Another Path to Degeneracy

One of my colleagues mentioned this in group (as in meeting, not therapy) — research that has formed a Bose-Einstein Condensate directly from laser cooling and trapping: Laser cooling to quantum degeneracy

In order to form a BEC, you have to get enough atoms together in a potential well, and with a low enough energy in order to trigger the transition that leaves the bulk of them in the ground state (that’s the condensation). Normally, atoms under such confinement will distribute themselves among the various quantized energy states, but under the proper circumstances the condensation will happen. The tipping point occurs when the phase space density — the number of atoms in a volume defined by the cube of the deBroglie wavelength (which is momentum dependent, so it’s related to the kinetic energy) is above some critical value. The trick, then, is getting enough atoms cold enough and confined such that their deBroglie waves are overlapping.

Up until now, the path to BECs have involved a stage of magnetic trapping, but since magnetic traps are shallow (energetically speaking) you laser-cooled and trapped atoms to load into the magnetic trap. The trap “walls” were then lowered, allowing the most energetic atoms to escape, much like the most energetic molecules leave a hot cup of water, so this is called evaporative cooling. It results in the average energy, and thus temperature, going down. A drawback is that you also lose atoms, but if the phase-space volume decreases faster than atoms are lost it results in the phase space density going up, and a condensate.

So why not simply laser-cool the atoms? The problem with that is that atoms absorbing the photons that are cooling them also re-emit those photons, and when the physical density is large, those emitted photons hit other atoms rather than escaping, and this limits both the density and the temperature. This interaction can’t be present where the condensate is formed.

So the researchers turned it off, in a manner of speaking. They added a second form of trap, a dipole trap, which doesn’t involve the absorption and emission of photons — the light is far from resonance, and you use the electric field gradient of the light to form the trap. In the region where this trap was present, the laser scattering was turned off by shifting the energy of the excited state, which you can do if you add an external field. This was accomplished by the addition of another laser, which gives rise to the AC Stark shift (Johannes, not Tony), or a light shift. The trapping light is no longer near resonance, so the scattering is greatly suppressed. It’s also localized, so it’s only in effect where you shine the laser.

The area where the trapping is still occurring still contains very cold atoms, so the atoms in the dipole trap can still collide and thermalize with them, but once that second trap fills sufficiently, the critical phase-space density is achieved and a BEC forms. Very neat.