Reusable and effective polyacrylic membranes for mecoprop and bentazon extractions

Water‑soluble pesticide formulations are widely used because they are readily mixed, rapidly absorbed by plants, and effective across many pest targets. However, their high solubility increases mobility via runoff and leaching, leading to contamination of water bodies and associated ecological and health risks. Two representative post‑emergence herbicides, mecoprop (MCP) and bentazon (BTZ), are water‑soluble, sorb at the leaf level, and have been detected in environmental matrices and even human hair, underscoring exposure concerns. The research question is whether a safe‑to‑handle, reusable, scalable membrane made from commercial monomers can efficiently extract such mobile pesticides at low concentrations from water, and what polymer–pesticide interactions govern this process. The study develops polyacrylic membranes bearing protonated aminoethyl side groups and evaluates structure–property–performance relationships, removal efficiency, transport, interaction mechanisms, and reusability.

Prior studies have explored diverse materials for removing MCP and BTZ, including photocatalysts (nanocrystalline TiO2, high RE under UV), adsorbents (activated carbons with high capacities, hydrotalcites, ferrihydrite), ion exchange resins (e.g., MIEX), phenolic resins, and filtration membranes (e.g., polyimide). Many of these materials are powders or resins that are difficult to recover from treated media, require complex conditions (e.g., UV irradiation), or are assessed at concentrations above legal limits. Reusability is limited in several cases (often ≤3 cycles). The present work positions a reusable, mechanically robust membrane operating effectively at low concentrations, with peak removal efficiency as concentrations diminish, as an advance over these prior approaches.

Materials: Commercial monomers 1‑vinyl‑2‑pyrrolidone (VP), methyl methacrylate (MMA), 2‑isocyanatoethyl methacrylate (NCO), crosslinker ethylene glycol dimethacrylate (E), initiator AIBN; pesticides: mecoprop (MCP) and bentazon (BTZ); interferents: urea and ammonium nitrate. Synthesis: Crosslinked membrane polymers (MEM1, MEM5, MEM10) were prepared by bulk radical copolymerization of VP/MMA/NCO at different NCO mol% (1, 5, 10) with 0.1 mol% E, initiated by AIBN at 60 °C overnight in oxygen‑free molds (90×120×0.1 mm). Post‑polymerization washing in water hydrolyzed isocyanate side groups to aminoethyl groups, with nascent CO2 protonating amines. Unwashed NCO‑bearing references are denoted MEM0_1/5/10. Discs (8 mm) were die‑cut for testing. A linear soluble analog (POL10; same composition as MEM10 but without crosslinker) was synthesized in dioxane (2 mol dm−3 monomers, AIBN 0.1 mol dm−3), polymerized at 60 °C, precipitated in hexane, Soxhlet‑washed, and water‑treated to generate aminoethyl side groups. Characterization: FTIR (ATR, mid‑IR) to monitor NCO hydrolysis/protonation and interaction signatures; 1H/13C NMR (POL10) for composition and MCP interaction titration; TGA (N2, 10 °C/min) and DSC (N2, 20 °C/min) for thermal properties; tensile testing for mechanics; N2 sorption (BET) for surface area (MEM10 pre/post hydrolysis); water swelling percentage (WSP) by gravimetry. Removal assays: UV‑Vis quantification (BTZ λmax 333 nm; MCP 280 nm) with calibrations 0–50 mg L−1. Standard tests used one 8 mm disc (~3.8 mg) in 2 mL pesticide solution at 10 mg L−1, 25 °C, 120 rpm, 24 h. Sorption metrics: qe = (C0−Ce)V/m; RE% = (C0−Ce)/C0. Transport properties: Permeation through membranes at 50 mg L−1 using a horizontal communicating vessel (A = 2 cm2), 25 °C, 250 rpm. Time‑lag method yielded apparent diffusion Dap = L2/(6θ) and permeability P = (V/A)S(L/Cpesticide); apparent partition Sap = P/Dap. Sorption kinetics and isotherms: Kinetics at 10 mg L−1 for 24 h, 25 °C; Fick’s 2nd law plane‑sheet model fit to estimate D and effective partition S. Isotherms (1–50 mg L−1; 24 h; 25 °C) fitted to Langmuir and BET models. NMR interaction study: Continuous variation (Job) method for MCP–POL10 in DMSO‑d6 to determine complex stoichiometry and equilibrium constant K, using NH3+ site concentration in POL10; BTZ trials were inconclusive. Interference/competitiveness/reusability: Competitive sorption from BTZ/MCP mixtures at fixed total molarity (0.07×10−3 mmol dm−3) and varying molar ratios. Interferent studies with urea and NH4NO3 at interferent:pesticide molar ratios 0.5–2 (pesticide at 10 mg L−1). Reusability over four sorption/desorption cycles (50 mg L−1, 24 h, 25 °C), desorbing with water or 0.1 M NaOH (followed by 0.1 M HCl to reprotonate). Molecular dynamics (MD) simulations: GROMACS 2022.1 with GROMOS 54a7, explicit SPC216 water, 5:5:1 VP/MMA/E polymer models representing MEM10 (with one protonated amino side group) and MEM (without amino group), paired with BTZ or MCP. Topologies via ATB (AM1‑BCC charges). Vacuum pre‑runs (20 ns) to obtain initial binding states; hydrated cubic boxes (5.5 nm), energy minimization, NVT/NPT equilibration (300 K, 1 bar), production 200 ns (final 50 ns for analysis). Nonbonded: LJ cutoff 1.2 nm, PME electrostatics; LINCS constraints. Binding enthalpies ΔH computed from Boltzmann‑averaged potential energies of complexes and components. Analyses included RMSD, water coordination numbers (RDF integration), clustering, and noncovalent interactions via Independent Gradient Method (IGM) using IGMPlot (δginter descriptor; sign(λ2)ρ to classify stabilizing/destabilizing interactions).

