Haoxiang Xu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Haoxiang Xu has a total of 58 co-authors. Here are the top co-authors by number of shared publications:
- Daojian Cheng
Beijing University of Chemical Technology
15 shared publications - Dapeng Cao
State Key Laboratory of Organic‐Inorganic Composites Beijing University of Chemical Technology Beijing P. R. China
6 shared publications - Hao Li
Advanced Institute for Materials Research (WPI-AIMR), Tohoku University, Sendai, 980-8577, Japan
5 shared publications Jin Liu
State Key Laboratory of Organic‐Inorganic Composites Beijing University of Chemical Technology Beijing China
5 shared publications- Run Jiang
State Key Laboratory of Organic-Inorganic Composites, Beijing Advanced Innovation Center for Soft Matter Science and Engineering Beijing University of Chemical Technology Beijing 100029 China
4 shared publications - Xiao Cheng Zeng
City University of Hong Kong
4 shared publications
Latest Publications
- 13 November 2026
- Published by Wiley in Journal of Computational Chemistry
- Vol. 47 (30)
OCCat : A Software Package for Optimization of Cluster Catalyst
Predicting global minimum (GM) structures of nanoalloy and oxide clusters remains challenging, and catalytic property evaluation typically requires disjoint manual workflows. Here, we present OCCat, an open‐source package unifying GM exploration with automated catalytic assessment. The engine employs a revised particle swarm optimization (RPSO) and a centered binary adjustment (CBA) method enforcing interatomic distance constraints with O(N·logN) complexity, improving upon conventional O(N 2 ) schemes. OCCat interfaces with empirical (LJ, Gupta), DFT (VASP), and neural network potentials (LASP) for multi‐fidelity screening. A downstream analyzer automates site identification, adsorption energy prediction via generalized coordination numbers (GCN), and turnover frequency evaluation via microkinetics. Benchmarks on Lennard‐Jones clusters, transition‐metal systems at the Gupta and DFT levels, and oxidized silver at the NNP level validate structural reliability and efficiency. The workflow is demonstrated through automated screening of CO oxidation, hydrogen evolution, and oxygen reduction on sub‐nanometer clusters.
- 13 November 2026
- Published by Wiley in Journal of Computational Chemistry
- Vol. 47 (30)
OCCat : A Software Package for Optimization of Cluster Catalyst
Predicting global minimum (GM) structures of nanoalloy and oxide clusters remains challenging, and catalytic property evaluation typically requires disjoint manual workflows. Here, we present OCCat, an open‐source package unifying GM exploration with automated catalytic assessment. The engine employs a revised particle swarm optimization (RPSO) and a centered binary adjustment (CBA) method enforcing interatomic distance constraints with O(N·logN) complexity, improving upon conventional O(N 2 ) schemes. OCCat interfaces with empirical (LJ, Gupta), DFT (VASP), and neural network potentials (LASP) for multi‐fidelity screening. A downstream analyzer automates site identification, adsorption energy prediction via generalized coordination numbers (GCN), and turnover frequency evaluation via microkinetics. Benchmarks on Lennard‐Jones clusters, transition‐metal systems at the Gupta and DFT levels, and oxidized silver at the NNP level validate structural reliability and efficiency. The workflow is demonstrated through automated screening of CO oxidation, hydrogen evolution, and oxygen reduction on sub‐nanometer clusters.
- 29 July 2026
- Published by American Chemical Society (ACS) in ACS Catalysis
- Vol. 16 (15) , 15303-15321
Why Is Alloying a Prerequisite for Interstitial Carbon Playing a Promotional Role in Selective Acetylene Hydrogenation?
