Fluorine and phosphorus heteronuclear coupling

The most commonly encountered coupling in NMR spectra is due to ¹H, but other nuclei can induce splitting of signals as well. The previous post discussed how coupling to ¹³C at natural abundance produces the small ¹³C satellite peaks. The other heteronuclear couplings that are most likely to be observed are due to ¹⁹F and ³¹P. Examples of these in 1D ¹H and  ¹³C spectra are shown below.

Trimethyl phosphate has three magnetically-equivalent methoxy groups attached to a central phosphate atom. The 1D ¹H spectrum taken in D₂O (shown below) shows what at first looks like a singlet, as might be expected if considering just the hydrogens. However, closer examination shows that the peak is a doublet with a coupling of 13.3 Hz, due to the ³J_HP coupling.

The 1D ¹³C spectrum (shown below) shows a similar situation. At first glance the spectrum appears to consist of a single peak, as one might expect considering only the magnetically-equivalent methoxy carbons. However, the peak shows a splitting of 6.0 Hz, which can be assigned to the ²J_CP coupling.

The compound p-fluorobenzoic acid contains a single fluorine atom attached to an aromatic ring. Its 1D ¹H spectrum acquired in D₂O (shown below) shows the expected two aromatic resonances, but the couplings of these peaks are complicated by the presence of the fluorine atom. The downfield peak at 7.88 ppm appears as a doublet of doublets with couplings of 5.6 and 8.8 Hz. The 8.8 Hz splitting is likely a three bond coupling to the neighbouring ¹H, but the 5.6 Hz coupling is too large for a four bond coupling to the symmetry related ¹H, and must be due to the ¹⁹F.

The upfield resonance at 7.16 ppm appears as a triplet with a coupling of 8.9 Hz. This matches the larger coupling on the downfield resonance, consistent with ³J_HH coupling to the neighbouring aromatic ¹H. The second coupling of 8.9 Hz, again, is unlikely to be a long range coupling to the symmetry related ¹H and must be a ³J_HF coupling from the fluorine. Since the upfield peak shows a larger ¹⁹F coupling than the other, the 7.16 ppm peak is likely due to the hydrogens ortho to the ¹⁹F and the 7.88 ppm peak to the meta hydrogens.

Turning to the 1D ¹³C spectrum of p-fluorobenzoic acid (shown below) we see six peaks, when there are only five magnetically distinct carbons in the molecule. The two largest peaks can be assigned to the symmetrical protonated carbons. The expansions show that both of these peaks are in fact doublets. The more upfield of these two peaks at 117.7 ppm has the larger coupling (21.9 Hz), suggesting this peak is from the carbons ortho to the fluorine, while the peak at 134.0 ppm with the 9.1 Hz coupling is from the carbons meta to the fluorine.

The remaining peaks must all be due to quaternary carbons. The small peak at 135.1 ppm shows a coupling of 2.8 Hz, likely a fluorine coupling, and the small magnitude suggests this carbon is further from the fluorine than the four protonated carbons. Assigning this peak to the carbon para to the fluorine, leaves three peaks, all singlets, to assign to two quaternary carbons. The peak at 177.5 ppm is typical for a carboxylic acid, thus the last two peaks at 168.2 and 166.2 ppm must be a doublet from the fluorinated carbon, with a ¹J_CF of 248 Hz.

Heteronuclei such as ¹⁹F and ³¹P can produce confusing spectra if one is not familiar with their effects. The large size of these couplings, compared to ¹H coupling, needs to be kept in mind. Unlike any other elements, the NMR active isotopes of fluorine and phosphorous are 100% abundant and have spin ½, making them easy to detect by NMR. After ¹H, ¹³C and ¹⁵N, these two nuclei are the ones most likely to be encountered by organic chemists.

Acknowledgements
Spectra were obtained from the BioMagResBank entries bmse000774 for trimethyl phosphate and bmse00739 for p-fluorobenzoic acid.