Most of the articles I've read on electric thrusters mention that Xenon is generally, with some exceptions, used as the propellant (or would it be termed reaction mass?). They never mention why though.
Whats so special about Xenon?
(I'd postulate its something to do with ionization energy vs atomic mass perhaps?)
You want a gas so you don't need to expend energy vaporising the propellant. You also want the gas to be as dense as possible so you can get as much impulse per unit volume of propellant as possible. It's also nice if the gas is inert and non-corrosive so you don't need to worry about it degrading or corroding whatever you're storing it in. Finally it's nice if it is easily ionised.
Xenon fits all these criteria. Its density is around 5.9 kg/m$^3$, and obviously it's inert because it's a noble gas. The first ionisation energy is about 12eV and the second is 20eV so it's relatively easy to produce Xe$^+$ ions.
The only gas I know of denser than xenon is sulphur hexafluoride, but it's only slightly denser and is more reactive. I suppose radon would be better still if only it weren't radioactive. The only trouble with xenon is that it's expensive, but then if you can afford to launch a probe to the asteroid belt you can probably afford the xenon to fuel it.
Noble gases have the advantage of being chemically inert, so that they are less likely to react with atoms in the electrostatic grids. Since ion thruster to date have been deployed only on unmanned transports, regular maintainance is not an option. Because of that, Noble gases are favoured over, say, hydrogen
One reason to pick Xenon over Argon though, might have to do with it having larger atomic mass per ionic charge (ionized plasmas in thrusters are mostly singly-ionised species) they have worse ISP, but higher thrust. But I'm not sure that is the only reason involved
