TEACHING CRYSTALLOGRAPHY TO THE CHEMIST

Richard Cooper ¹, Clare Keats ¹, Markus Neuburger ², Keith Prout ¹, David Watkin ¹

¹Chemical Crystallography Laboratory, 9 Parks Road, Oxford, UK, OX1 3PD.
²Laboratory for Chemical Crystallography, University of Basel, Switzerland.
E-mail richard.cooper@chem.ox.ac.uk

Keywords: crystallisation, crystal growth, teaching, computing, refinement.

The increasing use of area detectors for single crystal X-ray data collection has brought do-it-yourself structural analysis within the reach of the bench chemist. Tools are being developed to aid the novice crystallographer in all areas of crystal structure analysis from crystallisation to structure solution and refinement.

Techniques of crystallisation that can be easily employed by bench chemists are being developed. These include simple small-scale solvent evaporation using a selection of solvents that cover most of the solvent classes. A database has been designed to store information such as crystallisation technique, solvent properties, size and morphology of crystals obtained. A system of ranking the results of crystallisations has been devised. This can give a quick indication of promising systems and the likelihood of the compound exhibiting polymorphism.

Other techniques under development include an adaptable temperature cycling apparatus and prediction of vapour diffusion conditions.

Once a suitable single crystal has been obtained and the X-ray data collected, the chemist is presented with a sophisticated graphical user interface for structural analysis. A 3D visualisation of the structure is visible at all times which can be used to edit the structure and refinement conditions. In normal mode the user can command the program using menus which thoughtfully disable any options that are not possible. In scripted mode the program will guide the user through a refinement, evaluate its progress and advise the user if things go wrong. Dialog boxes can be created for entering options to common instructions such as TLS analysis, idealising fragment geometries, least squares restraints, and searches for false refinement minima. All scientific rationale is programmed in external ASCII text files so that it may be adjusted to suit local preferences.