Machine learning-assisted first-principles study of structural, electronic, optical, thermal, and mechanical properties of novel s-triazine-based organic framework monolayers

Research output: Contribution to journalArticleResearchpeer review

Authors

External Research Organisations

  • Persian Gulf University
  • University of Ostrava
View graph of relations

Details

Original languageEnglish
Article number121092
JournalCARBON
Volume248
Early online date24 Nov 2025
Publication statusPublished - 5 Feb 2026

Abstract

Nanoporous carbon-nitride covalent organic frameworks (CN-COFs) represent an emerging class of nanomaterials with tunable architectures and scalable synthesis routes. Recent advances introduced four novel CN-COFs nanosheets, featuring hybrid benzene/s-triazine cores. Using these advances as a foundation, we designed four CN-COFs with identical structures but slightly different chemistry. We evaluated structural stability, electronic/optical properties, mechanical strength, and thermal transport using a hybrid approach combining density functional theory (DFT) and machine learning interatomic potentials (MLIPs). Complex stable atomic configurations, along with thermal and mechanical properties, were efficiently identified using MLIPs, while electronic and optical properties were accurately analyzed through single-step DFT calculations. Structurally, four frameworks exhibit Kagome lattice symmetry, hosting unique electronic features like flat and Dirac bands, while corrugated configurations induce modified band structures. These semiconductors display band gaps ranging from 2.73 to 3.72 eV, allowing strong photon absorption across the UV–visible spectrum and aligning well with water redox potentials, making them promising for optoelectronic and photocatalytic applications. Despite their highly porous nature, CN-COFs demonstrate impressive mechanical resilience, sustaining strain levels up to around 0.3 and tensile strengths exceeding 10 GPa, significantly surpassing conventional polymers. Crucially, we demonstrate that fine-tuning the chemistry of the linkages allows for the occurrence of significan out-of-plane corrugations, resulting in ultralow lattice thermal conductivity, which is particularly attractive for thermoelectric and thermal insulating applications. Our comprehensive findings confirm the stability, mechanical robustness, ultralow thermal conductivity and appealing semiconducting nature of CN-COFs, highlighting their application prospects in flexible, high-performance optoelectronics and energy storage and conversion systems.

Keywords

    Carbon nitrides, Covalent organic frameworks, Density functional theory, Machine learning, Semiconductors

ASJC Scopus subject areas

Cite this

Machine learning-assisted first-principles study of structural, electronic, optical, thermal, and mechanical properties of novel s-triazine-based organic framework monolayers. / Mortazavi, Bohayra; Shojaei, Fazel; Shahrokhi, Masoud et al.
In: CARBON, Vol. 248, 121092, 05.02.2026.

Research output: Contribution to journalArticleResearchpeer review

Download
@article{8c52c48fceb04e2c9dde4ddcf9ecc199,
title = "Machine learning-assisted first-principles study of structural, electronic, optical, thermal, and mechanical properties of novel s-triazine-based organic framework monolayers",
abstract = "Nanoporous carbon-nitride covalent organic frameworks (CN-COFs) represent an emerging class of nanomaterials with tunable architectures and scalable synthesis routes. Recent advances introduced four novel CN-COFs nanosheets, featuring hybrid benzene/s-triazine cores. Using these advances as a foundation, we designed four CN-COFs with identical structures but slightly different chemistry. We evaluated structural stability, electronic/optical properties, mechanical strength, and thermal transport using a hybrid approach combining density functional theory (DFT) and machine learning interatomic potentials (MLIPs). Complex stable atomic configurations, along with thermal and mechanical properties, were efficiently identified using MLIPs, while electronic and optical properties were accurately analyzed through single-step DFT calculations. Structurally, four frameworks exhibit Kagome lattice symmetry, hosting unique electronic features like flat and Dirac bands, while corrugated configurations induce modified band structures. These semiconductors display band gaps ranging from 2.73 to 3.72 eV, allowing strong photon absorption across the UV–visible spectrum and aligning well with water redox potentials, making them promising for optoelectronic and photocatalytic applications. Despite their highly porous nature, CN-COFs demonstrate impressive mechanical resilience, sustaining strain levels up to around 0.3 and tensile strengths exceeding 10 GPa, significantly surpassing conventional polymers. Crucially, we demonstrate that fine-tuning the chemistry of the linkages allows for the occurrence of significan out-of-plane corrugations, resulting in ultralow lattice thermal conductivity, which is particularly attractive for thermoelectric and thermal insulating applications. Our comprehensive findings confirm the stability, mechanical robustness, ultralow thermal conductivity and appealing semiconducting nature of CN-COFs, highlighting their application prospects in flexible, high-performance optoelectronics and energy storage and conversion systems.",
keywords = "Carbon nitrides, Covalent organic frameworks, Density functional theory, Machine learning, Semiconductors",
author = "Bohayra Mortazavi and Fazel Shojaei and Masoud Shahrokhi and Xiaoying Zhuang",
note = "Publisher Copyright: {\textcopyright} 2025 The Authors.",
year = "2026",
month = feb,
day = "5",
doi = "10.1016/j.carbon.2025.121092",
language = "English",
volume = "248",
journal = "CARBON",
issn = "0008-6223",
publisher = "Elsevier Ltd.",