The thermodynamic−kinetic origin of why alloying is a prerequisite for interstitial carbon playing a promotional role in selective acetylene hydrogenation over Ni-based catalysts remains elusive. Moreover, catalytic mechanisms are frequently studied on pristine or bare surface models, overlooking the realistic surface states and coverage effects under operando conditions. Herein, we establish a realistic surface-state research paradigm, integrating global optimization, ab initio thermodynamic phase diagrams, constrained ab initio molecular dynamics, and coverage self-consistent calculations. We reveal that pure Ni thermodynamically rejects subsurface carbon (ΔEinc = −0.41 eV) with a prohibitively high kinetic barrier (3.75 eV), while Ga/Zn alloying switches the thermodynamics to favorable (ΔEinc = −3.61 eV) and lowers the penetration barrier by ∼0.95 eV via lattice expansion and electronic activation. Coverage-dependent calculations demonstrate that adsorbate−adsorbate interactions adjust the free energy profiles, and the coverage self-consistent microkinetic modeling and reactor simulations align well with experimental trends of Ni-based catalysts. The excellent performance of Ni3GaCx and Ni3ZnCx originates from interstitial carbon within the alloy matrix. Synergistic ligand and geometric effects tuning Ni-based carbic to the summit of a predictive selectivity volcano that simultaneously suppresses overhydrogenation and oligomerization, where electronic modulation optimizes intermediate adsorption strength while adjacent Ni atoms stretching creates selective advantage sites. This work transforms the understanding of interstitial carbon from a phenomenological observation into a quantitative, descriptor-based design map that can guide future rational catalyst design via controlled interstitial modification engineering.
- 29 July 2026
- Published by American Chemical Society (ACS) in ACS Catalysis
- Vol. 16 (15) , 15303-15321
Why Is Alloying a Prerequisite for Interstitial Carbon Playing a Promotional Role in Selective Acetylene Hydrogenation?
The thermodynamic−kinetic origin of why alloying is a prerequisite for interstitial carbon playing a promotional role in selective acetylene hydrogenation over Ni-based catalysts remains elusive. Moreover, catalytic mechanisms are frequently studied on pristine or bare surface models, overlooking the realistic surface states and coverage effects under operando conditions. Herein, we establish a realistic surface-state research paradigm, integrating global optimization, ab initio thermodynamic phase diagrams, constrained ab initio molecular dynamics, and coverage self-consistent calculations. We reveal that pure Ni thermodynamically rejects subsurface carbon (ΔEinc = −0.41 eV) with a prohibitively high kinetic barrier (3.75 eV), while Ga/Zn alloying switches the thermodynamics to favorable (ΔEinc = −3.61 eV) and lowers the penetration barrier by ∼0.95 eV via lattice expansion and electronic activation. Coverage-dependent calculations demonstrate that adsorbate−adsorbate interactions adjust the free energy profiles, and the coverage self-consistent microkinetic modeling and reactor simulations align well with experimental trends of Ni-based catalysts. The excellent performance of Ni3GaCx and Ni3ZnCx originates from interstitial carbon within the alloy matrix. Synergistic ligand and geometric effects tuning Ni-based carbic to the summit of a predictive selectivity volcano that simultaneously suppresses overhydrogenation and oligomerization, where electronic modulation optimizes intermediate adsorption strength while adjacent Ni atoms stretching creates selective advantage sites. This work transforms the understanding of interstitial carbon from a phenomenological observation into a quantitative, descriptor-based design map that can guide future rational catalyst design via controlled interstitial modification engineering.
- 22 June 2026
- Published by Elsevier in Chinese Journal of Chemical Engineering
- Vol. 96, 144-165
Research Progress of Domain-Specific Large Language Models for Chemical Materials
- 22 June 2026
- Published by Elsevier in Chinese Journal of Chemical Engineering
- Vol. 96, 144-165
Research Progress of Domain-Specific Large Language Models for Chemical Materials
- Open Access
- 27 April 2026
- Published by Wiley in Angewandte Chemie
- Vol. 138 (24)
Dual‐ Sabatier Optima: How Reaction Mechanism Determines Activity Volcano Map of Dual‐Atom Catalysts for Oxygen Reduction Reaction
Dual‐atom catalysts (DACs) have demonstrated superior potential in the oxygen reduction reaction (ORR). However, the single‐peak activity volcano derived from classical associative mechanism is contrast to the large‐scale experimental data from the Digital Catalysis Platform ( DigCat ). Herein, we studied ORR over 200 DACs from thermodynamic and kinetic perspectives, and found that the dissociative mechanism is generally dominant for DACs. By integrating potential‐related microkinetic modeling and machine learning (ML)‐derived interpretable structural descriptors, we discovered a dual‐ Sabatier optima volcano map against Δ G (OH*) (or structural descriptors), which was rigorously validated against available experimental data. Dual‐ Sabatier optima stem from the rate‐determining step of dissociative mechanism switching among three elementary reactions (O 2 dissociation → 2OH protonation → OH protonation), which can be extended across DACs containing transition metal, metal‐like, and non‐metal elements as center atoms. It opens a brand‐new perspective for rational design of DACs and atomically dispersed catalysts for other reactions beyond ORR, of which the dominant reaction mechanism may be different from single‐atom catalysts (SACs) and lead to diverse activity volcano maps. Most importantly, this work illustrates that new phenomenon can be identified from “old experimental data” under a large data scale, with the help of theoretical simulations integrated with interpretable ML .
