The structure of tetraiodoethylene at 4 K

Abstract

The crystal and molecular structure of tetraiodoethylene (C214) at 4 K has been determined by neutron powder diffraction with Rietveld's profile-refinement method. Three independent data sets at different wavelengths were collected and used in the calculations. The measurements were sufficiently accurate to locate the C atoms in the molecules, and to show that the C=C bonds in both molecules are fully ordered and aligned approximately parallel to a*. The dimensions of the unit cell at 4 K and several interatomic distances are reported. The estimated standard deviations derived by this method are discussed and a normalization and weighting scheme for combining results from multiple power-data sets is suggested.1765

Disorder in crystalline tetraiodoethylene; constrained refinements of neutron powder diffraction data

“…The present work consists of an analysis of roomtemperature powder neutron diffraction results and a reanalysis of the data already published by Haywood & Shirley (1977). A more positive result can be obtained than in the earlier work, made possible by the use of constraints in the refinement procedure.…”

“…We therefore conclude that at 4K one of the molecules [the second one as listed by Haywood & Shirley (1977)] is partially disordered whereas the other molecule is fully ordered.…”

“…Thus three parameters for each of the two molecular sites are required to orient all the molecules, and only two more parameters are necessary for the disordering on each site. The total number of positional, orientational and disorder parameters is thus 11, compared with 20 used by Haywood & Shirley (1977). Even this comparison is not accurate as the present model allows disorder of the whole molecule whereas Haywood & Shirley include it only on the C positions.…”

“…In this situation it is impossible to resolve the close pairs of atoms, and therefore these atoms cannot be refined independently. In the work of Haywood & Shirley (1977) the I atoms occur in such close pairing that they did not attempt to separate them. Any such attempt would result in an illconditioned least-squares matrix, probably causing large and unphysical positional shifts.…”

“…Instabilities arise when two parameters are trying to do the job of one, and it is quite clear that in the present case there is no such duplication of duty. ~cu~on counts Two refinements were penormed with the 4 K data of Haywood & Shirley (1977), the first assuming total order as they reported, and the second allowing disorder on both molecular sites. The results given in Table 1 indicate a better agreement with the disordered model, showing that one molecule is completely ordered whereas the other is disordered to about 21%.…”