}

Download

TY - JOUR

T1 - Machine learning-assisted first-principles study of structural, electronic, optical, thermal, and mechanical properties of novel s-triazine-based organic framework monolayers

AU - Mortazavi, Bohayra

AU - Shojaei, Fazel

AU - Shahrokhi, Masoud

AU - Zhuang, Xiaoying

N1 - Publisher Copyright: © 2025 The Authors.

PY - 2026/2/5

Y1 - 2026/2/5

N2 - Nanoporous carbon-nitride covalent organic frameworks (CN-COFs) represent an emerging class of nanomaterials with tunable architectures and scalable synthesis routes. Recent advances introduced four novel CN-COFs nanosheets, featuring hybrid benzene/s-triazine cores. Using these advances as a foundation, we designed four CN-COFs with identical structures but slightly different chemistry. We evaluated structural stability, electronic/optical properties, mechanical strength, and thermal transport using a hybrid approach combining density functional theory (DFT) and machine learning interatomic potentials (MLIPs). Complex stable atomic configurations, along with thermal and mechanical properties, were efficiently identified using MLIPs, while electronic and optical properties were accurately analyzed through single-step DFT calculations. Structurally, four frameworks exhibit Kagome lattice symmetry, hosting unique electronic features like flat and Dirac bands, while corrugated configurations induce modified band structures. These semiconductors display band gaps ranging from 2.73 to 3.72 eV, allowing strong photon absorption across the UV–visible spectrum and aligning well with water redox potentials, making them promising for optoelectronic and photocatalytic applications. Despite their highly porous nature, CN-COFs demonstrate impressive mechanical resilience, sustaining strain levels up to around 0.3 and tensile strengths exceeding 10 GPa, significantly surpassing conventional polymers. Crucially, we demonstrate that fine-tuning the chemistry of the linkages allows for the occurrence of significan out-of-plane corrugations, resulting in ultralow lattice thermal conductivity, which is particularly attractive for thermoelectric and thermal insulating applications. Our comprehensive findings confirm the stability, mechanical robustness, ultralow thermal conductivity and appealing semiconducting nature of CN-COFs, highlighting their application prospects in flexible, high-performance optoelectronics and energy storage and conversion systems.

AB - Nanoporous carbon-nitride covalent organic frameworks (CN-COFs) represent an emerging class of nanomaterials with tunable architectures and scalable synthesis routes. Recent advances introduced four novel CN-COFs nanosheets, featuring hybrid benzene/s-triazine cores. Using these advances as a foundation, we designed four CN-COFs with identical structures but slightly different chemistry. We evaluated structural stability, electronic/optical properties, mechanical strength, and thermal transport using a hybrid approach combining density functional theory (DFT) and machine learning interatomic potentials (MLIPs). Complex stable atomic configurations, along with thermal and mechanical properties, were efficiently identified using MLIPs, while electronic and optical properties were accurately analyzed through single-step DFT calculations. Structurally, four frameworks exhibit Kagome lattice symmetry, hosting unique electronic features like flat and Dirac bands, while corrugated configurations induce modified band structures. These semiconductors display band gaps ranging from 2.73 to 3.72 eV, allowing strong photon absorption across the UV–visible spectrum and aligning well with water redox potentials, making them promising for optoelectronic and photocatalytic applications. Despite their highly porous nature, CN-COFs demonstrate impressive mechanical resilience, sustaining strain levels up to around 0.3 and tensile strengths exceeding 10 GPa, significantly surpassing conventional polymers. Crucially, we demonstrate that fine-tuning the chemistry of the linkages allows for the occurrence of significan out-of-plane corrugations, resulting in ultralow lattice thermal conductivity, which is particularly attractive for thermoelectric and thermal insulating applications. Our comprehensive findings confirm the stability, mechanical robustness, ultralow thermal conductivity and appealing semiconducting nature of CN-COFs, highlighting their application prospects in flexible, high-performance optoelectronics and energy storage and conversion systems.

KW - Carbon nitrides

KW - Covalent organic frameworks

KW - Density functional theory

KW - Machine learning

KW - Semiconductors

UR - http://www.scopus.com/inward/record.url?scp=105023501232&partnerID=8YFLogxK

U2 - 10.1016/j.carbon.2025.121092

DO - 10.1016/j.carbon.2025.121092

M3 - Article

AN - SCOPUS:105023501232

VL - 248

JO - CARBON

JF - CARBON

SN - 0008-6223

M1 - 121092

ER -

By the same author(s)