- 27 April 2026
- Published by Wiley in Angewandte Chemie
- Vol. 138 (24)
Open AccessDual‐ Sabatier Optima: How Reaction Mechanism Determines Activity Volcano Map of Dual‐Atom Catalysts for Oxygen Reduction Reaction
Dual‐atom catalysts (DACs) have demonstrated superior potential in the oxygen reduction reaction (ORR). However, the single‐peak activity volcano derived from classical associative mechanism is contrast to the large‐scale experimental data from the Digital Catalysis Platform ( DigCat ). Herein, we studied ORR over 200 DACs from thermodynamic and kinetic perspectives, and found that the dissociative mechanism is generally dominant for DACs. By integrating potential‐related microkinetic modeling and machine learning (ML)‐derived interpretable structural descriptors, we discovered a dual‐ Sabatier optima volcano map against Δ G (OH*) (or structural descriptors), which was rigorously validated against available experimental data. Dual‐ Sabatier optima stem from the rate‐determining step of dissociative mechanism switching among three elementary reactions (O 2 dissociation → 2OH protonation → OH protonation), which can be extended across DACs containing transition metal, metal‐like, and non‐metal elements as center atoms. It opens a brand‐new perspective for rational design of DACs and atomically dispersed catalysts for other reactions beyond ORR, of which the dominant reaction mechanism may be different from single‐atom catalysts (SACs) and lead to diverse activity volcano maps. Most importantly, this work illustrates that new phenomenon can be identified from “old experimental data” under a large data scale, with the help of theoretical simulations integrated with interpretable ML .
- Open Access
- 27 April 2026
- Published by Wiley in Angewandte Chemie International Edition
- Vol. 65 (24) , e8386838
Dual‐ Sabatier Optima: How Reaction Mechanism Determines Activity Volcano Map of Dual‐Atom Catalysts for Oxygen Reduction Reaction
Dual‐atom catalysts (DACs) have demonstrated superior potential in the oxygen reduction reaction (ORR). However, the single‐peak activity volcano derived from classical associative mechanism is contrast to the large‐scale experimental data from the Digital Catalysis Platform ( DigCat ). Herein, we studied ORR over 200 DACs from thermodynamic and kinetic perspectives, and found that the dissociative mechanism is generally dominant for DACs. By integrating potential‐related microkinetic modeling and machine learning (ML)‐derived interpretable structural descriptors, we discovered a dual‐ Sabatier optima volcano map against Δ G (OH*) (or structural descriptors), which was rigorously validated against available experimental data. Dual‐ Sabatier optima stem from the rate‐determining step of dissociative mechanism switching among three elementary reactions (O 2 dissociation → 2OH protonation → OH protonation), which can be extended across DACs containing transition metal, metal‐like, and non‐metal elements as center atoms. It opens a brand‐new perspective for rational design of DACs and atomically dispersed catalysts for other reactions beyond ORR, of which the dominant reaction mechanism may be different from single‐atom catalysts (SACs) and lead to diverse activity volcano maps. Most importantly, this work illustrates that new phenomenon can be identified from “old experimental data” under a large data scale, with the help of theoretical simulations integrated with interpretable ML .
- 27 April 2026
- Published by Wiley in Angewandte Chemie International Edition
- Vol. 65 (24) , e8386838
Open AccessDual‐ Sabatier Optima: How Reaction Mechanism Determines Activity Volcano Map of Dual‐Atom Catalysts for Oxygen Reduction Reaction
Dual‐atom catalysts (DACs) have demonstrated superior potential in the oxygen reduction reaction (ORR). However, the single‐peak activity volcano derived from classical associative mechanism is contrast to the large‐scale experimental data from the Digital Catalysis Platform ( DigCat ). Herein, we studied ORR over 200 DACs from thermodynamic and kinetic perspectives, and found that the dissociative mechanism is generally dominant for DACs. By integrating potential‐related microkinetic modeling and machine learning (ML)‐derived interpretable structural descriptors, we discovered a dual‐ Sabatier optima volcano map against Δ G (OH*) (or structural descriptors), which was rigorously validated against available experimental data. Dual‐ Sabatier optima stem from the rate‐determining step of dissociative mechanism switching among three elementary reactions (O 2 dissociation → 2OH protonation → OH protonation), which can be extended across DACs containing transition metal, metal‐like, and non‐metal elements as center atoms. It opens a brand‐new perspective for rational design of DACs and atomically dispersed catalysts for other reactions beyond ORR, of which the dominant reaction mechanism may be different from single‐atom catalysts (SACs) and lead to diverse activity volcano maps. Most importantly, this work illustrates that new phenomenon can be identified from “old experimental data” under a large data scale, with the help of theoretical simulations integrated with interpretable ML .
