) Structure elucidation of minute quantities of a compound remains challenging. The best method for carbon skeleton determinations is a 2D INADEQUATE NMR spectrum. Despite recent method improvements, like our sensitivity enhancing analysis software marketed by Varian, Inc., sample sizes of 50 milligrams and acquisition times of several days are often required. Connectivity information from short-range and long-range proton-carbon correlation 2D NMR spectra can be used to derive molecular carbon skeletons. This method can increase sensitivity by three orders of magnitude compared to 2D INADEQUATE. The objective of this project is to replace the involved mental puzzle required to interpret this, usually ambiguous, data by an algorithm (AssembleIt) that lists, in order of decreasing agreement with constraints, all structures at least partially consistent with the spectral correlations. The method is computationally efficient because it generates only those structures that could satisfy the constraints, then attempts to resolve the ambiguities to the extent possible. The goal of the proposed research is to optimize the method to produce the true molecular structure by analyzing known compounds, to experimentally determine the limitations of this method, and to explore the use of additional NMR and other spectral information in the automated structure generation process.
NMR is a leading analytical tool and an over ten fold increase in its sensitivity for carbon skelton determinations allows solving structures from samples under 1 mg which are not solvable any other way. The resulting analysis complexities can be automated leading to a more efficient, cost-effective, and sensitive characterization of unknown molecules. Possible application areas include natural product characterization, metabolism studies, and organic synthesis control.
| Dunkel, Reinhard; Wu, Xinzi (2007) Identification of organic molecules from a structure database using proton and carbon NMR analysis results. J Magn Reson 188:97-110 |
