Analysis of Complex High-Resolution NMR Spectra by Sophisticated Evolutionary Strategies

Abstract

The NMR spectra of orientationally ordered molecules increases rapidly in complexity with the number of coupled nuclear spins. For example, the spectrum of n-pentane as solute in a nematic liquid-crystal solvent (12 coupled proton spins) consists of roughly 20,000 transitions, many of which overlap. The analysis of such complicated spectra by line-assignment techniques is fraught with difficulty. However, application of the ideas of genetic evolution via evolutionary strategies makes possible the analysis of spectra, which are quite probably not solvable by ordinary methods. In particular, the covariance matrix adaptation evolution strategy uses the idea of mutation and the results of previous trials to give solutions to quite complicated spectra. The techniques described herein have already led to solution of NMR spectra that were not achieved with older techniques, and show promise for cracking even harder problems in the future. An important aspect with NMR spectra of solutes in liquid-crystal solvents is the presence of a broad underlying baseline from the solvent protons. An automatic baseline removal method is described.