Thermodynamics, Adsorption Kinetics, and Interfacial Partitioning of Tetracycline Antibiotics on Biodegradable Polyester versus Polyethylene Microplastics

Abstract

Biodegradable mulch films, primarily poly­(lactic acid)/poly­(butylene adipate-co-terephthalate) (PLA/PBAT) blends, are increasingly applied as eco-friendly alternatives to conventional polyethylene (PE) films. However, the interfacial interactions and vector potential of commercial biodegradable mulch-derived microplastics toward typical agricultural antibiotics remain insufficiently understood. In this study, the adsorption thermodynamics, kinetics, and interfacial binding mechanisms of tetracycline (TC) and chlortetracycline (CTC) onto microplastics derived from two commercial biodegradable films (WHMPs and RAMPs) and a conventional PE film (PEMPs) were systematically investigated under controlled laboratory conditions. The results demonstrated that biodegradable microplastics exhibited markedly higher saturated adsorption capacities for both TC (5.20 mg/g for WHMPs and 6.18 mg/g for RAMPs) and CTC (11.46 mg/g for WHMPs and 15.01 mg/g for RAMPs) than those of PEMPs (1.22 mg/g for TC and 5.43 mg/g for CTC). Adsorption isotherms adhered closely to the Langmuir model, indicating a monolayer interfacial coverage. Thermodynamic analyses revealed that the adsorption was spontaneous and endothermic, driven predominantly by favorable entropy gain. The significantly higher adsorption affinity for CTC compared to TC was rationalized by its higher octanol–water partition coefficient and the electron-withdrawing induction of the C7–Cl substituent, which enhances localized dipole–dipole attractions with polymer ester carbonyls (CO). High-resolution X-ray photoelectron spectroscopy and FT-IR analyses revealed that noncovalent physical partitioning, hydrophobic interactions, and dipole–dipole attractions dominated the interfacial binding without irreversible covalent transitions. Multicondition desorption and dynamic leaching assays further revealed that biodegradable microplastics exhibit moderate desorption reversibility and sustained release potential. These findings provide fundamental thermodynamic and kinetic benchmarks for understanding antibiotic retention on commercial mulch-derived microplastics, highlighting the need for future field-scale investigations to evaluate their environmental behaviors under complex soil matrixes and chronic weathering conditions.

1. Introduction

Agricultural film is one of the most vital inputs in modern agricultural production, playing a crucial role in guaranteeing grain yields, stabilizing food supply, and mitigating famine. Traditional agricultural mulching film is a plastic film primarily composed of polyethylene (PE). By covering the soil surface, it serves multiple functions, including thermal insulation, moisture retention, weed suppression, pest and disease mitigation, and light regulation. However, the application of these films is typically unidirectional and single use. In agricultural environments, the films gradually age and fragment under the influence of external physical (e.g., friction and tearing), chemical (e.g., ultraviolet oxidation), and biological (e.g., microbial activity) forces, rendering the postuse films exceedingly difficult to recover and recycle. Consequently, unrecovered residual films continuously accumulate in farmlands, resulting in “white pollution” that disrupts the transport of soil moisture and nutrients, as well as the growth of microorganisms. Furthermore, these residual films undergo continuous fragmentation into microplastics (recognized as one of the four major emerging pollutants), posing potential ecological risks to agricultural environmental quality, food safety, and human health.

Against this backdrop, biodegradable agricultural films have emerged. They are primarily composed of biodegradable materials blended with additives such as plasticizers, antioxidants, and colorants. Under the action of microorganisms, biodegradable films can be completely decomposed to water and carbon dioxide. This not only fulfills the indispensable role of mulching films in agricultural production but also addresses the direct and indirect environmental problems caused by traditional agricultural film residues. With the advancement of science and technology, key properties of biodegradable films, such as tensile strength, durability, and light transmittance, have approached those of traditional plastic films, while their costs have continuously decreased, achieving a balance between environmental protection and practical utility. However, the degradation process of biodegradable films imposes stringent requirements on environmental conditions, including the temperature, humidity, and oxygen content. Under typical circumstances, farmland environments often fail to meet these specific conditions, leading to inadequate degradation. In such scenarios, biodegradable films will similarly generate substantial numbers of microplastic particles. It has been reported that, given the same mass, biodegradable films may generate relatively more microplastics than traditional PE films. , Consequently, microplastic pollution caused by biodegradable agricultural films in farmland environments has equally attracted widespread attention.

Notably, the environment harbors a multitude of pollutants, such as heavy metals, antibiotics, hormones, and pathogens. Due to their inherently small particle size, microplastics exhibit a unique small-size effect. Simultaneously, the surfaces of microplastics are endowed with abundant functional groups and negative charges. These characteristics collectively exert a profound influence on the migration and transformation of coexisting pollutants in the environment. For instance, Xu et al. discovered that microplastic exposure enhances the antibiotic and chlorine resistance of pathogens, thereby promoting their growth in drinking water supply systems. Guo et al. reported that microplastics enhanced adsorption and microbial transfer via biofilms, thereby inhibiting the degradation of oxytetracycline (with the degradation rate decreasing by 11.1%–20.2%). Li et al. revealed that Zn, acting as a UV protectant in microplastics, had minimal impact on key microbial diversity in maize-cultivated soils, and could even serve as a slow-release zinc fertilizer, with its accumulation order in maize being roots > stems ≈ leaves. Therefore, elucidating the interactive mechanisms between microplastics and specific pollutants is important for understanding and mitigating microplastic pollution.

Tetracyclines, particularly tetracycline (TC) and chlortetracycline (CTC), are among the most extensively administered veterinary antibiotics in livestock farming and animal husbandry worldwide. Due to their incomplete metabolic degradation in animal bodies, substantial fractions of administered TCs are excreted via feces and urine, frequently entering agroecosystems through organic manure fertilization and agricultural wastewater irrigation. Consequently, residual concentrations of TCs are widespread in agricultural soils, drainage ditches, and surface runoff, where they pose potential ecological concerns, including the propagation of antibiotic resistance genes and soil microbiota perturbation. Structurally, TC and CTC possess multiple ionizable functional groups (phenolic hydroxyl, dimethylamino, and enone systems), while CTC possesses an additional chlorine atom that alters its steric hindrance and electron-withdrawing nature. , These unique structural configurations render their interfacial adsorption behavior highly sensitive to the surface properties and functional motifs of the ambient particulate matrices.

With the extensive promotion of biodegradable agricultural mulch films (such as poly­(lactic acid)/poly­(butylene adipate-co-terephthalate) blends) as eco-friendly substitutes for conventional polyethylene (PE) films, the inevitable fragmentation and accumulation of biodegradable mulch-derived microplastics in agricultural soils have raised growing concerns. Recent studies have demonstrated that conventional PE microplastics typically exhibit limited affinity toward polar or ionizable hydrophilic contaminants due to their nonpolar aliphatic backbone and smooth hydrophobic surface. , However, the vast majority of existing literature on microplastic adsorption has predominantly relied on laboratory-grade virgin resin pellets or pristine analytical powders (e.g., pure synthesized PLA, PBAT, and PE particles). Such idealized models present critical limitations and fail to represent real-world agricultural scenarios because commercially manufactured agricultural mulch films undergo specialized industrial compounding, thermal blow-molding extrusion, and the incorporation of functional additives (e.g., plasticizers, stabilizers, and slip agents). These industrial manufacturing processes inevitably induce multiphase separation, localized microdomain heterogeneity, and structural defects. Consequently, upon fragmentation in farmlands, commercial film-derived microplastics develop distinct surface morphologies (crater-like microporous textures) and exposed interfacial chemistries (abundant CO/C–O ester linkages and terminal carboxyl groups) that differ fundamentally from smooth, dense, additive-free virgin resin pellets.

Despite these practical differences, a systematic and comparative assessment of how actual commercial biodegradable mulch microplastics differ from conventional PE film microplastics in governing the interfacial uptake, thermodynamic driving forces, and kinetic mass-transfer pathways of structurally diverse antibiotics remains largely unexplored. Crucial scientific questions, specifically, how commercial polyester-blend interfaces collaboratively regulate adsorption thermodynamics with substituent-induced electronic configurations (e.g., the strongly electronegative chlorine substituent in CTC versus TC) and whether physisorption or chemisorption dominates at the atomic scale, remain inadequately resolved. Furthermore, while ambient concentrations of antibiotics in bulk natural soils and open waters typically range from ng/L to μg/L, transiently elevated levels can occur in agricultural pollution hotspots, such as animal manure storage lagoons, fertilizer application zones, and agricultural runoff ditches. In this work, batch adsorption experiments were systematically designed across a broader concentration spectrum. This setting was critically selected to comprehensively map the entire adsorption isotherm up to theoretical plateau saturation, enabling the accurate determination of maximum monolayer adsorption capacity and thermodynamic parameters while providing sufficient interfacial mass loading for rigorous solid-phase characterization via high-resolution X-ray photoelectron spectroscopy (XPS), EDS mapping, and FT-IR.

