Resolution of the Cationic Distribution in Synthetic Germanite Cu22Fe8Ge4S32 by an Experimental Combinatorial Approach Based on Synchrotron Resonant Powder Diffraction Data: A Case Study and Guidelines for Analogous Compounds

The present work investigates the cationic distribution of a complex CuS compound belonging to a promising class of materials with significant prospects for thermoelectric generator and photovoltaic applications. We propose the use of powder resonant X-ray scattering in a combinatorial experimental approach with X-ray powder diffraction and single crystal , as a valid, general, and practical method to unravel the complex cation ordering encountered in the synthetic germanite Cu22Fe8Ge4S32. The generation and testing of all the possible structural models is rapid and rigorous as it is based on two modules written in python language: a module that takes care of model creation and ordering (permutation algorithm associated with some filtering and ordering algorithms) and a python parser for the refinement software (FullProf) input and output. Each possible cationic model is tested through combined Rietveld refinement of resonant X-ray data at selected edges and analyzed considering its global chi(2) (Bragg contribution) agreement factor. This approach can easily integrate information coming from other techniques, as in our case EXAFS spectroscopy and single crystal X-ray diffraction. In spite of the complexity of the germanite, we were able to define the main characteristics of the cationic ordering (space group P (4) over bar 3n): (i) the "interstitial" site 2a is fully occupied by Fe, (ii) Ge is located only on the 6d site (or the symmetry equivalent one 6c), and (iii) the remaining Fe atoms are located preferentially on the 12f site and possibly on the 6d (or 6c) site along with Ge. Moreover, we single out a probable enrichment of Ge at the expense of Fe. The approach developed in this case study can be used as a guideline for the crystal structure resolution of analogous compounds.