Per- and polyfluoroalkyl substances (PFAS) as contaminants in groundwater resources - A comprehensive review of subsurface transport processes.

Per- and polyfluorinated alkyl substances (PFAS) are persistent contaminants in the environment. An increased awareness of adverse health effects related to PFAS has further led to stricter regulations for several of these substances in e.g. drinking water in many countries. Groundwater constitutes an important source of raw water for drinking water production. A thorough understanding of PFAS subsurface fate and transport mechanisms leading to contamination of groundwater resources is therefore essential for management of raw water resources. A review of scientific literature on the subject of processes affecting subsurface PFAS fate and transport was carried out. This article compiles the current knowledge of such processes, mainly focusing on perfluoroalkyl acids (PFAA), in soil- and groundwater systems. Further, a compilation of data on transport parameters such as solubility and distribution coefficients, as well as, insight gained and conclusions drawn from the reviewed material are presented. As the use of certain fire-fighting foams has been identified as the major source of groundwater contamination in many countries, research related to this type of pollution source has been given extra focus. Uptake of PFAS in biota is outside the scope of this review. The review showed a large spread in the magnitude of distribution coefficients and solubility for individual PFAS. Also, it is clear that the influence of multiple factors makes site-specific evaluation of distribution coefficients valuable. This article aims at giving the reader a comprehensive overview of the subject, and providing a base for further work.

A systematic study of the competitive sorption of per- and polyfluoroalkyl substances (PFAS) on colloidal activated carbon.

Treatment of environmental media contaminated with per- and polyfluoroalkyl substances (PFAS) is crucial to mitigate mounting health risks associated with exposure. Colloidal activated carbon (CAC) has shown promise in treating contaminated soils, but understanding the interaction among PFAS during sorption is necessary for optimal remediation. This study investigated the extent to which PFAS of varying chain lengths and functional groups compete for sorption to CAC. Batch tests were conducted with natural soil and spiked water, using CAC at 0.2% w/w to remove seven PFAS with individual starting concentrations up to 0.05 mmol L-1. PFAS sorption to CAC was evaluated in three systems: a composite mixture of all studied compounds, a binary-solute system, and a single-solute system. The sorption experiments exhibited strong PFAS affinity to CAC, with removal rates between 41% and 100%, and solid/liquid partition coefficients (Kd) between 10 and 104 L kg-1. Differences were noticed among the various spiking mixtures, based on perfluorocarbon chain length, functional group, and the starting PFAS concentrations. Competition effects were detected when PFAS were in a multi-solute system, with an average 10% drop in removal, which can evidently become more relevant at higher concentrations, due to the observed non-linearity of the sorption process. The PFAS most vulnerable to competition effects in multi-solute systems were the short-chain perfluoropentanoic acid (PFPeA) and perfluorobutane sulfonic acid (PFBS), with an up to 25% reduction in removal. In bi-solute systems, perfluorooctane sulfonamide (FOSA) dominated over its ionisable counterparts, i.e. perfluorooctane sulfonic acid (PFOS) and perfluorononanoic acid (PFNA), indicating the importance of hydrophobic effects or layer formation in the sorption process. These results underscore the importance of considering competition in PFAS sorption processes when designing and implementing remediation techniques for PFAS-contaminated media.

Pore-Scale Mechanisms of Solid Phase Emergence During DNAPL Remediation by Chemical Oxidation.

In situ chemical oxidation (ISCO) has proven successful in the remediation of aquifers contaminated with dense nonaqueous phase liquids (DNAPLs). However, the treatment efficiency can often be hampered by the formation of solids or gas, reducing the contact between remediation agents and residual DNAPLs. To further improve the efficiency of ISCO, fundamental knowledge is needed about the complex multiphase flow and reactive transport processes as new solid and fluid phases emerge at the microscale. Here, via microfluidic experiments, we study the pore-scale dynamics of trichloroethylene degradation by permanganate. We visualize how the remediation evolves under the influence of solid phase emergence and explore the roles of injection rate, oxidant concentration, and stabilization supplement. Combining image processing, pressure analysis, and stoichiometry calculations, we provide comprehensive descriptions of the oxidant concentration-dependent growth patterns of the solid phase and their impact on the remediation efficiency. We further corroborate the stabilization mechanism provided by phosphate supplement, which is effective in inhibiting solid phase generation and thus highly beneficial for the oxidation remediation. This work elucidates the pore-scale mechanisms during remediation of chlorinated solvents with a particular context in the solid phase production and the associated effects, which is of general significance to understanding various processes in natural and engineered systems involving solid phase emergence or aggregation phenomena, such as groundwater and soil remediation.