Therefore, to bridge these critical research gaps, this study systematically investigates and compares the interfacial adsorption thermodynamics, multistage kinetics, and binding mechanisms of TC and CTC onto microplastics derived directly from two representative commercial biodegradable mulch films (WHMPs and RAMPs) and conventional PE films (PEMPs). By coupling high-resolution spectroscopic characterization (FT-IR and XPS) with multicondition desorption and dynamic leaching assessments, this work aims to elucidate the role of ester-rich interfacial motifs and structural defects in antibiotic vector behaviors, thereby providing a realistic theoretical foundation and valuable empirical support for the environmental safety and ecological risk assessment of biodegradable agricultural films in agroecosystems.

2. Materials and Methods

2.1. Materials and Pretreatment

Two types of commercially available biodegradable agricultural mulch films were procured from Hangzhou Wanhua Chemical Products Co., Ltd. (Hangzhou, Zhejiang, China) and Ruian Jiarun New Material Technology Co., Ltd. (Ruian, Zhejiang, China), respectively. A conventional low-density polyethylene (PE) agricultural film was obtained from Cangzhou New Material Co., Ltd. (Hebei, China). To represent practical agricultural plastic debris, the purchased pristine mulch films were manually shredded, mechanically pulverized, and sequentially sieved through standard stainless steel meshes (4, 5, 8, 10, and 20 mesh) to obtain film-derived microplastics with classified particle sizes of 4.0–5.0 mm, 2.5–4.0 mm, 2.0–2.5 mm, 0.9–2.0 mm, and less than 0.9 mm. The resulting film-derived microplastics were designated as WHMPs (Hangzhou Wanhua), RAMPs (Ruian Jiarun), and PEMPs (conventional PE), respectively. The chemical identities, surface functional groups, and structural compositions of WHMPs and RAMPs were systematically identified via FT-IR and XPS spectroscopy as aliphatic-aromatic copolyester blends (predominantly poly­(butylene adipate-co-terephthalate) (PBAT) and poly­(lactic acid) (PLA) formulations containing inorganic carbonate fillers). Tetracycline (TC) and chlortetracycline (CTC) were purchased from Sigma-Aldrich. Analytical-grade sodium hydroxide (NaOH) and nitric acid (HNO₃), utilized for adjusting the solution pH, were supplied by Sailanbo Technology Co., Ltd.

2.2. Adsorption of Antibiotics by Agricultural Film-Derived Microplastics

A certain mass of agricultural film-derived microplastics was weighed into 20 mL amber glass tubes (equipped with polytetrafluoroethylene liners), followed by the addition of 15 mL of TC or CTC solutions. The mixtures were incubated with constant shaking at a controlled speed of 250 rpm in the dark (ZD-85, Zhengzhou Yinggong Photoelectric Technology Co., Ltd., China). After the reaction, the suspensions were filtered through 0.45 μm filter membranes, and the concentrations of TC and CTC in the filtrates were determined using a UV–vis spectrophotometer (UV-2150, Unico, China) at wavelengths of 357 and 375 nm, respectively. All batch adsorption experiments were performed independently in triplicate. The residual aqueous concentrations of TC and CTC were quantified, and corresponding error bars representing the standard deviations of the triplicate measurements have been incorporated into all parametric influence plots. Specifically, the experimental parameters were controlled as follows: initial solution concentrations of 5, 10, 15, 20, 30, and 50 mg/L; reaction times of 1, 2, 4, 8, 10, 12, 24, 36, and 48 h; environmental temperatures of 30, 40, and 50 °C; solution pH values of 3, 5, 7, 9, and 11; microplastic dosages of 0.5, 1, 2, 3, and 4 g/L; microplastic particle size ranges of 0–0.9, 0.9–2.0, 2.0–2.5, 2.5–4.0, and 4.0–5.0 mm; and ionic strengths (NaCl) of 0, 0.05, 0.1, 0.2, and 0.3 mol/L. To eliminate potential background interference from water-soluble additives, oligomers, or leached plasticizers during incubation, comprehensive blank control experiments (containing only microplastics in background electrolyte solutions without antibiotics) and antibiotic matrix controls (containing antibiotic solutions without microplastics to assess abiotic photolysis or wall sorption) were performed in parallel under identical conditions. The microplastics were thoroughly prerinsed with ultrapure water prior to batch tests. Supernatants from the microplastic blank controls were scanned over 200–500 nm, which confirmed that microplastic leachate exhibited negligible absorbance (Abs < 0.005) at the target quantification wavelengths of TC (357 nm) and CTC (375 nm), as common polyester fragments and aliphatic/ester additives absorb predominantly in the deep UV region (<280 nm). The net equilibrium antibiotic concentrations were determined after subtracting the corresponding microplastic blank backgrounds.

2.3. Characterization of Physicochemical Properties of Agricultural Film-Derived Microplastics

The surface functional groups of the agricultural film-derived microplastics before and after antibiotic adsorption were detected using a Fourier transform infrared spectrometer (FT-IR, TENSOR 27, Bruker Corporation, USA). The microscopic morphology was observed utilizing a scanning electron microscope (SEM, Sigma 300, Carl Zeiss AG, Germany). Furthermore, the surface elemental composition and core-level chemical bonding states of the microplastics before and after antibiotic exposure were investigated by XPS (K-Alpha, Thermo Fisher Scientific, USA) using monochromatic Al Kα radiation (hν = 1486.6 eV). All core-level binding energies were charge-corrected using the adventitious carbon (C 1s) reference at 284.80 eV. The spectral data processing and peak deconvolution were executed using Avantage software, where Shirley background subtraction was applied, and component peaks were fitted using Gaussian–Lorentzian (GL(30)) line shapes with constrained full width at half-maximum (fwhm) values to guarantee reproducible peak area deconvolution. The point of zero charge (pH_pzc) of the microplastics was determined using the pH drift method (solid addition method). Briefly, 50 mL of 0.01 M NaCl solution was introduced into a series of 100 mL conical flasks. The initial pH values (pH_initial) of the solutions were adjusted across the range of 2.0 to 11.0 using 0.1 M HCl or 0.1 M NaOH solutions. Subsequently, 0.05 g of microplastics (PEMPs, WHMPs, or RAMPs) was added to each flask. The suspensions were sealed and agitated on a thermostatic shaker at 150 rpm and 25 °C for 48 h to ensure chemical equilibrium. The final pH (pH_final) of each supernatant was measured using a precision pH meter. The difference between final and initial pH (ΔpH = pH_final – pH_initial) was plotted against pHinitial, and the point where ΔpH = 0 was identified as the pH_pzc of the respective microplastic. The specific surface areas and pore structure characteristics of PEMPs, RAMPs, and WHMPs were determined by N₂ adsorption/desorption isotherms measured at 77 K using a volumetric adsorption analyzer (Micromeritics ASAP 2460). The Brunauer–Emmett–Teller (BET) method and Barrett–Joyner–Halenda (BJH) model were employed to calculate the specific surface areas and pore size distributions, respectively.

2.4. Data and Modeling Analysis

2.4.1. Calculation of Adsorption Capacity

The adsorption capacities of the agricultural film-derived microplastics for TC and CTC were calculated based on the initial concentration and the concentration at adsorption equilibrium in the solution (eq ):

$$Q_{\mathrm{e}}=\frac{(C_0-C_{\mathrm{e}})V}{m}$$ 1

where Q _e represents the equilibrium adsorption capacity of the antibiotics (mg/g); C ₀ represents the initial concentration of the antibiotics (mg/L); C _e represents the adsorption equilibrium concentration (mg/L); V represents the volume of the solution (L); and m represents the mass of the agricultural film-derived microplastics (g).

2.4.2. Adsorption Isotherm Models

The classical Langmuir, Freundlich, and Dubinin–Radushkevich (D–R) models were employed to fit the adsorption process of antibiotics by agricultural film-derived microplastics.

