In x-ray crystallography, the bond lengths are analyzed using inverse Fourier transform of the structure factors. If we simply consider the Fourier transform of the electron densities and then using inverse Fourier transform calculate the bond lengths, then we find that this length is not same as that obtained crystallographically.
Are you certain that crystal lattice of the crystal has been measured with sufficient accuracy. The electon density cell is dimensionless and relies on scaling by the measured cell dimension and that is ultimately determined by an accurate knowledge of the xray wavelength used in the experiment.
Thank you Miroslav, but within the accuracy limit, it has been found that the bond lengths are always shorter as obtained crystallographically than simple fourier transforms.
- In that case the only effect I am aware of is bond libration which has effect of shortening. I guess in refinement things like position and temperature factor are decoupled but all these things are combined together in a back fourier transform plus there are termination errors so the two methods are not equivalent. It is difficult to comment without knowledge of the structure and magnitude of errors.