Rate enhancement of catalytic reactions is a focus of chemical engineering research, with applications ranging from energy to materials fabrication and sustainability. There is recent experimental and simulation evidence indicating that kinetic limitations, such as the Sabatier maximum can be overcome when adsorbates change with time due to an externally enforced oscillation of the binding energies [1,2]. The resulting surface dynamics call for advanced numerical methods due to the increased number of parameters and overall complexity [3]. In this work, fixed point algorithms are employed to find entrained periodic solutions under reactor performance constraints for a broad range of dynamic parameters (such as the forcing frequency, amplitude or duty cycle). In fact, it is also possible, using a good set of basis functions, to parametrize entire families of imposed oscillation shapes. These descriptions (e.g. through radial basis functions) have more degrees of freedom, but also lead to more diverse entrained solutions. This can be exploited to “optimally sculpt" the shape of the forcing function in order to achieve (otherwise unattainable) selectivities. Bayesian Optimization can perform this task efficiently by adaptively constructing a surrogate model of the objective function as well as of the solution space [4]. Here, we demonstrate how this algorithmic pipeline can be used to optimally design forcing functions that drive the catalytic reactor to prescribed selectivities.
[1] M. A. Ardagh, O. A. Abdelrahman, P. J. Dauenhauer, â€Principles of Dynamic Heterogeneous Catalysis: Surface Resonance and Turnover Frequency Response†ACS Catalysis, 2019, 9(8), 6929-6937.
[2] J. Gopeesingh, M.A. Ardagh, M. Shetty, S. Burke, P. J. Dauenhauer, O.A. Abdelrahman, â€Resonance- Promoted Formic Acid Oxidation via Dynamic Electrocatalytic Modulation,†ACS Catalysis, 2020, 10(17), 9932-9942.
[3] A. M. Ardagh, T. Birol, Q. Zhang, O. A. Abdelrahman, P. J. Dauenhauer, â€Catalytic Resonance Theory: SuperVolcanoes, Catalytic Molecular Pumps, and Oscillatory Steady Stateâ€, Catalysis Science & Technology. 2019, 9, 5058-5076.
[4] B. Shahriari, K. Swersky, Z. Wang, R. P. Adams, N. de Freitas, "Taking the Human Out of the Loop: A Review of Bayesian Optimization", Proceedings of the IEEE, 2016, 104(1), 148-175.