The Langmuir model is expressed as (eq ):

$$Q_{\mathrm{e}}=\frac{Q_{\mathrm{m}}{K}_{\mathrm{l}}{C}_{\mathrm{e}}}{1+K_{\mathrm{L}}{C}_{\mathrm{e}}}$$ 2

where Q _m is the theoretical maximum adsorption capacity of the antibiotics (mg/g) and K _L is the Langmuir adsorption equilibrium constant (L/mg).

The Freundlich model is expressed as (eq ):

$$Q_{\mathrm{e}}=K_{\mathrm{F}}{C}_{\mathrm{e}}^{1/n}$$ 3

where K _F is the Freundlich adsorption constant, reflecting the strength of the adsorption capacity ((mg/g)­(L/mg)­1/n); and 1/n is a dimensionless constant related to the adsorption intensity.

The Dubinin–Radushkevich model is expressed as (eq ):

$$\ln Q_{\mathrm{e}}=\ln Q_{\mathrm{m}}-K_{\mathrm{D}\mathrm{R}}\times {\epsilon }^2$$ 4

where K _DR is the Dubinin–Radushkevich activity coefficient (mol²/kJ²) and ε is the Polanyi potential (kJ/mol), calculated via eq .

$$\epsilon =RT\ln (1+\frac{1}{C_{\mathrm{e}}})$$ 5

The mean free adsorption energy E (kJ/mol) per molecule is calculated using eq .

$$E=\frac{1}{{({2K}_{\mathrm{D}\mathrm{R}})}^{0.5}}$$ 6

In this study, the D–R model is evaluated strictly as an empirical mathematical expression to fit the adsorption curves.

2.4.3. Adsorption Kinetic Models

Pseudo-first-order (PFO), pseudo-second-order (PSO), and Elovich models were utilized to fit the adsorption process of antibiotics by agricultural film-derived microplastics.

The PFO kinetic model is expressed as (eq ):

$$Q_{\mathrm{t}}=Q_{\mathrm{e}}\times (1-{\mathrm{e}}^{-K_1\times \mathrm{t}})$$ 7

where Q _t is the adsorption capacity of the agricultural film-derived microplastics for antibiotics at time t (mg/g); K ₁ is the adsorption rate constant of the pseudo-first-order kinetics (h^–1); and t is the adsorption time (h).

The PSO kinetic model is expressed as (eq ):

$$Q_{\mathrm{t}}=\frac{Q_{\mathrm{e}}^2\times k_2\times t}{1+K_2\times Q_{\mathrm{e}}\times t}$$ 8

where k ₂ represents the adsorption rate constant of the pseudo-second-order kinetics (g/(mg·h)).

The Elovich model is expressed as (eq ):

$$Q_{\mathrm{t}}=\frac{1}{b}\ln (a^{*}b)+\frac{1}{b}\ln t$$ 9

where a represents the initial adsorption rate constant (mg/(g·h)) and b represents the constant related to the adsorption energy (mg/(g·h)).

2.4.4. Adsorption Thermodynamic Models

The thermodynamic parameters, including standard Gibbs free energy change (ΔG°), standard enthalpy change (ΔH°), and standard entropy change (ΔS°), were calculated according to the Van ’t Hoff eqs –:

$$K^{°}=\frac{Q_{\mathrm{e}}}{C_{\mathrm{e}}}\times \frac{{\rho }_{\mathrm{w}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{r}}}{1000}$$ 10

$${\mathrm{\Delta }G}^{°}=-RT\ln K_0$$ 11

$$\ln K_0=\frac{{\mathrm{\Delta }S}^{°}}{R}-\frac{{\mathrm{\Delta }H}^{°}}{RT}$$ 12

where K° is the dimensionless standard thermodynamic equilibrium constant obtained by normalizing the distribution coefficient (L/g) with water density (ρ_water ≈ 1000 g/L), R is the universal gas constant (8.314 J/(mol·K)), and T is the absolute temperature in kelvin (K). The values of ΔH° and ΔS° were determined from the slope and intercept of the linear regression plot of ln K ₀ versus 1/T.

2.5. Desorption and Leaching Experiments

To evaluate the mobility, reversibility, and secondary release risks of adsorbed antibiotics from microplastic vectors, desorption experiments were conducted following an equilibrium adsorption stage. The microplastics loaded with TC and CTC (pre-equilibrated at an initial concentration of 20 mg/L) were separated by centrifugation at 8000 rpm for 10 min, gently rinsed with deionized water to eliminate residual unadsorbed solution, and immediately resuspended in 25 mL of background electrolyte solutions with varying environmental chemistry conditions: (1) deionized water to determine the baseline desorption ratio; (2) 0.01 mol/L NaCl solution to evaluate the influence of background ionic strength; and (3) background electrolyte solutions (0.01 mol/L NaCl) adjusted to pH 5.0, 7.0, and 9.0 using 0.01 mol/L HCl or NaOH to evaluate pH-dependent release. The suspensions were agitated on a thermostatic shaker at 180 rpm and 298 K for 24 h to reach the desorption equilibrium. The desorbed TC and CTC concentrations in the supernatants were quantified using UV–vis spectrophotometry at 360 and 375 nm, respectively. The desorption efficiency (D, %) was calculated according to eq :

$$D(\%)=\frac{C_{\mathrm{d}\mathrm{e}\mathrm{s}}{V}_{\mathrm{d}\mathrm{e}\mathrm{s}}}{q_{\mathrm{e}}m}$$ 13

where C _des is the antibiotic concentration in the desorption medium (mg/L) and V _des is the volume of desorption solution (L).

To further simulate the continuous wash-off and sustained release of antibiotics under successive rainfall or irrigation events, multicycle leaching tests were performed. Microplastics preloaded with antibiotics were subjected to 4 consecutive leaching cycles using fresh background solution (0.01 mol/L NaCl, 25 mL per cycle) at 298 K with an agitation interval of 12 h per cycle. After each cycle, the mixture was centrifuged at 8000 rpm for 10 min, the supernatant was completely collected for concentration quantification, and an equal volume of a fresh leaching medium was immediately replenished for the subsequent cycle. The cumulative leaching efficiency (L _cum, %) across n consecutive cycles was calculated using eq :

$$L_{\mathrm{c}\mathrm{u}\mathrm{m}}(\%)=\frac{\mathrm{S}\mathrm{u}\mathrm{m}{C}_{\mathrm{l}\mathrm{e}\mathrm{a}\mathrm{c}\mathrm{h},i}V}{q_{\mathrm{e}}m}$$ 14

where C _leach,i is the antibiotic concentration measured in the supernatant at the ith leaching cycle (mg/L).

3. Results and Discussion

3.1. Physicochemical Properties and Microstructural Characterization of Agricultural Film-Derived Microplastics

3.1.1. Surface Morphology and Microstructure

The surface morphologies and microstructures of the three agricultural film-derived microplastics were characterized by SEM at different magnifications (1k× and 5k×), as shown in Figure . As illustrated in Figure a,b, PEMPs exhibited a relatively smooth, compact, and nonporous surface structure with distinct parallel microridges and shear bands, resulting from mechanical processing and shredding. High-magnification observation (Figure b) further confirmed the dense polymer matrix with minimal surface defects, suggesting a limited accessible contact area for interfacial interactions during solid–liquid contact. In contrast, the two biodegradable microplastics (WHMPs and RAMPs) displayed distinctly different surface characteristics. As shown in Figure c,d, WHMPs exhibited a rough, heterogeneous, and irregular surface texture with visible microcracks and crater-like depressions. Similarly, RAMPs (Figure e,f) presented a comparable eroded morphology, characterized by abundant globular protuberances and structural defects. These microscopic features are likely attributed to the phase separation and lower mechanical cohesion of the aliphatic-aromatic polyester blends. This distinct morphological difference indicates that the rough and defect-rich surfaces of WHMPs and RAMPs can provide more exposed contact regions and accessible active sites for solid–liquid interfacial contact compared to the smooth, dense PEMPs.

1.

Physicochemical and surface structural characterization of conventional (PEMPs) and biodegradable (WHMPs and RAMPs) agricultural film-derived microplastics: (a,b) SEM images of PEMPs at 1k× and 5k× magnifications; (c,d) SEM images of WHMPs at 1k× and 5k× magnifications; (e,f) SEM images of RAMPs at 1k× and 5k× magnifications; (g) FT-IR spectra; (h) XPS survey spectra; and high-resolution XPS spectra of (i) C 1s, (j) N 1s, and (k) O 1s.

