Tuesday, November 27, 2007

"Unparticles" and condensed matter

At the risk of contributing to what has recently been called the intellectual wasteland that is the physics blogosphere, I want to point out a nice review paper on the arxiv, and its connection to high energy physics. Subir Sachdev at Harvard has put up a relatively pedagogical review about quantum magnetism and criticality. Back when I was a grad student, I didn't appreciate that quantum magnetic systems were so interesting - I thought that they were a zoo or menagerie of semi-random compounds that happened to have effective model Hamiltonians of interest only to rather esoteric theorists. Now I understand the appeal - the relevant Hamiltonians can have some truly bizarre solutions that can be relevant (intellectually if not directly) to whole classes of systems. One class of such systems is the heavy fermion compounds that are non-Fermi liquids, and another comprises some exotic "spin liquids". The low energy excitations of these strongly correlated quantum systems are not readily described as particle-like or wave-like. They don't have simple quantum numbers and simple dispersion relations, and they result from complicated, correlated motion of electrons (or spins, or both). This has been known in condensed matter circles for some time, and is very neat. Much exciting theory work is being done to come up with good ways to treat such systems.

What I don't understand, and perhaps a reader can enlighten me, is how these ideas relate to "unparticles". Howard Georgi, also of Harvard, made a pretty big splash this past year by publishing a PRL (linked above in free form) about the possibility that there may be fundamental excitations of quantum fields (like the ones thought to be relevant in high energy physics) that are not simply described as particles. Since this paper came out, there are now 78 papers on the arxiv that deal with unparticles. So, is this a case of high energy physics reinventing an idea that's been known conceptually for some time in condensed matter? Or is there really a basic underlying difference here? I should point out that at present, while there is experimental evidence for non-particle-like excitations in condensed matter, there is not yet any evidence for such things in high energy experiments as far as I know.

Friday, November 23, 2007

Really? Seriously?

Sometimes I read a science article online or in the newspaper that I think is poor. This one I just don't know how to interpret. Lawrence Krauss is a solid guy, a very strong public advocate for science, and a very good popularizer of physics. Still, the idea that our observations of dark energy have somehow collapsed the quantum state of the entire universe is, umm, nuts on the same level as saying that the moon doesn't exist if no one is looking at it. There's no question that there are subtleties in worrying about applying quantum mechanics to the universe as a whole. Still, this carries the "spooky action at a distance" idea a bit far.

Saturday, November 17, 2007

This week in cond-mat

Two papers this week. I'll write about our own at a later date. These two are both connected to on-going long-term controversies in condensed matter/mesoscopic/nanoscale physics.

arxiv:0711.1810 - Capron et al., Low temperature dephasing in irradiated metallic wires
In the orthodox picture of metals (thought to be valid for relatively weak disorder), the quantum coherence time of electrons is expected to diverge toward infinity as the temperature approaches zero. Think about a two-slit experiment for electrons. If the electrons are well isolated from their environment, they can diffract off the slits and land on the screen, producing an interference pattern. If the electrons are coupled to environmental degrees of freedom that can change their state when the electron goes by, the relative phase of the electron wavefunctions going through each of the slits gets scrambled by that interaction, washing out the interference. In the usual 2-slit experiment, the degrees of freedom are those of detectors at the slits. Within a disordered metal, those environmental degrees of freedom can be lattice vibrations, other electrons, or magnetic impurities. For a decade now there has been an ongoing controversy about whether the coherence time (as inferred from some quantum correction to the classical electronic conductance) really does diverge, or whether it saturates as T -> 0. Intrinsic saturation would be a big deal - it would imply that the quasiparticle picture of electrons (Fermi liquid theory) fails at the low T limit. In this paper, the authors perform a very careful control experiment, looking at whether structural damage to silver nanowires can, by itself, introduce extra degrees of freedom that cause decoherence. They get this damage by ion-implanting Ag ions into Ag nanowires. The results show no sign of extra decoherence due to this irradiation.

arxiv:0711.1464 - Baenninger et al., Low-temperature collapse of electron localisation (sic) in two dimensions
Another ongoing brouhaha has been about whether electrons confined to two dimensions have an insulating or metallic ground state in the presence of any disorder. Without interactions, the "Gang of Four" (Anderson, Abrahams, Ramakrishnan, and Licciardello) showed that even infinitesimal disorder leads to localization and an insulating ground state for an infinite 2d system. Of course, real electrons do interact with each other, and real systems are of finite size. One big complication in this whole discussion is in trying to tell the difference between a real, uniform, insulating state and the breakup of your system into inhomogeneous "puddles" of electrons due to the disorder potential. The Cambridge group has done some careful experiments in mesoscopic samples of rather clean 2d electron gas, and they've found that small regions with higher temperature resistances far exceeding the quantum of resistance (~ h/e^2 ~ 26 kOhms) can show a crossover at low temperatures to what looks like a metallic state. I haven't been following this controversy in detail, but these data look very interesting, and I will have to read this closely.

