• reverse Monte Carlo techniques;
  • RMCProfile;
  • bond valence sums;
  • soft chemical constraints;
  • total scattering;
  • local crystalline disorder

The implementation of a new soft chemical constraint for the reverse Monte Carlo (RMC) program RMCProfile, based on bond valence sum (BVS) calculations, is described and its advantages for the analysis of `total scattering' diffraction data collected from disordered crystalline systems discussed. The inclusion of the BVS formalism proves particularly valuable in the early stages of the RMC fitting procedure, by avoiding the formation of regions containing chemically unreasonable local configurations which can become frozen in. Furthermore, this approach provides the fitting procedure with additional chemical information to differentiate between cation species that share the same crystallographic sites within the averaged unit cell and possess similar neutron scattering lengths. These issues are illustrated using total neutron scattering data collected at room temperature on the oxide-ion conductor Zr2Y2O7 and the nonlinear optical material KTiOPO4.