3.1.2. Surface Functional Groups

The surface functional groups of PEMPs, WHMPs, and RAMPs were characterized by FT-IR spectroscopy, as shown in Figure g. For PEMPs, the spectrum exhibited dominant aliphatic hydrocarbon absorption bands, where the prominent peaks at 2920 and 2840 cm^–1 corresponded to the asymmetric and symmetric stretching vibrations of –CH₂ groups, and the bands at 1460, 1420, and 730/640 cm^–1 were assigned to –CH₂ bending and rocking vibrations. In addition, detectable absorption peaks associated with O-containing groups were observed at 1730 cm^–1 (CO stretching), 1260 cm^–1 (C–O–C stretching), and 1180/1110 cm^–1 (C–O stretching), along with a weak broad band in the range of 3500–3200 cm^–1 (−OH/–NH), indicating the presence of oxygenated components likely originating from manufacturing additives or surface aging. , For the two biodegradable microplastics (WHMPs and RAMPs), their spectra displayed highly similar profiles with significantly stronger intensities for oxygen-containing functional groups. The sharp and intense band at 1730 cm^–1 was assigned to the ester carbonyl (CO) stretching vibration, which, together with the pronounced peaks at 1260 cm^–1 (C–O–C) and 1180/1110 cm^–1 (C–O), reflected the typical aliphatic-aromatic polyester backbone. , Furthermore, characteristic peaks at 940 (C–C), 1020 (C–H), and 870 cm^–1 (CO₃ ^2–, indicative of inorganic filler) were clearly identified. Overall, while PEMPs contain a certain amount of surface O-containing species, WHMPs and RAMPs exhibit a much higher abundance and density of ester and carbonyl groups, which are expected to offer more favorable polar sites and potential interaction centers during solid–liquid contact.

3.1.3. Surface Elemental Composition and Core-Level Electronic States

The surface elemental compositions and chemical bonding environments of the three microplastics were systematically analyzed by XPS, as displayed in Figure h–k. The survey spectra (Figure h) revealed that C, O, and N were present on the surfaces of all three microplastics. The relative atomic fractions of O 1s in WHMPs (15.29%) and RAMPs (15.68%) were substantially higher than that in PEMPs (10.08%), which is consistent with the stronger absorption bands of ester and carbonyl groups observed in the FT-IR spectra (Figure g). The presence of N 1s (1.84%–6.21%) and O 1s in PEMPs was likely derived from industrial additives (e.g., slip agents, antioxidants, or thermal stabilizers) or surface aging during processing. High-resolution C 1s deconvolution (Figure i) further elucidated the distinct chemical skeletons. The C 1s spectrum of PEMPs was fitted with components at 284.90 eV (CC, 89.38%), 285.96 eV (C–C/C–H, 4.46%), and 286.62 eV (C–O/C–N, 6.16%). In contrast, both WHMPs and RAMPs exhibited a distinct characteristic peak corresponding to ester carboxyl groups (O–CO) at 288.82 eV (8.83%) and 288.08 eV (8.80%), respectively, along with C–O/C–N components at 286.31 eV (14.18%) and 286.63 eV (13.04%), confirming the polyester backbone of the biodegradable matrices. The N 1s spectra (Figure j) were resolved into CN (399.55 eV, 65.27%) and OC–N/–NH (398.04 eV, 34.63%) for PEMPs, CN (399.45 eV, 76.85%) and C–N (402.02 eV, 23.15%) for WHMPs, and a dominant OC–N/–NH peak at 398.81 eV for RAMPs. The high-resolution O 1s spectra (Figure k) were deconvoluted into carbonyl oxygen (CO) and ether/ester oxygen (C–O). The proportion of C–O in WHMPs (533.35 eV, 41.08%) and RAMPs (532.70 eV, 34.29%) was considerably higher than that in PEMPs (532.84 eV, 19.78%), while CO accounted for 58.92% (531.84 eV) and 65.71% (531.10 eV), respectively. In summary, the XPS results corroborating the FT-IR observations demonstrate that WHMPs and RAMPs possess higher surface O fractions and prominent ester/carbonyl functionalities (O–CO, C–O, and CO), offering an abundance of electronegative and polar interfacial sites compared to PEMPs.

3.2. Parametric Influences on TC and CTC Adsorption by Conventional and Biodegradable Microplastics

3.2.1. Effect of Initial Concentration

The influence of initial concentration on the adsorption of TC and CTC by PEMPs, WHMPs, and RAMPs is illustrated in Figure a. As the initial antibiotic concentration increased, the equilibrium adsorption capacities (q _e) of all three microplastics exhibited a rapid initial increase, followed by a gradual leveling-off toward a saturation plateau. This pattern is primarily attributed to the enhanced concentration gradient at higher initial concentrations, which provides a stronger driving force to overcome the interfacial mass transfer resistance. As the concentration further increased, the accessible active sites on the microplastic surfaces were progressively occupied, eventually leading to site saturation. Throughout the tested concentration range, distinct differences in the adsorption performance were observed among the microplastics. The two biodegradable microplastics (WHMPs and RAMPs) showed comparable adsorption behaviors and exhibited substantially higher adsorption capacities than those of conventional PEMPs for both antibiotics. For instance, at maximum tested concentrations, the adsorption capacities of WHMPs and RAMPs reached 7.42–8.41 mg/g for CTC and 4.45–6.31 mg/g for TC, whereas PEMPs exhibited much lower values (2.81 mg/g for CTC and 0.88 mg/g for TC). This pronounced superiority of WHMPs and RAMPs is consistent with their rougher surface morphologies and higher abundance of O-containing functional groups (e.g., CO and C–O). Furthermore, all microplastics consistently exhibited higher affinity and uptake for CTC than for TC across the entire concentration range, indicating that the structural characteristics of the antibiotic molecules (e.g., the electron-withdrawing chlorine substituent on CTC) also play a critical role in interfacial binding.

2.

Effects of environmental factors and operating parameters on the adsorption of TC and CTC onto PEMPs, WHMPs, and RAMPs: (a) Initial concentration, (b) contact time, (c) ambient temperature, (d) solution pH, (e) ionic strength, (f) dosage, and (g) particle size (Error bars represent the standard deviation of triplicate experiments).

3.2.2. Effect of Contact Time

As illustrated in Figure b, the adsorption process of all microplastics exhibited a typical two-stage pattern consisting of an initial rapid uptake followed by a slower stage approaching equilibrium. During the initial stage (0–12 h), the adsorption capacities of WHMPs and RAMPs for both TC and CTC increased sharply. For the TC adsorption system, equilibrium was rapidly established within approximately 12 h, reaching plateau capacities of around 2.03–2.20 mg/g for WHMPs and RAMPs, whereas PEMPs exhibited a much lower uptake (<0.7 mg/g). In the case of CTC, the adsorption capacity on WHMPs and RAMPs reached a peak of approximately 3.65–3.68 mg/g at 24 h and stabilized thereafter, significantly outperforming that of PEMPs (1.08 mg/g at 48 h). The rapid uptake in the initial stage can be attributed to the abundant unoccupied surface active sites and the large concentration gradient between the solution and the microplastic surfaces. As the available sites were progressively occupied, the adsorption rate gradually leveled off until a dynamic equilibrium was achieved. Across the entire duration, the biodegradable microplastics (WHMPs and RAMPs) consistently demonstrated faster adsorption rates and substantially higher equilibrium capacities than PEMPs, with CTC exhibiting adsorption amounts higher than those of TC.

3.2.3. Effect of Ambient Temperature

As the temperature increased from 30 to 50 °C, the adsorption capacities of all three microplastics for both antibiotics exhibited a consistent upward trend, indicating that elevated temperatures favored the interfacial uptake process (Figure c). Specifically, for CTC, the adsorption capacities on WHMPs and RAMPs progressively increased from 3.27 to 3.32 mg/g at 30 °C to 3.78–3.82 mg/g at 50 °C, while that on PEMPs increased from 0.90 mg/g to 2.86 mg/g. A similar enhancement was observed for TC adsorption, with the capacities of WHMPs and RAMPs rising from 1.93 to 2.04 mg/g at 30 °C to 2.86–2.87 mg/g at 50 °C, and PEMPs increasing from 0.40 mg/g to 1.45 mg/g. The enhanced adsorption at higher temperatures is generally associated with the endothermic nature of the interaction, where thermal energy accelerates molecular thermal motion, enhances boundary layer mass transfer, and facilitates the interaction between antibiotic molecules and microplastic surface active sites. Across all tested temperatures, WHMPs and RAMPs consistently maintained substantially higher adsorption capacities than PEMPs, and the uptake of CTC remained superior to that of TC.

