Microplastics and nanoplastics barely enhance contaminant mobility in agricultural soils

The study addresses whether micro- and nanoplastics (MNP) facilitate the vertical transport of organic contaminants in agricultural soils to depths that could threaten groundwater. MNP enter soils through multiple agricultural pathways (biosolids, compost, mulching films) and can contain additives or sorbed contaminants. Concerns exist that MNP might act as vectors for otherwise immobile hydrophobic organic compounds. The research evaluates under what environmental and physicochemical conditions MNP-bound contaminants could be transported, focusing on particle mobility, contaminant properties, and desorption kinetics relative to transport time, using the Damköhler number framework to distinguish equilibrium versus decoupled transport.

Background work shows pervasive plastic inputs to soils and potential contaminant associations with MNP, including PAHs, PCBs, pesticides, and additives. Prior studies on colloid-facilitated transport and nanoparticle mobility provide transferable concepts for MNP. Particle mobility in soils depends on flow, geochemistry, and particle-soil interactions, with soils generally acting as effective filters. Desorption from plastics is governed by intra-particle diffusion (IPD) and aqueous boundary layer diffusion (ABLD), with ABLD typically rate-limiting in field conditions. The Damköhler number (Da) has been used to assess particle-facilitated transport of contaminants. Knowledge gaps include a lack of comprehensive evaluation of MNP as contaminant vectors in realistic soil settings, including the influence of flow regimes and polymer–contaminant combinations.

The authors used a modeling approach based on the Damköhler number (Da = λτ) to compare contaminant desorption times with MNP transport times over 1 m of soil depth. Two limiting desorption cases were considered: (1) IPD-limited desorption, characterized by apparent diffusion coefficient Dapp within spherical polymer particles (100 nm–1 mm diameter), and (2) ABLD-limited desorption, characterized by contaminant partitioning between polymer and water (Kpw) and diffusion across an aqueous boundary layer (ABL). Assumptions included spherical particles, high particle mobility with no particle–soil attachment/straining, and homogeneous soils. Two flow regimes were evaluated: slow matrix flow (1 m per year) and fast preferential flow (1 m per hour). For ABLD, laminar flow was confirmed via Reynolds numbers for grain diameters relevant to matrix and macropores; ABL thicknesses δ were set to 300 μm (slow flow) and 100 μm (fast flow). Aqueous diffusion coefficients were estimated with Worch’s equation, adopting D = 5×10⁻¹⁰ m² s⁻¹ (pyrene at 15 °C) for calculations, with sensitivity bounds 2.6×10⁻¹⁰–1.2×10⁻⁹ m² s⁻¹ producing <7% variability in logKpw thresholds. Dapp ranges (logDapp −30 to −8) and logKpw ranges (0–12) were scanned to delineate Da thresholds for decoupled (<0.01–0.1) versus equilibrium (>10–100) transport across particle sizes. Literature-reported values for Dapp (from film-stacking method) and Kpw for multiple polymer–contaminant pairs (PE, PS, PVC, PA, PDMS, tire materials vs PAHs, PCBs, pesticides, additives, pharmaceuticals) were overlaid to identify realistic scenarios enabling MNP-facilitated transport. Transport across a 1 m soil profile was assessed for each regime to evaluate potential for contaminants to remain particle-bound during transit.

