return to homepage
Today the structural analysis of small to medium sized molecules relies heavily on X-ray crystallography. The newest generation of diffractometers and software allow the determination of a structure within a few hours thus making crystallography into a routine experiment in many areas of chemistry.
However, this analysis still demands single crystals of the compound to be investigated and unfortunately many compounds do not crystallise easily to give samples for X-ray crystallography. Therefore, the art of growing suitable crystals becomes the limiting factor and challenge for structural analysis. The understanding behind the reasons for a particular arrangement of molecules in crystal structures is still very limited and the prediction of the crystal structure from a given molecular structure is generally not yet possible. The major driving forces for a preferred crystal structure are attractive and repulsive intermolecular interactions in addition to the need for dense packing to ensure as little "empty" space in the crystal as possible. However, the molecular structure of many compounds does not allow the ideal packing in a crystal lattice and additional molecules often from the solvent are incorporated into the crystal to create additional attractive interactions (e.g. hydrogen bonds) or simply to fill any "holes". Often the solvent may not have the ideal functional groups or shape for this purpose and crystals with large disorders result affording poor structural data.The ultimate aim is to find general trends to allow suggestions of suitable solvents or solvent additives in the crystallisation of certain compound classes.During the past seven years, I have used X ray crystallography to solve structural problems in collaboration with several groups. In many cases the solvent disorders have resulted in poor refinement. Investigation of the crystal structure and the solvent disorder has lead at least in one case to a proposal for a better solvent (R.D. Köhn, G. Kociok-Köhn, M. Haufe, Chem. Ber. 129 (1996) 25-27).
The crystal structures of the trisamides [M(N(SiMe3)2)3] (here M=Cr) are often plagued by severe disorder of solvent molecules in large channels between the amides. Originally, this channel appeared to be filled by CMe4 or similar hydrocarbons from petroleum ether solvent that was too small to fill the channel and was therefore severely disordered. Simple replacement of the solvent by the larger Me3SiSiMe3 resulted in a much better "fit" and a good refinement.The Chemistry Department of the University of Bath has recently obtained a new Nonius KappaCCD diffractometer and will get a second one early in 2001 and I have already started to work as a crystallographer in Bath as a visiting postdoctoral researcher. So far, about 280 structures out of a total of 27 structures showed disorder in the solvent molecules with the exception of the case cited above, none of these structures were reinvestigated with alternative solvents. For these, and future samples, suitable solvent molecules would be proposed that would help to lock the disorder. Recrystallisation of the compounds in those proposed solvents and subsequent X-ray crystallographic investigation should allow a comparison of the crystal structures with varying solvents.