3.2.4. Effect of Solution pH

Solution pH exerted distinct effects on the adsorption performance depending on the types of microplastics and antibiotics (Figure d). For the adsorption of CTC onto the biodegradable microplastics (WHMPs and RAMPs), the adsorption capacity was relatively low at pH 3.0, reached a maximum plateau under weakly acidic to neutral conditions (2.76–3.15 mg/g), and subsequently declined to 1.65–1.83 mg/g under alkaline conditions (pH 9.0–11.0). In contrast, the adsorption capacities of WHMPs and RAMPs for TC increased markedly from pH 3.0 and remained elevated across pH 5.0–9.0, peaking at pH 9.0 before slightly decreasing at pH 11.0. For PEMPs, the adsorption capacities for both TC and CTC were overall lower but increased progressively with rising pH, from less than 0.6 mg/g at pH 3.0 to 1.77–1.80 mg/g at pH 11.0.

The pH-dependent adsorption behavior is primarily governed by the electrostatic interplay between the speciation of antibiotic molecules and the surface charge properties of the microplastics, as characterized by their point of zero charge (pH_pzc). The measured pH_pzc values for WHMPs, RAMPs, and PEMPs were 3.2, 3.6, and 4.2, respectively. At pH 3.0, which is below or close to the pH_pzc of the biodegradable microplastics, the protonation of surface functional groups results in a nearly neutral or weakly positive surface charge, while TC and CTC predominantly exist as cationic species (TCH³⁺ and CTCH³⁺, pK _a1 ≈ 3.3). This leads to electrostatic repulsion and competitive proton binding, thereby suppressing antibiotic uptake. As the solution pH increases to 5.0–7.0 (pH > pH_pzc), deprotonation of the terminal carboxyl and ester groups imparts an increasingly negative surface charge to WHMPs and RAMPs. In this pH window, the antibiotics predominantly transition into zwitterionic forms (TCH^2± and CTCH^2±), where the positively charged dimethylamino groups establish strong electrostatic attraction with the negatively charged microplastic surfaces, concurrently optimizing hydrophobic partitioning and dipole–dipole interactions to achieve maximum adsorption capacities. Under alkaline conditions (pH > 9.0), TC and CTC dissociate into anionic species (TCH-/TC^2– and CTCH-/CTC^2–), resulting in strong electrostatic repulsion against the intensely negatively charged polyester surfaces, which leads to a marked decrease in CTC adsorption. In the case of PEMPs (pH_pzc = 4.2), the lack of polar O-containing functional groups makes hydrophobic interaction and surface wetting the predominant mechanisms, where the gradual deprotonation and slight structural expansion in solution facilitate moderate, continuous diffusion across the tested pH range. Overall, biodegradable microplastics exhibited favorable adsorption over a wide pH window compared to PEMPs.

3.2.5. Effect of Ionic Strength

The influence of ionic strength on the adsorption of TC and CTC onto the three microplastics was investigated across a NaCl concentration range of 0 to 0.3 mol/L, as shown in Figure e. An increase in ionic strength resulted in a monotonic decrease in the adsorption capacity across all microplastic and antibiotic systems. In the absence of an electrolyte (0 mol/L), WHMPs and RAMPs exhibited the highest adsorption capacities for CTC. As the NaCl concentration increased to 0.3 mol/L, their capacities progressively decreased to 2.35 and 2.28 mg/g, respectively. A corresponding reduction was observed for PEMPs, with CTC uptake dropping from 1.20 mg/g to 0.51 mg/g. For TC adsorption, the capacities of WHMPs and RAMPs similarly decreased from 2.21 and 1.91 mg/g at 0 mol/L to 1.19 and 1.16 mg/g at 0.3 mol/L, respectively, while PEMPs exhibited a decrease from 0.47 mg/g to 0.14 mg/g. This inhibitory effect is predominantly attributed to the electrostatic shielding effect induced by the high concentration of background ions (Na⁺ and Cl^–), which compresses the electrical double layer and weakens the electrostatic interactions between antibiotic molecules and microplastic surfaces. Additionally, competitive adsorption and the formation of hydrated ionic barriers may further hinder the interfacial binding of the antibiotic molecules. Notably, the adsorption capacities of the biodegradable microplastics (WHMPs and RAMPs) remained consistently superior to those of PEMPs at all tested ionic strengths.

3.2.6. Effect of Dosage

As shown in Figure f, across all adsorption systems, the unit adsorption capacity exhibited an overall declining trend with increasing microplastic dosage. For the CTC adsorption system, WHMPs and RAMPs demonstrated exceptionally high adsorption capacities at the lowest dosage of 0.5 g/L (8.28 and 8.45 mg/g, respectively). As the dosage increased to 4.0 g/L, the adsorption capacities progressively decreased to 1.93 and 1.92 mg/g. A corresponding reduction was observed for PEMPs, with the unit CTC uptake decreasing from 2.22 mg/g at 0.5 g/L to 0.77 mg/g at 4.0 g/L. For TC adsorption, the capacities of WHMPs and RAMPs ranged from 1.46 to 2.04 mg/g at moderate dosages (1.0–2.0 g/L) and decreased to 1.15–1.28 mg/g at 4.0 g/L, whereas PEMPs consistently exhibited the lowest uptake (<0.5 mg/g) throughout the tested dosage range. The decline in the unit adsorption capacity with increasing adsorbent dosage is primarily attributed to the split of a fixed total amount of antibiotic molecules over an excess of available surface active sites, resulting in site unsaturation. Furthermore, partial particle agglomeration at elevated solid concentrations may shield active surface sites. Across all dosages, the biodegradable microplastics (WHMPs and RAMPs) maintained distinctly higher adsorption capacities than PEMPs, with CTC consistently exhibiting a higher uptake than TC.

3.2.7. Effect of Particle Size

The influence of microplastic particle size on the adsorption of TC and CTC was investigated across five size fractions (0–0.9 mm, 0.9–2.0 mm, 2.0–2.5 mm, 2.5–4.0 mm, and 4.0–5.0 mm), as shown in Figure g. Across all adsorption systems, the adsorption capacity increased monotonically as the particle size decreased from 4.0 to 5.0 mm to 0–0.9 mm. For CTC adsorption, the capacities of WHMPs and RAMPs rose from 3.00 and 2.93 mg/g at 4.0–5.0 mm to 3.69 and 3.67 mg/g at 0–0.9 mm, respectively, while that of PEMPs increased from 0.53 mg/g to 0.99 mg/g. A similar size-dependent enhancement was observed for the TC adsorption. The uptake on WHMPs and RAMPs increased from 1.88 and 1.45 mg/g at 4.0–5.0 mm to 2.40 and 2.09 mg/g at 0–0.9 mm, respectively, and PEMPs showed an increase from 0.20 mg/g to 0.58 mg/g. The enhanced adsorption with decreasing particle size is primarily attributed to the increased geometric external surface area and the greater exposure of accessible active surface sites per unit mass, which also reduces the interfacial mass-transfer distance for antibiotic molecules. Across all particle size fractions, the biodegradable microplastics (WHMPs and RAMPs) consistently demonstrated superior adsorption performance compared to PEMPs, with CTC exhibiting higher overall adsorption amounts than TC.

3.3. Isotherm, Kinetic, and Thermodynamic Analysis

3.3.1. Adsorption Isotherm Modeling

The equilibrium adsorption data of TC and CTC onto PEMPs, WHMPs, and RAMPs were fitted using the Langmuir, Freundlich, and Dubinin–Radushkevich (D–R) isotherm models (Figure a–c), with the corresponding fitting parameters summarized in Table .

3.

Adsorption isotherms and kinetics of TC and CTC on PEMPs, WHMPs, and RAMPs: (a) Langmuir isotherm, (b) Freundlich isotherm, (c) Dubinin–Radushkevich (D–R) isotherm, (d) PFO model, (e) PSO model, and (f) Elovich model.

1. Adsorption Isotherm Model Fitting Parameters for the Adsorption of Antibiotics by Agricultural Film-Derived Microplastics .

