Welcome to Prof. Jin-Xun Liu's Research Group

Our group has expertise in first-principles computational modeling and machine learning approaches to understand and predict catalysts for use in energy storage, sustainable chemical production and reducing the environmental impact of products. Developing and using new theoretical approaches, we devoted to establish the structure-activity relationship of the catalytic materials at atomic scale to realize the rational design of better catalysts. The interests of our group include: (1) Integrated Genetic Algorithm - Grand Canonical Monte Carlo - Microkinetics simulations approaches together to study the active structure of clusters catalysts, and to clarify the influence of the dynamic evolution of structures on catalytic performance; (2) The theoretical investigation of crystal phase effect on heterogeneous catalysis and electrocatalysis; (3) Using machine learning approach to accelerate catalyst and materials design. Our group has published more than 40 peer-reviewed papers, including Nat. Commun., Natl. Sci. Rev., J. Am. Chem. Soc., Angew. Chem. Int. Ed. and so on. We welcome students of all ages, backgrounds and beliefs to join our group. 

latest news&upcoming events

Mechanisms of Ethylene Epoxidation over Silver from Machine Learning-Accelerated First-Principles Modeling and Microkinetic Simulations

    We employed machine learning-augmented density functional theory (DFT) thermodynamic calculations to assess the stability of different AgOx structures under catalytic ethylene epoxidation reaction conditions. We found that there are multiple AgOx surface motifs that could co-exist under the relevant conditions. These included Ag surface oxides (e.g., AgO_p(4 × 4) and Ag1.83O) and atomic oxygen-covered Ag(111) surfaces. Furthermore, we employed DFT calculations to evaluate the energetics of different reaction mechanisms by which ethylene and oxygen can react on these surfaces. These studies revealed several energetically viable reaction pathways for ethylene epoxidation. Microkinetic modeling analyses, based on the DFT-calculated reaction pathways, showed that ethylene epoxidation can proceed on all surfaces and that multiple pathways, including those involving Langmuir–Hinshelwood and Eley–Rideal mechanisms, could be involved in selective and unselective reactions. The diversity of mechanisms that we discovered in the context of the relatively simple ethylene epoxidation reaction on Ag suggests that the richness and complexity of surface chemistry are most likely a rule rather than an exception in heterogeneous catalytic chemical transformations on metal surfaces and that the concept of a single or even a dominant mechanism and reaction intermediates might need to be revisited for many reactions.