• Under slow matrix flow (1 m a⁻¹):
  • IPD-limited: MNP ≥1 mm could facilitate transport only if logDapp < −17. Reported logDapp values are > −16 for all combinations; thus, NPs and MPs <100 μm reach equilibrium within 1 year. For 100 μm–1 mm MPs, some combinations (e.g., additives in PE/PS; PCBs in PVC) with logDapp ≈ −14 to −16 fall into kinetic regime; PDMS and tire materials exhibit high logDapp and are negligible vectors.
  • ABLD-limited: MPs <100 μm do not facilitate transport for contaminants with logKpw < 5.5 (most compounds). Transport of PAHs, PCBs, and a few pesticides by PE/PDMS MPs <100 μm is kinetic but not decoupled. MPs >1 mm only decouple for highly hydrophobic PCBs with logKpw > 6.5. Overall, MNP rarely enhance mobility under slow flow, with few exceptions for large MPs.
• Under fast preferential flow (1 m h⁻¹):
  • IPD-limited: NPs <1 μm still do not enhance mobility (logDapp values > −16). MPs <100 μm require logDapp ≤ −12 to facilitate transport; many PE/PS/PVC with PAHs, PCBs, pesticides, and additives (logDapp −15 to −12) are kinetic; decoupled transport appears for larger (>100 μm) PE/PS/PVC MPs with contaminants where logDapp < −13.
  • ABLD-limited: NPs <1 μm decouple only for logKpw > 7 (primarily certain PCBs). Most contaminants (logKpw < 4) remain equilibrium with NPs. MPs <10 μm reach equilibrium within 1 h if logKpw < 3.5 (typical for pharmaceuticals, many pesticides, additives). MPs >100 μm can facilitate transport for logKpw > 4 (e.g., PAHs with logKow > 5, hydrophobic PCBs, certain organochlorine pesticides). For MPs >1 mm, decoupled transport is possible for logKpw > 3 across many polymers and contaminant classes except less hydrophobic pharmaceuticals.
• Polymer dependence: Glass-like polymers (PE, PS, PVC) more prone to vector behavior than rubber-like PDMS and tire materials (styrene-butadiene rubber), which show faster diffusion or strong sorption but still seldom achieve decoupling under realistic flows.
• Overall: MNP rarely enhance contaminant mobility in agricultural soils; potential relevance is limited to preferential flow conditions, larger microplastics, and very hydrophobic contaminants (logKow > 5).

The findings indicate that contaminant desorption from MNP is generally too fast relative to soil transport times for nanoplastics and small microplastics, leading to equilibrium with soil phases within the rooting zone. Only under preferential flow, when travel times are very short and particle retention minimal, can larger microplastics act as vectors for highly hydrophobic, slowly desorbing contaminants. However, real soils strongly filter and retain particles via attachment, straining, and heteroaggregation (e.g., with EPS and minerals), further reducing MNP mobility compared to the high-mobility assumption used in the model. Irregular shapes and fibers experience enhanced straining and lower mobility; geometry can also affect IPD times. Environmental aging (e.g., UV embrittlement) can alter polymer porosity and sorption characteristics; biofilms may increase apparent desorption times, but still typically not enough to change overall conclusions. Moreover, natural mobile colloids (organic matter, clays) are abundant and often more influential vectors than plastics; soil organic matter has higher sorption coefficients than most plastics, limiting contaminant uptake by MNP. Consequently, the scenarios in which MNP contribute substantially to vertical contaminant transport beyond the root zone are rare and require a convergence of specific conditions: preferential flow paths, mobile large MPs, glassy polymers, and very hydrophobic contaminants.

The study systematically delineates when MNP could act as vectors for organic contaminants in agricultural soils using a Damköhler number framework. It shows that nanoplastics and small microplastics rapidly equilibrate with soil, and MNP-facilitated contaminant transport is generally negligible under typical matrix flow. Only larger microplastics might add to the transport of very hydrophobic contaminants (logKow > 5) under preferential flow, and even then particle retention often limits mobility. Thus, MNP are unlikely to enhance vertical contaminant mobility sufficiently to threaten groundwater. Management should prioritize reducing plastic inputs (complete recovery of mulching films, regulation of plastics in biosolids/compost) and focus research on risks within the root zone, including MNP-derived additives, plant uptake, and impacts on soil microbiomes. Future studies should integrate particle–soil interactions, realistic particle geometries, aging, biofilms, and comparative roles of natural colloids into transport models.

• The model assumes spherical particles, high mobility, and no particle–soil interactions (no attachment/straining), likely overestimating MNP transport.
• Biological and physicochemical processes (bioturbation, EPS, heteroaggregation) and irregular particle shapes/fibers are not explicitly modeled and generally reduce mobility.
• Parameterization relies on literature Dapp and Kpw values, which are incomplete for all polymer–contaminant pairs; thresholds might shift with new data.
• Only two idealized flow regimes and a 1 m travel distance are considered; spatial heterogeneity and longer pathways are not explicitly modeled.
• ABL thickness and aqueous diffusion coefficients are estimated; although sensitivity was assessed, uncertainties remain.
• Competing natural colloids and soil organic matter that dominate sorption/transport are discussed but not mechanistically included in the Da framework.