ORDER AND DISORDER IN SOLID OXYGEN - NEUTRON-SCATTERING INVESTIGATIONS

Citation
F. Dunstetter et al., ORDER AND DISORDER IN SOLID OXYGEN - NEUTRON-SCATTERING INVESTIGATIONS, Fizika nizkih temperatur, 22(2), 1996, pp. 140-157
Citations number
83
Categorie Soggetti
Physics, Applied
Journal title
ISSN journal
01326414
Volume
22
Issue
2
Year of publication
1996
Pages
140 - 157
Database
ISI
SICI code
0132-6414(1996)22:2<140:OADISO>2.0.ZU;2-L
Abstract
Solid oxygen exists at zero pressure in three crystallographic phases, denoted alpha, beta, gamma. Various results have been obtained by neu tron scattering in the two disordered beta and gamma phases. Plastic g amma-O-2 has a A15 structure with 8 molecules per unit cell. The molec ules, located at 2a and 6c sites, exhibit two different types of disor der of molecular axes: a spherical-like disorder for the 6c molecules, and a disc-iike disorder for the 6c molecules. A precise diffraction analysis on large single crystals shows that the disorder of molecular axis of the molecules can be interpreted in terms of localized forbid den orientations, or orientational holes, with a wide disorder among t he other orientations. An important contribution of the translation-ro tation coupling must be taken into account for the 6c molecules. Analy sis of diffuse scattering by both an approximate analytical model and molecular dynamics simulation shows an important contribution to the o rientational correlations between 6c-6c and 2a-6c molecules. Magnetic disorder can be investigated by neutron polarization analysis of the p aramagnetic scattering. The 3D long range antiferromagnetic order of t he alpha phase becomes a 2D short range helicoidal order in the beta p hase, with a very short 0.5 nm correlation range. In the gamma phase, one observes, as in the liquid, a strong antiferromagnetic correlation , with a suggestion of a 1D order along the linear chains of 6c molecu les. In all the measured phases, the variation of the scattering inten sity with scattering vector q cannot be reproduced at high q by the kn own magnetic form factor. This suggests the need for a better descript ion of the pi orbital and the theoretical magnetic form factor of the molecule.