Advances in Computational Modeling and Machine Learning of Cellulosic Biopolymers: A Comprehensive Review
Biomimetics · 01.12.2025 · 10 (12) · 802
DOI: 10.3390/biomimetics10120802 ↗Kurz gefasst
Übersichtsarbeit in der Zeitschrift Biomimetics, online erschienen am 01.12.2025. Verfasst von Sharmi Mazumder und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Mechanical Engineering, Baylor University, Waco, TX 76706, USA. Die Arbeit wurde bisher 16-mal zitiert (Crossref) und nennt 173 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: U.S. National Science Foundation, National Science Foundation.
Abstract (Originalsprache)
The hierarchical structure and multifunctional properties of bio-based cellular materials, particularly cellulose, hemicellulose, and lignin, have attracted increasing attention and interest due to their sustainability and versatility. Recent advances in computational modeling and machine learning strategies have provided transformative insights into the molecular, mechanical, thermal, and electronic behaviors of these biopolymers. This review categorizes the conducted studies based on key material properties and discusses the computational methods utilized, including quantum mechanical approaches, atomistic and coarse-grained molecular dynamics, finite element modeling, and machine learning techniques. For each property, such as structural, mechanical, thermal, and electronic, we have analyzed the progress made in understanding inter- and intra-molecular interactions, deformation mechanisms, phase behavior, and functional performance. For instance, atomistic simulations have shown that cellulose nanocrystals exhibit a highly anisotropic elastic response, with axial elastic moduli ranging from approximately 100 to 200 GPa. Similarly, thermal transport studies have shown that the thermal conductivity along the chain axis (≈5.7 W m−1 K−1) is nearly an order of magnitude higher than that in the transverse direction (≈0.7 W m−1 K−1). In recent years, this research area has also experienced rapid advancement in data-driven methodologies, with the number of machine learning applications for biopolymer systems increasing more than fourfold over the past five years. By bridging multiscale modeling and data-driven approaches, this review aims to illustrate how these techniques can be integrated into a unified framework to accelerate the design and discovery of high-performance bioinspired materials. Eventually, we have discussed emerging opportunities in multiscale modeling and data-driven discovery to outline future directions for the design and application of high-performance bioinspired materials. This review aims to bridge the gap between molecular-level understanding and macroscopic functionality, thereby supporting the rational design of next-generation sustainable materials.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Mechanical Engineering, Baylor University, Waco, TX 76706, USA.
Förderung
- U.S. National Science Foundation
- National Science Foundation
Zitieren
APA
Mazumder, S., Golbabaei, M. H., & Zhang, N. (2025). Advances in Computational Modeling and Machine Learning of Cellulosic Biopolymers: A Comprehensive Review. Biomimetics, 10(12), 802. https://doi.org/10.3390/biomimetics10120802
Vancouver
Mazumder S, Golbabaei MH, Zhang N. Advances in Computational Modeling and Machine Learning of Cellulosic Biopolymers: A Comprehensive Review. Biomimetics. 2025;10(12):802. doi:10.3390/biomimetics10120802
BibTeX
@article{Mazumder2025,
author = {Mazumder, Sharmi and Golbabaei, Mohammad Hossein and Zhang, Ning},
title = {Advances in Computational Modeling and Machine Learning of Cellulosic Biopolymers: A Comprehensive Review},
journal = {Biomimetics},
year = {2025},
volume = {10},
number = {12},
pages = {802},
doi = {10.3390/biomimetics10120802}
}
EUVIMED