- 20 April 2026
- Published by Elsevier in Chinese Journal of Chemical Engineering
- Vol. 96, 97-108
DFT study on mechanism for size-dependence of Pt catalysts for propane dehydrogenation
- 20 April 2026
- Published by Elsevier in Chinese Journal of Chemical Engineering
- Vol. 96, 97-108
DFT study on mechanism for size-dependence of Pt catalysts for propane dehydrogenation
- Top Cited PapersOpen Access
- 24 February 2026
- Published by Wiley in Angewandte Chemie International Edition
- Vol. 65 (21) , e24612
Catalysis AI Agent Guides Discovering the Universal Design Principle of Cu‐Based Single‐Atom Alloy Catalysts for CO 2 Electroreduction
Copper (Cu)‐based single‐atom alloys (SAAs) represent a promising strategy for optimizing the electroreduction of CO 2 (CO 2 R) to multi‐carbon products (C 2+ ). However, the diverse enhancement degrees of C 2+ selectivity brought about by various dopants have not yet been rationalized, which lead to the absence of guidelines for further designing desired Cu‐based SAAs. Herein, guided by the Catalysis AI Agent developed based on large‐scale data + large language model, as well as the Digital Catalysis Platform (the DigCat experimental database), we performed first‐principles calculations to evaluate C 2+ products selectivity trends through identifying the energy barrier of rate‐determining step (RDS) among diverse C‐C coupling pathways. With first‐principles results fed back, Catalysis AI Agent reveals that the element classification in the periodic table of guest metal dopant is essential for establishing robust structure‐selectivity correlations among Cu‐based SAAs. A structural descriptor (φ) is developed and helps to establish a strong correlation among the electronic‐scale structural features, the adsorption strength of C‐C coupling precursors, and the macroscopic C 2+ products selectivity. A universal design principle based on φ for Cu‐based SAAs enables the rapid and qualitative evaluation of C 2+ selectivity, which is fully supported by most of the experimental references and our experimental verification.
- 24 February 2026
- Published by Wiley in Angewandte Chemie International Edition
- Vol. 65 (21) , e24612
Top Cited PapersOpen AccessCatalysis AI Agent Guides Discovering the Universal Design Principle of Cu‐Based Single‐Atom Alloy Catalysts for CO 2 Electroreduction
Copper (Cu)‐based single‐atom alloys (SAAs) represent a promising strategy for optimizing the electroreduction of CO 2 (CO 2 R) to multi‐carbon products (C 2+ ). However, the diverse enhancement degrees of C 2+ selectivity brought about by various dopants have not yet been rationalized, which lead to the absence of guidelines for further designing desired Cu‐based SAAs. Herein, guided by the Catalysis AI Agent developed based on large‐scale data + large language model, as well as the Digital Catalysis Platform (the DigCat experimental database), we performed first‐principles calculations to evaluate C 2+ products selectivity trends through identifying the energy barrier of rate‐determining step (RDS) among diverse C‐C coupling pathways. With first‐principles results fed back, Catalysis AI Agent reveals that the element classification in the periodic table of guest metal dopant is essential for establishing robust structure‐selectivity correlations among Cu‐based SAAs. A structural descriptor (φ) is developed and helps to establish a strong correlation among the electronic‐scale structural features, the adsorption strength of C‐C coupling precursors, and the macroscopic C 2+ products selectivity. A universal design principle based on φ for Cu‐based SAAs enables the rapid and qualitative evaluation of C 2+ selectivity, which is fully supported by most of the experimental references and our experimental verification.
- 5 November 2025
- Published by Wiley in ChemCatChem
- Vol. 17 (23)
Review on the Carbon‐Supported Metal Catalysts with Coexistence of Single Atoms and Nanoparticles/Clusters for Electrocatalysis
Combining homonuclear or heteronuclear nanoparticles and clusters with single‐atom catalysts (i.e., NMP‐SACs) has become a strategy for developing efficient supported metal catalysts. In recent years, NMP‐SACs have garnered increased attention from various fields due to their innovative structures and superior catalytic performance. Compared to conventional single‐atom catalysts, NMP‐SACs achieve enhanced activity and stability through multiple effective mechanisms. For example, they can optimize kinetic processes via synergistic effects, or NMPs can promote charge redistribution at single‐atom (SA) sites. Emerging studies demonstrate that metal‐center electronic configurations in NMP‐SACs can be precisely tailored through metal‐support interactions, enabling optimized adsorption energetics of reactive intermediates. This electronic engineering strategy critically enhances reaction kinetics for fundamental processes, including oxygen evolution (OER), hydrogen evolution reactions (HER), and other electrocatalytic reactions. This review systematically summarizes recent advancements in NMP‐SACs across diverse catalytic scenarios. We first critically assess synthetic strategies and characterization methodologies, followed by a mechanistic analysis of their electrocatalytic performance in representative reactions. The origin of their superior activity is elucidated through electronic structure‐property correlations. Finally, we discuss persistent challenges and propose future research directions for optimizing NMP‐SACs.
- 5 November 2025
- Published by Wiley in ChemCatChem
- Vol. 17 (23)
Review on the Carbon‐Supported Metal Catalysts with Coexistence of Single Atoms and Nanoparticles/Clusters for Electrocatalysis
Combining homonuclear or heteronuclear nanoparticles and clusters with single‐atom catalysts (i.e., NMP‐SACs) has become a strategy for developing efficient supported metal catalysts. In recent years, NMP‐SACs have garnered increased attention from various fields due to their innovative structures and superior catalytic performance. Compared to conventional single‐atom catalysts, NMP‐SACs achieve enhanced activity and stability through multiple effective mechanisms. For example, they can optimize kinetic processes via synergistic effects, or NMPs can promote charge redistribution at single‐atom (SA) sites. Emerging studies demonstrate that metal‐center electronic configurations in NMP‐SACs can be precisely tailored through metal‐support interactions, enabling optimized adsorption energetics of reactive intermediates. This electronic engineering strategy critically enhances reaction kinetics for fundamental processes, including oxygen evolution (OER), hydrogen evolution reactions (HER), and other electrocatalytic reactions. This review systematically summarizes recent advancements in NMP‐SACs across diverse catalytic scenarios. We first critically assess synthetic strategies and characterization methodologies, followed by a mechanistic analysis of their electrocatalytic performance in representative reactions. The origin of their superior activity is elucidated through electronic structure‐property correlations. Finally, we discuss persistent challenges and propose future research directions for optimizing NMP‐SACs.
- 27 August 2025
- Published by American Chemical Society (ACS) in ACS Catalysis
- Vol. 15 (18) , 15765-15779
Reconstruction-Dependent Coordination Descriptor for Selectivity toward C2+ Products of CO2 Electroreduction over Cu Catalysts
- 27 August 2025
- Published by American Chemical Society (ACS) in ACS Catalysis
- Vol. 15 (18) , 15765-15779
Reconstruction-Dependent Coordination Descriptor for Selectivity toward C2+ Products of CO2 Electroreduction over Cu Catalysts
- 21 July 2025
- Published by American Chemical Society (ACS) in Accounts of Chemical Research
- Vol. 58 (16) , 2535-2549
Structural Descriptor Bridging the Microstructural Feature and Catalytic Reactivity for Rational Design of Metal Catalysts
ConspectusMetal heterogeneous catalysis is the workhorse of the chemical industry, driving the conversion of reactants to desirable products. Traditional design approaches for metal catalysts rely on trial-and-error tests and take a lot of time to identify promising catalytic active species from the large candidate space. Over the decades, much focus has been placed on identifying the factors affecting the active sites, which, in turn, guides the design and preparation of more active, selective, and stable catalysts. In the context of theoretical design method for catalysts, the concept of the energy descriptor strategy provides correlations between the adsorption energy of key intermediates and catalytic reactivity. Such energy descriptors for catalytic reactivity can be used to predict the activity of candidate catalysts and understand trends among different catalysts.However, a more efficient descriptor strategy is still attractive and needed that avoids density functional theory calculation on the adsorption energy of each candidate and possesses the guidance power for the rational design of microstructural characteristics of catalytic active species. In this regard, bridging the gap between the electronic/atomic-level descriptions of the microscopic properties of the catalytic active species and the macroscopic catalytic performance of the desirable reaction, that is, the microscopic-to-macroscopic relationship, remains intriguing yet challenging, toward which progress leads to revolutionizing catalyst design.In this Account, we propose a structural descriptor strategy that for the first time maps the quantitative relationship between microstructural features and catalytic performances for metal catalysts, as well as its application in the high-throughput screening and rational design of catalytic active species. We begin with the analysis of the microstructural characteristics of the reaction center and its coordination environment and extract key feature parameters to build a mathematical expression of the structural descriptor. Next, through regression fitting, a mathematical correlation is built between the structural descriptor and the energetics involved with the reaction pathway. Finally, substituting the above statistical correlations into the rate equation derived from microkinetic model offers the structural descriptor-based prediction model for metal catalysts. The use of easily accessible structural descriptors has proven to be a powerful method to advance and accelerate the discovery and design of metal catalysts, including atomically dispersed metal catalysts, metal alloy catalysts, and metal cluster catalysts. Overall, the structural descriptor strategy not only demonstrates much potential to elucidate the quantitative interplay between microstructural features of catalytic active species and intrinsic catalytic reactivity but also provides a new approach in kinetics analysis to rationalize metal catalyst design. We conclude with an outlook for further constructing a universal structural descriptor and accelerating predictions on catalytic performance of metal catalysts by leveraging material databases and machine learning.
- 21 July 2025
- Published by American Chemical Society (ACS) in Accounts of Chemical Research
- Vol. 58 (16) , 2535-2549
Structural Descriptor Bridging the Microstructural Feature and Catalytic Reactivity for Rational Design of Metal Catalysts
ConspectusMetal heterogeneous catalysis is the workhorse of the chemical industry, driving the conversion of reactants to desirable products. Traditional design approaches for metal catalysts rely on trial-and-error tests and take a lot of time to identify promising catalytic active species from the large candidate space. Over the decades, much focus has been placed on identifying the factors affecting the active sites, which, in turn, guides the design and preparation of more active, selective, and stable catalysts. In the context of theoretical design method for catalysts, the concept of the energy descriptor strategy provides correlations between the adsorption energy of key intermediates and catalytic reactivity. Such energy descriptors for catalytic reactivity can be used to predict the activity of candidate catalysts and understand trends among different catalysts.However, a more efficient descriptor strategy is still attractive and needed that avoids density functional theory calculation on the adsorption energy of each candidate and possesses the guidance power for the rational design of microstructural characteristics of catalytic active species. In this regard, bridging the gap between the electronic/atomic-level descriptions of the microscopic properties of the catalytic active species and the macroscopic catalytic performance of the desirable reaction, that is, the microscopic-to-macroscopic relationship, remains intriguing yet challenging, toward which progress leads to revolutionizing catalyst design.In this Account, we propose a structural descriptor strategy that for the first time maps the quantitative relationship between microstructural features and catalytic performances for metal catalysts, as well as its application in the high-throughput screening and rational design of catalytic active species. We begin with the analysis of the microstructural characteristics of the reaction center and its coordination environment and extract key feature parameters to build a mathematical expression of the structural descriptor. Next, through regression fitting, a mathematical correlation is built between the structural descriptor and the energetics involved with the reaction pathway. Finally, substituting the above statistical correlations into the rate equation derived from microkinetic model offers the structural descriptor-based prediction model for metal catalysts. The use of easily accessible structural descriptors has proven to be a powerful method to advance and accelerate the discovery and design of metal catalysts, including atomically dispersed metal catalysts, metal alloy catalysts, and metal cluster catalysts. Overall, the structural descriptor strategy not only demonstrates much potential to elucidate the quantitative interplay between microstructural features of catalytic active species and intrinsic catalytic reactivity but also provides a new approach in kinetics analysis to rationalize metal catalyst design. We conclude with an outlook for further constructing a universal structural descriptor and accelerating predictions on catalytic performance of metal catalysts by leveraging material databases and machine learning.