Per- and polyfluoroalkyl substance (PFAS) retention by colloidal activated carbon (CAC) using dynamic column experiments.

Developing effective remediation methods for per- and polyfluoroalkyl substance (PFAS)-contaminated soils is a substantial step towards counteracting their widespread occurrence and protecting our ecosystems and drinking water sources. Stabilisation of PFAS in the subsurface using colloidal activated carbon (CAC) is an innovative, yet promising technique, requiring better understanding. In this study, dynamic soil column tests were used to assess the retardation of 10 classical perfluoroalkyl acids (PFAAs) (C5-C11 perfluoroalkyl carboxylic acids (PFCAs) and C4, C6, C8 perfluoroalkane sulfonates (PFSAs)) as well as two alternative PFAS (6:2 and 8:2 fluorotelomer sulfonates) using CAC at 0.03% w/w, to investigate the fate and transport of PFAS under CAC treatment applications. Results showed high retardation rates for long-chain PFAS and eight times higher retardation for the CAC-treated soil compared to the non-treated reference soil for the ∑PFAS. Replacement of shorter chain perfluorocarboxylic acids (PFCAs), such as perfluoropentanoic acid (PFPeA), by longer chained PFAS was observed, indicating competition effects. Partitioning coefficients (Kd values) were calculated for the CAC fraction at ∼103-105 L kg-1 for individual PFAS, while there was a significant positive correlation (p < 0.05) between perfluorocarbon chain length and Kd. Mass balance calculations showed 37% retention of ∑PFAS in treated soil columns after completion of the experiments and 99.7% higher retention rates than the reference soil. Redistribution and elution of CAC were noticed and quantified through organic carbon analysis, which showed a 23% loss of carbon during the experiments. These findings are a step towards better understanding the extent of CAC's potential for remediation of PFAS-contaminated soil and groundwater and the limitations of its applications.

Electrokinetic remediation for removal of per- and polyfluoroalkyl substances (PFASs) from contaminated soil.

Uncontrolled use and disposal of per- and polyfluoroalkyl substances (PFASs) in recent decades has resulted in extensive soil and groundwater contamination, necessitating counteraction. Electrokinetic remediation (EKR) offers a promising approach to in-situ soil remediation. Two novel modifications to conventional EKR were tested for the first time in a laboratory-scale study, to explore the capacity of EKR for PFAS removal. The first modification was a two-compartment setup designed for PFAS extraction from soil to an electrolyte-filled chamber. The second was a single-compartment setup designed to transport and confine contaminants in a chamber filled with granular activated carbon (GAC), thus, combining extraction with stabilisation. Electromigration varied for individual compounds, based mainly on perfluorocarbon chain length and functional group. The results indicated up to 89% concentration and extraction of ∑PFASs for the two-compartment setup, with removal efficiency reaching 99% for individual PFASs with C ≤ 6. Removed PFASs were concentrated adjacent to the anode at the anion exchange membrane, while short-chain compounds were extracted in the anolyte. The single-compartment setup achieved 75% extraction and accumulation of ∑PFASs in GAC. This demonstrates, for the first time, good effectiveness of coupling EKR with AC stabilisation for PFAS removal from soil. Perfluorocarbon chain length was a dominant factor affecting treatment efficiency in both setups, with very high removal rates for short-chain PFASs.

Modified MODFLOW-based model for simulating the agglomeration and transport of polymer-modified Fe0 nanoparticles in saturated porous media.

