Abstract:Catalysis involves complex reactions with dynamic changes in catalyst morphology, challenging the capabilities of traditional Density Functional Theory (DFT) methods. To address this, we present the Catalytic Large Atomic Model (CLAM), a machine-learning-based interatomic potential designed for heterogeneous catalysis. Trained on a comprehensive dataset that includes metals, alloys, oxides, clusters, zeolites, 2D materials, and small molecules, CLAM ensures high accuracy across diverse catalytic systems. We also introduce a "local fine-tuning" algorithm that enhances the model’s applicability by accelerating structural optimizations and transition state searches while maintaining precision. Additionally, CLAM facilitates rapid reaction network construction and efficient kinetic analysis. This work advances computational catalysis by providing a universal and robust tool for catalyst design and mechanism exploration.

What problem does this paper attempt to address?

### Problems the Paper Aims to Solve This paper aims to address the challenges posed by the complex reactions and dynamic changes in catalyst morphology during catalytic processes. Specifically: 1. **Limitations of Traditional Methods**: While Density Functional Theory (DFT) is very useful in revealing reaction mechanisms, it has limitations in accurately describing the complex changes in catalyst morphology and active sites. Although first-principles molecular dynamics (AIMD) simulations based on DFT can capture these dynamic behaviors, their high computational cost limits their application in large-scale systems. 2. **Development of Machine Learning Potentials (MLIPs)**: To overcome these limitations, the authors propose a machine learning-based interatomic potential (Catalytic Large Atomic Model, CLAM). CLAM is trained on a comprehensive dataset covering diverse catalytic systems such as metals, alloys, oxides, clusters, zeolites, 2D materials, and small molecules. Additionally, a "local fine-tuning" algorithm is introduced to accelerate structure optimization and transition state search while maintaining accuracy. 3. **Dataset Construction**: The paper details the process of constructing the dataset used to train CLAM, including the selection and processing methods of different material systems. By using consistent computational parameters, the high quality of the dataset is ensured, and VASP is utilized for structural relaxation and AIMD simulations. 4. **Model Training and Evaluation**: The CLAM model is trained using the DPA1 model from DeePMD-kit and the GemNet-OC model, with a detailed explanation of the training strategies. The accuracy of the trained model in energy and force prediction is validated by evaluating its performance on different datasets. In summary, the main goal of this paper is to develop an efficient and versatile tool for catalyst design and reaction mechanism exploration, particularly in the field of heterogeneous catalysis.

Catalytic Large Atomic Model (CLAM): A Machine-Learning-Based Interatomic Potential Universal Model

Catalytic Large Atomic Model (CLAM): A Machine-Learning-Based Interatomic Potential Universal Model

Machine Learning Interatomic Potentials for Catalysis

Machine-learning-accelerated Discovery of Single-Atom Catalysts Based on Bidirectional Activation Mechanism

Examining Generalizability of AI Models for Catalysis

Machine-learning atomic simulation for heterogeneous catalysis

Fine-structure sensitive deep learning framework for predicting catalytic properties with high precision

Predicting binding motifs of complex adsorbates using machine learning with a physics-inspired graph representation

An Atomistic Machine Learning Package for Surface Science and Catalysis

Machine learning models predict calculation outcomes with the transferability necessary for computational catalysis

CataLM: Empowering Catalyst Design Through Large Language Models

Developing General Reactive Element-Based Machine Learning Potentials as the Main Computational Engine for Heterogeneous Catalysis

Data-efficient modeling of catalytic reactions via enhanced sampling and on-the-fly learning of machine learning potentials

Open Challenges in Developing Generalizable Large Scale Machine Learning Models for Catalyst Discovery

Machine Learning Potentials for Heterogeneous Catalysis

AdsorbML: A Leap in Efficiency for Adsorption Energy Calculations using Generalizable Machine Learning Potentials

Machine Learning Molecular Dynamics Shows Anomalous Entropic Effect on Catalysis via Surface Pre‐melting of Nanoclusters

Machine Learning-Assisted Screening of Stepped Alloy Surfaces for C 1 Catalysis

Machine Learning for Atomic Simulation and Activity Prediction in Heterogeneous Catalysis: Current Status and Future

Applying Large Graph Neural Networks to Predict Transition Metal Complex Energies Using the tmQM_wB97MV Dataset

Open Challenges in Developing Generalizable Large-Scale Machine-Learning Models for Catalyst Discovery