Recent first principles calculations find that (H2)i is sited at tetrahedral positions in the silicon lattice aligned along [100] or [111], and possessing a molecular stretch frequency directly comparable to the 3789 cm-1 mode observed by infrared and Raman spectroscopy. The observation of a single IR and Raman active mode is consistent with the molecule being aligned along [110], and unable to rotate. If rotation were not hindered, then, for Td symmetry, a minimum of at least two IR lines would be expected in each of the H2, and D2 treated samples. The [100] alignment can be also ruled out, since only a single Raman active mode would be observed in this case, and no observable infrared signal. The [111] alignment would give a single Raman and IR active mode, but this alignment can be discounted since this line would split in HD samples.