(703f) Determining the Effect of Adsorbate-Adsorbate Intermolecular Interactions in the Adsorption and Decomposition of Binary Carboxylic Acid Mixtures on Pt(111)

Biorefinery effluent is an excellent potential supply of green hydrogen as the hydrocarbons within the effluent can be catalytically decomposed. However, the diversity of chemicals within the effluent causes intermolecular interactions that reduce the resulting catalytic performance from that predicted from studies on ideal, mono-component feedstocks. Experiments performed on binary mixtures of carboxylic acids (acetic (AA), propionic (PA), and glycolic acid (GA)) over Pt/C showed that, depending on the constituent compositions, one component’s catalytic conversion would be inhibited. Here, we seek to construct a nanoscale picture of the competitive adsorption and decomposition of carboxylic acids on a representative Pt catalyst surface (i.e. Pt(111)) using density functional theory and ab initio molecular dynamics (AIMD) to clarify the role of intermolecular interactions and chemical structure on the upgrading of carboxylic acid mixtures to green hydrogen (Figure 1). Acid coverage was varied from near zero to saturation, and the possible configurations at each coverage were sampled using AIMD simulations. The average adsorption energy for each acid becomes more negative as coverage increases, indicative of strongly attractive intermolecular interactions that come about due to hydrogen bonding between adjacent carboxylic acid functionalities. An examination of the most stable dimer configurations for all three acids shows a clear trend in the intermolecular interaction strength from AA < PA < GA, which offers insights into the competitive adsorption effects observed experimentally. Furthermore, the calculation of the interaction energies for both pure and mixed acid dimers shows that mixed dimers have weaker interactions compared to pure dimers. Overall, the most stable acid configurations are those that maximize hydrogen bonding. Because the low-coverage results are unable to account for experimental findings that the coverage-dependent model adequately supports, this work demonstrates the importance of accounting for intermolecular interactions when studying reactions in mixed feedstock streams.