The solute transport model MODFLOW has become a standard tool in risk assessment and remediation design. However, particle transport models that take into account both particle agglomeration and deposition phenomena are far less developed. The main objective of the present study was to evaluate the feasibility of adapting the standard code MODFLOW/MT3D to simulate the agglomeration and transport of three different types of polymer-modified nanoscale zerovalent iron (NZVI) in one-dimensional (1-D) and two-dimensional (2-D) saturated porous media. A first-order decay of the particle population was used to account for the agglomeration of particles. An iterative technique was used to optimize the model parameters. The model provided good matches to 1-D NZVI-breakthrough data sets, with R 2 values ranging from 0.96 to 0.99, and mass recovery differences between the experimental results and simulations ranged from 0.1 to 1.8 %. Similarly, simulations of NZVI transport in the heterogeneous 2-D model demonstrated that the model can be applied to more complicated heterogeneous domains. However, the fits were less good, with the R 2 values in the 2-D modeling cases ranging from 0.75 to 0.95, while the mass recovery differences ranged from 0.7 to 6.5 %. Nevertheless, the predicted NZVI concentration contours during transport were in good agreement with the 2-D experimental observations. The model provides insights into NZVI transport in porous media by mathematically decoupling agglomeration, attachment, and detachment, and it illustrates the importance of each phenomenon in various situations. Graphical Abstract ᅟ.

Spectral induced polarization of Na-montmorillonite dispersions.

Montmorillonite (Mt) clays have a high specific surface area and surface charge, which confer them remarkable adsorption properties. Nevertheless, their electrochemical and aggregation behavior are not completely elucidated because of the complexity of their microstructural and interfacial properties. In this work, the conductive and dispersive properties of Na-Mt suspensions of weight fractions 0.5-5.2% were investigated for the first time using the spectral induced polarization method. A four-electrode system was used to reduce errors introduced by electrode polarization and contact resistances. Complex conductivity spectra in the low-frequency range of 0.1Hz to 45kHz were successfully described using a triple layer model of the basal surface of Mt and a complex conductivity model that considers conduction of the diffuse layer and polarization of the Stern layer. Aggregate size distributions were inferred from inverted relaxation time distributions. We found that the negative and permanent surface charge of the basal plane of Na-Mt controls its quadrature (imaginary) conductivity, which is not very sensitive to pH and salinity (NaCl) in the 100Hz to 45kHz frequency range. For lower frequencies, the sudden increase of the quadrature conductivity at the highest salinities was explained by considering coagulation of Na-Mt particles.

Transport of surface-modified carbon nanotubes through a soil column.

Carbon nanotubes (CNTs) are widely manufactured nanoparticles, which are being utilized in a number of consumer products, such as sporting goods, electronics and biomedical applications. Due to their accelerating production and use, CNTs constitute a potential environmental risk if they are released to soil and groundwater systems. It is therefore essential to improve the current understanding of environmental fate and transport of CNTs. The transport and retention of CNTs in both natural and artificial media have been reported in literature, but the findings widely vary and are thus not conclusive. There are a number of physical and chemical parameters responsible for variation in retention and transport. In this study, a complete procedure of selected multiwalled carbon nanotubes (MWCNTs) is presented starting from their surface modification to a complete set of laboratory column experiments at critical physical and chemical scenarios. Results indicate that the stability of the commercially available MWCNTs are critical with their attached surface functional group which can also influence the transport and retention of MWCNT through the surrounding medium.

Dissolution of dense non-aqueous phase liquids in vertical fractures: effect of finger residuals and dead-end pools.

Understanding the dissolution behavior of dense non-aqueous phase liquids (DNAPLs) in rock fractures under different entrapment conditions is important for remediation activities and any related predictive modeling. This study investigates DNAPL dissolution in variable aperture fractures under two important entrapment configurations, namely, entrapped residual blobs from gravity fingering and pooling in a dead-end fracture. We performed a physical dissolution experiment of residual DNAPL blobs in a vertical analog fracture using light transmission techniques. A high-resolution mechanistic (physically-based) numerical model has been developed which is shown to excellently reproduce the experimentally observed DNAPL dissolution. We subsequently applied the model to simulate dissolution of the residual blobs under different water flushing velocities. The simulated relationship between the Sherwood number Sh and Peclet number Pe could be well fitted with a simple power-law function (Sh=1.43Pe°·4³). To investigate mass transfer from dead-end pools, another type of trapping in rock fractures, entrapment and dissolution of DNAPL in a vertical dead-end fracture was simulated. As the entrapped pool dissolves, the depth of the interface between the DNAPL and the flowing water increases linearly with decreasing DNAPL saturation. The interfacial area remains more or less constant as DNAPL saturation decreases, unlike in the case of residual DNAPL blobs. The decreasing depth of the contact interface changes the flow field and causes decreasing water flow velocity above the top of the DNAPL pool, suggesting the dependence of the mass transfer rate on the depth of the interface, or alternatively, the remaining mass percentage in the fracture. Simulation results show that the resultant Sherwood number Sh is significantly smaller than in the case of residual blobs for any given Peclet number, indicating slower mass transfer. The results also show that the Sh can be well fitted with a power-law function of Pe and remaining mass percentage. The obtained relationships of dimensionless groups concerning the mass transfer characteristics at the level of individual fractures can be further used in predictive modeling of dissolution at a larger (fracture network) scale.

