Simulations of thermal decomposition and film growth from the group VI metal hexacarbonyls

Abstract

In an effort to link the macroscopic chemical composition of films grown from the Cr, Mo, and W hexacarbonyls on metal and oxide surfaces to detailed, molecular level surface reactions of the precursors, we propose a simple, semiquantitative model of the surface kinetics involved in chemical vapor deposition (CVD). Kinetics data are obtained by simulation of temperature programmed desorption (TPD) spectra from the experimental literature. All CVD simulations are carried out under surface kinetics-controlled conditions. Two distinct models for TPD and CVD surface decomposition kinetics are investigated: one which assumes a kinetic barrier to CO dissociation on the surface and one which assumes no kinetic barrier to CO dissociation. The stochastic kinetics simulations show that the two models lead to very different results. The model that includes a kinetic barrier to CO dissociation is the only one that predicts deposition of essentially pure metal over a temperature range of 500-900 K, as is observed in laser-assisted thermal deposition experiments, or at >800 K, as found in conventional blanket CVD. The model with no kinetic barrier, on the other hand, is the only one that reproduces the temperature programmed desorption experiments. These results indicate that there are fundamental differences between the surface reaction conditions present at monolayer coverages and under chemical vapor deposition conditions: CO decomposition reactions leading to film contamination appear to be kinetically inhibited during film growth but dominant at low coverage. This suggests that model studies of CVD precursor surface chemistry should be interpreted cautiously. The CVD model with the kinetic barrier also correctly reproduces the experimentally observed formation of a metal oxycarbide interfacial layer when deposition occurs on oxide substrates. The features and limitations of the CVD model are described, together with areas where experimental test and new data are needed. © 1995 American Chemical Society.