N396_12c 149..152

8 collimations 109-209-209-209 we obtained a wavevector resolution of dQ=Q ˆ 0:0020. Inelastic neutron scattering. A 50-g powder sample was sealed with He exchange gas in a thin-walled aluminium can for the inelastic neutronscattering experiments at the NIST centre for neutron research. For energy transfers ,40 meV we used a direct geometry time-of̄ight (TOF) spectrometer in the neutron energy gain mode with a PG(002) double crystal monochromator set at Ei ˆ 3:55 meV and a cooled beryllium ®lter in the incident beam. The elastic energy resolution was 0.2 meV. Low temperature (10 K) neutron energy gain data served as a direct measure of the background. For energy transfer from 40 to 200 meV we used a ®lter analyser (FA) spectrometer with a Cu(220) monochromator surrounded by 609-409 horizontal collimation and combined with a cooled polycrystalline beryllium ®lter as the analyser. The relative energy resolution of this instrument was ,8% in the energy range probed. The fast neutron background in the FA experiment was measured and subtracted. We also subtracted a constant background arising from the ®nite transmission through the FA of elastically scattered thermal neutrons. The scale factor between the TOF and FA data was determined by comparing the integrated intensity associated with the 37-meV peak in the phonon spectrum of ZrW2O8 (Fig. 1). Absolute normalization of the TOF data was accomplished by measuring the count rate associated with incoherent elastic scattering from a known amount of vanadium. Analysis of neutron scattering data. In the limit of large wavevector transfer, Q, the one-phonon double differential scattering cross-section can be approximated by: