Towards solution and refinement of organic crystal structures by fitting to the atomic pair distribution function

Abstract

A method towards the solution and refinement of organic crystal structures by fitting to the atomic pair distribution function (PDF) is developed. Approximate lattice parameters and molecular geometry must be given as input. The molecule is generally treated as a rigid body. The positions and orientations of the molecules inside the unit cell are optimized starting from random values. The PDF is obtained from carefully measured X-ray powder diffraction data. The method resembles `real-space' methods for structure solution from powder data, but works with PDF data instead of the diffraction pattern itself. As such it may be used in situations where the organic compounds are not long-range-ordered, are poorly crystalline, or nanocrystalline. The procedure was applied to solve and refine the crystal structures of quinacridone (β phase), naphthalene and allopurinol. In the case of allopurinol it was even possible to successfully solve and refine the structure in P1 with four independent molecules. As an example of a flexible molecule, the crystal structure of paracetamol was refined using restraints for bond lengths, bond angles and selected torsion angles. In all cases, the resulting structures are in excellent agreement with structures from single-crystal data.

Orientational disorder of monomethyl-quinacridone investigated by Rietveld refinement, structure refinement to the pair distribution function and lattice-energy minimizations

“…R PDF values cannot be compared with R values from single-crystal or Rietveld refinements. In our experience a R wp PDF value of 30-40% for an organic compound denotes a reasonable fit (Prill et al, 2016). It is remarkable that some of the models gave a very good fit to the PDF data [e.g.…”

“…The structures match very well. Obviously, the geometrical accuracy of the structure from the PDF fit is quite high, as was already observed by Prill et al (2016). In contrast, the structural model G [ Fig.…”

“…In contrast, disordered molecular systems have only occasionally been investigated by PDF methods: examples include C 60 (Elschner et al, 2011), SF 6 (Tucker et al, 2007) and sodium p-chlorobenzenesulfonate monohydrate (Wolf et al, 2012). A refinement of atomic coordinates by fitting to the PDF curve has hardly been performed for organic compounds (Rademacher et al, 2012;Prill et al, 2015Prill et al, , 2016.…”

Oscail, a program package for small-molecule single-crystal crystallography with crystal morphology prediction and molecular modelling

“…It has recently been shown that the application of a pair distribution function, PDF, transformation to synchrotron XRPD data can be used to extract structural information from nanocrystalline samples of organic compounds which do not give Bragg diffraction (Prill et al, 2016). Below the structure determination level, the application of a PDF transform to XRPD data obtained using copper-based laboratory X-ray equipment improves the classification of amorphous samples and the principal component analysis, PCA, of sample processing history (Nollenberger et al, 2009;Boetker et al, 2012).…”

xINTERPDF: a graphical user interface for analyzing intermolecular pair distribution functions of organic compounds from X-ray total scattering data

“…) which are the building blocks of organic compounds. The atomic pair distribution function (PDF) calculated from X-ray (and/or neutron, electron) total scattering has been demonstrated to be a valuable tool for investigating structures of disordered and amorphous organic compounds (Shi et al, 2017;Prill et al, 2015Prill et al, , 2016Benmore & Weber, 2011;Rademacher et al, 2012;Chen et al, 2014;Gorelik et al, 2015). Although existing tools such as DiffPy-CMI (Juhá s et al, 2015) and XISF (Mou et al, 2015) can be used for this problem, a new software program that provides a user-friendly graphical user interface (GUI, as opposed to script-based programming in DiffPy-CMI) and analyzes the data in real space (as opposed to in reciprocal space in XISF) is still of great value.…”