were exposed to microwaves, the overall
longitudinal resistance could vary widely -- for example, increasing by an
order of magnitude or extending down to zero, forming a
zero-resistance state, depending on the relation between microwave
frequency and the strength of the applied magnetic field (for
background, see Physics Today, April 2003).
Some theorists proposed that in such zero-resistance state, the
resistance would actually have been less than zero: the swirling
electrons would have drifted backwards against the applied voltage.
However, this rearwards motion would be difficult to observe because
of an instability in the current flow -- that is, the current
distribution becomes inhomogeneous so as to yield a vanishing
voltage drop.
A Utah/Minnesota/Rice/Bell Labs group has now
tested this hypothesis in a clever bichromatic experiment using
microwaves at the two frequencies. Michael Zudov (now at the
University of Minnesota, zudov@physics.umn.edu, 612-626-0364) and
Rui-Rui Du (now at Rice University) sent microwaves of two different
frequencies at the electrons, observing that for nonzero-resistance
states the resultant resistance was the average of the values
corresponding to the two frequencies separately. On the other hand,
when the measurements included frequencies that had yielded a zero
resistance, the researchers observed a dramatic reduction of the
signal.
Judging from the average resistance observed for non-zero
measurements, they deduce that whenever zero resistance was
detected, the true microscopic resistance had actually been less
than zero. In other words, an observed zero resistance was masking
what was in fact an unstable negative- resistance state.
Zudov et al., Physical Review Letters,
16 June 2006
Contact Michael Zudov,
University of Minnesota
zudov@physics.umn.edu, 612-626-0364
Enlarge text
Shrink text
Print
E-mail
Digg this
Del.icio.us
Furl