Tuesday, August 08, 2006

This week in cond-mat

Two papers for now....
cond-mat/0608069 - Zhou et al., First direct observations of Dirac fermions in graphite
This paper is also in press at Nature Physics. The authors take angle-resolved photoemission spectroscopy (ARPES), and apply it to high purity graphite. ARPES is a very impressive technique - a really nice (highly collimated, bright, well-controlled energy - like from a synchrotron) x-ray beam is incident in a carefully controlled geometry on a sample, and the photoelectrons kicked out of the material are detected in an angularly resolved way. Applying conservation of momentum and energy lets one use this method to extract (2d) band structure information about the material. In high Tc compounds, for example, ARPES has contributed greatly to the understanding of "Fermi Arcs" and so forth. Anyway, these folks look at graphite, and find that massless Dirac fermions really do describe well the 2d band structure of this material. They also see some "boring" carriers in there, with parabolic dispersion (that is, energy proportional to the square of carrier momentum, indicating that the effective mass is a well-defined concept). Finally, they see signs that impurities and defects lead to electrons sitting in there. So, the electronic transport physics in this stuff is "rich", meaning very complicated. This is a good example of applying a highly refined tool to a new (yet very old) material system.

cond-mat/0608159 - Sellier et al., Transport spectroscopy of a single dopant in a gated silicon nanowire
The authors here have done a very elegant experiment. They've taken doped Si on insulator, and etched it to form an "island" with source, drain, and gate leads. That island contains a single dopant atom, and by performing low temperature conductance measurements, including significant magnetic fields, they've been able to look at two charge states of that single dopant, and compare with long-held models (D0 and D- configurations) of how dopants sit in Si. The single arsenic donor acts like an extremely small quantum dot, having electron addition energies exceeding 15 meV. This is the kind of experiment that is conceptually simple, but actually doing the work has real experimental challenges.


Alison Chaiken said...

We've known for perhaps 40 years that the dispersion of pi band carriers near the corner of the graphite Brillouin zone is linear. See for example this 1964 paper.

While I can see why it's amusing that carriers in solids would have dynamics described by the Dirac equation, I can't figure out why it's news right now. Perhaps like "colossal magnetoresistance" in magnetic oxides, this is merely a field whose time has come around again.

Doug Natelson said...

Sure. I'm pretty sure I have a textbook someplace (one of Harrison's?) that derives the linear dispersion using tight binding. I think the fact that people can now do transport measurements in single sheets is what has brought all this into trendiness. It's a bit like Bose-Einstein condensation. Noone seriously doubted that achieving the right density and temperature combination would lead to BEC of dilute gases of bosons, but now it's spawned a whole subfield because of the experimental and calculational prowess people now have.

Alison Chaiken said...

The great interest in BEC is due to the fundamental tests of QM and even phase transitions in solids that they have made possible. BEC are perfect solids or fluids on which long envisioned experiments can be performed for the first time.

To me the exciting aspect of BEC is the idea of using them to synthesize novel materials. Already you can "crystallize" a lattice of ions or even neutral atoms. You can also orient the bonds and select the atomic excitation state. Now suppose you brought two lattices of different species together. Could you make a metastable solid that is dynamically inaccessible through normal reaction methods? If so, is it possible that you could form materials that would persist when the lasers are turned off, in analogy to new phases made in diamond-anvil cells?

Anonymous said...

I'm not really sure what this graphite paper shows...

The WHOLE point of this recent work on single layer graphene is that new physics arises in an isolated sheet which is manifestly not there in bulk samples.

If I understand them correctly, these photoemission measurements were done on bulk crystals where it is known that small interlayer tunneling causes the bands to become parabolic at low energy scales (and more importantly to lose the pseudospin degeneracy).

So all these measurements demonstrate to me are the limitations of this photoemission technique. It is no great surprise that bulk graphite has bands which are at least approximately described by band theory. So, the technique obviously doesn't have the precision to see the small details - and the small details are where the interesting physics is. The rest we have known since 1964 or thereabouts.

Doug Natelson said...

Alison - Interesting timing that you should mention optical lattices + trapped gases as a means of implementing condensed matter systems. There's definitely a convergence of the two subfields these days. (Glad you're reading this blog, btw. I hope all is well these days. Still at HP?)

Anon. - I know what you're saying (if band theory was wildly wrong about bulk graphite we would have known long ago), though I did find the observation of the impurity-based component interesting.

Alison Chaiken said...

Interesting timing that you should mention optical lattices + trapped gases as a means of implementing condensed matter systems. There's definitely a convergence of the two subfields these days.

I was first inspired to think about this topic by Dave Wineland, who gave a talk 10 or 15 years ago about using a trap to make a 3D lattice of polystyrene spheres. The experiment was a crude demonstration of synthesizing a solid but nonetheless hinted at what might be possible.

The obstacle to ready synthesis of solids in traps is that the wavelengths of light that are typically used are much longer than bond lengths in ordinary solids. So to make a stable crystal in a trap, you'd have to figure out how to bring the atoms/ions together. I'm not an expert on lattices, but it must be possible to manipulate potentials in order to accomplish this convergence.

Glad you're reading this blog, btw.

I'm glad that a CMP is blogging in addition to all those cosmologists and string theorists.

I hope all is well these days. Still at HP?

There are real questions about the long-term future of research at HP Labs, but for the moment life there is sweet.