June 3, 2021
Analysis of nonequilibrium catalytic activity by integrating experiments, mathematics and machine learning
A research group from RIKEN CSRS acquired experimental data supporting a new theory of catalytic activity.
The conventional theory of catalysis has been established based on assuming an environment near equilibrium where reactions proceed slowly. In reality however, catalysts are used under nonequilibrium conditions where reactions proceed actively, posing a challenge to accurately predict their activity. To solve this, in 2019, the research team proposed a new theory on the catalytic activity under nonequilibrium conditions by expanding the conventional theory.
To verify this new theory experimentally, the research team focused on the hydrogen production activity during water electrolysis on a platinum (Pt) catalyst. They measured the activity of hydrogen evolution in the nonequilibrium state, and analyzed the obtained data using a combined approach of mathematics and machine learning. They found that the binding energy of the catalyst and hydrogen atom was 0.094±0.002 eV. This result is the first experimental support to the new theory which predicts the existence of a highly active material whose binding energy deviates from zero.
The finding from this study is expected to promote the development of highly active catalytic materials and improve the precision of catalytic informatics which has been accelerating in recent years.
ACS Catalysis doi:10.1021/acscatal.1c01018
H. Ooka, M. E. Wintzer, R. Nakamura,
"Non-Zero Binding Enhances Kinetics of Catalysis: Machine Learning Analysis on the Experimental Hydrogen Binding Energy of Platinum".
Hideshi Ooka; Research Scientist
Marie E. Wintzer; Technical Staff I
Ryuhei Nakamura; Team Leader
Biofunctional Catalyst Research Team