(389f) Advancing Non-Thermal Plasma Catalytic Material Design through Multiscale Simulation and Interpretable Machine Learning Integration

This study presents a transformative approach in the design of non-critical catalytic materials for hydrogen production through the integration of multiscale modeling, machine learning (ML), and experimental validations. Targeting the efficient conversion of ammonia cracking into hydrogen using non-thermal plasma at low temperatures, this research addresses the urgent needs for decarbonizing the maritime shipping industry [1]. Our methodology circumvents the environmental and economic drawbacks of traditional hydrogen production methods, which are dependent on rare, costly catalysts (e.g., Ru) and are significant CO₂ emitters [2]. By leveraging density functional theory (DFT)-based microkinetic modeling with interpretable ML algorithm development, we explored the interactions between non-thermal plasma and catalytic reactions and identified the optimal earth-abundant, non-critical materials for ammonia cracking to generate hydrogen under mild non-thermal plasma conditions [3].

Our multi-scale simulation results show that under non-thermal plasma conditions, Co is the best catalyst for ammonia decomposition rather than Ru under conventional heating, which is consistent with experimental validations (Fig. 1). The agreement between theoretical and experimental results not only validates our results but further reinforces the reliability of our integrated model. We also identified the N^* binding energy as a crucial descriptor influencing ammonia decomposition rates under non-thermal plasma conditions. Subsequently, utilizing this descriptor combined with d-band theory, we developed interpretable machine learning algorithm and obtained various optimal bimetallic catalysts from earth-abundant resources. This machine learning-guided multi-scale simulations presents a scalable and cost-effective pathway to design plasma catalysts for green hydrogen production, significantly contributing to the sustainability of energy systems.

References

[1] Yuan Y and Zhou L, et al., Science, 2022, 378(6622):889-93.

[2] Hansgen, D. A. et al., Nature chemistry, 2010, 2 (6), 484-489.

[3] Andersen, J. et al., Chemical Engineering Science, 2023, 271, 118550.