Friday, September 2, 2022

Extracting 1H-1H coupling constants

When publishing NMR data 1H-1H couplings are often reported. These are most easily obtained from a 1D 1H spectrum. Measuring the couplings can be made easier, and more accurate, if the spectrum is processed to enhance resolution.

The figure below shows two expansions of a doublet of doublets in a 1D 1H spectrum. The upper red spectrum was processed with the standard parameters, an exponential window with 0.3 Hz line broadening. The lower blue spectrum is the same data reprocessed with a resolution enhancing gaussian window using gaussian broadening of 0.7 and -3.0 Hz line broadening. The measured couplings in Hz are shown above the peaks. The gaussian window greatly reduces the signal intensity so the lower spectrum is scaled up by a factor of 16.

The benefits of the resolution enhancing processing are immediately obvious. While the large coupling in the doublet of doublets is easily identified in both spectra, the smaller coupling could be missed in the upper spectrum. More importantly, the values of the measured couplings are different. In the upper spectrum the broader peaks overlap more and shift their maxima closer to each other, resulting in reduced values. While overlap may not be completely removed in the lower, resolution-enhanced spectrum, it is greatly reduced, and the measured couplings will be closer to their true values.

Resolution enhancement may also allow measurement of couplings in peaks that would normally be labelled as "multiplets". The peak at 3.425 ppm shows no measurable splitting in the upper spectrum, but in the lower spectrum some of the lines are resolved. 

When measuring 1H-1H couplings from 1D 1H spectra I recommend recording a good quality spectrum with high signal-to-noise. The spectrum should then be processed at least twice; firstly, with a sensitivity enhancing window such as the exponential multiplication; and then with one or more resolution enhancing windows to obtain more accurate coupling constants.

Another approach is lineshape fitting of the experimental spectra. This gives the most accurate values for peak positions and linewidths, but is the most involved and likely only carried out for peaks of particular interest in a spectrum.

1 comment:

  1. Good topic and blogpost, Brendan. Re-apodization of your FID will not only resolve very small splittings in 1H NMR, but also very fine differences in chemical shifts in 13C NMR, e.g., the 35Cl/37Cl isotope shift of C-Cl signals. Here, the loss of S/N will be more noticeable as 13C NMR, of course, are often noisy to begin with. Also, we typically zero-fill the 1D FID before apodization and FT.

    ReplyDelete