	Langmuir model			Freundlich model			D–R model
	K L (L/mg)	Q m (mg/g)	R 2	K F ((mg/g) (L/mg)1/n )	1/n	R 2	Q m	E (kJ/mol)	R 2
TC adsorption
PEMPs	0.062	1.22	0.962	0.13	0.52	0.888	0.73	0.36	0.821
WHMPs	0.14	5.20	0.980	1.04	0.43	0.906	3.64	0.61	0.777
RAMPs	0.083	6.18	0.983	0.59	0.65	0.967	3.76	0.50	0.743
CTC adsorption
PEMPs	0.028	5.43	0.938	0.23	0.69	0.921	2.04	0.33	0.706
WHMPs	0.12	11.46	0.957	1.90	0.44	0.828	6.39	0.74	0.925
RAMPs	0.086	15.01	0.969	1.83	0.50	0.846	7.10	0.71	0.894

^aNote: Parameters for the D–R model are presented strictly as empirical curve-fitting constants.

As summarized in Table , the Langmuir isotherm model exhibited high correlation coefficients (R ² = 0.938–0.983), showing better consistency than the Freundlich model (R ² = 0.828–0.967) and D–R model (R ² = 0.706–0.925). This suggested that the interfacial uptake of TC and CTC on both conventional and biodegradable microplastics predominantly followed monolayer adsorption behavior with localized site saturation. The theoretical maximum monolayer adsorption capacity (q _m) followed the sequence of RAMPs, followed by WHMPs, and was substantially higher than that of PEMPs for both antibiotics. Specifically, the q _m values for TC were 6.18, 5.20, and 1.22 mg/g for RAMPs, WHMPs, and PEMPs, respectively, whereas substantially higher q _m values were achieved for CTC (15.01, 11.46, and 5.43 mg/g, respectively). Furthermore, the higher Langmuir constant (K _L) values observed for WHMPs (0.12–0.14 L/mg) and RAMPs (0.083–0.086 L/mg) relative to those of PEMPs (0.028–0.062 L/mg) demonstrate the stronger surface binding affinity of biodegradable microplastics toward tetracyclines.

The notably superior adsorption capacity observed for CTC over TC across all three microplastics can be rigorously rationalized by their intrinsic molecular descriptors, hydrophobic partition driving forces, and functional group electronics. First, CTC possesses a higher octanol–water partition coefficient (logK _ow = −0.62 to −0.45) compared to TC (logK _ow = −1.30 to −1.19), indicating a markedly lower hydration energy and stronger intrinsic hydrophobicity. Under aqueous conditions, the nonpolar backbone of PEMPs and the aromatic/aliphatic segments of PBAT/PLA exert a stronger hydrophobic exclusion effect toward the less hydrophilic CTC molecules, driving their preferential interfacial accumulation. Second, the structural disparity between the two antibiotics originates fundamentally from the substitution at the C7 position of the naphthacene core, where CTC possesses a strongly electronegative and polarizable chlorine atom (–Cl, Pauling electronegativity = 3.16) in place of the hydrogen atom in TC. The robust electron-withdrawing induction of the C7–Cl substituent accentuates localized charge asymmetry and enhances the molecular dipole moment across the conjugated phenolic diketone system. This electronic configuration promotes stronger dipole–dipole attractions and n–π/π–π interfacial interactions with the ester carbonyl motifs (CO) and aromatic rings on the biodegradable polyester matrices. Under weakly acidic to neutral conditions (pH 5.0–7.0) where both antibiotics predominantly exist as zwitterions (pK _a1 = 3.3, pK _a2 = 7.4–7.7), electrostatic repulsion is minimized, allowing these hydrophobic and dipole-induced driving forces to dominate the interfacial mass loading.

Although the D–R isotherm model was also fitted to the experimental data to obtain empirical curve parameters, its derived adsorption energy (E) values are discussed here solely as mathematical metrics rather than rigorous thermodynamic indicators, in view of the theoretical and dimensional limitations associated with liquid-phase D–R formulations. , The robust monolayer adsorption behavior and underlying mechanisms are instead comprehensively supported by the high goodness of fit of the Langmuir model.

To elucidate whether the uptake of TC and CTC is governed by surface adsorption or bulk polymer dissolution, N₂ adsorption/desorption measurements were conducted (Figure and Table ). As shown, all tested microplastics exhibited restricted specific surface areas (ranging from 0.083 to 0.37 m²/g) and very low pore volumes. Notably, conventional polyethylene (PEMPs) possessed the largest specific surface area (0.37 m²/g) yet exhibited the lowest adsorption capacities for both antibiotics. In contrast, biodegradable microplastics (WHMPs and RAMPs) with smaller surface areas showed substantially higher uptake capacities. This inverse relationship clearly demonstrates that internal pore diffusion or bulk matrix dissolution and/or partitioning is not the dominant mechanism. Instead, the accumulation of antibiotics is primarily controlled by surface-specific chemical interactions (such as hydrogen bonding and dipole–dipole attractions) occurring at active interfacial sites.

4.

N₂ adsorption/desorption isotherms of (a) PEMPs, (b) RAMPs, and (c) WHMPs, and (d) their corresponding pore diameter distributions.

2. Porosity Properties of the MPs.

	specific surface area (m2/g)	mesoporous area (m2/g)	micropore area (m2/g)	total pore volume (cm3/g)	mesoporous volume (cm3/g)	micropore volume (cm3/g)	average pore diameter (nm)
PEMPs	0.37	0.24	0.13	0.00075	0.0003	0.000004	8.02
RAMPs	0.15	0.10	0.044	0.0004	0.00024	0.000012	10.88
WHMPs	0.23	0.09	0.14	0.0004	0.00023	0.000053	6.99

3.3.2. Adsorption Kinetics Modeling

To investigate the dynamic adsorption process and potential rate-limiting mechanisms, the experimental data as a function of contact time were fitted using the PFO, PSO, and Elovich kinetic models (Figure d–f). The corresponding kinetic parameters, along with experimental equilibrium capacities (q _e,exp), are summarized in Table .

3. Kinetic Parameters for Antibiotic Adsorption by Agricultural Film-Derived Microplastics.

		PFO model			PSO model			Elovich model
	q e,exp	q e (mg/g)	k 1 (1/h)	R 2	q e (mg/g)	k 2 (g/(mg·h))	R 2	a (mg/g)	b (mg/(g·h))	R 2
TC adsorption
PEMPs	0.65	0.40	2.30	0.012	0.50	0.59	0.486	0.33	0.033	0.246
WHMPs	2.20	2.02	0.34	0.858	2.25	0.19	0.898	0.87	0.36	0.851
RAMPs	2.16	2.06	0.83	0.840	2.18	0.50	0.759	0.15	0.20	0.542
CTC adsorption
PEMPs	1.08	1.06	0.080	0.983	1.40	0.050	0.992	0.010	0.27	0.990
WHMPs	3.31	3.38	0.16	0.929	3.93	0.029	0.864	0.32	0.90	0.777
RAMPs	3.44	3.53	0.21	0.971	3.74	0.071	0.915	0.79	0.82	0.806

The kinetic parameters listed in Table demonstrate distinct fitting preferences across different microplastic–antibiotic pairs. For CTC adsorption onto biodegradable microplastics (WHMPs and RAMPs), the PFO model yielded superior determination coefficients (R ² = 0.929–0.971), and the calculated equilibrium adsorption capacities (q _e = 3.38 mg/g for WHMPs and 3.53 mg/g for RAMPs) well matched the experimentally observed values (q _e,exp = 3.31 and 3.44 mg/g). Mechanistically, the PFO model describes an interface-controlled, concentration-gradient-driven physical partitioning process, where the adsorption rate is governed by the occupancy of external surface-active sites without high activation barriers. The strongly electronegative and polarizable chlorine substituent (–Cl) on the CTC core markedly facilitates dipole–dipole attractions and hydrophobic interactions with the abundant ester linkages (CO and C–O) of biodegradable polyesters. Consequently, CTC rapidly undergoes reversible physisorption on the accessible film interface. In contrast, for TC systems where secondary surface rearrangement and multisite electrostatic or hydrogen-bonding competition occur among the polar phenolic and amide functionalities, the PSO model provided a slightly better correlation (R ² = 0.898 for WHMPs-TC).

