A theoretical study of standard heat of formation of systems involving in the zinc reduction of silicon tetrachloride

The gas phase zinc reduction of silicon tetrachloride produces the silicon for solar cells. While this reaction provides a new low-cost production route for silicon materials for photovoltaic cells, little is known about the chemistry of this process. Theoretical methods, based on quantum chemistry predictions, in the gas phase, are now fully capable of providing molecular thermochemistry and kinetic parameters with sufficient accuracy for modeling purposes. This kind of kinetic information is crucial for reactor design and scale-up of reaction systems. In this spirit, we have developed two test sets, one for silicon and another for zinc compounds, for evaluating the performance of various computational methods: density functional theory (B3LYP, BH and HLYP, BMK, and M05-2X), and composite methods (G3 and CBS-QB3). The new generation of DFT methods BMK and M05-2X and the composite CBS-QB3 method allow to predict the standard heat of formation, Delta H-f(0), of the silicon compounds with MAD of, respectively, 7, 13, and 15 kJ mol(-1), whereas the previous DFT methods are less reliable. At least triple zeta, for basis set, is needed in order to predict correctly the standard heat of formation. For the zinc compounds, BMK, B3LYP, and CBS-QB3 are the best performing methods for the calculation of Delta H-f(0) with MADs of 24, 25, and 28 kJ mol(-1), respectively. We recommend BMK and CBS-QB3 methods for investigating the new solar silicon process.