Core‐level shifts and the choice of Auger parameter

The approximation that the change in the Wagner Auger parameter α* is equal to twice the change in the extra‐atomic relaxation energy depends on the assumption of equal core‐level binding‐energy shifts. It is shown from the available published data that, in general, this assumption is incorrect, and that α* cannot be used to estimate relaxation energies accurately. If accurate relaxation energies are required, then Auger parameters such as ζ of Hohlneicher et al. , or ζ′ of Sodhi and Cavell, based on the generalized Auger parameter ζ of Lang and Williams, must be used. However, for purely empirical chemical state identification in metals and semi‐metals, the various versions of α* are perfectly adequate, and indeed it matters little which version is used, since the changes Δα* with chemical state are the same for all of them within experimental error. For non‐metals there are as yet insufficient data to be certain just how useful α* might be, but for those two elements, sulphur and phosphorus, for which there are adequate data, it seems that α* on its own may not be as helpful as for the metals. In the latter case there is much less ambiguity in identification if the Wagner ‘chemical state’ diagram is used; this is particularly so for sulphur, for which the α* and β variations with chemical state seem complex.