Modeling Catalyst Preparation: The Structure of Impregnated–Dried Copper Chloride on γ-Alumina at Low Loadings

Abstract

The structure of uncalcined atomically dispersed copper(II) chloride on γ-alumina is modeled with Density Functional Theory (DFT). Calculations are performed for the (110), (100), and (111) surfaces at several levels of hydration. The importance of the hydration and the uncertainties in the dehydration temperatures of the various γ-alumina surfaces are discussed. We find that both Cu2+ and Cl– do not adsorb on the (111) surface of γ-alumina. Moreover, we show that additional water molecules are present on the copper ions, which are not part of the alumina hydration layer. A correct modeling of adsorbed metal ions should therefore include these additional water molecules. Regarding the chloride ions, we find that these prefer “sitting” on the copper ions, in contrast with reported results for uncalcined copper chloride. We explain this difference in terms of kinetic accessibility and predict the formation of CuCl+ and CuCl2 species during calcination. Validation laboratory experiments using temperature programmed desorption (TPD) and reduction (TPR) confirm the predicted repositioning of chloride ions upon moderate heating. Depending on the placement of the chlorides on the alumina surface or on the copper ions, the calculated structures compare well with experimental results before and after heating, respectively. The copper location and coordination in the former structure also applies to samples impregnated with copper(II) nitrate.