This is a beautiful piece of data re-interpretation. A few years ago, there was a report published about a compound which supposedly demonstrated the only instance of a solution O-H...F-C hydrogen bond. This seemed to provide some support for hydrogen bonding involving fluorine.
Now a Spanish group has published a nice paper in CC that provides a refutation and re-interpretation of the data that along with some calculations, indicates that the observed data is not due to a C-F...H-O hydrogen bond, but simply due to steric hindrance that makes the three fluorines of a CF3 group non-equivalent. In the former interpretation, it was assumed that the non-equivalence of two Fs of this group with the third F indicated that the third F was involved in hydrogen bonding. The new interpretation says that it is steric hindrance that prevents rapid rotation of the CF3 group, and makes the three Fs non-equivalent. Calculations support the interpretation.
Thus, now we will have to look for other instances in which there is bonafide C-F...H-O bonding. Quite a neat piece of careful data analysis supported by crystallography and quantum chemical caclulations.
Another fact mentioned in the paper reminds me of one of the more memorable papers that I have read; Stanford's Eric Kool's demonstration that difluorotoluene- an isostere of thymine in which Ns are replaced by Cs and Os are replaced by Fs- behaves like thymine when DNA polymerase inserts it opposite adenine.
Reference:
Is there any bona fide example of O–HF–C bond in solution? The cases of HOC(CF3)2(4-X-2,6-C6H2(CF3)2) (X = Si(i-Pr)3, CF3)
Chem. Commun., 2007, 4384 - 4386, DOI: 10.1039/b710304b
The term hydrogen bond is already a misnomer. Frankly, any sharing of a partial-positive hydrogen with even a partial electronegative atom could be termed "hydrogen bonding".
ReplyDeleteMitch
Yes, that's why some authors have proposed resurrecting the old term "hydrogen bridge"
ReplyDeleteI am observing a rise in use of computer in chemistry. Now it seems that everything you find in experiment has been calculated in a paper long long ago, which also calculated/predicted the other three or four situations at the same time.
ReplyDeleteYes, the problem with computer modeling still is that you get multiple solutions, especially in case of things like docking and pharmacophores. But in some cases you can narrow it down to one or two. It certainly helps. In any case, modeling always has to be done by a sound chemist and not by a "black box-er".
ReplyDelete