Homodecoupling

Homodecoupling is a technique for selectively removing coupling in 1D NMR spectra. Often 1D ¹H spectra show complex splitting patterns that are difficult to interpret. Selectively eliminating one coupling may allow a multiplet to be interpreted or enable measurement of couplings that could not otherwise be quantified. Recently, we used homodecoupling to assign the stereochemistry of a pair of enantiomers.

A reaction to produce the synthetic intermediate shown below gave two compounds. Before proceeding to the next synthetic step it was necessary to identify the stereoisomers. Selective 1D NOE experiments were inconclusive so homodecoupling was tried.

One of the stereoisomers was used to record the spectra shown in the figure below. The reference 1D ¹H spectrum is in blue at the bottom. It shows a doublet at 4.79 ppm due to the amine, a doublet of doublets at 4.63 ppm from the unknown stereocenter, and a quartet of doublets at 4.40 ppm from the (S) stereocenter.

Homodecoupling at 4.63 ppm gave the middle, red spectrum. Artifacts are present where the homodecoupling was applied, but the multiplet at 4.40 ppm has lost the small coupling and now appears just as a quartet. In the upper, green spectrum, where homodecoupling was applied at 4.40 ppm, the multiplet at 4.63 ppm has now lost the small coupling and appears as a simple doublet.

Since homodecoupling the multiplets at 4.40 and 4.63 ppm removed the same coupling, this small coupling must be due to their interaction with each other. The size of the coupling indicates that the two protons are on the same face of the molecule and thus identifies the geometry of the unknown stereocenter as (S). A similar set of experiments was recorded on the other reaction product and found to support the assigned stereochemistry.

To homodecouple NMR spectra composite pulse decoupling is applied during acquisition. This scrambles the magnetisation, effectively removing its influence from the spectrum. To selectively apply the decoupling the power level must be chosen so that it only affects a small region of the spectrum. The power level is typically chosen from a calibration curve. We have calibration curves available for both our spectrometers.

Acknowledgments
Thanks to Mitch Christy from the Siegel lab for use of his data.