• Membrane composition and activation: FTIR showed disappearance of NCO band (~2270 cm−1) and appearance of NH3+ bands (~1560 cm−1) upon water washing, confirming hydrolysis and protonation. MEM10 BET surface area doubled after hydrolysis (0.76±0.05 to 1.60±0.17 m2 g−1), favoring sorbate–sorbent interactions.
• Swelling and mechanics: WSP values were ~63% (MEM1) and 57% (MEM5, MEM10). Young’s modulus decreased slightly with higher NCO content, indicating added flexibility; membranes maintained good manageability and durability. Glass transition temperatures increased with NCO content and after hydrolysis (e.g., MEM10 Tg ~170 °C), indicating stronger intermolecular interactions.
• Screening vs blank: Presence of NH3+‑bearing monomer was key for sorption; increasing NH3+ percentage beyond low levels did not proportionally increase qe or RE. Considering performance and cost, MEM1 was selected for detailed study.
• Transport and kinetics: Permeation time‑lag analysis at 50 mg L−1 gave Dap on the order of 10−9 cm2 s−1 for both pesticides: BTZ 0.7±0.2×10−9 and MCP 1.3±0.2×10−9 cm2 s−1 (BTZ exhibited higher time‑lag, i.e., greater resistance). Fickian fits to sorption kinetics yielded diffusion and partition coefficients; early‑time behavior showed non‑Fickian contributions, suggesting coupled diffusional/relaxational mechanisms due to sorbate–polymer interactions.
• Isotherms: BTZ followed a Langmuir model with monolayer adsorption and film saturation (reported capacity ~18 mg g−1), indicating ordered sorption. MCP followed a BET multilayer model with monolayer capacity qm ~3.9 mg g−1 and progression to multilayer sorption; a maximum solution sorption level of ~112.6 mg L−1 was noted, consistent with stronger sorbate–sorbate interactions and non‑Fickian kinetics.
• Removal efficiency: RE increased as initial concentration decreased, reaching >90% for BTZ and >99% for MCP at low concentrations (e.g., 10 mg L−1).
• Molecular interactions: 1:1 MCP–POL10 complex by Job’s plot; equilibrium constant K = 57±4 dm3 mol−1 from 1H‑NMR titration. MD binding enthalpies: MEM10:BTZ ΔH = −40 kJ mol−1; MEM10:MCP ΔH = −15 kJ mol−1; neutral MEM complexes were unfavorable (BTZ +441; MCP +694 kJ mol−1), highlighting the crucial role of protonated amino groups. MD and IGM analyses indicated dominant hydrophobic/dispersion interactions with occasional H‑bonding (N–H…N for BTZ), and polymer adopting a “scorpion‑like” conformation enveloping the pesticide; water solvation shells modulated direct charge–charge contacts.
• Selectivity and interference: In equimolar BTZ/MCP mixtures, MEM10 favored BTZ (RE ~72%) over MCP (~27%). Interferents: BTZ removal was robust and even enhanced with urea (to ~97% at BTZ:urea 1:2), while MCP RE decreased by ~17% (urea) and ~42% (NH4NO3).
• Reusability: Membranes were reusable for at least 4 cycles. Water washing released ~50% of loaded pesticides and for MCP led to a ~23% performance increase in subsequent cycles (attributed to pH effects), whereas 0.1 M NaOH washing enabled full desorption and stable performance across cycles.
• Practical attributes: Membranes are easy to handle, resistant, and scalable; safe operation without PPE is feasible due to mechanical robustness.

