Laser-Cooled Atoms: Cesium

Element: Cesium (Cs)

Atomic Number: 55

Mass: One stable isotope, mass 133 amu.

Laser cooling wavelength: 854nm, but see below.

Doppler cooling limit: 125 μK.

Chemical classification: Yet another alkali metal, column I of the periodic table.
This one isn’t greyish, though! It’s kind of gold color. Still explodes violently in water, though.

Other properties of interest: The definition of the second in the SI system of units is in terms of the microwave transition between hyperfine ground states in Cs– 9,192,631,770 oscillations to one second, to be precise. Has a really large scattering length compared to the other alkalis. In particular, it has a huge “inelastic” collision rate in a magnetic field, meaning that a pair of colliding atoms easily flip the spin state of one, releasing a lot of energy and kicking both out of the trap.

History: This is the penultimate alkali (which, by the way, is the name of my next band)– the only one we haven’t talked about is potassium, and we may yet get there. Cesium was a very early target for laser cooling experiments, because of its importance as the standard reference for the SI second. Back in the 60’s or so, Jerrold Zacharaias had pointed out that a lot of problems with atomic clocks could be made better by orienting the clock vertically, and letting the atoms go up and fall back down through a single microwave cavity. Unfortunately, if you’re using room-temperature atoms, this isn’t really practical– the atoms all fly away off to the sides, and don’t make it back down.

Laser cooling allows you to change that, and make a real “fountain” clock— if you can cool the atoms to microkelvin temperatures, or even colder, then the sideways expansion isn’t really a problem. And a few other issues also go away completely, making laser cooling a great asset for clock development. Which is why so much of the work was done at and funded by NIST and the Navy– they have an intense professional interest in clocks. Happily, cesium has a laser cooling transition in the near infrared, at 854 nm, which is a relatively convenient laser wavelength. And, indeed, the best primary clocks in the world are now laser-cooled fountain clocks; the first came on-line in Paris in the 1990’s, and all the major national standards labs have fountains these days.

Unfortunately, a lot of the other properties of cesium kind of suck. The collision rate is huge relative to other alkalis, meaning there’s a large density-dependent shift, so if you try to boost your clock signal by packing more atoms in, the frequency changes. The giant inelastic collision rate also means it’s really difficult to get BEC by the easiest methods– when you try to pack cesium atoms into a magnetic trap at the densities needed for BEC, they collide at huge rates and your whole sample goes *poof*. (That’s a technical term…)

This didn’t stop people from trying– the middle image in the set above is from this very old page at Stanford, showing a college classmate of mine (at least, I think that’s Jamie) wedged into a tiny space to make adjustments to the cesium apparatus they were installing in Steve Chu’s lab. Jamie’s not a big guy, so that’s really an impressively cramped space, even by AMO physics standards… A lot of people tried a lot of different tricks to get BEC in Cs before Rudy Grimm’s group succeeded (Grimm, by the way, is someone who gets overlooked a lot (incredibly, because he might be bigger than I am), but has been involved in an amazing number of “firsts” in laser cooling…).

Once the basic trick for condensing cesium was worked out, it turns out to have some fairly nice properties for basic BEC stuff– the phenomenon that makes the annoying collision rate happen in the first place also makes it fairly easy to change the collisional parameters, so it’s used for a lot of investigations of changing coupling strengths. There are also some nice applications for cesium in molecules, with several people pursuing applications involving heteronuclear diatomics, mostly RbCs.

Random fun things: The melting point of cesium is right around human body temperature, so if you hold one of the glass ampules it comes in for a while, you can melt the stuff, which is kind of cool. It’s a little risky if you’re holding the ampule while waiting to load it into a vacuum system, though, as you can easily melt it by accident then end up making a mess.

The British spell it with an extra “a” (“caesium,”), which I find surprisingly annoying for no immediately obvious reason.

Art: The cartoon version of cesium is a naked woman on fire. The Comic Book Periodic Table comes up empty, alas.