Effects of single-fracture aperture statistics on entrapment, dissolution and source depletion behavior of dense non-aqueous phase liquids.

Understanding of the entrapment and dissolution behavior of dense non-aqueous phase liquids (DNAPLs) in single fractures is important for modeling contaminant flux generation from fractured sites. Here a systematic numerical study is presented to investigate the effect of fracture aperture statistics on DNAPL migration, entrapment and dissolution within individual, variable-aperture fractures. Both fractures with open and closed bottom boundaries were considered. For the simulation a continuum-based two-phase model was used with a capillary pressure function which calculates the entry pressure based on the local aperture. Prior to application the model was compared against the invasion percolation approach and found more suitable for the present study, in particular as it allows a more versatile presentation of boundary conditions. The results showed that increasing aperture standard deviation and/or decreasing correlation length lead to larger amounts of entrapped DNAPL (due to the fact that larger standard deviation produces more distinct contrast between small and large aperture regions and the fact that longer correlation length provides more possible channels through the fracture) as well as larger maximum and average sizes of DNAPL blobs, and subsequently lead to longer times for complete dissolution. To understand the relationship between the solute flux and the remaining mass, a simplified source depletion function which links the outflow concentration to the DNAPL saturation was found adequate to describe the dissolution process for the case where the bottom boundary is open for DNAPL migration and thus the DNAPL does not accumulate to form a pool. The parameters in this function were not very sensitive to variations in correlation length but were sensitive to aperture standard deviation. The same average entrapped DNAPL saturation produced considerably smaller solute concentrations in cases with larger aperture variability due to the larger average size of DNAPL blobs (i.e., smaller contact area for DNAPL dissolution). Boundary conditions had a significant impact on DNAPL entrapment and dissolution. A closed boundary at the bottom led to DNAPL pooling (i.e., large continuous blobs) which causes significant tailing in the dissolution breakthrough curve due to water bypassing.

Polymer-modified Fe0 nanoparticles target entrapped NAPL in two dimensional porous media: effect of particle concentration, NAPL saturation, and injection strategy.

Polymer-modified nanoscale zerovalent iron (NZVI) particles are delivered into porous media for in situ remediation of nonaqueous phase liquid (NAPL) source zones. A systematic and quantitative evaluation of NAPL targeting by polymer-modified NZVI in two-dimensional (2-D) porous media under field-relevant conditions has not been reported. This work evaluated the importance of NZVI particle concentration, NAPL saturation, and injection strategy on the ability of polymer-modified NZVI (MRNIP2) to target the NAPL/water interface in situ in a 2-D porous media model. Dodecane was used as a NAPL model compound for this first demonstration of source zone targeting in 2-D. A driving force for NAPL targeting, the surface activity of MRNIP2 at the NAPL/water interface was verified ex situ by its ability to emulsify NAPL in water. MRNIP2 at low particle concentration (0.5 g/L) did not accumulate in or near entrapped NAPL, however, MRNIP2 at moderate and high particle concentrations (3 and 15 g/L) did accumulate preferentially at entrapped NAPL, i.e., it was capable of in situ targeting. The amount of MRNIP2 that targets a NAPL source depends on NAPL saturation (S(n)), presumably because the saturation controls the available NAPL/water interfacial area and the flow field through the NAPL source. At effective S(n) close or equal to 100%, MRNIP2 bypassed NAPL and accumulated only at the periphery of the entrapped NAPL region. At lower S(n), flow also carries MRNIP2 to NAPL/water interfaces internal to the entrapped NAPL region. However, the mass of accumulated MRNIP2 per unit available NAPL/water interfacial area is relatively constant (∼0.8 g/m(2) for MRNIP2 = 3 g/L) from S(n) = 13 to ∼100%, suggesting that NAPL targeting is mostly controlled by MRNIP2 sorption onto the NAPL/water interface.

