Is there a hydrogen bond radius? Evidence from microwave spectroscopy, neutron scattering and x-ray diffraction results

Abstract

Intermolecular distances in D-H…A hydrogen bonded systems have usually been interpreted in terms of the van der Waals radii of D and A. In this work, X-ray and neutron diffraction data from the Cambridge Crystal Structure Database (CSD) and the electrostatic potential of A, have been used to define hydrogen bond radii for OH, NH and CH groups. For OH, X-ray and neutron diffraction both give comparable results, validating the X-ray data for defining a hydrogen bond radius. The hydrogen bond radii determined for C≡CH and OH groups from CSD analysis are comparable to those determined from the gas phase rotational spectroscopic data for HCCH and H₂O complexes. For NH as a proton donor, gas phase structural data are scarce and a hydrogen bond radius has been determined by using X-ray diffraction data only. For the CH group, the histogram of hydrogen bond distances shows a peak recognizable as a hydrogen bond only if it is acidic such as CCl₃H, OCH (aldehydic) or CCH (acetylenic). The hydrogen bond radii for OH, NH and acidic CH groups are 0.60 ± 0.15, 0.76 ± 0.15 and 1.10 ± 0.20 Å, respectively. For C-CH₃ and CH₂CH₃, though a peak in the histogram of distances is not found, the distribution of hydrogen bond angles unambiguously shows that the preferred geometry is linear. It appears that a CH group without any electronegative substituents could have a radius larger than 1.2 Å when involved in hydrogen bonding.