3.3.3. Adsorption Thermodynamics Analysis

As shown in Table , all linear fits demonstrated high correlation coefficients (R ² = 0.943–0.994). The ΔH° values for all adsorption systems were positive (16.05–73.12 kJ/mol), confirming that the adsorption of TC and CTC onto both biodegradable and conventional microplastics is inherently endothermic. , This thermodynamic finding is consistent with the enhanced adsorption capacities observed at elevated temperatures (Figure c). Notably, the ΔH values for WHMPs and RAMPs (16.05–30.96 kJ/mol) were markedly lower than those for PEMPs (62.42–73.12 kJ/mol), indicating lower energy barriers for antibiotic uptake on biodegradable microplastic surfaces. Across the tested temperature range, ΔG° values for all systems decreased monotonically with increasing temperature from 303 to 323 K. For CTC adsorption onto WHMPs and RAMPs, ΔG° transitioned to negative values at 313 and 323 K (ranging from −0.16 to −1.28 kJ/mol), indicating that the adsorption became thermodynamically spontaneous and favorable at higher temperatures. Furthermore, all ΔS° values were positive (53.33–222.30 J/(mol·K)), which is primarily attributed to the release of hydrated water molecules during the interfacial binding process (desolvation), leading to increased degrees of freedom and randomness at the microplastic–solution interface. Consequently, the adsorption of tetracyclines onto the three microplastics is an endothermic and entropy-driven process.

4. Thermodynamic Parameters for Antibiotic Adsorption by Agricultural Film-Derived Microplastics.

	ΔG° (kJ/mol)
	303 K	313 K	323 K	ΔH° (kJ/mol)	ΔS° (J/(mol·K))	R 2
TC adsorption
PEMPs	7.87	6.66	4.26	62.42	179.43	0.943
WHMPs	2.68	1.96	1.07	26.92	79.92	0.994
RAMPs	2.91	1.86	1.06	30.96	92.72	0.988
CTC adsorption
PEMPs	5.55	3.98	1.08	73.12	222.30	0.947
WHMPs	0.02	–0.46	–1.28	19.76	64.96	0.955
RAMPs	–0.16	–0.55	–1.23	16.05	53.33	0.943

3.4. Potential Adsorption Mechanisms

3.4.1. Surface Chemical State Analysis

To elucidate the interfacial interaction mechanisms of TC and CTC on conventional versus biodegradable microplastics, high-resolution XPS analysis was performed on PEMPs and RAMPs before and after antibiotic uptake (Figure ). The wide-scan survey spectra revealed distinct elemental variations between the two microplastic systems (Figure a,e). For RAMPs, the surface O and N atomic contents increased from 15.68% and 1.84% to 21.30% and 2.96% after TC loading and to 17.31% and 3.12% after CTC loading, respectively, accompanied by a corresponding decline in C content. This marked enrichment of heteroatoms directly confirms the successful immobilization of TC and CTC molecules onto RAMP surfaces. In contrast, PEMPs exhibited negligible changes in elemental composition following TC adsorption (O content shifted from 10.08% to 9.89% and N from 6.21% to 6.23%), which is consistent with its minimal adsorption capacity. The deconvoluted C 1s spectra showed that while PEMPs maintained relatively stable carbon chemical states before and after adsorption (Figure b), RAMPs displayed noticeable shifts in binding energy (Figure f). The CC and C–C/C–H components of RAMPs shifted from 284.08 and 285.63 eV to 284.00 and 285.69 eV for TC and to 284.11 and 285.81 eV for CTC, respectively. Because these shifts are small, they should be interpreted cautiously and are not used as a standalone proof of specific chemical bond formation. Instead, they are considered to be supportive of a changed local interfacial environment after adsorption.

5.

XPS wide-scan survey and high-resolution spectra of PEMPs and RAMPs before and after TC and CTC adsorption: (a) wide-scan survey of PEMPs, (b) C 1s spectra of PEMPs, (c) N 1s spectra of PEMPs, (d) O 1s spectra of PEMPs, (e) wide-scan survey of RAMPs, (f) C 1s spectra of RAMPs, (g) N 1s spectra of RAMPs, and (h) O 1s spectra of RAMPs.

In the N 1s spectra of RAMPs (Figure g), pristine RAMPs displayed only a single OC–N/–NH peak at 398.81 eV. Following adsorption, a new CN/protonated nitrogen component emerged at 401.13 eV (19.42%) for TC and 401.02 eV (23.37%) for CTC, while the primary OC–N/–NH peak shifted to lower binding energies (398.57–398.74 eV). This evolution suggests that the dimethylamino and amide groups of tetracyclines are involved in interfacial interactions with the polar surface domains of RAMPs, although the assignment is treated cautiously given the small magnitude of the shifts. In the O 1s spectra of RAMPs (Figure h), the relative area of CO species increased substantially from 65.71% in pristine RAMPs to 88.79% (TC) and 86.82% (CTC), accompanied by a positive binding energy shift from 531.10 eV to 531.36–531.38 eV (+0.26–0.28 eV). This O 1s variation is consistent with the involvement of surface carbonyl and ester groups in the adsorption process, but it is discussed as corroborative evidence rather than definitive proof of a specific bonding mode. For PEMPs (Figure c,d), the shifts in N 1s and O 1s peaks were comparatively smaller and primarily reflected nonspecific surface deposition. The higher adsorption affinity of biodegradable RAMPs over conventional PEMPs is jointly governed by physical surface accumulation, dipole–dipole attractions, and weak hydrogen-bonding networks. The XPS results are interpreted together with FT-IR, BET, and thermodynamic analyses to support a surface-controlled adsorption mechanism.

3.4.2. Morphological, Elemental Mapping, and FT-IR Spectroscopic Characterization

To obtain direct microstructural and vibrational evidence of antibiotic uptake on biodegradable microplastics, the surface morphologies, elemental distributions, and functional groups of RAMPs before and after TC and CTC adsorption were comprehensively characterized (Figure ). Compared with the relatively smooth, intact, and compact laminated surface of pristine RAMPs (Figure ), the SEM micrographs of RAMPs after antibiotic exposure displayed pronounced microstructural modifications (Figure a–d). Following TC adsorption, the surface became remarkably rough, accompanied by noticeable microflaking and fine particulate coatings. For RAMPs-CTC, the surface exhibited more severe structural looseness, conspicuous cracks, and dense aggregate deposition. This morphological evolution directly confirms that antibiotic molecules are effectively captured and accumulated on the microplastic surface, with the greater surface modification in CTC systems matching its higher theoretical maximum adsorption capacity (q _m) determined in the isotherm modeling.

6.

Surface morphology, elemental distributions, and FT-IR spectra of RAMPs before and after TC and CTC adsorption: (a,b) SEM images of RAMPs-TC at 1k and 5k magnifications; (c,d) SEM images of RAMPs-CTC at 1k and 5k magnifications; (e–h) elemental mapping of C, N, O, and layered image for RAMPs-TC; (i–l) elemental mapping of C, N, O, and layered image for RAMPs-CTC; (m,n) EDS spectra and quantitative elemental compositions of RAMPs-TC and RAMPs-CTC; and (o) FT-IR spectra of RAMPs before and after TC and CTC adsorption.

The EDS analysis verified substantial surface elemental accumulations (Figure m,n). Specifically, RAMPs-TC exhibited 45.56 at. % C, 42.48 at. % O, and 11.96 at. % N, while RAMPs-CTC showed 42.36 at. % C, 47.68 at. % O, and 9.96 at. % N. Furthermore, the corresponding elemental mapping images (Figure e–l) demonstrated that C, N, and O were highly dispersed and uniformly distributed across the microplastic matrix without localized phase segregation. This spatially homogeneous elemental distribution provides direct visual confirmation of the Langmuir monolayer adsorption mechanism on energetically uniform active sites.

The FT-IR spectra further verified the specific chemical binding behaviors (Figure o). After adsorption, a broad absorption band emerged in the 3500–3200 cm^–1 region corresponding to –NH and –OH stretching vibrations, which directly originate from the abundant phenolic hydroxyl, alcohol hydroxyl, and amide moieties of the adsorbed tetracyclines. , Additionally, characteristic absorption peaks associated with the ester backbone of RAMPs, including CO (1730 cm^–1), C–O–C (1260 cm^–1), C–O (1180 and 1110 cm^–1), and aliphatic CH₂ bands (2920, 2840, 1460, and 1420 cm^–1), exhibited prominent intensity enhancements following TC and CTC loading. These spectral changes confirm that the O- and N-containing groups of tetracyclines actively engage with the ester/carbonyl functionalities of RAMPs via intermolecular hydrogen bonding and dipole–dipole interactions.

