On the effect of neglecting anharmonic nuclear motion in charge density studies.

The effect of neglecting anharmonic nuclear motion when it is definitely present is studied. To ensure the presence of anharmonic nuclear motion a model was used that was previously refined against experimental data including anharmonic nuclear motion, and these calculated structure factors were used as observed data for a multipole refinement. It was then studied how the neglect of anharmonic nuclear motion and noise in the data affects the usual crystallographic quality measure R, the density parameters and the residual density distribution. It is demonstrated that the neglect of anharmonic nuclear motion leads to a characteristic imprint onto the residual density distribution in terms of residual density peaks and holes, in terms of the whole residual density distribution and in terms of the number, location and strength of valence shell charge concentrations (VSCCs). These VSCCs differ from that of the input model in a way which heavily influences and misleads the chemical interpretation of the charge density. This imprint vanishes after taking anharmonic nuclear motion into account. Also the input model VSCCs are restored. The importance of modeling anharmonic nuclear motion is furthermore shown by the characteristic imprint on the residual density distribution, even in the case of a numerically almost unaffected R value.