Single-Atom Alloy Dataset for Machine Learning

OverviewThis dataset, contained within <code>Database.csv</code>, is a comprehensive collection tailored for machine learning applications in the field of catalysis and materials science, focusing on single-atom alloys. It encompasses a wide array of data with 10,950 entries, each featuring 85 intrinsic descriptors alongside novel information on the predicted C-H dissociation energy barriers and reaction rates. These intrinsic descriptors include a variety of element and surface properties extracted from renowned databases like the Materials Project and Pymatgen, as well as surface structural features and characteristics derived through expert knowledge.Intrinsic DescriptorsThe 85 intrinsic descriptors provided in this dataset offer a detailed insight into the properties of single-atom alloys. These descriptors cover:<b>Element Properties:</b> Extracted from the Materials Project and Pymatgen databases, these properties include atomic size, electronegativity, and other elemental characteristics critical for the study of material properties.<b>Surface Properties:</b> Features related to the surface characteristics of the alloys, which play a significant role in their catalytic behavior and interaction with reactants.<b>Surface Structural Features:</b> Detailed information on the structural aspects of the alloy surfaces, which can influence the material's catalytic activity and stability.<b>Expert-Derived Features:</b> A set of features developed through expert knowledge, combining various data points to form comprehensive descriptors for machine learning applications.Predicted Properties<b>C-H Dissociation Energy Barrier:</b> A key metric for evaluating the catalytic efficiency of single-atom alloys, particularly in processes involving hydrocarbons.<b>Reaction Rates:</b> Provides valuable insights into the kinetics of reactions facilitated by single-atom alloys, crucial for the development and optimization of catalytic processes.