Group Additivity for Aqueous Phase Thermochemical Properties of Alcohols on Pt(111)

Abstract

International audienceDespite progress in theoretical tools, the influence of solvation in heterogeneous catalysis remains poorly understood and predicted due to the large computational burden. In this work, we show that the inclusion of the solvation by water using a continuum model thermodynamically inhibits the O–H bond scissions involved in the ethanol aqueous phase reforming reaction over Pt(111), while it tends to favor the C–H, C–C, and C–O scissions. Then, we present a novel group additivity scheme for the free energy of adsorbates at the Pt/water interface that is able to capture this solvent effect. The mean absolute error (MeanAE) for the Gibbs free energy of formation is 3.3 kcal/mol over the investigated set of 200 species at the interface and the MeanAE for the 151 reaction free energies of ethanol aqueous phase reforming is 2.8 kcal/mol. Regarding the effect of solvation, our scheme is able to predict it with a MeanAE of about 1 kcal/mol. Together, the scheme promises to be accurate enough for narrowing down the most important reaction pathways in complex reaction networks as encountered in biomass conversion

Force Field for Water over Pt(111): Development, Assessment, and Comparison

“…While implicit solvation methods are, without any doubt, the most convenient and efficient ones for large-scale applications, 45,46 several approaches have been devised to replace the implicit solvent by an effective field (or solvation energy) obtained from molecular mechanics based molecular dynamics (MMMD) simulations or related methods. [47][48][49][50][51] In contrast to AIMD, the MMMD simulations can easily access the necessary time scales (ns range for equilibration) and length scales that are required to equilibrate these interfaces and to 3 avoid spurious effects due to 2D periodic repetitions of water configurations.…”

Can microsolvation effects be estimated from vacuum computations? A case-study of alcohol decomposition at the H2O/Pt(111) interface

“…where ∆G PCM inter represents the interaction energy between the continuum solvent and the solute, which is composed of the electrostatic interaction ∆G elec and the non-electrostatic term ∆G cavitation which depends on the area of the cavity and captures an average of positive and negative contributions (Pauli repulsion and dispersion interactions). This second term is often called cavitation energy and was set to 0 for our study to increase computational efficiency (see ref 49 for a discussion of the small effect of the cavitation energy). Therefore, the solvent model only assess dielectric interactions.…”

“…A reasonable way of dealing with large and complex reaction networks in silico is to follow a multiscale approach. 45,46 Starting with density functional theory (DFT) on smaller reactions networks, one can build linear scaling relationships and group additivity schemes for estimating reaction energies, [47][48][49] and Brønsted-Evans-Polanyi (BEP) relationships for estimating reaction barriers. [50][51][52] Together, these fast approximations allow to perform micro-kinetic simulations for extended reaction networks to identify the most important pathways.…”

Developments in the Atomistic Modelling of Catalytic Processes for the Production of Platform Chemicals from Biomass

“…[90] and [91]. The main results concern the wetting/non‐wetting character of the layers;, the stability of different water arrangements on the surface,, as distinctive patterns appear at sub‐monolayer coverages depending on the metal; and the possibility of fitting new potentials that could be used in electrochemical simulations . There are two ways of including the effect of solvation on theoretical simulations, either by explicit water molecules, or by replacing the vacuum region with an implicit solvation, Scheme .…”

“…This means that once solvation effects are considered for the intermediates, their effect on the activation energies can be predicted. Recently, a thermochemical model based on group additivity quantified the solvation effects . Moreover, the scaling relationships drawn for surrogates can be transferred to higher alcohols provided that a few rules are preserved.…”