(176c) Liquid Phase Effects on Adsorption Processes over Metal Surfaces
AIChE Annual Meeting
2022
2022 Annual Meeting
Catalysis and Reaction Engineering Division
Catalysis in Liquid Media II: Impact of Liquid Phase on Adsorption and Conversion in Catalysis
Monday, November 14, 2022 - 4:06pm to 4:24pm
Despite the ubiquitous use of solvents in heterogeneously catalyzed reactions, an accurate description of the associated free energy changes remains challenging. Computationally efficient implicit models fail to address solvent effects on metallic surface systems1. Alternatively, techniques that rely on a quantum mechanical (QM) description of solvent molecules, i.e., ab initio molecular dynamics (AIMD), are only applicable to systems restricted in size.
Consequently, we formulated a hybrid QM/MM free energy perturbation approach to compute solvation effects on adsorption free energy (∆∆Gads) via an appropriately constructed thermodynamic cycle. We utilized our methodology to study the solvent effects on aqueous phase adsorption of various biomass model molecules over transition metal catalysts. Adsorption of phenol and furfural on Pt at 298 K was found to be destabilized in aqueous solution relative to the gas phase by about 0.78 ± 0.02 eV and 0.94 ± 0.08 eV, respectively. Values that agree qualitatively with the experimental findings of Singh et al. where they measured ∆∆Gads to be in the range of 1.1-1.5 eV for phenol and 0.96-1.03 eV for furfural2 However, the application of QM/MM methods is limited to systems for which classical force fields are sufficiently reliable.
To address the lack of such parameters for water-ruthenium interactions we developed a Neural Network Potential (NNP) via application of an Artificial Neural Network (ANN) and atom centered symmetry functions. We used the potential together with classical force fields within our QM/MM-FEP approach to study aqueous phase adsorption of phenol over a reduced Ru(0001) surface and we predict a strong endergonic solvation free energy of 0.96 ± 0.02 at 298 K.
Consequently, we formulated a hybrid QM/MM free energy perturbation approach to compute solvation effects on adsorption free energy (∆∆Gads) via an appropriately constructed thermodynamic cycle. We utilized our methodology to study the solvent effects on aqueous phase adsorption of various biomass model molecules over transition metal catalysts. Adsorption of phenol and furfural on Pt at 298 K was found to be destabilized in aqueous solution relative to the gas phase by about 0.78 ± 0.02 eV and 0.94 ± 0.08 eV, respectively. Values that agree qualitatively with the experimental findings of Singh et al. where they measured ∆∆Gads to be in the range of 1.1-1.5 eV for phenol and 0.96-1.03 eV for furfural2 However, the application of QM/MM methods is limited to systems for which classical force fields are sufficiently reliable.
To address the lack of such parameters for water-ruthenium interactions we developed a Neural Network Potential (NNP) via application of an Artificial Neural Network (ANN) and atom centered symmetry functions. We used the potential together with classical force fields within our QM/MM-FEP approach to study aqueous phase adsorption of phenol over a reduced Ru(0001) surface and we predict a strong endergonic solvation free energy of 0.96 ± 0.02 at 298 K.
References
1. M. Saleheen, M. Zare, M. Faheem, A. Heyden, J. Phys. Chem. C. 123, 19052–19065 (2019).
2. J. Akinola, I. Barth, B. R. Goldsmith, N. Singh, ACS Catal. 10, 4929–4941 (2020).