August 27, 2008
by Arvin Moser, Team Manager, Application Scientists, ACD/Labs
Like driving a car, the driver navigates the vehicle within the designated lanes on a roadway, thus using the lane markers as a reference. The same can be said when analyzing spectral data. Correctly referencing multiple NMR data is critical in a structure elucidation process. For the case where 1D and 2D NMR data is present, the typical approach is to reference a correlation in the 2D NMR data to a resonance from a 1D NMR spectrum.
The spectral data below shows two 1H-13C HMBC experiments (optimized for 8 Hz) differing in the reference point. From the correctly referenced HMBC, there are 4 correlations indicating H to C bond separation of 2, 3 or 4 bonds away: 1H resonance 3.71 ppm to carbons 127.1 and 140.5 ppm and 1H resonance 3.86 ppm to carbons 132.7 and 142.1 ppm (illustrated by the green arrows on the structure).
The incorrectly referenced HMBC shows the same 4 correlations, however, the H-C bond separation is unreasonable for data. Although the 2 and 3-bond separation is coincidently workable for the 1H resonance at 4.14 ppm to carbons 127.1 and 140.5 ppm, it is the 1H resonance at 4.29 ppm to the carbon resonances at 132.7 and 142.7 ppm that shows unreasonable 5 and 6-bond separations. Ultimately, if the 2D NMR data is not correctly aligned to the 1D NMR data, then the corresponding fragments most likely cannot be pieced together to complete a structure.