3.4.3. Desorption and Dynamic Leaching Behaviors of TC/CTC

To comprehensively elucidate the environmental mobility, interfacial reversibility, and potential carrier risks of antibiotic-loaded microplastics in agroecosystems, static batch desorption and multicycle dynamic leaching tests were conducted for PEMPs, WHMPs, and RAMPs pre-equilibrated with TC and CTC (Figure ). As depicted in Figure a,b, the single-batch desorption efficiency varied markedly depending on the polymer matrix, antibiotic substituent structure, and solution chemistry. Across all investigated desorption media, PEMPs exhibited substantially higher desorption efficiencies (18.0%–55.1% for TC and 18.0%–42.0% for CTC) than the two biodegradable microplastics (8.0%–42.0% for WHMPs and 10.0%–42.0% for RAMPs). This distinct contrast arises from the fundamentally different interfacial binding mechanisms: the smooth, nonpolar surface of PEMPs relies solely on weak nonspecific van der Waals interactions, allowing adsorbed molecules to detach readily upon dilution. Conversely, the abundant oxygenated ester linkages (CO, C–O) and porous microstructures of WHMPs and RAMPs provide localized multipoint binding sites and steric entrapment, thereby contributing to pronounced adsorption hysteresis. Furthermore, CTC consistently displayed lower desorption ratios than TC across all matrices, which is attributed to the presence of the electron-withdrawing chlorine atom (–Cl) on the CTC skeleton that enhances its molecular hydrophobicity and dipole interaction with the polymer surfaces.

7.

Desorption and dynamic leaching behaviors of TC and CTC on PEMPs, WHMPs, and RAMPs: (a) single-batch desorption efficiency of TC and (b) CTC across different solution media (deionized water, unadjusted 0.01 mol/L NaCl, and 0.01 mol/L NaCl adjusted to pH 5.0, 7.0, and 9.0); (c) cumulative leaching efficiency of TC and (d) CTC across five consecutive leaching cycles using fresh 0.01 mol/L NaCl background solution (initial loading concentration = 20 mg/L, T = 298 K, n = 3). Error bars represent standard deviations of triplicate measurements.

Solution chemistry exerted a critical modulating effect on the antibiotic detachment. Compared to deionized water, an unadjusted 0.01 mol/L NaCl medium slightly enhanced the desorption efficiency due to electrostatic shielding and competitive displacement by hydrated Na⁺ ions. When the background ionic strength was fixed at 0.01 mol/L of NaCl, increasing the solution pH from 5.0 to 9.0 resulted in a monotonic elevation in desorption efficiency for all three microplastics. At pH 5.0, TCs predominantly exist as neutral zwitterions with minimal net charge, and the low deprotonation of carboxyl/ester groups on biodegradable microplastics maintains stable hydrogen bonding, yielding the lowest desorption ratio (10.0%–16.9%). Conversely, at alkaline pH (pH 9.0), both antibiotic species (fully converted into divalent anions TC^2–/CTC^2–) and the microplastic surfaces (abundant deprotonated –COO^– moieties) carry intense negative charges. The resulting severe electrostatic repulsion destabilizes the interfacial complexes, accelerating the detachment of the bound antibiotics.

The multicycle dynamic leaching profiles (Figure c,d) further illustrate the sustained release potential under continuous rinse conditions (0.01 mol/L NaCl). All microplastic systems exhibited a characteristic two-stage leaching pattern: a rapid initial release in cycle 1 (accounting for 50%–70% of total leached amounts) corresponding to the detachment of loosely bound outer-sphere molecules, followed by a sustained and plateauing phase from cycle 2 to cycle 5 governed by intraparticle diffusion resistance. Notably, while the cumulative leaching ratios of WHMPs (36.0% for TC, 22.0% for CTC) and RAMPs (34.0% for TC, 24.0% for CTC) after 5 cycles were significantly lower than those of PEMPs (63.0% for TC, 54.1% for CTC), the absolute mass of antibiotics released from biodegradable microplastics was substantially higher due to their 3–4 fold greater initial adsorption capacities. These observations demonstrate that while biodegradable microplastics exhibit superior pollutant retention and retarding capabilities in neutral to acidic soils, they still possess considerable secondary release potential under fluctuating environmental chemistries (e.g., rainfall leaching, irrigation flushes, or alkaline drainage), functioning as dynamic vectors that may facilitate the long-distance migration and prolonged ecological exposure of antibiotics in agricultural environments.

3.5. Environmental Relevance, Matrix Effects, and Practical Implications

It is important to note that the present batch adsorption and desorption experiments were conducted under controlled aqueous conditions using a relatively broad antibiotic concentration range. This experimental design is methodologically essential for fully mapping nonlinear adsorption isotherms, identifying plateau capacities, and determining fundamental thermodynamic parameters. However, in real agricultural soil and aquatic environments, antibiotic concentrations typically reside at trace levels (μg/L to mg/kg), and microplastic interfacial behaviors are modulated by complex environmental matrices. In natural farmland soils, dissolved organic matter (DOM, such as humic and fulvic acids) may compete with antibiotics for interfacial hydrophobic and dipole binding sites or form surface coatings that alter particle charge and pore availability. Meanwhile, multivalent soil cations (Ca²⁺, Mg²⁺, and Fe³⁺) can form cation bridges between negatively charged polymer ester groups and antibiotic functional moieties, potentially enhancing retention. Furthermore, agricultural microplastics undergo chronic environmental weathering (UV irradiation, mechanical abrasion, and biological fragmentation) and biofilm colonization (plastisphere formation), which introduce abundant oxygenated defects, increase surface roughness, and dynamically alter surface hydrophobicity. Although these matrix complexities may alter the absolute adsorption/desorption kinetics in the field, the fundamental interfacial affinity and the greater carrier risk demonstrated by biodegradable microplastics over conventional PE remain critically valid. Future investigations should incorporate realistic soil-porewater systems, chronic weathering matrices, and microbial biofilm interactions to comprehensively track the multiscale fate of antibiotic-loaded biodegradable microplastics in agroecosystems.

4. Conclusions

This study provides a systematic and comparative investigation into the interfacial adsorption behaviors, thermodynamic driving forces, and release dynamics of TC and CTC on commercial biodegradable mulch microplastics (WHMPs and RAMPs) versus those of conventional polyethylene microplastics (PEMPs). The core findings and scientific insights are summarized as follows:

First, commercial biodegradable microplastics exhibited substantially greater adsorption capacities for both antibiotics compared to conventional PEMPs, with maximum monolayer capacities reaching 5.20–6.18 mg/g for TC and 11.46–15.01 mg/g for CTC on biodegradable matrices, compared to 1.22 mg/g for TC and 5.43 mg/g for CTC on PEMPs. Across all microplastics, CTC exhibited 2.4- to 4.4-fold higher adsorption than TC, which is fundamentally attributed to its higher hydrophobicity and the strong electron-withdrawing induction of the C7–Cl moiety that reinforces interfacial dipole–dipole attractions.

Second, the adsorption processes were thermodynamically spontaneous and endothermic with entropy increases serving as the principal driving force. The experimental equilibrium data conformed excellently to the Langmuir monolayer isotherm model (R ² = 0.938–0.983). The adsorption kinetics followed pseudo-first-order or pseudo-second-order models depending on specific adsorbate–adsorbent electronic properties, reaching equilibrium within 24–48 h.

Third, high-resolution XPS core-level spectra (C 1s, O 1s, N 1s) and FT-IR characterizations suggested that the interfacial retention on biodegradable polyesters is predominantly governed by reversible physisorption, hydrophobic partitioning, and weak dipole–dipole/hydrogen-bonding interactions with ester linkages (CO/C–O), without forming rigid covalent bonds. Dynamic desorption and multicycle leaching tests demonstrated that the adsorbed antibiotics possess sustained release potential in aqueous environments.

Finally, while commercial biodegradable mulch films effectively reduce macroscopic nondegradable plastic debris in agriculture, their fragmented microplastics possess a higher intrinsic affinity and vector capacity for agricultural antibiotics than conventional PE under controlled conditions. These results establish vital physical and chemical baselines. Future studies should focus on real farmland matrices (e.g., DOM, multivalent soil cations, and chronic photo-oxidative weathering) to comprehensively evaluate the multiscale ecological fate and carrier risks of biodegradable agricultural plastics in agroecosystems.

Data will be made available on request.

Lei Yu: Methodology, Investigation. Beier Qian: Methodology, Investigation. Rui Kang: Methodology, Investigation. Qichang Wang: Methodology, Investigation. Luoyi Cai: Methodology, Investigation. Zhanghong Wang: Supervision, Methodology, Investigation, Writing-original draft & Writing-review and editing.

This work was supported by Chizhou University High level Talent Research Start up Fund (CZ2024YJRC34), National Natural Science Foundation of China (52260018), the Guizhou Provincial Science and Technology Projects (QKHZC[2025]­YB104), the Department of education of Guizhou Province (No. QianJiaoJi[2023]­034).

The authors declare no competing financial interest.