COMPUTATIONAL MODELING OF NANOSTRUCTURED METAL SURFACES AND THEIR ELECTRON IC AND CATALYTIC PROPERTIES

I will present results of our work in two different topics in computational catalysis, whose underlying theme lies in the possibility of creating catalysts with rationally designed nanoscale surface features. The former direction deals with nanoporous platinum structures obtained by electrochemical leaching Ni-Pt alloy and their extraordinary activity in the oxygen reduction reaction (ORR). We focus on nanoparticle and nanowires and show how it is possible to achieve insight into the relationships between nanoscale atomistic surface features and ORR performance via first-principles-based theory and simulations [1]. Enhanced ORR activity by a factor of 50 with respect to commercial platinum catalysts are achieved in the nanowire case. Comparison/validation with experiment is further pursued via operando vibrational frequencies and XPS shifts at the electrode-electrolyte interface (EEI). Second, I will switch to regular metal nanoparticle catalysts. I will report an extensive study of ammonia synthesis via the Haber-Bosch (HB) process over Fe bcc(111) and bcc(211). Our computational protocol combines Density-Functional Theory plus dispersion to unveil the mechanistic steps at the atomic level with kinetic Monte Carlo (kMC) modeling to predict steady-state catalytic reaction rates under realistic conditions, thus allowing us to validate our predictions against experimental kinetic data from classic surface science literature [2a]. We then use the detailed knowledge derived for this system to consider modifications of the catalyst such as alloying the first few surface layers. In order to span the largest possible set of catalyst modifications simultaneously minimizing computational effort, we derive a Hierarchical High-Throughput Screening (HHTS) approach to catalyst design. The HHTS approach singles out the most promising alloying elements and configurations as a function of catalyst structure and alloying site. In the case of HB over top-layer substitutional 0.25 ML alloyed Fe-bcc(111), HHTS indicates Rh as the most promising dopant. The approach is validated by reconstructing the complete free-energy diagram for the Rh-doped system, conducting a full kinetic analysis, and comparing the results from those estimated on the basis of HHTS, finding very good agreement [2b]. Other doping sites and strategies are discussed, enabling catalytic activity enhanced by up to 2 orders of magnitude with respect to the pure Fe case. This represents a new strategy that can be used to optimize catalyst performance for complex reactions involving 10 to 20 potential rate determining steps, where the simple Sabatier-principle-based volcano relationships in terms of a single controlling parameter no longer apply. [1] (a) A. Fortunelli, W. A. Goddard, et al. Chem. Sci. 6 (2015) 3915-3925; (b) M. Li, et al. Science 354 (2016) 1414-1419; (c) T. Cheng, A. Fortunelli, W. A. Goddard PNAS (2019) 201821709. [2] (a) J. Qian, Q. An, A. Fortunelli, R. J. Nielsen, W. A. Goddard, J. Am. Chem. Soc. 140 (2018) 6288-6297; (b) Q. An, Y. Shen, A. Fortunelli, W. A. Goddard J. Am. Chem. Soc. 140 (2018) 17702-17710.