Kinetic models based on first principles are becoming common place in heterogeneous catalysis because of their ability to interpret experimental data, identify the rate-controlling step, guide experiments and predict novel materials. To overcome the tremendous computational cost of estimating parameters of complex networks on metal catalysts, approximate quantum mechanical calculations are employed that render models potentially inaccurate. Here, by introducing correlative global sensitivity analysis and uncertainty quantification, we show that neglecting correlations in the energies of species and reactions can lead to an incorrect identification of influential parameters and key reaction intermediates and reactions. We rationalize why models often underpredict reaction rates and show that, despite the uncertainty being large, the method can, in conjunction with experimental data, identify influential missing reaction pathways and provide insights into the catalyst active site and the kinetic reliability of a model. The method is demonstrated in ethanol steam reforming for hydrogen production for fuel cells.

We present a theoretical and computational analysis to elucidate the relation of the Brønsted−Evans−Polanyi (BEP) and transition state scaling (TSS) correlations. We find that the TSS correlation is an approximation of the BEP correlation. The BEP correlation allows for the straightforward identification of homologous series via standard statistical tests and has better error properties than the TSS correlations. We find that the unit cell size used in the DFT calculations does not have a significant effect on the correlation parameters; however, the zero point energy correction does have a significant effect on the correlation parameters, especially for (de)hydrogenation reactions. We propose a method for using this information to estimate zeropoint-energy-corrected activation energies without resorting to calculations of the zero point energy correction for every species. Finally, we find that BEP correlations derived for one parent molecule may be applicable to other molecules beyond the training set.

Density functional theory (DFT) calculations for the thermal decomposition and oxidative dehydrogenation of ethanol, mechanistic aspects of water–gas shift reaction, and experimental kinetic data are integrated so as to develop and assess a comprehensive DFT-based microkinetic model of low temperature ethanol steam reforming on Pt catalysts. The DFT calculations show (1) that the C–C scission should occur late in the dehydrogenation sequence, (2) that the C–C scission barriers in highly dehydrogenated intermediates are comparable to early C–H abstraction barriers, and (3) that the oxidative dehydrogenation reactions should not be important under steam reforming conditions. The DFT-parametrized model shows good qualitative agreement with experiments, with reasonable deviations attributed to modeling only the metal chemistry (i.e., excluding support effects). Both the model and the experiments show that the overall mechanism is simply thermal decomposition of ethanol followed by incomplete water–gas shift. The most abundant surface species in the model are the decomposition products CO, H, and free sites, while the key reactive intermediates are present in much lower amounts. Unlike findings of simplified previous models, the rate determining step was identified as the initial dehydrogenation of ethanol, while the selectivity to C1 products is controlled by the C–C cracking of CHCO. Brønsted–Evans–Polanyi (BEP) correlations for the oxidative dehydrogenation reactions are developed and the effect of coadsorption on BEPs is discussed.