'Correlated disorder' is not disorder and can change lattice energies by >2.5 kcal mol −1

Read full text Summarise with VIRAT
Authors: Jacco van de Streek
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

X-ray diffraction experiments measure an electron density that is a time and space average, and quantum-mechanical calculations can be used to deconvolute this time and space average into a detailed picture of molecular crystal structures at the atomic level.While continuing to validate our temperature-dependent lattice free energies [1], which have carefully calibrated error bars of approximately 0.25 kcal/mol for ordered, Z'=1 crystal structures of chemical compounds containing between 40 and 70 atoms, we encountered a number of puzzling outliers for crystal structures that had been published as 50:50 disordered.We spent several years searching for examples of disordered molecular crystal structures in the literature and calculating crystal structures, powder diffraction patterns, lattice energies and lattice free energies with a variety of structural models, and we realised that the outliers were due to a type of disorder that is largely unknown and fundamentally misunderstood in much of the crystallographic community: the disorder is only visible in the average crystal structure, the crystal structure has to be doubled in one direction and is then fully ordered in at least that one dimension, often even in two dimensions; we have dubbed this "correlated disorder".We were able to show that although the experimental average electron density is modelled correctly (so all published experimental crystal structures are still correct), some molecular crystal structures published as 50:50 "disordered" are on a local level not disordered at all, but require a doubling of the unit cell resulting in a fully ordered atomistic model that in terms of atomic coordinates, unit-cell parameters and temperaturedependent lattice free energy allows for excellent agreement between density functional theory calculations and experiment whereas any traditional disorder model does not [2] (see Fig. 1).For most crystal structures published as disordered, we were able to prove the exact nature of the disorder (correlated or not) through the use of quantum-mechanical calculations of the crystal structures, the lattice energies and/or the lattice free energies.Some cases were so subtle that we have not yet been able to establish beyond reasonable doubt what the nature of the disorder is.

Fields of Study & Subjects

Crystallography and molecular interactions X-ray Diffraction in Crystallography Crystal structures of chemical compounds

Publication details

Journal / source
Acta Crystallographica Section A Foundations and Advances
Volume
81
Issue
a1
Pages
e104–e104
Publication date
30 March 2026
Language
EN