Monday, November 12, 2007


A small selection of links....

This game is very addictive, educational, and as you play, you feed the hungry (albeit extremely slowly).

Now this is a nanotube radio! Rather than having the nanotube just be the nonlinear element responsible for demodulating the AM signal on the carrier wave, this one has the nanotube acting as the antenna and amplifier as well, effectively. I heard Alex Zettl get interviewed on NPR about it.

The FSP has an interesting post about ambition. Physics as a discipline has issues with this, with an historical attitude that anything less than a tenured job at Harvard is somehow inadequate - a notion that's wrongheaded and sociologically unhealthy.

Schlupp has a post about a little frustrating science journalism. It is a shame that sometimes the media can't tell the difference between good science or engineering and crackpottery. On a plane last week I had someone (who realized I was a physicist from my reading material) ask me about the guy who can get hydrogen from seawater by hitting it with microwaves. Kind of cool, yes. Source of energy? Of course not - it takes more microwave power to break the water into hydrogen and oxygen than you can get back by burning the resulting hydrogen. It's called thermodynamics.

Monday, November 05, 2007

This week in cond-mat

Several entries from the arxiv this week. My descriptions here are a bit brief b/c of continued real-world constraints.

arxiv:0711.0343 - Dietl, Origin and control of ferromagnetism in dilute magnetic semiconductors and oxides
arxiv:0711.0340 - Dietl, Origin of ferromagnetic response in diluted magnetic semiconductors and oxides
These are two review articles by Tomasz Dietl, one of the big names in the dilute magnetic semiconductor (DMS) game. DMS are semiconductor materials that exhibit ferromagnetic order usually because of doping with transition metal atoms that contain unpaired d electrons, such as manganese. The idea of integrating magnetic materials directly with semiconductor devices, and ideally controlling magnetism via electrical or optical means, is quite appealing. However, it is very challenging to achieve high magnetic ordering temperatures (e.g., room temperature) and decent electronic properties at the same time. In many systems the high doping levels required for the magnetism go hand in hand with lots of disorder, in part because crystal growth must be performed under nonequilibrium conditions to force enough transition metal atoms to sit on the appropriate lattice sites. Anyway, these articles (one coming out in J. Phys.: Cond. Matt.
and the other coming out in J. Appl. Phys.) should give you plenty of reading material if you're interested in this area.

arxiv:0711.0218 - Leek et al., Observation of Berry's phase in a solid state qubit
In basic quantum mechanics we learn that particles are described by a complex wavefunction that has a phase factor. Propagation of a particle in space racks up phase at a rate proportional to the particle's momentum. As Feynman would tell us, each possible trajectory of a particle from A to B then contributes some complex amplitude (with a phase). The total probability of finding the particle at B is the squared magnitude of the sum of all of those amplitudes, rather than the classical sum of the probabilities of each path. Phase differences between paths lead to interference terms, and are the sort of thing responsible for electro diffraction, for example. Besides propagating through space, there are other ways of accumulating phase. In the case of the Aharanov-Bohm effect, the vector potential leads to an additional phase factor that depends on trajectory. In the general case of Berry's Phase, the slow variation of some external parameters (such as electric fields) can lead to a similar geometrical phase factor. The intro to this paper gives a nice discussion of the classical analog of this in terms of moving a little vector on the surface of a sphere. Anyway, this team has used a solid-state superconducting qubit to demonstrate this geometric phase explicitly. Quite nice.

arxiv:0710.5515 - Castelnovo et al., Magnetic monopoles in spin ice
One of the things that I find so interesting about condensed matter physics is the idea of emergent degrees of freedom. For example, phonons (quantized sound waves) are quantum mechanical quasiparticles in solids that can have well-defined quantum numbers, and arise because of the collective motion of large numbers of atoms. In a more exotic example, Cooper pairs in ordinary superconductors are objects with spin 0, charge -2e, yet are "built" out of electrons plus phonons. In a very exotic example, the quasiparticles in the fractional quantum Hall effect can have fractional charges and obey exotic statistics. In an even more extreme case, these authors propose that there are quasiparticle excitations in a kind of magnetically ordered insulator that act like magnetic monopoles. It seems that magnetic monopoles do not exist as elementary particles. Indeed, they would require a modification of Maxwell's equations. (In this solid state system the argument is that they exist as monopole/antimonopole pairs, so that the net divergence of the magnetic field is still zero). "Forbidden" particles emerging from the collective action of many electrons - a very neat idea, and it would appear that there may even be some experimental evidence for this already.