When the President of the IUCr gave the introduction to the 35th International School on Crystallography at Erice, Sicily, in June 2004, he listed particular divisions of the subject as X-ray Crystallography, Neutron Crystallography and Electron Crystallography. This list, whilst probably not intended to be exhaustive, is now rather obviously defective in not containing the increasingly important area of NMR crystallography.
NMR is an exceedingly important tool for chemical structure determination and also for the study of molecular-level mobility. It is accepted as by far the most powerful technique for such investigations for chemical compounds in the solution state. This situation arises from the fact that NMR spectra show exceedingly high resolution, thus distinguishing between atoms in closely similar chemical sites.
The situation is very different for solids, which, without the use of special techniques, generally give rise to broad unresolved resonances. Therefore, in the period 1955-1975 chemists largely ignored NMR of solids. However, the situation was revolutionised in 1976 by the development of the suite of techniques for 13C NMR (and subsequently for other nuclei) comprising cross-polarisation (CP), magic angle spinning (MAS) and high-power proton decoupling (HPPD). Since that date these techniques have become widely used and many refinements and developments have been produced. The result is that much crystallographic information can now be derived from NMR and the technique is increasingly important in this area.
NMR methods are sufficiently advanced that complete crystal structures have now been determined for the first time without other information. There is now truly a subject which may be called “NMR Crystallography”. This is generally complementary to diffraction crystallography, as demonstrated by a number of the above points. Unfortunately, NMR efforts in the investigation of crystallography have been very slow to be integrated with those of the rest of the crystallographic community, so that many “diffraction-based” crystallographers do not consider solid-state NMR as a natural part of their suite of techniques. Whilst it is mildly encouraging to see that the program for the Florence Congress this August contains sub-sections under two of the Congress main topics (though both, for some inexplicable reason, appear to be limited to macromolecules), these are the only mentions of NMR and there is nothing under the main topic number 01 “Instrumentation and Experimental Techniques”.
Perhaps now is the time to consider constituting a division of NMR crystallography within IUCr. Certainly a formal or semiformal link with the solid-state NMR community is desirable.
Robin K. Harris, University of Durham, UK
A few points about NMR and its potential applications to crystallography are as follows:
•NMR responds to the short-range environment of relevant atoms and is not directly influenced by long-range order.
•It can therefore be applied to amorphous materials as well as crystalline ones, though with broader lines for the former to encompass the variations in nuclear environments.
•It can be readily used to determine the chemical nature of a solid compound, including crystallographically important information such as conformation and tautomeric form.
•Chemical shifts give information about intermolecular interactions.
•Inter- and intra-molecular hydrogen-bond linkages can be identified.
•Information on crystallographic asymmetric units is especially readily available, usually merely by counting lines.
•Polymorphs are usually easily distinguished.
•Phase transitions can be monitored.
•Crystallographic disorder is detectable, and distinctions between spatial and temporal disorder can be made.
•Motions such as internal rotation and ring inversion can be detected and their rates obtained, even in cases of mutual exchange (e.g. 180° ringflips of phenyl groups).
•More general information about molecular-level mobility can be obtained by measurements of relaxation times.
•Measurement of dipolar coupling constants yields through-space inter-atomic (i.e. internuclear) distances, though these will be modulated by local mobility.
•NMR data can be used as restraints in carrying out full structure determination from powder diffraction data.
•Heterogeneous materials can be studied and selective spectra for particular domains obtained by the use of special pulse sequences.
•Intensities can be made quantitative (e.g. for polymorph ratios or crystallinity proportions).
•NMR is a multinuclear technique, each relevant isotope and element having its own specific frequency range.
•NMR experiments can be tailored to produce particular results by suitable choice of pulse sequences.