Improved quantitative crystal-structure comparison using powder diffractograms via anisotropic volume correction

Abstract

Crystal structure prediction (CSP) aims to determine the experimentally isolable crystal structure(s) of a molecule given only its 2D molecular diagram. The ability to match candidate structures to known experimental structures is critical in benchmarking CSP methods. In this work, a new approach to improve comparison of crystal structures using their calculated powder X-ray diffractograms (PXRD) is presented. The protocol involves anisotropic volume correction of the compared structure to that of the target. Its ability to distinguish matching structures from other candidates is assessed using the submissions to the 6th CSP blind test. The anisotropic volume correction is found to surpass currently available methods of PXRD comparison in its ability to separate similar from dissimilar structures. This is demonstrated by its ability to distinguish a polytype from a target structure, and by the identification of two uncredited matching structures in the 6th CSP blind test. The developed method yields a quantitative measure that is as useful as the root-mean-square deviation (RMSD) in atomic positions for structure comparison.