Monday, July 03, 2006

This week in cond-mat

Just returned from the Electronic Materials Conference. Interesting, and generally much more oriented toward engineering than pure physics, but fun nonetheless. I'll be out of commission for the next week or so, so this blog entry will have to tide over my dedicated readership :-)

cond-mat/0606742 - Camino et al., Transport in the Laughlin quasiparticle interferometer: Evidence for topological protection in an anyonic qubit
In the fractional quantum Hall effect, in very clean two-dimensional electron systems (typically formed at the interface between GaAs and AlGaAs layers) at very low temperatures and particular large magnetic fields, the "normal" metallic state of the electrons is unstable. The particular values of magnetic field are those for which the ratio of magnetic flux through the sample (in units of h/e, the so-called flux quantum) to the density of electrons (number of electrons per cm^2) takes on special values, such as three or five halves (corresponding, respectively, to three flux quanta for each electron, and five flux quanta for each pair of electrons). At these special values of magnetic field, the electrons form a correlated state named after Bob Laughlin, who first wrote down a trial many-body wave function to describe it. In a Laughlin state, the electrons can't be treated as nearly independent, as in a normal metal. Instead, when one tries to probe the electronic system, one finds collective excitations (rather than simple electron-like excitations in a normal metal). These collective excitations have very funky properties: they can have fractional charge (in the three flux quanta per electron case, the excitations have charge 1/3 e) and obey fractional statistics.
Fractional statistics are funky. Swap two electrons, and the total wave function picks up a factor of exp(i pi) = -1. Swap two bosons (like two 4He atoms), and the total wave function of the boson system picks up a factor of exp(i 2pi) = 1. Swap two Laughlin quasiparticles, and the total wave function picks up a factor of exp(i alpha), where alpha depends on precisely which fractional state the system is in. Generically alpha can be anything, earning the nickname anyons for particles that obey such statistics.
This paper looks at conductance oscillations as a function of magnetic field in a patch of Laughlin electron fluid that should exhibit fractional statistics and fractional charge of 1/5 e. The authors claim that these oscillations are surprisingly robust as temperature is increased, and that this is evidence of special stability of that state due to topological considerations. I'm not sure I believe the final conclusions, which seem to depend in great detail on precisely knowing the electron temperature. It's a neat experiment, though, and gives real insight into some exotic quantum effects that people think might be useful for building a quantum computer.

cond-mat/0606802 - Costache et al., Spin accumulation probed in multiterminal lateral all-metallic devices.
The authors in this paper look in detail at the magnetoresistive properties of a little piece of aluminum connected to four separate cobalt electrodes. It turns out fortuitously that each of the four cobalt leads can have its magnetization switched independently of the others, and this lets the authors study effects that arise from pumping certain spin polarizations of electrons into the aluminum island. Since aluminum is a low atomic number material, spin-orbit scattering is pretty weak in there, so electrons can maintain their spin polarization for a while. These experiments require extremely clean interfaces between the Co and the Al to work, and provide concrete numbers for spin lifetimes and diffusion lengths in practical materials.

No comments: