Learning from machine learning: the case of band-gap directness in semiconductors
Abstract Having a direct or indirect band gap can influence the potential applications of a semiconductor, for indirect band gap materials are usually not suitable for optoelectronic devices. Even though this is a fundamental property of semiconducting materials, discussed in textbooks, no unified t...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2024-02-01
|
| Series: | Discover Materials |
| Online Access: | https://doi.org/10.1007/s43939-024-00073-x |
| _version_ | 1797272829627990016 |
|---|---|
| author | Elton Ogoshi Mário Popolin-Neto Carlos Mera Acosta Gabriel M. Nascimento João N. B. Rodrigues Osvaldo N. Oliveira Fernando V. Paulovich Gustavo M. Dalpian |
| author_facet | Elton Ogoshi Mário Popolin-Neto Carlos Mera Acosta Gabriel M. Nascimento João N. B. Rodrigues Osvaldo N. Oliveira Fernando V. Paulovich Gustavo M. Dalpian |
| author_sort | Elton Ogoshi |
| collection | DOAJ |
| description | Abstract Having a direct or indirect band gap can influence the potential applications of a semiconductor, for indirect band gap materials are usually not suitable for optoelectronic devices. Even though this is a fundamental property of semiconducting materials, discussed in textbooks, no unified theory exists to explain why a material has a direct or indirect band gap. Here we used an interpretable machine learning model, the multiVariate dAta eXplanation (VAX) method, to gather information from a dataset of materials extracted from the Materials Project. The dataset contains more than 10000 entries, and atomic properties such as the number of electrons, electronic affinity and orbital energies were used as features to build random forest models that successfully explain the directness of the band gaps. Our results indicate that symmetry is an important feature that dictates the target property, which is the reason why our analysis is made based on sub-groups with similar structures. These sub-groups include materials with zincblende, rocksalt, wurtzite, and perovskite structures. Besides the symmetry of the materials, the existence or not of d bands and the relative energy of atomic orbitals were found to be important in defining whether a material’s band gap is direct or indirect. In conclusion, interpretable machine learning methods such as VAX can be useful in obtaining physical interpretation from materials databases. |
| first_indexed | 2024-03-07T14:35:11Z |
| format | Article |
| id | doaj.art-81dd9583cd514070bab796011b28a22d |
| institution | Directory Open Access Journal |
| issn | 2730-7727 |
| language | English |
| last_indexed | 2024-03-07T14:35:11Z |
| publishDate | 2024-02-01 |
| publisher | Springer |
| record_format | Article |
| series | Discover Materials |
| spelling | doaj.art-81dd9583cd514070bab796011b28a22d2024-03-05T20:41:22ZengSpringerDiscover Materials2730-77272024-02-014111410.1007/s43939-024-00073-xLearning from machine learning: the case of band-gap directness in semiconductorsElton Ogoshi0Mário Popolin-Neto1Carlos Mera Acosta2Gabriel M. Nascimento3João N. B. Rodrigues4Osvaldo N. Oliveira5Fernando V. Paulovich6Gustavo M. Dalpian7Center for Natural and Human Sciences, UFABCFederal Institute of São Paulo, IFSPCenter for Natural and Human Sciences, UFABCCenter for Natural and Human Sciences, UFABCCenter for Natural and Human Sciences, UFABCSão Carlos Institute of Physics, USPDepartment of Mathematics and Computer Science, Eindhoven University of TechnologyCenter for Natural and Human Sciences, UFABCAbstract Having a direct or indirect band gap can influence the potential applications of a semiconductor, for indirect band gap materials are usually not suitable for optoelectronic devices. Even though this is a fundamental property of semiconducting materials, discussed in textbooks, no unified theory exists to explain why a material has a direct or indirect band gap. Here we used an interpretable machine learning model, the multiVariate dAta eXplanation (VAX) method, to gather information from a dataset of materials extracted from the Materials Project. The dataset contains more than 10000 entries, and atomic properties such as the number of electrons, electronic affinity and orbital energies were used as features to build random forest models that successfully explain the directness of the band gaps. Our results indicate that symmetry is an important feature that dictates the target property, which is the reason why our analysis is made based on sub-groups with similar structures. These sub-groups include materials with zincblende, rocksalt, wurtzite, and perovskite structures. Besides the symmetry of the materials, the existence or not of d bands and the relative energy of atomic orbitals were found to be important in defining whether a material’s band gap is direct or indirect. In conclusion, interpretable machine learning methods such as VAX can be useful in obtaining physical interpretation from materials databases.https://doi.org/10.1007/s43939-024-00073-x |
| spellingShingle | Elton Ogoshi Mário Popolin-Neto Carlos Mera Acosta Gabriel M. Nascimento João N. B. Rodrigues Osvaldo N. Oliveira Fernando V. Paulovich Gustavo M. Dalpian Learning from machine learning: the case of band-gap directness in semiconductors Discover Materials |
| title | Learning from machine learning: the case of band-gap directness in semiconductors |
| title_full | Learning from machine learning: the case of band-gap directness in semiconductors |
| title_fullStr | Learning from machine learning: the case of band-gap directness in semiconductors |
| title_full_unstemmed | Learning from machine learning: the case of band-gap directness in semiconductors |
| title_short | Learning from machine learning: the case of band-gap directness in semiconductors |
| title_sort | learning from machine learning the case of band gap directness in semiconductors |
| url | https://doi.org/10.1007/s43939-024-00073-x |
| work_keys_str_mv | AT eltonogoshi learningfrommachinelearningthecaseofbandgapdirectnessinsemiconductors AT mariopopolinneto learningfrommachinelearningthecaseofbandgapdirectnessinsemiconductors AT carlosmeraacosta learningfrommachinelearningthecaseofbandgapdirectnessinsemiconductors AT gabrielmnascimento learningfrommachinelearningthecaseofbandgapdirectnessinsemiconductors AT joaonbrodrigues learningfrommachinelearningthecaseofbandgapdirectnessinsemiconductors AT osvaldonoliveira learningfrommachinelearningthecaseofbandgapdirectnessinsemiconductors AT fernandovpaulovich learningfrommachinelearningthecaseofbandgapdirectnessinsemiconductors AT gustavomdalpian learningfrommachinelearningthecaseofbandgapdirectnessinsemiconductors |