A Multiple Filter Based Neural Network Approach to the Extrapolation of Adsorption Energies on Metal Surfaces for Catalysis Applications

Chowdhury, Asif; Yang, Wenqiang; Abdelfatah, Kareem; Zare, Mahdi; Heyden, Andreas; Terejanu, Gabriel

doi:10.1021/acs.jctc.9b00986

Cited by 26 publications

(31 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the context of computational materials science and chemistry this usually translates to identifying catalysts for reactions of interest (for instance CO 2 reduction) and their respective activity and selectivity. The computational demand of such studies can be drastically decreased by employing ML, and a wide range of such investigations has been performed [155], [156], [157], [158], [159], [160], [161], [162], [163], [164], [165], [166], [167], [168], [169], [170], [171], [172], [173], [174], [175], [176], [177], [178], [179], [180], [181], [182], [183], [184], [185], [186], [187], [188], [189], [190], [191], [192], [193], [194], [195], [196], [197], [198], [199], [200], …”

Section: Chemical Reactionsmentioning

confidence: 99%

A Deep Dive into Machine Learning Density Functional Theory for Materials Science and Chemistry

Fiedler¹,

Shah²,

Bußmann³

et al. 2021

Preprint

View full text Add to dashboard Cite

With the growth of computational resources, the scope of electronic structure simulations has increased greatly. Artificial intelligence and robust data analysis hold the promise to accelerate large-scale simulations and their analysis to hitherto unattainable scales. Machine learning is a rapidly growing field for the processing of such complex datasets. It has recently gained traction in the domain of electronic structure simulations, where density functional theory takes the prominent role of the most widely used electronic structure method. Thus, DFT calculations represent one of the largest loads on academic high-performance computing systems across the world. Accelerating these with machine learning can reduce the resources required and enables simulations of larger systems. Hence, the combination of density functional theory and machine learning has the potential to rapidly advance electronic structure applications such as in-silico materials discovery and the search for new chemical reaction pathways. We provide the theoretical background of both density functional theory and machine learning on a generally accessible level. This serves as the basis of our comprehensive review including research articles up to December 2020 in chemistry and materials science that employ machine-learning techniques. In our analysis, we categorize the body of research into main threads and extract impactful results. We conclude our review with an outlook on exciting research directions in terms of a citation analysis.

show abstract

Section: Chemical Reactionsmentioning

confidence: 99%

A Deep Dive into Machine Learning Density Functional Theory for Materials Science and Chemistry

Fiedler¹,

Shah²,

Bußmann³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Previous studies − have shown that non-linear ML models outperform linear models in this case. It has also been shown , that flat molecular fingerprints based on SMILES notation give good predictive results when training and testing sets contain similar-sized molecules.…”

Section: Introductionmentioning

confidence: 99%

Comparative Study on the Machine Learning-Based Prediction of Adsorption Energies for Ring and Chain Species on Metal Catalyst Surfaces

et al. 2021

Self Cite

View full text Add to dashboard Cite

Computation of adsorption and transition-state energies for a large number of surface intermediates for numerous active site models poses significant computational overhead in computational screening of catalysts. Machine learning (ML) techniques can be used to predict part of these energies. To predict the energies, ML models need to be fed appropriate metal and species descriptors. For complex surface chemistries, the structures of the intermediate species can vary greatly. In this paper, working with the hydrodeoxygenation of succinic acid on six different metal surfaces, we have studied the effect of linear and non-linear ML models used along with pen-and-paper-based species descriptors and two categories of metal descriptors on two different categories of intermediate species: chain and ring. More specifically, our computations include the prediction of chain species when trained on only chain species and also when trained on both chain and ring species. Similar computations were performed for predictions of ring species. In each case, the results of linear ML models were compared with kernel-based non-linear models. Our results indicate that ring species data do not improve the prediction of chain species. Similarly, chain species data do not improve the prediction of ring species. The use of non-linear ML models, however, did help to minimize the prediction errors compared to the linear models. The study also shows that electronic or adsorption energybased metal descriptors along with bond count-based species fingerprints can achieve a mean absolute error (MAE) of less than 0.2 eV for complex chain molecules when used with an appropriate machine learning model.

show abstract

Section: ■ Introductionmentioning

confidence: 99%

“…Primary features generally originate from accumulation knowledge of experimental consequences, for example, typical size dependence as a prevalent phenomenon in numerous catalytic reactions . The primary features could be roughly categorized into three aspects based on geometric, electronic, and energetic effects, which are enveloping atomic radius, shape, cluster size, coordination number, graph representations, , shape parameters from sp band and d band in the electronic density of states, − scaling relationship between adsorbates, − and Brønsted−Evans−Polanyi relationship between activate barrier energy and adsorption energy. , In the past, Calle-Vallejo et al witnessed the scaling relations of adsorption energy for adsorbed atoms and their hydrogenated species when bound similarly to the surface from the dataset of near-surface alloys of Pt. Further, due to that the difference of transition metals primarily originates from d band, , d center and combinations of d band shape parameters (such as width, skewness, kurtosis, area) have been successfully applied to correlate adsorption energy (H, O, C, N, and their hydrogenated species) in some specific alloy datasets. , Further bringing in multi-primary features could promote the performance of models.…”

Section: Introductionmentioning

confidence: 99%

Feature Analysis: Revealing the Importance of Bulk Nearest-Neighbor Distance Feature for CO Adsorption Energy on Bimetallic Alloys

2021

J. Phys. Chem. C

View full text Add to dashboard Cite

An effective machine learning model FAIR (feature analysis based on linear regression) of high interpretation was proposed to predict CO adsorption energy on 105 kinds of bimetallic alloys with high accuracy (root-mean-square error (RMSE) 0.17 eV and R 2 0.97). We assessed 524 288 kinds of combinations as input features based on 19 primary features from elemental features and surface electronic features of bimetallic alloys. The results demonstrated that the model not only displayed excellent prediction performance but also captured the hidden physical meaning, implying that the distance between nearest atoms in the bulk of the element in the first layer is the most important feature that determines the accuracy of models; in addition, the emergence of d band center also tends to make some improvements. Furthermore, the FAIR model could be independent of massive amount of database, when the number of train database is far less than that of test database with the ratio of only 3:7, the RMSE still reached 0.13 eV. Taking CO2 electrochemical reduction reaction into account, several promising alloys, especially Au/Rh/Au(111), Au/Pt/Au(111), and Au/Ni/Au(111), are screened as promising candidates capable of transforming CO2 into valuable hydrocarbons based on reasonable CO adsorption energy range. Besides, the FAIR model as an effective tool for property prediction and feature analysis could also be applied to any system with explicit primary features and target.

show abstract

A Multiple Filter Based Neural Network Approach to the Extrapolation of Adsorption Energies on Metal Surfaces for Catalysis Applications

Cited by 26 publications

References 36 publications

A Deep Dive into Machine Learning Density Functional Theory for Materials Science and Chemistry

A Deep Dive into Machine Learning Density Functional Theory for Materials Science and Chemistry

Comparative Study on the Machine Learning-Based Prediction of Adsorption Energies for Ring and Chain Species on Metal Catalyst Surfaces

Feature Analysis: Revealing the Importance of Bulk Nearest-Neighbor Distance Feature for CO Adsorption Energy on Bimetallic Alloys

Contact Info

Product

Resources

About