High-field NMR has overcome the problem of spectral ambiguity in nitrogen-rich compounds, thanks to efforts by Canadian scientists. The team has studied two anhydrous polymorphs of the stimulant, caffeine, and has found that, despite extensive disorder, both caffeine polymorphs reveal the characteristic structural signatures of crystalline compounds and show them to be different from the earlier powder diffraction results and the molecular modelling structures.

Polymorphism, variations on the crystalline theme within the same chemical formula, has become a rather important attribute of chemical structures, particularly in the pharmaceutical arena. Indeed, SpectroscopyNOW covered previously the issues surrounding the polymorphs of the almost ubiquitous drug aspirin and how these might be exploited in new formulations and applications. Closer inspection of various compounds have revealed previously reported crystal structure to have been at best misinterpreted, while a study of almost 15000 crystal structures published in major academic journals and lodged in the Protein Data Bank revealed worrying numbers of structures to be incorrect. Errors were most common in those journals usually considered to be the most prestigious.

Now, Gary Enright, Victor Terskikh, Darren Brouwer, and John Ripmeester, of the Steacie Institute for Molecular Sciences, at the National Research Council Canada, in Ottawa have used two techniques - single-crystal X-ray diffraction and ultrahigh-field solid-state ¹³C NMR spectroscopy - to reveal that the structure of one of the most well-known compounds, caffeine, betrayed earlier investigators. The polymorphs of caffeine studied cannot be classified as glasses or plastic phases, the researchers say. Unique positional disorder observed in these two caffeine polymorphs is best described as rotator phase crystals. As in aspirin, hydrogen bonding seems to play a crucial role in defining the packing motif in caffeine crystals.

As most crystallographers know, X-ray structures are not infallible when it comes to polymorphism, but NMR can help corroborate results or moreover provide unique insights into a structure that is otherwise ambiguous or inaccessible to diffraction techniques.

"One of the most attractive points of this research were results shown in Figure 8 in our paper," Terskikh told SpectroscopyNOW.

"Many organic compounds contain nitrogen," says Terskikh, "caffeine, for instance, has four nitrogen atoms. The problem is, that when you study these with solid-state NMR at low magnetic fields, ¹³C resonances from carbons attached to nitrogens are always broadened and split." This complicates the interpretation of such spectra significantly, he adds. However at higher fields this broadening and splitting "magically" disappears with no additional efforts from the spectroscopist. "As a result spectra now are very highly resolved," says Terskikh, "Of course this works not only for caffeine, but for many other systems, such as tetracycline."

Terskikh points out that the spectrum of a tetracycline, although recorded in the solid-state, is more reminiscent of a spectrum recorded in the solution phase. At lower magnetic fields at least four lines will be broadened simply because of the bonding to nitrogen. "This may help in assigning these lines," he explains, "but when there are several components or polymorphs in the mixture, things can quickly get very messy, at higher magnetic field, once again, this issue is resolved. "Our results may encourage other researchers to try high magnetic fields in their nitrogen-rich systems too," adds Terskikh.

Related links:

• Cryst Growth Design, 2007, 7, 1406-1410 (cover article)
• Dr John Ripmeester Page
• National Ultra-high Field NMR Facility for Solids
• C&EN report on crystal structure quality 

Article by David Bradley