THE METASTABLE ORANGE MODIFICATION OF HgI2
Marc Hostettler, Henrik Birkedal, and Dieter Schwarzenbach
Institute of Crystallography, University of Lausanne, BSP,
Dorigny, CH-1015 Lausanne, Switzerland
Marc.Hostettler@ic.unil.ch
Keywords: Diffuse Scattering, Stacking Faults, Twinning, Simulation, Disorder
Mercuric Iodide (HgI2) crystallises in three different modifications at room temperature: the thermodynamically stable red modification, the yellow modification, which is related to the high temperature phase, and the orange modification. The orange modification appears to be metastable over the entire temperature range. The orange crystals are mechanically unstable, and generally change to the red form in the course of some days. However, with the use of an Image Plate diffraction system it has been possible to collect a full set of data. Undistorted reciprocal layers have been reconstructed from the images. From the reconstructed layers it is seen that the diffraction pattern contains a wealth of diffuse scattering superimposed on Bragg reflections.
An idealized model proposed by Schwarzenbach [1] described the structure in terms of stacking faults characterized by Markov chain probabilities. The new data show features not observed previously and not explained by the initial model. We here propose a more detailed model which consists of two different structures stacked along a crystallographic axis. The stacking faults in this direction are responsible for the diffuse lines with periodic intensity distributions. These diffuse streaks are explained by the change of the atomic positions of mercury from one structure to the other. The previously unobserved features are explained by small displacements of the mercury atoms from the ideal positions supposed by Schwarzenbach.
Simulated diffraction patterns obtained with DISCUS [2] are presented. The program is used to calculate the Fourier Transform of a stack of the two structures, where each structure has its own lattice and space group. This powerful tool allows testing different stacking probabilities versus the experimental data.
Currently we are developing a program for the refinement of several structures on the same set of diffraction data. All sharp features in the diffraction patterns are indexed in one cell and integrated as if arising from pure Bragg diffraction. The program treats the structures as independent, i.e. the scattered intensity is taken as the volume weighted sum of the scattered intensity of each contributing structure. The present state of such refinements will be presented and discussed.