(175a) Nonlinear Optimization-Based Approaches for Modeling and Design of Heterogeneous Catalytic Systems

Microkinetic modeling, with input from computational chemistry calculations, is a key step in parameter estimation, experimental design, and rational catalyst design of heterogeneous catalytic systems. All these applications are, however, nonlinear optimization problems with an embedded microkinetic model that captures the dynamics of the reaction system. We, therefore, propose a generic methodology that combines nonlinear programming and differential algebraic systems solution for modeling and first-principles based design of heterogeneous catalytic systems.

For any given system, a comprehensive reaction network and DFT-derived thermochemical parameters of all species and transition states are taken as inputs. Subsequently, three classes of nonlinear optimization problems can be formulated and solved: (i) parameter estimation based on a simultaneous or sequential approach using activation barriers of reactions and binding energies of species as parameters [1, 2] to elucidate the detailed reaction mechanism, (ii) identification of optimal experimental conditions that maximize the coverage of or flux through an experimentally observable and mechanistically relevant surface intermediate, and  (iii) determination of "Sabatier-optimal" binding energies of surface species and transition state energies of reactions that maximize turnover frequencies and/or selectivity of desired products.

The formulation of the optimization problems, selection of a relevant set of parameters, solution approaches, and multi-start schemes to identify a near-global minimum will be presented and discussed in the context of methanol synthesis from CO and CO₂ on transition metals.

References:

1. P. Rubert-Nason, M. Mavrikakis, C. T. Maravelias, L. C. Grabow, and L. T. Biegler., AIChE J 2014, 60, 1336 – 1346
2. L. T. Biegler, Nonlinear Programming: Concepts, algorithms, and applications to chemical processes. SIAM 2010