Direct prediction of electrical properties of grain boundaries from photoluminescence profiles using machine learning

We present a machine learning model to directly predict the carrier recombination velocity, vGB, at the grain boundary (GB) from the measured photoluminescence (PL) intensity profile by training it with numerical simulation results. As the training dataset, 1800 PL profiles were calculated with a co...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Applied physics letters 2021-07, Vol.119 (3)
Hauptverfasser: Kutsukake, Kentaro, Mitamura, Kazuki, Usami, Noritaka, Kojima, Takuto
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue 3
container_start_page
container_title Applied physics letters
container_volume 119
creator Kutsukake, Kentaro
Mitamura, Kazuki
Usami, Noritaka
Kojima, Takuto
description We present a machine learning model to directly predict the carrier recombination velocity, vGB, at the grain boundary (GB) from the measured photoluminescence (PL) intensity profile by training it with numerical simulation results. As the training dataset, 1800 PL profiles were calculated with a combination of random values of four material properties—vGB, the GB inclination angle, and the carrier diffusion lengths in the grains on both sides of the GB. In addition, the measured noise was modeled artificially and applied to the simulated profiles. A neural network was constructed with the inputs of the PL profile and the outputs of the four properties. This served as the solver of the reverse problem of the computational simulation. The coefficient of determination and the root mean squared error of vlog, which is the common logarithm of vGB, for the test dataset were 0.97 and 0.245, respectively. This prediction error was sufficiently low for the practical estimation of vGB. Moreover, the calculation time was reduced by a factor of 198 000 compared to conventional numerical optimization of repeating the computational simulations. By utilizing this fast prediction method, continuous evaluation of vGB along a GB was demonstrated. The finding is expected to advance scientific investigation of the electrical properties of local defects.
doi_str_mv 10.1063/5.0049847
format Article
fullrecord <record><control><sourceid>proquest_cross</sourceid><recordid>TN_cdi_crossref_primary_10_1063_5_0049847</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2554099355</sourcerecordid><originalsourceid>FETCH-LOGICAL-c327t-f5821dd36e61f538795e2d7838dc7c8b0928047168262eca2beb3eea629012003</originalsourceid><addsrcrecordid>eNqdkE1LxDAQhoMouK4e_AcFTwpd89E06VHWT1jwoueQppPdLG1Tk1bw35uyC949hXnfZzIzL0LXBK8ILtk9X2FcVLIQJ2hBsBA5I0SeogXGmOVlxck5uohxn0pOGVug6dEFMGM2BGicGZ3vM28zaJMWnNFtMvwAYXQQZ2MbtOuz2k99o8Os2eC7bNj50bdT53qIBnoDc5d1bfKn6Ppt1mmzS2bWgg59Ei7RmdVthKvju0Sfz08f69d88_7ytn7Y5IZRMeaWS0qahpVQEsuZFBUH2gjJZGOEkTWuqMSFIKWkJQWjaQ01A9AlrTCh6eIlujn8m_b5miCOau-n0KeRinJe4KpinCfq9kCZ4GMMYNUQXKfDjyJYzakqro6pJvbuwEbjRj3n9T_424c_UA2NZb8vBodu</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2554099355</pqid></control><display><type>article</type><title>Direct prediction of electrical properties of grain boundaries from photoluminescence profiles using machine learning</title><source>AIP Journals Complete</source><source>Alma/SFX Local Collection</source><creator>Kutsukake, Kentaro ; Mitamura, Kazuki ; Usami, Noritaka ; Kojima, Takuto</creator><creatorcontrib>Kutsukake, Kentaro ; Mitamura, Kazuki ; Usami, Noritaka ; Kojima, Takuto</creatorcontrib><description>We present a machine learning model to directly predict the carrier recombination velocity, vGB, at the grain boundary (GB) from the measured photoluminescence (PL) intensity profile by training it with numerical simulation results. As the training dataset, 1800 PL profiles were calculated with a combination of random values of four material properties—vGB, the GB inclination angle, and the carrier diffusion lengths in the grains on both sides of the GB. In addition, the measured noise was modeled artificially and applied to the simulated profiles. A neural network was constructed with the inputs of the PL profile and the outputs of the four properties. This served as the solver of the reverse problem of the computational simulation. The coefficient of determination and the root mean squared error of vlog, which is the common logarithm of vGB, for the test dataset were 0.97 and 0.245, respectively. This prediction error was sufficiently low for the practical estimation of vGB. Moreover, the calculation time was reduced by a factor of 198 000 compared to conventional numerical optimization of repeating the computational simulations. By utilizing this fast prediction method, continuous evaluation of vGB along a GB was demonstrated. The finding is expected to advance scientific investigation of the electrical properties of local defects.