[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-121076-en":3,"doc-seo-121076-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},121076,8796095462418,"Noah","https://ap-avatar.wpscdn.com/avatar/80000253c1241d02b47?x-image-process=image/resize,m_fixed,w_180,h_180&k=1778826106357471780",8,"Research & Report","Modelling atomic and nanoscale structure in the silicon–oxygen system through active machine learning","Silicon–oxygen compounds underpin minerals, semiconductors, and catalysis, yet their nanoscale heterogeneity is hard to capture beyond atomic length scales. This work develops a unified computational description of the full Si–O system using atomistic machine learning integrated with an active-learning workflow. The approach builds reference datasets of local atomic environments, fits accurate ML potentials, and demonstrates high-accuracy models for very-high-pressure silica, interfaces such as surfaces and aerogels, and amorphous silicon monoxide.","Article [https://doi.org/10.1038/s41467-024-45840-9](https://doi.org/10.1038/s41467-024-45840-9)  \nModelling atomic and nanoscale structure in the silicon–oxygen system through active machine learning  \nReceived: 6 July 2023  \n\n| Accepted: 2 February 2024 |\n| --- |\n|  |\n| Check for updates |\n\nLinus C. Erhard 1, Jochen Rohrer1 , Karsten Albe 1  & Volker L. Deringer 2   \nSilicon–oxygen compounds are among the most important ones in the natural sciences, occurring as building blocks in minerals and being used in semiconductors and catalysis. Beyond the well-known silicon dioxide, there are phases with different stoichiometric composition and nanostructured composites. One of the key challenges in understanding the Si–O system is therefore to accurately account for its nanoscale heterogeneity beyond the length scale of individual atoms. Here we show that a unified computational description of the full Si–O system is indeed possible, based on atomistic machine learning coupled to an active-learning workflow. We showcase applications to very-high-pressure silica, to surfaces and aerogels, and to the structure of amorphous silicon monoxide. In a wider context, our work illustrates how structural complexity in functional materials beyond the atomic and few-nanometre length scales can be captured with active machine learning.  \nElemental silicon and its oxide, silica (SiO2), are widely studied building blocks of the world around us1: from minerals in geology to siliconbased computing architectures; thin-film solar cells in which amorphous silicon is the active material2; or zeolite catalysts based on the SiO2 parent composition3. Some of these materials have a single phase and are precisely defined on the atomic scale, whereas others show longer-ranging, hierarchical structures and varying degrees of disorder. For example, silica aerogels contain pores with sizes of 5–100 nm, leading to very low thermal conductivity and making aerogels promising candidates for thermal insulation4. Under pressure, SiO2 shows amorphous–amorphous transitions to structures exceeding sixfold coordination5, crystallisation from the amorphous phase under shock compression6, and conversely the formation of complex disordered phases from crystalline SiO27. Beyond fundamental studies, there is much technological importance in silicon–oxygen phases with nanoscale structure—the interface between Si and SiO2 is essential in silicon metal-oxide semiconductors, and defects at this interface have been investigated for decades8–10.  \nA material in the binary silicon–oxygen system which is in fact dominated by such interfaces is the so-called silicon monoxide (SiO). The structure of SiO was controversially discussed for long11,12; today, it is known as a nanoscopic mixture of amorphous Si and SiO213–15. Initial applications of SiO have been in protective layers for mirrors16 or dielectrics for thin-film capacitors17; more recently, the same material has emerged as a promising anode material for lithium-ion batteries18,19. However, to be able to fully exploit SiO in nextgeneration energy-storage solutions, it would be valuable to understand the features ofthe nanoscopic structure on an atomistic level.  \nTo develop atomic-scale models ofcomplex materials such as SiO, molecular-dynamics (MD) computer simulations have become a central research tool. While there are now plenty of interatomic potentials for silicon20–22 and silica23–25, the number of potentials for the mixed (i.e., full binary) system is limited due to its chemical complexity26–29. Alongside established, empirically fitted potentials based on physical models, alternatives based on large datasets and machine learning (ML) have emerged in recent years. These models have been fitted for  \n1Institute of Materials Science, Technische Universität Darmstadt, Otto-Berndt-Strasse 3, D-64287 Darmstadt, Germany. 2Department of Chemistry, Inorganic  \nChemistry Laboratory, University of Oxford, Oxford OX1 3QR, United Kingdom. e","cbCaiu0RWHHAGCA1","https://ap.wps.com/l/cbCaiu0RWHHAGCA1","pdf",3547888,1,12,"English","en",105,"# Background and motivation\n## Challenges of modelling Si–O nanoscale heterogeneity\n## Importance of interfaces and amorphous SiO\n# Method and active-learning workflow\n## Building datasets and quantum-mechanical reference data\n## Fitting ML potentials with active learning\n# Results\n## Accuracy across configurational space\n## 10-nm-scale atomistically resolved SiO structure models\n# Applications\n## High-pressure silica, surfaces, and aerogels","[{\"question\":\"What problem does the work address in modelling the Si–O system?\",\"answer\":\"It targets the difficulty of representing nanoscale structural heterogeneity in the Si–O system beyond individual-atom length scales.\"},{\"question\":\"How does the method combine machine learning with active learning?\",\"answer\":\"It extracts representative local atomic environments from large simulations, embeds them in an amorphous matrix to sample environments, and uses an active-learning workflow to select what to include for training ML potentials.\"},{\"question\":\"What systems and length scales does the final model enable?\",\"answer\":\"The model achieves high accuracy for high-pressure silica, silica surfaces and mixtures, and it produces fully atomistically resolved ~10-nm-scale structure models of amorphous and partially crystalline silicon monoxide.\"}]","Modelling atomic and nanoscale structure in the silicon–oxygen system through active machine learning | PDF",1785733597,30,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"modelling-atomic-and-nanoscale-structure-in-the-siliconoxygen-system-through-active-machine-learning","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/modelling-atomic-and-nanoscale-structure-in-the-siliconoxygen-system-through-active-machine-learning/121076/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-03",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does the work address in modelling the Si–O system?","Question",{"text":75,"@type":76},"It targets the difficulty of representing nanoscale structural heterogeneity in the Si–O system beyond individual-atom length scales.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the method combine machine learning with active learning?",{"text":80,"@type":76},"It extracts representative local atomic environments from large simulations, embeds them in an amorphous matrix to sample environments, and uses an active-learning workflow to select what to include for training ML potentials.",{"name":82,"@type":73,"acceptedAnswer":83},"What systems and length scales does the final model enable?",{"text":84,"@type":76},"The model achieves high accuracy for high-pressure silica, silica surfaces and mixtures, and it produces fully atomistically resolved ~10-nm-scale structure models of amorphous and partially crystalline silicon monoxide.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":29,"slug":121},"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]