The study addresses the need for practical, reusable materials to remove mobile, water‑soluble herbicides at low concentrations. By incorporating protonated aminoethyl side groups into a VP/MMA polyacrylic matrix, the membranes exhibit strong affinity toward MCP and BTZ, as shown by high partitioning, favorable binding enthalpies (for MEM10 complexes), and high removal efficiencies that improve at lower concentrations. Transport and kinetic analyses confirm that, while average diffusion is comparable for both pesticides, BTZ experiences greater membrane resistance and follows monolayer Langmuir sorption, whereas MCP exhibits multilayer BET behavior and non‑Fickian kinetics, reflecting different interaction regimes. MD simulations and IGM analysis revealed that hydrophobic dispersion interactions dominate, with the polymer reorganizing into a scorpion‑like conformation around the pesticide and occasional specific H‑bonding (notably for BTZ). The charged amino groups enhance binding energetics and overall affinity, consistent with FTIR/NMR evidence. The membranes perform robustly in the presence of common nitrogen fertilizers, preferentially remove BTZ in competitive mixtures, and can be fully regenerated with mild base, demonstrating realistic reusability. Compared with prior adsorbents and photocatalysts, these membranes offer operational simplicity (solid, retrievable format), safety, and sustained performance at low target concentrations, advancing practical remediation strategies for pesticide‑contaminated waters.

Reusable polyacrylic membranes made from commercial monomers and bearing protonated aminoethyl side groups efficiently extract mecoprop and bentazon from water, with removal efficiencies exceeding 95% and improving at lower concentrations. Comprehensive physicochemical, transport, and sorption analyses, supported by NMR and molecular simulations, elucidate a primarily hydrophobic interaction mechanism and polymer conformational adaptation (scorpion‑like) that stabilize pesticide binding; charged amino groups are pivotal to performance. The membranes are mechanically robust, thermally stable, safe to handle, and reusable for at least four cycles, with complete regeneration using dilute base. These attributes, combined with scalability, position the materials as promising tools for water remediation. Future work could assess performance in real water matrices with complex ion/organic backgrounds, extend durability testing to more cycles and environmental conditions, optimize composition for broader contaminant spectra, and scale device designs for field deployment.

• Tests were conducted in Milli‑Q water to avoid salt interactions; performance in natural waters with complex ionic and organic backgrounds remains to be demonstrated.
• Reusability was validated for four cycles; longer‑term durability and fouling resistance were not assessed.
• Interference studies were limited to urea and ammonium nitrate; broader interferent panels and pH/ionic strength effects warrant evaluation.
• NMR interaction analysis was successful for MCP but inconclusive for BTZ in the chosen conditions.
• BET surface area analyses were only conclusive for MEM10; low‑NCO formulations yielded insufficient signal for definitive textural analysis.
• Experiments focused on two herbicides; generalizability to other pesticides needs validation.