(488c) Free Energy Surface of Nitrogen Dissociation on Ru(0001) from First Principles Molecular Dynamics with Enhanced Sampling Methods

The effect of thermal energy on the reaction free energies and barriers of elementary chemical reactions on surfaces is often calculated using harmonic transition state theory (TST) due to the computational expense of first principles quantum chemistry methods and the simplicity of harmonic analysis. However, often these methods rely on system-specific intuition and assumptions regarding which configurations and types of anharmonicities may be important. As chemical reaction models become increasingly large and complex, more generalizable approaches to calculate free energy surfaces of surface reactions are needed. This motivates the development and application of more rigorous, scalable and efficient free energy enhanced sampling methods that can be coupled to first principles molecular dynamics simulations.

In this work, we study a chemical reaction catalyzed by a metal surface by coupling state-of-the art enhanced sampling methods, including adaptive biasing force and recently developed machine-learning-based approach [1] to Density Functional Theory (DFT)-based molecular dynamics. The dissociation/association of nitrogen on the close packed ruthenium is chosen as prototypical example due to its relevance as a critical step in ammonia synthesis and the large body of previous theoretical and experimental studies. We compare free energy surfaces from these enhanced sampling methods to those predicted using harmonic TST. We find that the free energy surfaces obtained from enhanced sampling methods and harmonic TST differ. We discuss the origins of their differences and implications on predicting the most probable reaction pathway. This work is a step toward the application of free energy enhanced sampling to predict chemical reactions in increasingly complex systems.

[1] E. Sevgen, A. Z. Guo, H. Sidky, J. K. Whitmer, J. J. de Pablo. “Combined Force-Frequency Sampling for Simulation of Systems Having Rugged Free Energy Landscapes”. J. Chem. Theory Comput. 2020, 16, 3, 1448-1455; https://github.com/MICCoM/SSAGES-public