The GC results look interesting. However, in a typical thermal conductivity gas analysis, the peaks for a standard of hydrogen are negative going as hydrogen has a higher thermal conductivity than the He carrier gas. Instruments are set-up to invert the peaks in a certain time frame. Did the authors have the instrument invert peaks in a window - i.e. around H2 elution but not before or after? If a window, than perhaps the window was too narrow and for some reason the retention times of the gases shifted slightly. Thus, the split peak could be an instrumental artifact where the system is only reversing the polarity in a too narrow of a window. Why they should have a positive going peak before the negative peak would not be consistent with this explanation unless the positive peak is deuterium or HD or switching polarities causes an electrical artifact. The peak shapes of the sample is not the same as the standards, which causes me concern. The presentation mentions running the TCD at 60C to make the peaks positive. I an too ignorant of TCD detectors to know if that statement is correct. A brief reading of the literature indicates that hydrogen should retain its thermal conductivity difference over He at most temperatures so to me this statement does not make sense.
If the authors are observing hydrogen by MS, then what is the expected binding of the electrons in this shrunken hydrogen? The MS typically has an electron energy of 70eV (mean) so if the ionization potential is above this, then the MS should not show the shrunken hydrogen. Coupling the MS to the GC would have been helpful and varying the electron energy of the MS would also have been helpful to show that the expected gas is somehow different from ordinary hydrogen gas in the appearance of the H2+ peak. i.e. plot electron energy vs. signal level for the two gases and show that the onset energies are different. (This is best done with specialized equipment but most MS can vary this number and if the Ionization potential of the molecules are quite different then you should see this difference. You cannot get and absolute number without comparison to a standard. However, a relative number would suffice in this case and maybe just using the methane or CO2 as a reference would work)