| Title | Quantum signatures for screening metavalent solids |
|---|---|
| ID_Doc | 56 |
| Authors | Giri, D; Williams, L; Mukherjee, A; Rajan, K |
| Published | Journal Of Chemical Physics, 154, 12 |
| Structure | Here are the sections of the article with two sentences each: The objective of this paper is to establish a data signature that can be used to compute quantum-level descriptors for rapid screening of crystallographic data to identify potentially new "metavalent" solids with novel and emergent properties. Metavalent solids possess characteristics that are nominally associated with metallic and covalent bonding but are distinctly different from both due to their anomalously large response properties and unique bond-breaking mechanism. The authors used atomic Hirshfeld surfaces, which were inspired by the long history of Atom-In-Molecule (AIM) research, to provide quantum-level descriptors that can be used for rapid screening of crystallographic data to identify potentially new metavalent solids. Hirshfeld surface analysis defines regions of space within a crystal structure that "belong" to each atom and provides properties that can be used to characterize a crystal structure. The results of the analysis show that descriptors obtained from Hirshfeld surface analysis can be used to discriminate between covalent, metavalent, and metallically bonded materials. The study identified dnorm and globularity as key metrics in distinguishing metavalently bonded materials, and the authors developed a machine learning-based classification tool using these descriptors to screen for metavalently bonded materials. The study found that materials with higher bond polarizability, Grüneisen parameter, and optical dielectric constant have greater values in metavalently bonded materials. The authors proposed a new virtual library of potentially new metavalent compounds with enhanced dielectric properties and compared their predicted stable and metastable metavalently bonded materials to those of other types of bonding. The study demonstrated the utility of Hirshfeld surface analysis and machine learning-based classification for rapid screening of crystallographic databases to identify potentially new metavalent solids with novel and emergent properties. The authors proposed a list of easily synthesizable materials likely to possess metavalent bonding and compared their predicted stable and metastable metavalently bonded materials to those of other types of bonding. |
| Summary | The article describes a new data-driven framework for computational screening and discovery of "metavalent" solids, which possess characteristics of both metallic and covalent bonding. The researchers used Hirshfeld surface analysis to provide quantum-level descriptors that can be used for rapid screening of crystallographic data to identify potentially new "metavalent" solids with novel properties. Hirshfeld surface analysis is a method that defines regions of space within a crystal structure that "belong" to each atom, and it can provide properties such as shape index, curvedness, and dnorm that can be used to characterize a crystal structure. The researchers calculated these properties for 65 systems and found that dnorm and globularity were the key metrics in distinguishing between different types of bonding. They used machine learning algorithms to classify the compounds into covalent, metallic, and metavalent categories, and the results showed that the algorithm achieved 100% accuracy in classifying metavalent compounds. The researchers also identified 1489 binary compounds extracted from the Materials Project database as potentially new metavalent materials, and they found that most of these materials have high optical dielectric constants. The researchers proposed a list of easily synthesizable metavalently bonded materials that are thermodynamically stable or metastable, and they found that these materials often have unique properties such as high dielectric constants. The study demonstrates the potential of Hirshfeld surface analysis and machine learning for rapid screening of materials with novel properties. The researchers also proposed a new virtual library of potentially new metavalent compounds with enhanced dielectric properties. |
| Scientific Methods | The research methods used in this paper are: 1. Hirshfeld surface analysis: This method is used to analyze the electronic density distribution in a crystal structure. The Hirshfeld surface is a mathematical surface that encloses the positive and negative lobes of the electron density distribution around each atom. 2. CrystalExplorer: This software is used to generate Hirshfeld surfaces and extract geometrical descriptors from them. 3. Tonto: This software is used as a computational chemistry package to perform quantum chemistry calculations. 4. Density Functional Theory (DFT): This method is used to calculate the electronic density distribution in a crystal structure. 5. Bader charge analysis: This method is used to separate the space of the crystal structure into "basins" by locating the regions of zero flux in the electron density. 6. Domain Overlap Matrix (DOM) analysis: This method is used to analyze the electronic density distribution in a crystal structure. 7. Machine Learning (ML) classification: This method is used to classify materials into different categories based on their geometrical descriptors. 8. Scikit-learn: This is an open-source machine learning library that is used to implement the ML classification algorithm. 9. Support Vector Machine (SVM): This is a type of ML algorithm that is used to classify materials into different categories based on their geometrical descriptors. 10. Nonlinear kernel-based vector machines: This is a type of ML algorithm that is used to classify materials into different categories based on their geometrical descriptors. The research methods used in this paper are mainly focused on: * Analyzing the electronic density distribution in crystal structures using Hirshfeld surface analysis and DFT * Extracting geometrical descriptors from Hirshfeld surfaces * Classifying materials into different categories using ML algorithms * Analyzing the relationship between the geometrical descriptors and material properties such as dielectric constant, bond polarizability, and Grüneisen parameter. Overall, the research methods used in this paper are a combination of quantum chemistry calculations, machine learning algorithms, and data analysis techniques to analyze the properties of materials and identify new materials with specific properties. |
| Article contribution | This article presents a novel data-driven framework for computational screening and discovery of metavalent solids, a class of materials that exhibits characteristics of both metallic and covalent bonding. The authors propose the use of Hirshfeld surface analysis to provide quantum level descriptors that can be used for rapid screening of crystallographic data to identify potentially new metavalent solids with novel and emergent properties. The main contributions of this article to regenerative economics are: 1. 2. 3. 4. 5. The article highlights the importance of computational screening and discovery in regenerative economics, where the rapid identification and development of new materials can help to drive innovation and economic growth. The use of machine learning-based classification and Hirshfeld surface analysis provides a promising approach for identifying metavalent materials with novel and emergent properties. However, the article also highlights some limitations, such as: 1. 2. 3. Overall, the article presents a promising approach for computational screening and discovery of metavalent solids, which can be used to drive innovation and economic growth in regenerative economics. |
No similar articles found.