Abstract:Computational screening is indispensable for the efficient design of high-entropy alloys (HEAs), which hold considerable potential for catalytic applications. However, the chemical space of HEAs is exponentially vast with respect to the number of constituent elements, making even machine learning-based screening calculations time-intensive. To address this challenge, we propose a rapid method for predicting HEA properties using data from monometallic systems (or few-component alloys). Central to our approach is the newly introduced local surface energy (LSE) descriptor, which captures local surface reactivity at atomic resolution. By applying linear regression, we successfully screened adsorption energies of molecules on HEAs based on LSEs derived from monometallic systems. Furthermore, we developed high-precision models by employing both classical and quantum machine learning. Our method enabled CO adsorption-energy calculations for 1000 quinary nanoparticles, comprising 201 atoms each, within a few days, considerably faster than density functional theory, which would require hundreds of years or neural network potentials, which would have taken hundreds of days. The proposed approach accelerates the exploration of the vast HEA chemical space, facilitating the design of novel catalysts.

What problem does this paper attempt to address?

### Problems the paper attempts to solve This paper aims to solve the computational efficiency problem in high - entropy alloy (HEAs) catalyst screening. Specifically: 1. **Exponential growth of chemical space**: High - entropy alloys are composed of five or more elements, and the concentration range of each element is between 5 and 35 atomic percent. This combination leads to an exponential growth of the chemical space, making screening calculations even using machine - learning methods very time - consuming. 2. **Limitations of traditional computational methods**: - **First - principles calculations (such as density functional theory DFT)**: Although it can describe chemical reactions on solid surfaces with high precision, the amount of calculation is huge, especially for the complex surface environment of HEAs. - **Neural network potentials (NNPs)**: Although it improves computational efficiency, it still requires a large amount of computational resources, especially when dealing with multi - element systems. 3. **Deficiencies of existing descriptors**: Existing machine - learning models based on descriptors can predict adsorption energy, but their effectiveness is limited when dealing with the complex composition of HEAs and they rely on a large number of first - principles calculations. ### Solutions To solve the above problems, the paper proposes a new method to predict the molecular adsorption energy of HEAs by introducing the local surface energy (LSE) descriptor. The specific steps are as follows: 1. **Define the LSE descriptor**: \[ \text{LSE} \equiv E_{\text{surf,atom}} - E_{\text{bulk,atom}} \] where \( E_{\text{surf,atom}} \) and \( E_{\text{bulk,atom}} \) represent the energies of atoms in the surface and bulk environments respectively. LSE describes the energy loss of a single atom when exposed from a single - element bulk environment to a single - or multi - element surface, and can quantify the surface stability in a complex environment. 2. **Construct a prediction model**: - Use data from single - metal systems to construct a linear regression model to predict the molecular adsorption energy of multi - element surfaces. The form of the model is: \[ E_{\text{ad}}^{\text{Predict}}=\alpha_M\times\text{LSE}+\beta_M \] where \( \alpha_M \) and \( \beta_M \) are the regression coefficient and the constant term respectively, corresponding to each element \( M \). 3. **Verify the accuracy and efficiency of the model**: - Verify the reliability of the LSE descriptor and the accuracy of the prediction model by comparing with DFT calculation results. - Demonstrate the advantage of the LSE method in computational efficiency by comparing the calculation time. ### Main contributions 1. **Improve computational efficiency**: The LSE method can complete the CO adsorption energy calculation of 1,000 five - element nanoparticles (each containing 201 atoms) within a few days, which is much faster than DFT (which takes hundreds of years) and NNP (which takes hundreds of days). 2. **Expand the scope of application**: The LSE descriptor is not only applicable to single - element systems, but also can effectively predict the adsorption energy of multi - element HEAs, providing a powerful tool for the design of new catalysts. Through these methods, the paper has successfully accelerated the exploration of the HEAs chemical space and promoted the efficient design of new catalysts.

Lowering the Exponential Wall: Accelerating High-Entropy Alloy Catalysts Screening using Local Surface Energy Descriptors from Neural Network Potentials

Machine-Learning-Driven High-Entropy Alloy Catalyst Discovery to Circumvent the Scaling Relation for CO2 Reduction Reaction

Neural Network-Assisted Development of High-Entropy Alloy Catalysts: Decoupling Ligand and Coordination Effects

Ab Initio to Activity: Machine Learning-Assisted Optimization of High-Entropy Alloy Catalytic Activity

Rational design of high-entropy alloy nanoparticle catalysts through high-throughput composition space screening

Machine-Learning-Augmented Chemisorption Model for CO2 Electroreduction Catalyst Screening.

Developing Cheap but Useful Machine Learning-Based Models for Investigating High-Entropy Alloy Catalysts

Supervised AI and Deep Neural Networks to Evaluate High-Entropy Alloys as Reduction Catalysts in Aqueous Environments

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

Recent Progress in High‐entropy Nanocatalysts

Machine Learning-Assisted Screening of Stepped Alloy Surfaces for C1 Catalysis

Developing Cheap but Useful Machine Learning based Models for Investigating High-Entropy Alloy Catalysts

High-entropy alloy catalysts: high-throughput and machine learning-driven design

High-throughput and machine-learning accelerated design of high entropy alloy catalysts

Machine-Learning Adsorption on Binary Alloy Surfaces for Catalyst Screening

Machine-learning-assisted discovery of highly efficient high-entropy alloy catalysts for the oxygen reduction reaction

Machine Learning-Based Optimization Method for the Oxygen Evolution and Reduction Reaction of the High-Entropy Alloy Catalysts

Optimization of High Entropy Alloy Catalyst for Ammonia Decomposition and Ammonia Synthesis

Efficient screening of single phase forming low-activation high entropy alloys

Amorphous Catalysis: Machine Learning Driven High-Throughput Screening of Superior Active Site for Hydrogen Evolution Reaction