Extraction of Individual Spectra from Mixture Data Based on High-Resolution 13C-13C NMR Correlation Spectrum and DECODE Procedure.

Abstract

In NMR analysis of complex organic molecules, low natural abundance of 13C and the low resolution of two-dimensional (2D) experiments are significant difficulties. Also challenging is the analysis of a mixture spectrum without separation, which has been limited to simple molecules. Through nonuniform sampling using modified heteronuclear multiple bond correlationcombined with indirect covariance, a high-resolution13C-13C correlation spectrum was obtained with 1H sensitivity. Built on the thus-obtained 13C-13C connectivities, deconvolution of the mixture spectra was achieved through a new signal-processing procedure, termed DECODE, tailored to the indirect covariance eigendecomposition. When applied to a complex natural product mixture of rotenone and brucine with many quaternary carbons, the method resolved very close carbon peaks and extracted clean individual spectra. Essentially providing molecule-wide 13C connectivities for complex molecules from 1H-detected 2D spectra, our approach should prove useful in many areas of NMR analysis.