Mercury has a very low axial tilt, and substantial surface cratering. Any crater close enough to the poles is going to be in perpetual darkness, and thus very c

Mercury has a very low axial tilt, and substantial surface cratering. Any crater close enough to the poles is going to be in perpetual darkness, and thus very cold. Any volatiles vaporized elsewhere naturally travel until they form deposits on the very cold inner surfaces of such craters. Based on MESSENGER radar data, the Mercury south pole probably has more water than Mars. The temperatures at most of the craters would also support CO2 ice, although our current data can't prove it's existence or rule it out. Nitrogen supply would probably be rough. One of the interesting reasons to consider Mercury for colonization is that a colony on the inside rim of a crater would have very easy access to energy in many forms. Want natural-ish lighting? Raise a diffuse reflector on a pole. Want to melt something? Raise a mirror.