No consensus has been reached so far about the hydrogen anomaly problem in Compton scattering of neutrons, although strongly reduced H cross-sections were first reported almost 20 years ago. Over the years, this phenomenon has been observed in many different hydrogen-containing materials. Here, we use yttrium hydrides as test objects, YH2, YH3, YD2 and YD3, Y(HxD1−x)2 and Y(HxD1−x)3, for which we observe H anomalies increasing with transferred momentum q. We also observe reduced deuteron cross-sections in YD2 and YD3 and have followed those up to scattering angles of 140° corresponding to high momentum transfers. In addition to data taken using the standard Au-197 foils for neutron energy selection, the present work includes experiments with Rh-103 foils and comparisons were also made with data from different detector setups.
The H and D anomalies are discussed in terms of the different models proposed for their interpretation. The 'electron loss model' (which assumes energy transfer to excited electrons) is contradicted by the present data, but it is shown here that exchange effects in scattering from two or more protons (or deuterons) in the presence of large zero-point vibrations, can explain quantitatively the reduction of the cross-sections as well as their q-dependence. Decoherence processes also play an essential role. In a scattering time representation, shake-up processes can be followed on the attosecond scale. The theory also shows that large anomalies can appear only when the neutron coherence lengths (determined by energy selection and detector geometry) are about the same size as the distance between the scatterers.