Automatic Recognition of Multiplet Patterns and Measurement of Coupling Constants in NMR and EPR Spectra through the Application of Maximum-Entropy Deconvolution

Abstract

Abstract Complex spin-coupled multiplets in EPR and NMR spectra have been deconvolved in an iterative manner using quantified maximum entropy (MaxEnt) as a pattern-recognition tool. This approach leads both to the identification of the coupling multiplicity and to a measurement of the values of the coupling constants. The technique has been applied to simulated electron paramagnetic resonance spectra of free radicals in isotropic solution and leads to the establishment of the size and multiplicity of the coupling constants present in a complex spectrum, together with their nuclear-spin values. This is achieved through the use of a series of progressively more complex peak-profile models. The method therefore provides a successive simplification of complex spectra by removing several couplings of a given multiplicity (e.g., doublets or triplets) at each stage of the deconvolution. Deconvolution of a complex spectrum, ultimately to a single line, yields a significant improvement in signal–noise, and this could be useful in situations such as EPR imaging. The approach has been shown to be applicable at typical experimental signal–noise levels. It has also been applied to the analysis of experimental NMR data exemplified by interpretation of the highly coupled 19 F NMR multiplets in perfluoro-6H-hex-1-ene, leads to complete interpreta- tion of the resonances from the olefinic fluorines, and also demonstrates the detection of effects caused by magnetic nonequivalence.