Thermal diffuse scattering for molecular crystals: errors in X-ray diffraction intensities and atomic parameters

First-order TDS calculations have been made for naphthalene at 100 K to judge the suitability of different models for calculating TDS corrections used in practice. It appears that a long-wave model which takes account of the non-linearity of the frequency dispersion, gives about the same corrections for structural parameters as the lattice dynamical model for rigid body molecules and harmonic vibrations. Longwave models with linear frequency dispersion give an overcorrection of the errors. For second-order TDS no explicit comparison of the models has been possible, but rough estimates show that the comparison made for first-order TDS is also valid in good approximation for the total TDS.