Empirical correlations to estimate agglomerate size and deposition during injection of a polyelectrolyte-modified Fe0 nanoparticle at high particle concentration in saturated sand.

Controlled emplacement of polyelectrolyte-modified nanoscale zerovalent iron (NZVI) particles at high particle concentration (1-10 g/L) is needed for effective in situ subsurface remediation using NZVI. Deep bed filtration theory cannot be used to estimate the transport and deposition of concentrated polyelectrolyte-modified NZVI dispersions (>0.03 g/L) because particles agglomerate during transport which violates a fundamental assumption of the theory. Here we develop two empirical correlations for estimating the deposition and transport of concentrated polyelectrolyte-modified NZVI dispersions in saturated porous media when NZVI agglomeration in porous media is assumed to reach steady state quickly. The first correlation determines the apparent stable agglomerate size formed during NZVI transport in porous media for a fixed hydrogeochemical condition. The second correlation estimates the attachment efficiency (sticking coefficient) of the stable agglomerates. Both correlations are described using dimensionless numbers derived from parameters affecting deposition and agglomeration in porous media. The exponents for the dimensionless numbers are determined from statistical analysis of breakthrough data for polyelectrolyte-modified NZVI dispersions collected in laboratory scale column experiments for a range of ionic strength (1, 10, and 50mM Na(+) and 0.25, 1, and 1.25 mM Ca(2+)), approach velocity (0.8 to 55 × 10(-4)m/s), average collector sizes (d(50)=99 µm, 300 µm, and 880 µm), and polyelectrolyte surface modifier properties. Attachment efficiency depended on approach velocity and was inversely related to collector size, which is contrary to that predicted from classic filtration models. High ionic strength, the presence of divalent cations, lower extended adsorbed polyelectrolyte layer thickness, decreased approach velocity, and a larger collector size promoted NZVI agglomeration and deposition and thus limited its mobility in porous media. These effects are captured quantitatively in the two correlations developed. The application and limitations of using the correlations for preliminary design of in situ NZVI emplacement strategies is discussed.

Particle size distribution, concentration, and magnetic attraction affect transport of polymer-modified Fe(0) nanoparticles in sand columns.

The effect of particle concentration, size distribution (polydispersity) and magnetic attractive forces (Fe(0) content) on agglomeration and transport of poly(styrene sulfonate) (PSS) modified NZVI was studied in water-saturated sand (d(p) = 300 microm) columns. Particle concentrations ranged from 0.03 to 6 g/L in 5 mM NaCl/5 mM NaHCO3 at a pore water velocity of 3.2 x 10(-4) m/s. Three NZVI dispersions with different intrinsic particle size distributions obtained from sequential sedimentation are compared. The influence of magnetic attraction (Fe(0) content) on NZVI agglomeration and deposition in porous media is assessed by comparing the deposition behavior of PSS-modified NZVI (magnetic) having different Fe(0) contents with PSS-modified hematite (nonmagnetic) with the same surface modifier. At low particle concentration (30 mg/L) all particles were mobile in sand columns regardless of size or magnetic attractive forces. At high concentration (1 to 6 g/L), deposition of the relatively monodisperse dispersion containing PSS-modified NZVI (hydrodynamic radius (R(H)) = 24 nm) with the lowest Fe(0) content (4 wt%) is low (attachment efficiency (alpha) = 2.5 x 10(-3)), insensitive to particle concentration, and similar to PSS-modified hematite. At 1 to 6 g/L, the attachment efficiency of polydisperse dispersions containing both primary particles and sintered aggregates (R(H) from 15 to 260 nm) of PSS-modified NZVI with a range of Fe(0) content (10-60%) is greater (alpha = 1.2 x 10(-2) to 7.2 x 10(-2) and is sensitive to particle size distribution. The greater attachment for larger, more polydisperse Fe(0) nanoparticles with higher Fe(0) content is a result of their agglomeration during transport in porous media because the magnetic attractive force between particles increases with the sixth power of particle/agglomerate radius. A filtration model that considers agglomeration in porous media and subsequent deposition explains the observed transport of polydisperse PSS-modified NZVI at high concentration.