{"id":{"repo_id":"uic","oai_identifier":"oai:figshare.com:article/31451479"},"canonical_url":"https://search.dev.ndltd.org/etd/uic/oai:figshare.com:article/31451479","repository":{"repo_id":"uic","name":"University of Illinois - Chicago","base_url":"https://api.figshare.com/v2/oai"},"display":{"title":"AI-Driven Discovery and Multiscale Computational Frameworks for Carbon-Based Materials","abstract":"This dissertation develops a unified, data-driven framework for carbon-based materials discovery that integrates generative artificial intelligence, multiscale simulations, and reactive dynamics to accelerate research central to carbon-to-X technologies. Chapter 2 establishes a one-directional workflow in which a generative AI model autonomously proposes novel metal–organic frameworks (MOFs) for carbon capture, followed by multiscale physics-based screening. 120,000 MOF AI-generated candidates are evaluated through an array of screening methods that validate structure and property of MOF, yielding 6 structures with CO2 uptakes exceeding 2 mol/kg at 0.1 bar and 300 K—demonstrating that generative design coupled with high-throughput validation can efficiently navigate the vast MOF design space. Chapter 3 extends this framework into a closed-loop, property-guided generative system. Here, an oracle-based scoring function provides feedback from machine learning predicted adsorption metric, dynamically steering the generative model toward desirable regions of the property space. This self-optimizing loop improves both discovery efficiency, physical fidelity, and design success rate. Chapter 4 transitions to nanocarbon systems, employing reactive molecular dynamics and machine-learning prediction to investigate the temperature- and pressure-dependent graphitization of nanodiamond surfaces, revealing atomistic pathways for sp3-to-sp2 transformation. Chapter 5 explores the co-pyrolysis of cellulose–graphene mixtures through reactive dynamics, coupled with non-equilibrium Green’s function (NEGF) formalism to quantify key molecular junctions for electron transport. Together, these studies unify porous, sp2, and hybrid carbon materials within a coherent methodological and conceptual framework. By combining forward generative design, feedback-driven optimization, and transport-level modeling, this work contributes a scalable paradigm for AI-accelerated carbon-materials innovation aligned with global carbon capture, utilization, and conversion goals.","abstract_html":"This dissertation develops a unified, data-driven framework for carbon-based materials discovery that integrates generative artificial intelligence, multiscale simulations, and reactive dynamics to accelerate research central to carbon-to-X technologies. Chapter 2 establishes a one-directional workflow in which a generative AI model autonomously proposes novel metal–organic frameworks (MOFs) for carbon capture, followed by multiscale physics-based screening. 120,000 MOF AI-generated candidates are evaluated through an array of screening methods that validate structure and property of MOF, yielding 6 structures with CO2 uptakes exceeding 2 mol/kg at 0.1 bar and 300 K—demonstrating that generative design coupled with high-throughput validation can efficiently navigate the vast MOF design space. Chapter 3 extends this framework into a closed-loop, property-guided generative system. Here, an oracle-based scoring function provides feedback from machine learning predicted adsorption metric, dynamically steering the generative model toward desirable regions of the property space. This self-optimizing loop improves both discovery efficiency, physical fidelity, and design success rate. Chapter 4 transitions to nanocarbon systems, employing reactive molecular dynamics and machine-learning prediction to investigate the temperature- and pressure-dependent graphitization of nanodiamond surfaces, revealing atomistic pathways for sp3-to-sp2 transformation. Chapter 5 explores the co-pyrolysis of cellulose–graphene mixtures through reactive dynamics, coupled with non-equilibrium Green’s function (NEGF) formalism to quantify key molecular junctions for electron transport. Together, these studies unify porous, sp2, and hybrid carbon materials within a coherent methodological and conceptual framework. By combining forward generative design, feedback-driven optimization, and transport-level modeling, this work contributes a scalable paradigm for AI-accelerated carbon-materials innovation aligned with global carbon capture, utilization, and conversion goals.","abstract_has_math":false,"creators":["Xiaoli Yan (420602)"],"institution":null,"degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2025,"date_issued":"2025-12-01T00:00:00Z","date_published":"2025-12-01T00:00:00Z","updated_at":"2026-07-27T21:34:28Z","subjects":["Engineering","Materials Science"],"languages":[],"rights":["In Copyright"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://doi.org/10.25417/uic.31451479.v1","outbound_label":"DOI","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:creator","label":"Author","values":["Xiaoli Yan (420602)"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2025-12-01T00:00:00Z"]},{"key":"dc:relation","label":"Dc