There is growing evidence that the usual approach to modelling chemical events at surfaces is incomplete — an important c oncern in studies of the many catalytic processes that involve surface reactions