Recently I was asked to help assign a 19F spectrum. The Skaggs NMRs are not capable of recording 19F spectra so I have not had much experience interpreting them. The extent of my 19F knowledge was an awareness of a large chemical shift range and large scalar couplings. Nevertheless, what follows is our rationalisation of the spectrum. Please let me know if you have a better explanation!
The figure below shows the full 19F spectrum recorded on the Department of Chemistry's JEOL 400. The large peaks near -65 ppm are a compound added for chemical shift referencing. The two peaks of interest appear near -107 ppm and in the lower inset an expansion is shown where it can be seen they appear as triplets. Integrations of the two triplets are nearly equal. Both triplets show the same coupling of 6.6 Hz and are separated by 238.3 Hz. The structure of the compound is shown in the upper inset.
Looking at the structure we first assumed that the fluorine atoms would be in magnetically equivalent environments due to symmetry. The triazopyrimidine substituent is likely perpendicular to the fluorinated ring and, if sufficiently planar, would leave the aromatic ring symmetrical. The two 19F peaks could then be the result of long range 19F-19F coupling. Coupling to the protons ortho and para to each fluorine might then create the observed triplets. In this explanation the 4JFF coupling would be 238.3 Hz and the 3JHF and 5JHF couplings would both be 6.6 Hz.
I have collected from the literature some typical 19F coupling values on the Facility website but none of these were helpful in assigning this example. The best source of small molecule NMR data I have found on the web is The Reich Collection. On this page of The Reich Collection we found reported values for 4JFF on an aromatic ring of 6.6 Hz, 8.9 Hz for 3JHF and 0.22 Hz for 5JHF - none of which supported our first rationalisation.
Assuming that the aromatic ring is not symmetrical and each triplet is due to one fluorine atom, the splitting pattern could be explained as the result of 4JFF coupling and 3JHF coupling, both of 6.6 Hz. The fact that the two couplings are coincidentally the same makes the peaks appear as triplets. In this rationalisation, the 5JHF coupling from the protons para to the fluorines would be too small to be detected. This explanation is more consistent with the reported J couplings found on The Reich Collection. It is also supported by the 13C spectrum which shows a signal for each carbon atom in the fluorinated aromatic ring, indicating that it is not magnetically symmetric.
Acknowledgments
Thanks to Thibault Alle in the Ballatore group for allowing use of his spectrum.
Brendan, I agree with your second interpretation which is consistent with atropisomerism: restricted rotation about the aryl-aryl C-C bond due to steric clash between the F and Cl atoms and/or EtNH and F groups. Consequently, the molecule is chiral (at least on the NMR time scale) and the two F nuclei are diastereotopic. The first interpretation is eliminated by a way-to-large separation in resonant frequencies (4J(FF) as you point out, should be no more than ~7 Hz). The energy barrier to rotation (Ea) determines whether separate chemical shifts will be observed (in this case, the two F atoms) at room temperature (rt). Sometimes, in natural products, atropisomerism is observed. For example, in the 1H NMR spectrum of the tetra-O-methyl achiral bastadin-12* (500 MHz, CDCl3, rt) the symmetrical 1,4-disubstituted-3,5-dibromophenyl ring is a ‘free rotor’ – the H-2 and H-6 aryl H signals appear as a two-proton singlet at 7.65 ppm. At –30 ˚C, rotation is slowed to the point that the aryl signals are diastereotopic and observed at 7.43 and 7.85 ppm, with a mutual 4J(H-H) = 1.8 Hz. Interestingly, the diastereotopism is observed for a difluoromethyl group (CHF2) in chiral molecules. In the 19F NMR (470 MHz) of chiral 13-debromo-13-difluoromethyl-agelastatin A, (compound 1m) we observed CHF2 with 2J(F-F) = 308 Hz and 2J(H-F) = 55 Hz. [Stout, et al. J. Med. Chem. 2014, 5085-93]. Presumably, the 2J(F-F) has a negative sign.
ReplyDelete* Andersen, et al. J. Nat. Prod. 1990, 1441-6]. Here, the parent compound (structure 2) is named ‘bastadin-9’, but for other reasons, later renamed ‘bastadin-12’.
" Interestingly, diastereotopism is observed for a difluoromethyl group (CHF2) in chiral molecules. In the 19F NMR (470 MHz) of chiral 13-debromo-13-difluoromethyl-agelastatin A, (compound 1m) we observed CHF2 with 2J(F-F) = 308 Hz and 2J(H-F) = 55 Hz. [Stout, et al. J. Med. Chem. 2014, 5085-93].”
ReplyDeleteIt was Paige Stout who brought this to my attention when she was trying to interpret the 1H NMR (500 MHz) of her synthetic compound, 1m, where the CHF2 signal appeared as a triplet (dd, 2J (H-F) = 55 Hz)
BTW, for the same reasons, the H atoms in the difluoroaryl ring (Thibault’s compound) are diastereotopic: two signals will appear in the 1H NMR spectrum, each a dd (3J(H-F) ~9 Hz and 4J(H-H) ~2 Hz), although the chemical shift differences will be small.
P.S. I noticed in your 19F NMR spectrum the downfield signal is shorter in height than the upfield which suggests, to me, that there is an unresolved H-F coupling which is broadening the former signal. Perhaps resolution enhancement through apodization will reveal this as 5J(H-F) or as different 4J(H-F) magnitudes for each of the 19F signals?
P.P.S. On a JEOL 400, 19F NMR resonates at 376 MHz.
I did try enhanced apodisation to try and resolve additional small couplings but didn't get anything. I was actually looking at the upfield triplet as the left most line has a notch in the top, but I could not resolve anything more than that.
DeleteHmm, might just be dynamical broadening rather than unresolved splitting, although I’d expect the two F signals to behave similarly as they are part of the same rotor.
DeleteHere’s an interesting article on 19F NMR of CF3- groups in amides of what we now call ‘Mosher’s acid’, PhC(CF3)(OMe)COOH (doi:10.1139/v82-354). The CF3 group is ‘frozen out’ at –100 ˚C so the 19F NMR shows an ABC pattern with crisply resolved 2J(F-F) couplings (average ~110 Hz), larger in magnitude than what we observed for CHF2-C in 1m. In other words, CF3- becomes a ‘chiral methyl’ analog at low temperature.
The salient point in these examples is access to measured 2J(F-F) values which is non-obvious in simpler fluorinated compounds.
Does the aromatic protons also show as two sets of peaks?
ReplyDeleteThe compound is likely to have a locked conformation like bi-aryl atropisomer. If heated up to high enough temperature I think you may see either the peaks start to merge together or at least some chemical exchange signals in 19F-19F NOE between the two 19F chemical shifts (or 1H-1H NOE between the two Ar-H if they are separated).
I don't know for certain that the aromatic 1H peaks were not symmetrical as well, but I am pretty certain they were. I suspect the temperature needed to cause coalescence of these two peaks would be rather high.
Delete