My source was MFMP, having been told by Piantelli.
I'm not clear on how the hydrogen was used in this case it seems, after re-reading the article.
I'm guessing that the nickel sample was previously exposed to hydrogen. Subsequent low temperature exposure to liquid helium would then cause the hydrogen to desorb in atomic form from the sample, which is where something unexpected could happen.
In later patents by Piantelli et al. like for example EP2754156B1 it's hinted that as long as hydrogen atoms are dissociated and ionized in the process, a transition metal lattice is not strictly required to observe the effect, which is mostly about creating a special reactive form of hydrogen. Piantelli calls it "H-", but it might be something different than the name suggests.