No matter how expensive the spectrometer, we seem never to have enough resolution. Even ‘small’ molecules can pose problems, especially in mixtures. Multiplet structure is a mixed blessing: it gives us valuable structural information, but all too often it leads to a thicket of overlapping signals. Pure shift NMR, in which we temporarily sacrifice set aside the information contained in spin-spin couplings, suppresses the effects of homonuclear couplings to leave just one peak per chemical shift – hence the name. For proton NMR this can improve resolution by up to an order of magnitude, and a wide range of pure shift methods are already in use. Some recent developments, published1-3 and unpublished, will be presented, including pure shift methods involving nuclei other than 1H.
Selective excitation methods offer another route to improve resolution, but once again multiplet structure poses problems: ideally we would like to excite a single chemical shift, not a single frequency. The GEMSTONE variation on chemical shift selective filtration (CSSF)4 allows us to do this particularly efficiently, opening up a range of ultraselective 1D analogues of 2D experiments.5-7 Finally, data processing methods can both optimise the precision of measurements and take us beyond normal instrumental limits on resolution.