Predictive Design of Sustainable Biobased Packaging via Machine Intelligence for Improved Postharvest Preservation
Date
2025-05-26
Authors
Chen, Po-Yen
Li, Yang
Chen, Tianle
Ma, Peihua
Chung, Tsai-Chun
Mukta, Shahnaz
Wu, Lianping
Little, Joshua
Whitley, Hayden
Zhou, Bin
Major Professor
Advisor
Committee Member
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The widespread use of petrochemical-based plastics in food packaging raises environmental concerns and lacks antimicrobial properties, limiting protection against microbial contamination. While biobased nanocomposites offer a sustainable alternative, optimizing their formulations remains challenging due to a vast materials library, inefficient trial-and-error experimentation, and complex multi-property requirements. Herein, a data-driven workflow integrating robotic automation, machine learning predictions, density functional theory (DFT) simulations, and life cycle assessment (LCA) is developed to accelerate the discovery of sustainable biobased packaging materials, enabling enhanced postharvest preservation with a reduced environmental footprint. An automated pipetting robot formulates 2,420 biobased nanocomposites, and their film quality data train an artificial neural network classifier, defining a design space. Within this space, 16 active learning loops iteratively fabricate and characterize 343 biobased nanocomposites, generating a high-quality experimental dataset. Leveraging this dataset and DFT simulations, a prediction model is constructed to explore ~1 billion formulations, identifying biobased nanocomposites with superior mechanical resilience and tunable transparency. Among them, a Cu2+-incorporated, chitosan-rich film further demonstrates moisture absorption, oxygen impermeability, and antimicrobial performance, outperforming conventional plastic wraps and extending the shelf life of postharvest produce. To further enhance sustainability, LCA-informed feedback is integrated into predictive modeling, refining nanocomposite formulations to minimize environmental impact. Additionally, a data-sharing platform is created, featuring forward prediction and inverse design capabilities to promote community adoption. This integrative approach significantly accelerates the development of high-performance biobased packaging and paves the way for a more sustainable and antimicrobial alternative to conventional plastics.
Series Number
Journal Issue
Is Version Of
Versions
Series
Academic or Administrative Unit
Type
preprint
Comments
This preprint is published as Po-Yen Chen, Yang Li, Tianle Chen et al. Predictive Design of Sustainable Biobased Packaging via Machine Intelligence for Improved Postharvest Preservation, 26 May 2025, PREPRINT (Version 1);https://doi.org/10.21203/rs.3.rs-6215151/v1.
Supplementary Files - Supplementary Information (chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/https://assets-eu.researchsquare.com/files/rs-6215151/v1/79e3c0438b940c1222092bdb.pdf); Supplementary Video 1- https://assets-eu.researchsquare.com/files/rs-6215151/v1/af6f8ef1afbcf5e94c819037.mp4.; Supplementary Video 2 - https://assets-eu.researchsquare.com/files/rs-6215151/v1/36d567b4b2c3427076874964.mp4
Supplementary Files - Supplementary Information (chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/https://assets-eu.researchsquare.com/files/rs-6215151/v1/79e3c0438b940c1222092bdb.pdf); Supplementary Video 1- https://assets-eu.researchsquare.com/files/rs-6215151/v1/af6f8ef1afbcf5e94c819037.mp4.; Supplementary Video 2 - https://assets-eu.researchsquare.com/files/rs-6215151/v1/36d567b4b2c3427076874964.mp4
Rights Statement
This work is licensed under a CC BY 4.0 License