The CCDC has been at the forefront of the systematic analysis of crystal structures for many years. As one might expect, a particular focus has been on the analysis of symmetry in crystal structures. Fundamentally, the CCDC has been striving to determine why molecules pack in the way that they do. The complex interplay of molecular shapes, molecule interactions and the influence of symmetry have all been addressed. Another strand of research has been the assessment of both the accuracy and precision of molecular structures arising from crystallographic experiments.
The Cambridge Structural Database (CSD) now contains over 600,000 structures and in this vast collection of structures is a wealth of data on the standard uncertainties of parameters reported in crystallographic studies. Previous studies have shown that when unit cell parameter standard uncertainties are analysed over a statistically significant number of crystallographic structures, these uncertainties seem to be quoted to a much higher degree of accuracy than is realistically achieved. Using the extensive and comprehensive data stored in the CSD, a much larger statistical analysis may be made, than was ever previously possible. Given the volume of data available it will be possible to amass a large set of the identical structures allowing very accurate normalisation factors for the cell dimensions. This will provide crystallographers with a better measure of the true standard uncertainties and allow us to investigate how the uncertainties affect more fundamental values such as bond lengths and angles within a crystal structure.
Recent software developments have focussed on the need to compare, match or filter similarities between crystal structures, offering valuable tools to the Crystal Structure Prediction and the Crystal Engineering communities. A current focus is on extending these tools to investigate packing relationships that might exist across the crystallographic landscape. A method has been developed to decompose molecules into a representation of their moment of inertia tensor. This allows the packing of crystal structures to be overlaid like-for-like, irrespective of chemical composition, to quickly reveal similarities and differences in real space. A research project, building on the theme of understanding similarity to tease out important factors that direct molecular crystallisation is underway. It is hoped that molecular packing descriptors will be developed that allow some degree of prediction of structure from molecular properties.
A further research project, in collaboration with Chris Hunter at the University of Sheffield will be to investigate the important connection between structure and thermodynamics. In general, methods to analyse structures at the CCDC are based on interactions geometries and probabilities. The approach developed at Sheffield is based on the thermodynamics of solution phase non-covalent chemistry.
The aim is to establish correlations between the solid state structural parameters and solution phase thermodynamic parameters that allow us to combine aspects of the two approaches and use parameters from different sources interchangeably.