Generalizing Performance Equations in Heterogeneous Catalysis from Hybrid Data and Statistical Learning

Abstract

Activity equations trying to mimic experimental catalytic performance derived from reaction profiles and microkinetic models have been the state of the art in modeling in the last decades. This approach has been able to reproduce semiquantitatively activity volcano plots leading to successful catalyst optimization through the use of descriptors. As systems become more complex (both catalysts and reactants), these methods face increasing limitations. Statistical Learning (SL) techniques can overcome these limitations and improve the search for descriptor-based performance equations. However, the black-box nature of SL techniques makes physical interpretation of the so-obtained models difficult. To advance in the integration of these methodologies to real problems, we have merged experimental activity and selectivity presented as a function of chemical descriptors from Density Functional Theory for the catalyzed hydrodehalogenation of CH2X2 (for X = Br, Cl) leading to three main products. The employed Bayesian procedure is able to identify robust equations for activity and selectivity as a function of only two descriptors. This work provides a starting point to solve complex reaction networks using a set of statistical learning tools and hybrid data.

Universal Principles for the Rational Design of Single Atom Electrocatalysts? Handle with Care

“…While useful in principle to screen new potential catalysts, the complexity of the relation, the number of variables, and the fact that some are not easily defined limit our understanding of the physical reasons that determine the chemical behavior of a given SAC. Nevertheless, useful examples of applications of machine learning techniques exist . Recently an interesting hybrid approach has been proposed for a specific hydrodehalogenation reaction where experimental activity and selectivity are analyzed as a function of only two chemical descriptors from DFT …”

“…Nevertheless, useful examples of applications of machine learning techniques exist. 77 Recently an interesting hybrid approach has been proposed for a specific hydrodehalogenation reaction where experimental activity and selectivity are analyzed as a function of only two chemical descriptors from DFT. 77…”

Efficient Transfer Hydrodehalogenation of Halophenols Catalyzed by Pd Supported on Ceria

“…size. [65][66][67] Thus, the smallest halogen F presents the lowest degree of covalency, the weakest bonding to the metal surface and the highest disruption of the metal's work function. All these results support the idea that the differences in reactivity between the substrates is controlled by the removal of the halide from the Pd catalyst (Fig.…”

What do we talk about, when we talk about single-crystal termination-dependent selectivity of Cu electrocatalysts for CO2reduction? A data-driven retrospective

“…Notwithstanding this limitation, machine learning inference on experimental data enabled to infer robust composition-property relationships for the oxidative coupling of methane, 33–36 the 5-ethoxymethylfurfural hydrogenation, 37 the oxidative dehydrogenation of butane to 1,3-butadiene, 38 the hydrodehalogenation of CH 2 Br 2 and CH 2 Cl 2 , 39 thermocatalytic CO 2 hydrogenation into methanol, 40 and higher alcohol synthesis. 41 Such composition-property relationship were then, oftentimes, exploited to discover heterogeneous catalysts with improved activity and/or selectivity.…”