</description><identifier>ISSN: 0003-6951</identifier><identifier>EISSN: 1077-3118</identifier><identifier>DOI: 10.1063/5.0049847</identifier><identifier>CODEN: APPLAB</identifier><language>eng</language><publisher>Melville: American Institute of Physics</publisher><subject>Applied physics ; Carrier recombination ; Computer simulation ; Crystal defects ; Datasets ; Electrical properties ; Grain boundaries ; Inclination angle ; Machine learning ; Material properties ; Mathematical models ; Neural networks ; Optimization ; Photoluminescence ; Training</subject><ispartof>Applied physics letters, 2021-07, Vol.119 (3)</ispartof><rights>Author(s)</rights><rights>2021 Author(s). Published under an exclusive license by AIP Publishing.</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c327t-f5821dd36e61f538795e2d7838dc7c8b0928047168262eca2beb3eea629012003</citedby><cites>FETCH-LOGICAL-c327t-f5821dd36e61f538795e2d7838dc7c8b0928047168262eca2beb3eea629012003</cites><orcidid>0000-0002-4141-3587 ; 0000-0001-7878-347X ; 0000-0003-3784-5667 ; 0000-0002-0602-2847</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://pubs.aip.org/apl/article-lookup/doi/10.1063/5.0049847$$EHTML$$P50$$Gscitation$$H</linktohtml><link.rule.ids>314,776,780,790,4498,27901,27902,76126</link.rule.ids></links><search><creatorcontrib>Kutsukake, Kentaro</creatorcontrib><creatorcontrib>Mitamura, Kazuki</creatorcontrib><creatorcontrib>Usami, Noritaka</creatorcontrib><creatorcontrib>Kojima, Takuto</creatorcontrib><title>Direct prediction of electrical properties of grain boundaries from photoluminescence profiles using machine learning</title><title>Applied physics letters</title><description>We present a machine learning model to directly predict the carrier recombination velocity, vGB, at the grain boundary (GB) from the measured photoluminescence (PL) intensity profile by training it with numerical simulation results. As the training dataset, 1800 PL profiles were calculated with a combination of random values of four material properties—vGB, the GB inclination angle, and the carrier diffusion lengths in the grains on both sides of the GB. In addition, the measured noise was modeled artificially and applied to the simulated profiles. A neural network was constructed with the inputs of the PL profile and the outputs of the four properties. This served as the solver of the reverse problem of the computational simulation. The coefficient of determination and the root mean squared error of vlog, which is the common logarithm of vGB, for the test dataset were 0.97 and 0.245, respectively. This prediction error was sufficiently low for the practical estimation of vGB. Moreover, the calculation time was reduced by a factor of 198 000 compared to conventional numerical optimization of repeating the computational simulations. By utilizing this fast prediction method, continuous evaluation of vGB along a GB was demonstrated. The finding is expected to advance scientific investigation of the electrical properties of local defects.</description><subject>Applied physics</subject><subject>Carrier recombination</subject><subject>Computer simulation</subject><subject>Crystal defects</subject><subject>Datasets</subject><subject>Electrical properties</subject><subject>Grain boundaries</subject><subject>Inclination angle</subject><subject>Machine learning</subject><subject>Material properties</subject><subject>Mathematical models</subject><subject>Neural networks</subject><subject>Optimization</subject><subject>Photoluminescence</subject><subject>Training</subject><issn>0003-6951</issn><issn>1077-3118</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><recordid>eNqdkE1LxDAQhoMouK4e_AcFTwpd89E06VHWT1jwoueQppPdLG1Tk1bw35uyC949hXnfZzIzL0LXBK8ILtk9X2FcVLIQJ2hBsBA5I0SeogXGmOVlxck5uohxn0pOGVug6dEFMGM2BGicGZ3vM28zaJMWnNFtMvwAYXQQZ2MbtOuz2k99o8Os2eC7bNj50bdT53qIBnoDc5d1bfKn6Ppt1mmzS2bWgg59Ei7RmdVthKvju0Sfz08f69d88_7ytn7Y5IZRMeaWS0qahpVQEsuZFBUH2gjJZGOEkTWuqMSFIKWkJQWjaQ01A9AlrTCh6eIlujn8m_b5miCOau-n0KeRinJe4KpinCfq9kCZ4GMMYNUQXKfDjyJYzakqro6pJvbuwEbjRj3n9T_424c_UA2NZb8vBodu</recordid><startdate>20210719</startdate><enddate>20210719</enddate><creator>Kutsukake, Kentaro</creator><creator>Mitamura, Kazuki</creator><creator>Usami, Noritaka</creator><creator>Kojima, Takuto</creator><general>American Institute of Physics</general><scope>AAYXX</scope><scope>CITATION</scope><scope>8FD</scope><scope>H8D</scope><scope>L7M</scope><orcidid>https://orcid.org/0000-0002-4141-3587</orcidid><orcidid>https://orcid.org/0000-0001-7878-347X</orcidid><orcidid>https://orcid.org/0000-0003-3784-5667</orcidid><orcidid>https://orcid.org/0000-0002-0602-2847</orcidid></search><sort><creationdate>20210719</creationdate><title>Direct prediction of electrical