Mixture analysis is crucial in many areas of chemistry, and a wide variety of separation methods are in use. A common method for physical separation is high-performance liquid chromatography (HPLC), but resolution is a problem: chemically similar species coelute. An alternative approach is diffusion-ordered NMR spectroscopy (DOSY), in which the signals of mixture components are separated according to the diffusion coefficient. Again, separation is limited if species diffuse similarly or have overlap in their NMR spectra. Using the two techniques in combination can resolve both NMR spectra and the elution profiles of individual components, even where both techniques fail when used in isolation. Recording diffusion NMR data as a function of HPLC retention time gives a three-dimensional (3D) data set that can be analyzed using multiway statistical methods. PARAFAC analysis of diffusion NMR data measured from HPLC eluate for commercial "monoacetin" (a mixture of glycerol and its mono-, di-, and triacetates) yielded fully resolved and quantitative NMR spectra and elution profiles for all four components, whereas neither HPLC nor diffusion NMR applied independently was able to resolve the components.
Read full abstract