Machine learning correlates the photo oxidation state and pyrolysis behavior of photo cured waste plastics, providing new ideas for aging characterization

Research Background and Methods: UV cured materials undergo continuous exposure to ultraviolet radiation, oxygen, and heat during outdoor service, gradually accumulating photo oxidation products on the surface and inside, forming an unquantifiable “pre aging” state. This state not only affects the secondary processing and recycling of materials, but also directly determines the distribution of their pyrolysis products. The study titled ‘Linking the pre exiting photooxidative state of photocured waste to pyrolysis via machine learning and time scale normalization’ was published in Volume 254 of Polymer Degradation and Stability (December 2026) Hongnan Zhang、Yadong Ge、Yunan Sun、Junyu Tao、Guandong Chen、Hesameddin Fatehi、Beibei Yan、Zhanjun Cheng、Guanyi Chen After completion, try to link this pre aging state with pyrolysis behavior in a data-driven manner. The key to the research lies in the combination of “time scale normalization” and machine learning: the photooxidation process spans different time scales, and direct comparison of data under different aging conditions is prone to distortion. Normalization processing allows experimental data from different sources to be modeled in a unified framework, thereby identifying the photooxidation features most relevant to pyrolysis behavior.

The significance of polymer aging and stabilization technology

From the perspective of polymer aging, the value of this study lies not in providing a specific formula, but in providing a method to transform the “aging history” into a moldable variable.

. Traditionally, evaluating weather stability performance relies on mechanical, color difference, and carbonyl index indicators after xenon lamp or UV accelerated aging, which lack a quantitative bridge between these indicators and subsequent pyrolysis, recovery, or reprocessing behaviors. This study suggests that the oxygen-containing functional groups and degree of chain breakage formed during photooxidation can serve as intermediate variables linking aging state and thermal behavior. For researchers engaged in the design of antioxidant and photostable systems, this approach has two implications. Firstly, the essence of the interaction between HALS and UVA is to inhibit or delay the generation of photo oxidation products. If a corresponding relationship can be established between the stabilizer formula and the photo oxidation characteristic spectrum, it is possible to judge the long-term weather resistance performance earlier in the formula screening stage. Secondly, the effectiveness of complex systems is often reflected in the synergistic inhibition of multiple aging pathways, which cannot be fully reflected by a single indicator. Multi feature modeling methods are more suitable for evaluating the comprehensive efficiency of complex systems.

Extended thinking on detection characterization and processing stability

At the level of detection characterization, this study emphasizes time scale normalization, which corresponds to the extrapolation problem between accelerated aging and natural exposure commonly found in aging research.

. For processing stability, photo cured waste already has a pre oxidized state before pyrolysis or reprocessing, and its thermal behavior differs from that of fresh materials. This means that the strategy of adding antioxidant systems in recycled material processing needs to refer to its aging history, rather than simply applying new material formulas. It should be pointed out that this study focuses on the relationship between the photo oxidation state and pyrolysis of photo cured waste, and does not directly involve the specific mechanism of action or formulation data of additives such as HALS and UVA. The significance of its methodology lies in providing a transferable modeling approach for polymer aging research. Whether it can be extended to the weathering stability evaluation of mainstream materials such as polyolefins in the future still needs to be verified.

Research Information

• Journal: Polymer Degradation and Stability, Volume 254
• Publication Date: December 2026
• Author: Hongnan Zhang, Yadong Ge, Yunan Sun, Junyu Tao, Guandong Chen, Hesameddin Fatehi, Beibei Yan, Zhanjun Cheng, Guanyi Chen
• Source link:

Information source:Linking the pre-existing photooxidative state of photocured waste to pyrolysis via machine learning and time-scale normalization， Organized by Herun Chemical.