Relation","values":["https://figshare.com/articles/thesis/AI-Driven_Discovery_and_Multiscale_Computational_Frameworks_for_Carbon-Based_Materials/31451479"]},{"key":"dc:type","label":"Dc Type","values":["Text","Thesis"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Engineering","Materials Science"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:rights","label":"Dc Rights","values":["In Copyright"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["10.25417/uic.31451479.v1"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["This dissertation develops a unified, data-driven framework for carbon-based materials discovery that integrates generative artificial intelligence, multiscale simulations, and reactive dynamics to accelerate research central to carbon-to-X technologies. Chapter 2 establishes a one-directional workflow in which a generative AI model autonomously proposes novel metal–organic frameworks (MOFs) for carbon capture, followed by multiscale physics-based screening. 120,000 MOF AI-generated candidates are evaluated through an array of screening methods that validate structure and property of MOF, yielding 6 structures with CO2 uptakes exceeding 2 mol/kg at 0.1 bar and 300 K—demonstrating that generative design coupled with high-throughput validation can efficiently navigate the vast MOF design space. Chapter 3 extends this framework into a closed-loop, property-guided generative system. Here, an oracle-based scoring function provides feedback from machine learning predicted adsorption metric, dynamically steering the generative model toward desirable regions of the property space. This self-optimizing loop improves both discovery efficiency, physical fidelity, and design success rate. Chapter 4 transitions to nanocarbon systems, employing reactive molecular dynamics and machine-learning prediction to investigate the temperature- and pressure-dependent graphitization of nanodiamond surfaces, revealing atomistic pathways for sp3-to-sp2 transformation. Chapter 5 explores the co-pyrolysis of cellulose–graphene mixtures through reactive dynamics, coupled with non-equilibrium Green’s function (NEGF) formalism to quantify key molecular junctions for electron transport. Together, these studies unify porous, sp2, and hybrid carbon materials within a coherent methodological and conceptual framework. By combining forward generative design, feedback-driven optimization, and transport-level modeling, this work contributes a scalable paradigm for AI-accelerated carbon-materials innovation aligned with global carbon capture, utilization, and conversion goals."]},{"key":"dc:title","label":"Title","values":["AI-Driven Discovery and Multiscale Computational Frameworks for Carbon-Based Materials"]}]}],"canonical_facts":{"dc:creator":["Xiaoli Yan (420602)"],"dc:date":["2025-12-01T00:00:00Z"],"dc:description":["This dissertation develops a unified, data-driven framework for carbon-based materials discovery that integrates generative artificial intelligence, multiscale simulations, and reactive dynamics to accelerate research central to carbon-to-X technologies. Chapter 2 establishes a one-directional workflow in which a generative AI model autonomously proposes novel metal–organic frameworks (MOFs) for carbon capture, followed by multiscale physics-based screening. 120,000 MOF AI-generated candidates are evaluated through an array of screening methods that validate structure and property of MOF, yielding 6 structures with CO2 uptakes exceeding 2 mol/kg at 0.1 bar and 300 K—demonstrating that generative design coupled with high-throughput validation can efficiently navigate the vast MOF design space. Chapter 3 extends this framework into a closed-loop, property-guided generative system. Here, an oracle-based scoring function provides feedback from machine learning predicted adsorption metric, dynamically steering the generative model toward desirable regions of the property space. This self-optimizing loop improves both discovery efficiency, physical fidelity, and design success rate. Chapter 4 transitions to nanocarbon systems, employing reactive molecular dynamics and machine-learning prediction to investigate the temperature- and pressure-dependent graphitization of nanodiamond surfaces, revealing atomistic pathways for sp3-to-sp2 transformation. Chapter 5 explores the co-pyrolysis of cellulose–graphene mixtures through reactive dynamics, coupled with non-equilibrium Green’s function (NEGF) formalism to quantify key molecular junctions for electron transport. Together, these studies unify porous, sp2, and hybrid carbon materials within a coherent methodological and conceptual framework. By combining forward generative design, feedback-driven optimization, and transport-level modeling, this work contributes a scalable paradigm for AI-accelerated carbon-materials innovation aligned with global carbon capture, utilization, and conversion goals."],"dc:identifier":["10.25417/uic.31451479.v1"],"dc:relation":["https://figshare.com/articles/thesis/AI-Driven_Discovery_and_Multiscale_Computational_Frameworks_for_Carbon-Based_Materials/31451479"],"dc:rights":["In Copyright"],"dc:subject":["Engineering","Materials Science"],"dc:title":["AI-Driven Discovery and Multiscale Computational Frameworks for Carbon-Based Materials"],"dc:type":["Text","Thesis"]},"updated_at":"2026-07-27T21:34:28Z"}