Tritium has a half life of around 12 years, and they saw 53 excess counts over a year. To be explained by tritium decay, there would only need to be 1000 atoms of contamination. These types of low background experiments are build from some of the purest and cleanest objects ever manufactured. However, electron recoils weren't a focus in the design of the experiment, much of which is focused on reducing backgrounds. Tritium is just hydrogen chemically, which has a tendency to exchange with hydrogen in organic compounds, and can be innadvertantly captured by metals and later released. This is a pretty pernicious problem to deal with and i really feel for the students. If it is tritium, letting everything run for a long time and looking for a time dependent effect is a possible approach, but this isn't very compatible with phD timelines, and the xenon will probably be in demand elsewhere.
