"If there is any hydrogen or ice caught by sunlight, even for a short while, the hydrogen would have disappeared a long time ago."

As you probably know, there is hydrogen and even very small quantities of water is in the lunar regolith which is in full sunlight (for about 14 days) per day on the Moon. There is also other volatiles in lunar regolith.
As far as hydrogen at lunar poles I don't think it's cold enough to be held there unless it's trapped in regolith like it is in the more equatorial regions. And don't know why significantly more would be trapped in lunar polar regions as in comparison to equatorial regions. As far as ice being exposed to sunlight, it would depend how much this sunlight warmed the area (the water molecules). If only part of the sunlight from the sun (say a sliver of the top of the Sun) reach the area it would take longer to warm up the water molecules, if the duration of this partial sunlight exposure was say limited to a couple hours per month, this might not allow enough time to raise the temperature of the ice so it evaporates. And of course with reflected sunlight one is also dealing with with intense energy from the Sun- though with a large enough reflecting area and long enough duration it could cause temperatures to get high enough. I would also wonder if large enough solar flares could influence the temperature. And I also wonder about high energy particles in general & how it would affect water ice in polar region.

It must be deep in permanent shadow. Perhaps there is a different IR signature between regolith, anorthosite, brecchias and any hydrogenous composition, under the same darkness.

I'm sure there is, the question is can this IR camera detect these differences with sufficent resolution- it's cold, therefore in IR light it's going to be dark- are they expecting to see large patches of ice on the surface?