Abstract. The interaction of H2 with materials is relevant to areas such as catalysis, hydrogen storage and fuel cells. Inelastic neutron scattering has been extensively used for such studies, which have generally utilised the J 01 rotational line at 14.5 meV of parahydrogen. Unfortunately, this is very difficult to model. The problem would be mitigated by studying the fundamental H-H stretch vibration at 516 meV, because the calculation becomes a conventional lattice dynamics problem, which is tractable with density functional theory. Here, we have chosen to investigate solid and liquid H2 and D2 as a test of whether it is possible to observe the H-H and D-D stretch. This is only possible with a direct geometry spectrometer at a spallation neutron source, since only this combination of facilities gives access to high energy neutrons with a wide range of momentum transfer. The measurements were successful and, in addition to several rotational states of the ground vibrational state, a feature recoiling from an origin value of 516 meV for H2 and from 370 meV for D2 is observed in both the solid and liquid states. These are assigned to the H-H and D-D stretch respectively.
IntroductionThe study of molecular hydrogen (H2, dihydrogen) in, or on, materials is relevant to areas such as catalysis [1], hydrogen storage [2] and fuel cells [3]. Inelastic neutron scattering (INS) is well-suited to studies of such systems because the large 1 H cross section and the transparency of most materials to neutrons means that most of the information obtained directly relates to the adsorbed dihydrogen. These studies are almost invariably carried out with the J 01 rotational line at 14.5 meV of parahydrogen as the probe of choice [4]. The reasons for this are that not only its transition energy [4] but also its underlying line shape [5,6] are sensitive probes of the local environment and the instruments generally have their best resolution in this energy region.The disadvantage of studying the rotational transition is that it is very difficult to model. This has been done with either a phenomenological model of how the rotational energy levels behave in an applied potential [4] or have required very sophisticated multi-dimensional ab initio calculations [7]. These difficulties could be reduced by studying the fundamental H-H stretch vibration at 516 meV, because the calculation is now a conventional lattice dynamics problem, which is tractable with density functional theory [8].