(93b) Integrating the Identification of Active-Site Ensembles within Descriptor-Based Screening Protocols for Bimetallic Catalysts

Authors: 
Choksi, T. S., Stanford University
Streibel, V., Stanford University
Roling, L., Stanford University
Snider, J., SLAC National Accelerator Laboratory
Yang, A. C., Stanford University
Gallo, A., Stanford University
Jaramillo, T. F., Stanford University
Cargnello, M., Stanford University
Abild-Pedersen, F., SLAC National Accelerator Laboratory
AlJama, H., Stanford University
Given their broad tunability in structure and composition, bimetallic nanoparticles (NPs) are promising materials for catalytic applications. Their vast design space can be readily explored by combining density functional theory (DFT) with descriptor-based screening and microkinetic modeling resulting in descriptor-activity maps known as volcano plots. We use this screening approach in concert with in-situ catalyst characterization in an effort to reveal the promotional effect of In-Pd intermetallic compounds on In2O3 catalysts for methanol synthesis from CO2 and H2. While our screening study shows that InPd and In3Pd2 do have intrinsically high methanol synthesis activities, experimental characterization reveals that it is the presence of an indium oxide/In-Pd intermetallic compound interface that maximizes methanol activity and selectivity.[1]

In further developing prevailing volcano-based screening protocols, we need to address two key questions: First, how can we explicitly consider the inherently dynamic nature of catalysts under working conditions so as to obtain a precise determination of active-site ensembles? Second, how can we directly map optimal catalytic descriptors to morphological and compositional features of active sites at the atomic level? We have recently addressed these questions by forging a three-way connection between the structure of active sites, their thermodynamic stability, and the energy space of reaction intermediates in coordination-based and site-specific scaling models.[2]–[5] Our models rely on only few parameters derived from simple slab calculations, but are nonetheless directly transferable to NPs spanning a broad structure and composition space.

We demonstrate the significance of our models by identifying active-site ensembles that catalyze propene combustion on PtPd nanoalloys.[6] In this project, our experimental findings led us to hypothesize that OH* adsorption causes catalyst restructuring under reaction conditions. In our theoretical analysis, we use our coordination-based model to predict adsorption site stabilities that we correlate with OH* adsorption energies through site-specific scaling relations. This approach allows us to screen a large composition and structure space without needing to perform individual DFT calculations. We then use the information of our models to rapidly determine the impact of OH* adsorption on the surface energies of different surface terminations to simulate reaction conditions. We find that OH* adsorption stabilizes undercoordinated surface sites and corresponding Wulff constructions reveal substantial restructuring of the initially highly faceted nanoparticles. By counting the density of potential active-site ensembles on a library of uniform PdPt nanoparticles of different sizes, we find a 1:1 correlation of the density of 7-7 coordinated bridge sites with the turnover frequencies measured in experiment as the nanoparticle size is varied. Since these undercoordinated sites also have lower C=C scission barriers than terrace sites, we conclude that step sites dynamically forming under reaction conditions are active for breaking C=C bonds during propene combustion.

Having demonstrated both that intermetallic compounds like InPd adhere to linear scaling relations and that our coordination-based model reliably predicts active-site motifs, we integrate these tools with transition state scaling relations to create site-specific activity volcanoes. These volcano plots enable us to screen intermetallic compounds in terms of activity, selectivity, and stability in our endeavor to identify the next generation of methanol synthesis catalysts.

[1] J. L. Snider, V. Streibel, M. A. Hubert, T. S. Choksi, E. Valle, D. C. Upham, J. Schumann, M. S. Duyar, A. Gallo, F. Abild-Pedersen, and T. F. Jaramillo, ACS Catal., 3399–3412, 2019.

[2] L. T. Roling, L. Li, and F. Abild-Pedersen, J. Phys. Chem. C, 121, 23002–20010, 2017.

[3] L. T. Roling and F. Abild-Pedersen, ChemCatChem, 10, 7, 1643–1650, 2018.

[4] L. T. Roling, T. S. Choksi, and F. Abild-Pedersen, Nanoscale, 11, 10, 4438–4452, 2019.

[5] T. S. Choksi, L. T. Roling, V. Streibel, and F. Abild-Pedersen, J. Phys. Chem. Lett., 1852–1859, 2019.

[6] A.-C. Yang, H. Aljama,V. Streibel, T. S. Choksi, C. Wrasman, L. T. Roling, S. Bare, E. Goodman, D. Thomas, R. Sanchez, Y. Li, F. Abild-Pedersen and M. Cargnello “Revealing the Geometric Active Site Ensemble for Propene Combustion Using Uniform Pd and Pt Nanocrystal Catalysts” (submitted)