Scholarlink

<p>Evaluating the environmental fate of poly(ε-caprolactone) (PCL), a marine-biodegradable plastic, requires understanding its degradation behavior under realistic marine conditions. Previous studies have separately examined the effects of crystallinity and environmental conditions on PCL degradation, but in actual ocean environments where conditions vary spatiotemporally, the relative contributions of these factors remain insufficiently quantified. In this study, long-term coastal field tests in Japan were combined with machine learning analysis to construct a regression framework relating PCL degradation rates to crystallinity and environmental variables. PCL sheets with controlled crystallinity were exposed for 6–15 months at six coastal locations, and degradation rates were derived from mass loss. Among the regression models tested, CatBoost showed the best performance (test <i>R</i>² = 0.60). Within the present dataset and coastal seafloor deployment conditions, the SHapley Additive exPlanations (SHAP) analysis ranked water temperature highest in terms of mean absolute contribution, followed by depth and total nitrogen (TN), whereas crystallinity contributed moderately. Depth and TN may partly capture location-level background differences rather than isolated variable effects. These results suggest that environmental factors contributed substantially to the variation in observed degradation rates. These findings offer a basis for field-relevant environmental compatibility assessment and for interpreting the role of crystallinity in material design.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Relative contributions of crystallinity and environmental factors to the degradation rate of poly(ε-caprolactone) (PCL) under coastal field conditions

  • 期刊论文

  • 发表日期：

    2026年2月19日

查看全文
  • Hironori Taguchi,
  • Takako Kikuchi,
  • Yoshifumi Amamoto,
  • Hiroshi Morita,
  • Keiji Tanaka

摘要

Evaluating the environmental fate of poly(ε-caprolactone) (PCL), a marine-biodegradable plastic, requires understanding its degradation behavior under realistic marine conditions. Previous studies have separately examined the effects of crystallinity and environmental conditions on PCL degradation, but in actual ocean environments where conditions vary spatiotemporally, the relative contributions of these factors remain insufficiently quantified. In this study, long-term coastal field tests in Japan were combined with machine learning analysis to construct a regression framework relating PCL degradation rates to crystallinity and environmental variables. PCL sheets with controlled crystallinity were exposed for 6–15 months at six coastal locations, and degradation rates were derived from mass loss. Among the regression models tested, CatBoost showed the best performance (test R² = 0.60). Within the present dataset and coastal seafloor deployment conditions, the SHapley Additive exPlanations (SHAP) analysis ranked water temperature highest in terms of mean absolute contribution, followed by depth and total nitrogen (TN), whereas crystallinity contributed moderately. Depth and TN may partly capture location-level background differences rather than isolated variable effects. These results suggest that environmental factors contributed substantially to the variation in observed degradation rates. These findings offer a basis for field-relevant environmental compatibility assessment and for interpreting the role of crystallinity in material design.