You may have heard that scientists have discovered ice on the moon.  Here’s what is going on.  Any molecules of water, oxygen, or any other thing floating around the moon perform ballistic orbits.  A molecule might bounce a dozen times before settling down or flying on out of the gravity well.  BUT if it randomly goes by the north or south pole, it slows down, because it’s cold there.  At the poles, the Sun is low on the horizon, and solar irradiation is long slant rays.  But the Moon is covered with craters!  This means that there are places sunlight never reaches.  Here’s a great video showing this, using what appears to be radar data to generate the 3d map including the bottoms of the craters – and the sun and shadows.

Regular photography could never show the interior of craters which were eternally dark, unless we set off a flare.

Deep in the Moon, the temperature is constant, and this modulates the surface temperature during the lunar day and night.  Subsurface temperature of -110C is typical.  But from the video, you can see that the interior of some craters never sees any sunlight, so the only heat there is that conducted up to the surface from below, and then radiated to space.  Measurements show that the crater near the North pole, Hermite, has “cold traps” as low as 26K – low enough for most gases to be solid ice which would never sublimate.  Helium and hydrogen would not be found, liquifying at 4K and 20K respectively.

There’s been speculation that man could modify the cold traps at the lunar poles so as to use them for his own purposes.  For example, we could put a big wall circular around one of the lunar poles.  This would completely block sunlight, so the entire area would become a cold trap and various useful volatiles would accumulate.  In a more practical scheme, amounts of regolith would simply be moved around to provide shading for areas in which scientists might want to enhance the cold.  There could be an area, typically a crater, which had good siting and which for engineering and navigation reasons was optimal, but which say had one side of the crater wall fallen in, and hence did not, without modifications, act as a cold trap.