Sunday, October 27, 2013

Two striking results on the arxiv

This past week, I noticed two preprints on the arxiv that particularly got my attention.

First, "How many is different?" In condensed matter physics, we like to point to Phil Anderson's famous 1972 paper "More is different" (pdf) as an important tract about the nature of our field:  Fascinating, rich, profound physics can emerge from collective, interacting, many-particle systems.  As I've said before, one water molecule is not wet.  A large ensemble of water molecules can be.  The preprint at hand looks specifically at a classic statistical mechanics problem, the ideal Bose gas, and summarizes results that show that it takes, for a cubic box, exactly 7616 particles for there to be a real thermodynamic liquid-vapor-like transition.  (Note that in this case there are no interparticle interactions; it's the identical nature of the particles that leads to this effect.)   A spherical box would lead to a different criterion.  Still, it's pretty neat to me that sometimes we can come up with a direct answer to "how many particles does it take before we see an emergent collective response?"

The other paper is "Reaching Fermi degeneracy by universal dipolar scattering".  Truly cooling the atoms is a major challenge in the trapped-atomic-gasses community, particularly those people who want to use optical lattices filled with fermionic atoms as a way to simulate electronic phases relevant to condensed matter.  A phenomenon called "Pauli blocking" is a real hindrance.  As a Fermi gas cools, there are fewer and fewer empty single-particle states into which you want to scatter the excited ("hot") atoms.  (Fermionic atoms can only get into empty states, because the Pauli exclusion principle forbids multiple occupancy.)  These authors are able to leverage magnetic dipole-dipole scattering within a cloud of erbium atoms to get to temperatures below 0.2 times the Fermi temperature - quite cold for these folks.  If this can be adapted to other systems and optical lattices, that would be very exciting.  (For reference, room temperature is something like 0.005 times the Fermi temperature of copper, and lots of exciting solid state electronic physics requires such low temperatures.) 

1 comment:

Tobias said...

Along the lines of the first paper is a recently published work from Selim Jochim's group (paywalled): His apparatus is very interesting, as he can "dial in" the exact number of Fermions that are interacting in his trap, to study emergent many-body behavior.