有機化合物於微塑膠上的吸附：薈萃分析與建模

Abstract

微塑膠（MPs）可作為有機污染物的載體，透過吸收或吸附作用影響其在環境中的分佈、毒性與生物可利用性。儘管已有大量關於有機化合物在不同微塑膠上的吸附行為之研究與實驗，但由於微塑膠與有機化合物的多樣性，以及實驗條件的差異，至今仍難以形成一致而全面的理解。 本研究旨在透過分析與建模不同微塑膠與有機化合物的實驗吸附係數，建立對有機物—微塑膠吸附行為的整體認識。我們整理了一個包含13種微塑膠與31種有機化合物的有機碳歸一化吸附係數（KOC）資料庫，共收錄3654組數據，來自83篇實驗研究。時間尺度分析顯示，約49% 的吸附數據（主要是玻璃態微塑膠）可能未達到吸附平衡。有機化合物在玻璃態微塑膠中的擴散速率極慢，擴散係數約為10-14 cm2/s，遠低於橡膠態微塑膠中的值（約10-10 cm2/s），這意味著若要達到平衡吸附，玻璃態微塑膠可能需長達一年以上的吸附時間。與天然有機質（NOM）相比，96% 的微塑膠–化合物對的log KOC小於或等於（69%等於）NOM–化合物對的log KOC，顯示NOM的分配行為可作為微塑膠吸附的上限估值。這一趨勢在不同類型的微塑膠、化合物以及可電離化合物（如四環黴素與三氯生）中皆一致。 進一步分析顯示，微塑膠的log KOC主要受化合物特性所主導，而非微塑膠本身的性質所決定。化合物的疏水性（如log KOW、極性表面積）在89%的微塑膠中與log KOC顯著相關，反之微塑膠的極性、結晶度、密度及與水接觸角等性質僅具弱或可忽略的影響。線性溶劑化能關係（LSER）模型僅能針對PE與PCL建立（RMSE:0.06–0.36），其餘微塑膠因未達平衡或資料不足而無法建模。模型中McGowan體積係數（v=2.5–2.8）在不同模型間保持穩定，顯示吸附行為主要受空腔形成作用主導，此結果與NOM–MP比較及log KOC與化合物性質的相關性結果一致。 整體而言，本研究指出微塑膠的化學組成在有機化合物吸附中僅扮演次要角色，而NOM的分配行為可作為評估微塑膠污染水環境中有機污染物遷移性與可利用性之實用上限參考，無需區分微塑膠種類。未來仍需進一步探討化合物種態與電離行為對微塑膠log KOC的影響，以及玻璃態、擴散受限微塑膠中化合物的結合機制。

Microplastics (MPs) can serve as carriers of organic pollutants and influence their distribution, toxicity, and bioavailability in the environment through absorption or adsorption. Although the sorption of organic chemicals to different MPs has been extensively studied and experimented, the diverse nature of MPs and organic chemicals and varying experimental conditions have prevented a coherent understanding of the phenomenon from emerging. This study aims to develop a holistic understanding of organic-MP binding by analyzing and modeling experimental sorption coefficients of different MPs and organic sorbates. A sorption database of organic-carbon normalized sorption coefficient (KOC) (n=3654) encompassing 13 MPs and 31 organic chemicals is curated from 83 primary experimental studies. Timescale analysis reveals that 49% of the sorption data, mostly associated with glassy MPs, likely has not reached chemical equilibrium. Organic sorbates diffuse through glassy MPs very slowly with diffusivities in the order of 10-14 cm2/s, which are orders of magnitude lower than those in rubbery MPs (~10-10 cm2/s). This implies that a yearly incubation time is needed to achieve sorption equilibrium with glassy MPs. When compared against natural organic matter (NOM), 96% of MPs-sorbate pairs have log KOC less than or equal (69%) to those of NOM-sorbate pairs, suggesting NOM partitioning may serve as an upper estimate of log KOC in MPs. This is observed across MPs, chemicals, and ionizable compounds (tetracycline and triclosan). Analysis reveals that log KOC of MPs is dominated by chemical properties rather than MP properties. Chemical hydrophobicity (e.g., log KOW, polar surface area) significantly correlate with log KOC in 89% of MPs, whereas MP properties (polarity, crystallinity, density, and water contact angle) have weak or negligible influence. Linear solvation energy relation (LSER) models (RMSE: 0.06–0.36) are constructed only for PE and PCL; the remaining MPs cannot be modeled due to non-equilibrium or insufficient data. The McGowan volume coefficient remains conserved across models (v=2.5–2.8), suggesting sorbate-MP association is dominated by cavity formation, which is consistent results from NOM-MP comparison and log KOC and sorbate properties correlation. Overall, this study suggests that MP chemistry plays a secondary role in the sorption of organic chemicals. Furthermore, NOM partitioning may serve as a practical upper bound reference for assessing mobility and available of organic pollutants in MP-contaminated aquatic environments without the need to differentiate plastic type. Further investigations are needed to clarify the role of speciation/ionization on log KOC of MPs and the binding of chemicals with glassy diffusion-limited MPs.