properties of grain boundaries from photoluminescence profiles using machine learning</title><author>Kutsukake, Kentaro ; Mitamura, Kazuki ; Usami, Noritaka ; Kojima, Takuto</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c327t-f5821dd36e61f538795e2d7838dc7c8b0928047168262eca2beb3eea629012003</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Applied physics</topic><topic>Carrier recombination</topic><topic>Computer simulation</topic><topic>Crystal defects</topic><topic>Datasets</topic><topic>Electrical properties</topic><topic>Grain boundaries</topic><topic>Inclination angle</topic><topic>Machine learning</topic><topic>Material properties</topic><topic>Mathematical models</topic><topic>Neural networks</topic><topic>Optimization</topic><topic>Photoluminescence</topic><topic>Training</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Kutsukake, Kentaro</creatorcontrib><creatorcontrib>Mitamura, Kazuki</creatorcontrib><creatorcontrib>Usami, Noritaka</creatorcontrib><creatorcontrib>Kojima, Takuto</creatorcontrib><collection>CrossRef</collection><collection>Technology Research Database</collection><collection>Aerospace Database</collection><collection>Advanced Technologies Database with Aerospace</collection><jtitle>Applied physics letters</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Kutsukake, Kentaro</au><au>Mitamura, Kazuki</au><au>Usami, Noritaka</au><au>Kojima, Takuto</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Direct prediction of electrical properties of grain boundaries from photoluminescence profiles using machine learning</atitle><jtitle>Applied physics letters</jtitle><date>2021-07-19</date><risdate>2021</risdate><volume>119</volume><issue>3</issue><issn>0003-6951</issn><eissn>1077-3118</eissn><coden>APPLAB</coden><abstract>We present a machine learning model to directly predict the carrier recombination velocity, vGB, at the grain boundary (GB) from the measured photoluminescence (PL) intensity profile by training it with numerical simulation results. As the training dataset, 1800 PL profiles were calculated with a combination of random values of four material properties—vGB, the GB inclination angle, and the carrier diffusion lengths in the grains on both sides of the GB. In addition, the measured noise was modeled artificially and applied to the simulated profiles. A neural network was constructed with the inputs of the PL profile and the outputs of the four properties. This served as the solver of the reverse problem of the computational simulation. The coefficient of determination and the root mean squared error of vlog, which is the common logarithm of vGB, for the test dataset were 0.97 and 0.245, respectively. This prediction error was sufficiently low for the practical estimation of vGB. Moreover, the calculation time was reduced by a factor of 198 000 compared to conventional numerical optimization of repeating the computational simulations. By utilizing this fast prediction method, continuous evaluation of vGB along a GB was demonstrated. The finding is expected to advance scientific investigation of the electrical properties of local defects.</abstract><cop>Melville</cop><pub>American Institute of Physics</pub><doi>10.1063/5.0049847</doi><tpages>6</tpages><orcidid>https://orcid.org/0000-0002-4141-3587</orcidid><orcidid>https://orcid.org/0000-0001-7878-347X</orcidid><orcidid>https://orcid.org/0000-0003-3784-5667</orcidid><orcidid>https://orcid.org/0000-0002-0602-2847</orcidid></addata></record>
fulltext fulltext
identifier ISSN: 0003-6951
ispartof Applied physics letters, 2021-07, Vol.119 (3)
issn 0003-6951
1077-3118
language eng
recordid cdi_crossref_primary_10_1063_5_0049847
source AIP Journals Complete; Alma/SFX Local Collection
subjects Applied physics
Carrier recombination
Computer simulation
Crystal defects
Datasets
Electrical properties
Grain boundaries
Inclination angle
Machine learning
Material properties
Mathematical models
Neural networks
Optimization
Photoluminescence
Training
title Direct prediction of electrical properties of grain boundaries from photoluminescence profiles using machine learning
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-09T04%3A14%3A35IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_cross&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Direct%20prediction%20of%20electrical%20properties%20of%20grain%20boundaries%20from%20photoluminescence%20profiles%20using%20machine%20learning&rft.jtitle=Applied%20physics%20letters&rft.au=Kutsukake,%20Kentaro&rft.date=2021-07-19&rft.volume=119&rft.issue=3&rft.issn=0003-6951&rft.eissn=1077-3118&rft.coden=APPLAB&rft_id=info:doi/10.1063/5.0049847&rft_dat=%3Cproquest_cross%3E2554099355%3C/proquest_cross%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2554099355&rft_id=info:pmid/&rfr_iscdi=true