Influence of Impurities on Acid-base Data for Oxide Minerals – Analysis of »Observable« Surface Charge and Proton Affinity Distributions and Model Calculations for Single Crystal Samples*

Acid-base data for oxide minerals are fundamental building blocks for adsorption models. Although the influence of impurities is often discussed in a qualitative way, a quantitative analysis is largely missing. A notable exception is the discussion of the possible effect of surface impurities on the results of mass titrations. In the present contribution potential contaminations are discussed in a generic way, with respect to their origin and their effects. Also dynamic aspects of such a contamination are considered in some detail. Besides, two examples are quantitatively analysed in more detail. In this analysis, one potential surface impurity (sulphate impurities on the iron oxyhydroxide goethite) and one possible source of »environmental« impurities (carbonate contamination on the iron(III) oxide hematite) are evaluated. The contaminant effect is a function of the ratio between the amount of contaminant and exposed surface area and also depends on the ionic strength of the sulphate-goethite system, since the adsorption of sulphate is sensitive to changes in ionic strength. While for a surface contamination like sulphate, the threshold ratio is mainly affected by the total amount of goethite at a given ionic strength, for an »environmental« contaminant like carbonate, the absolute value of the exposed surface area of the mineral is most important. From the hematite calculations the model surface potentials are obtained. These are of interest, because this most important quantity has recently been measured for a range of solids. One outcome of the calculations is that model surface potentials in the absence of a contamination are not in agreement with the measured values. The calculations for a contaminated system suggest that carbon dioxide adsorption will not significantly affect the surface potential contrary to the big effect obtained for the diffuse layer potential. The predictions like the ones in the present paper have the capability of avoiding experimental problems, if the calculations are carried out prior to the experimental work, or to evaluate potential effects a posterior for example in data selection for data base development.