ARCADE: Automated and Interpretable Design of Complex Catalysts

Abstract The rational design of high-performance catalysts is pivotal for modern industry but is often hindered by the time-consuming nature of trial-and-error methods. While computational and data-driven approaches have accelerated materials discovery, significant challenges remain, especially for complex systems like multielement disordered nanoparticles. These challenges include immense computational costs, difficulties in modeling stability and heterogeneous active sites, and a lack of interpretability in “black-box” AI models. To address these multifaceted challenges, herein, we introduce ARCADE (Automated Rational Catalyst Design), a generalizable and transferable framework for the automated and interpretable design of complex catalysts. This highly cohesive pipeline introduces key methodological innovations in realistic structure generation, autonomous active-site processing, and short-range order guided interpretability, seamlessly coupled with state-of-the-art pH-dependent microkinetic modeling. Notably, ARCADE streamlines the structural modeling of irregular nanoparticles by transforming intractable combinatorial enumeration into a targeted iterative search; robustly identifies active sites and subsequently resolves ubiquitous adsorbate drift issues arising during structural relaxations; and employs geometric short-range order descriptors derived contribution scores to rapidly decode catalytic mechanisms, thereby efficiently unraveling the complex synergistic effects inherent in multielement catalysts. Using the challenging design of multielement nanoparticle catalysts for CO2 reduction reaction (CO2RR) as a representative system, we demonstrate that ARCADE can systematically navigate the design process, from evaluating material stability to identifying site-specific activity and selectivity. Our results show that this methodology substantially accelerates the development cycle for complex catalysts, confirming its feasibility and broad applicability. Crucially, this work establishes a fully integrated paradigm for the rational design and discovery of complex materials for challenging chemical reactions.

Authors

Institutions

Publication Details

Journal
ACS Catalysis
Published
2026-10-29
DOI
https://doi.org/10.1021/acscatal.6c06280
Primary Topic
Machine Learning in Materials Science
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

ARCADE: Automated and Interpretable Design of Complex Catalysts

Yufei Ding, Hui Li, Weiwei Li, Yuhang Wang et al.
3 links
ACS Catalysis
Machine Learning in Materials Science
article

ARCADE: Automated and Interpretable Design of Complex Catalysts

Yufei Ding, Hui Li, Weiwei Li, Yuhang Wang, Xujian Wang, Yongfeng Guo, Jialu Li, Jianzhuo Wu, Honglin Li, Xue Jia, Guangsheng Liu, Yu Li
article en

Abstract

Abstract The rational design of high-performance catalysts is pivotal for modern industry but is often hindered by the time-consuming nature of trial-and-error methods. While computational and data-driven approaches have accelerated materials discovery, significant challenges remain, especially for complex systems like multielement disordered nanoparticles. These challenges include immense computational costs, difficulties in modeling stability and heterogeneous active sites, and a lack of interpretability in “black-box” AI models. To address these multifaceted challenges, herein, we introduce ARCADE (Automated Rational Catalyst Design), a generalizable and transferable framework for the automated and interpretable design of complex catalysts. This highly cohesive pipeline introduces key methodological innovations in realistic structure generation, autonomous active-site processing, and short-range order guided interpretability, seamlessly coupled with state-of-the-art pH-dependent microkinetic modeling. Notably, ARCADE streamlines the structural modeling of irregular nanoparticles by transforming intractable combinatorial enumeration into a targeted iterative search; robustly identifies active sites and subsequently resolves ubiquitous adsorbate drift issues arising during structural relaxations; and employs geometric short-range order descriptors derived contribution scores to rapidly decode catalytic mechanisms, thereby efficiently unraveling the complex synergistic effects inherent in multielement catalysts. Using the challenging design of multielement nanoparticle catalysts for CO2 reduction reaction (CO2RR) as a representative system, we demonstrate that ARCADE can systematically navigate the design process, from evaluating material stability to identifying site-specific activity and selectivity. Our results show that this methodology substantially accelerates the development cycle for complex catalysts, confirming its feasibility and broad applicability. Crucially, this work establishes a fully integrated paradigm for the rational design and discovery of complex materials for challenging chemical reactions.

ACS Catalysis
University of Pittsburgh (US), Tohoku University (JP), University of California San Diego (US), Advanced Institute for Materials Research, Tohoku University (JP)
Hellman Foundation, Japan Society for the Promotion of Science, Basic Energy Sciences
Openalex Percentile: Top 29%
Machine Learning in Materials Science
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon