The hydrogen bond between N—H or O—H and organic fluorine: favourable yes, competitive no

This is the authors' abstract. We don't add key points for this paper.

A study was made of X —H...F—C interactions ( X = N or O) in small-molecule crystal structures. It was primarily based on 6728 structures containing X —H and C—F and no atom heavier than chlorine. Of the 28 451 C—F moieties in these structures, 1051 interact with X —H groups. However, over three-quarters of these interactions are either the weaker components of bifurcated hydrogen bonds (so likely to be incidental contacts) or occur in structures where there is a clear insufficiency of good hydrogen-bond acceptors such as oxygen, nitrogen or halide. In structures where good acceptors are entirely absent, there is about a 2 in 3 chance that a given X —H group will donate to fluorine. Viable alternatives are X —H...π hydrogen bonds (especially to electron-rich aromatics) and dihydrogen bonds. The average H...F distances of X —H...F—C interactions are significantly shorter for C R 3 F ( R = C or H) and C sp 2 —F acceptors than for C R F 3 . The X —H...F angle distribution is consistent with a weak energetic preference for linearity, but that of H...F—C suggests a flat energy profile in the range 100–180°. X —H...F—C interactions are more likely when the acceptor is C sp 2 —F or C R 3 F, and when the donor is C—NH 2 . They also occur significantly more often in structures containing tertiary alcohols or solvent molecules, or with Z ′ > 1, i.e. when there may be unusual packing problems. It is extremely rare to find X —H...F—C interactions in structures where there are several unused good acceptors. When it does happen, there is often a clear reason, e.g. awkwardly shaped molecules whose packing isolates a donor group from the good acceptors.