Enhanced adsorption of perfluoro alkyl substances for in situ remediation

Yousof H. Aly ab, Daniel P. McInnis b, Samuel M. Lombardo c, William A. Arnold ac, Kurt D. Pennell d, James Hatton e and Matt F. Simcik *ab
aWater Resources Sciences Program, University of Minnesota, 1985 Buford Ave, St. Paul, MN 55108, USA. E-mail: msimcik@umn.edu
bSchool of Public Health Division of Environmental Health Sciences, University of Minnesota, 420 Delaware St., Minneapolis, MN 55455, USA
cDepartment of Civil, Environmental, and Geo- Engineering, University of Minnesota, 500 Pillsbury Dr. SE, Minneapolis, MN 55455, USA
dSchool of Environmental Engineering, Brown University, Providence, RI 02912, USA
eJacobs Engineering Group, 9191 South Jamaica Street, Englewood, CO 80112, USA

Received 21st May 2019 , Accepted 27th August 2019

First published on 30th August 2019


Numerous groundwater sites around the globe have been contaminated by aqueous film forming foam (AFFF) as a result of firefighting, fire training activities and the storage and accidental spillage of AFFF. AFFF contains numerous per- and polyfluoroalkyl substances (PFAS), including perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS) for which the U.S. Environmental Protection Agency has set a combined health advisory level of 70 ng L−1 in drinking water. One approach for in situ remediation of PFAS-impacted groundwater is to enhance the adsorption of PFAS by introducing adsorbents, which serve to increase the total adsorption capacity of aquifer solids. This paper describes a method for increasing the sorption of PFAS to aquifer solids by adding commercially available drinking water coagulants. These coagulants are the cationic polymers polydiallyldimethyl ammonium chloride (polyDADMAC) and polyamine (a co-polymer of epichlorohydrine and dimethyl amine). The six PFAS on the Unregulated Contaminant Monitoring Rule 3 (UCMR3) list were studied including perfluorobutane sulfonate (PFBS), perfluorohexane sulfonate (PFHxS), PFOS, perfluoorbutanoic acid (PFBA), PFOA and perfluorononanoic acid (PFNA). Treatability tests were conducted with natural soil excavated from an Air Force Base located in the south central United States. In completely mixed batch reactors studies, the coagulants increased the sorption capacity of the soil for PFAS by a factor of 2.0–6.1. One-dimensional columns pre-loaded with coagulant delayed the breakthrough of PFAS by as much as 20 pore volumes with an applied PFAS concentration of 100 μg L−1. Following loading with PFAS and coagulant the columns were then flushed with PFAS-free simulated groundwater to assess desorption behavior. For all PFAS the retention on the column showed hysteresis where only 1 to 20% of the PFAS was recovered from the column after flushing with 30 pore volumes of simulated groundwater. The observed increase in PFAS adsorption could not be accounted by solely by the increase in organic carbon content resulting from the addition of adsorption enhancers, suggesting that exchange interactions contributed to PFAS retention. These results indicate that amending soils and aquifer solids with cationic polymers acting as sorption enhancers holds promise as a viable method for in situ PFAS sequestration.



Water impact

Numerous groundwater resources are contaminated with per- and polyfluoroalkyl substances (PFAS) as a result of firefighting activities and industrial processes. PFAS are incredibly persistent, which makes destroying them in situ extremely difficult. An alternative to destruction is sequestration. This paper presents a method for sequestration by increasing adsorption through the introduction of commercially available drinking water coagulants.

Introduction

Per- and polyfluoro-alkyl substances (PFAS) have received increasing attention in recent years due to their capacity for bioaccumulation, long-range transport, toxicity and persistence in the environment.1–4 Aqueous film forming foam (AFFF), used in fighting fires and fire training activities, contains many identified and unidentified PFAS, some of which are precursors.5 These chemicals are often introduced to groundwater after the use of AFFF at a given site.6–9 Their relatively high aqueous solubilities and low sorption affinity results in transport from the site of AFFF application, which may lead to contamination of nearby water resources, and ultimately to human exposure.2,7 Recently, there has been substantial effort to develop in situ remediation technologies for PFAS-impacted groundwater as an alternative to pump and treat methods. In situ chemical oxidation (ISCO) has been investigated for PFAS remediation, but has been ineffective for sulfonates at mg L−1 levels and can require high temperatures to be activated.10 Laboratory-scale investigations of PFAS degradation using chemical oxidants have yielded varying results. For instance, a laboratory scale column study found that that the uses of three common ISCO reagents (permanganate, activated persulfate and catalyzed hydrogen peroxide) were not able to degrade perfluoroalkyl acids (PFAAs).11 Other laboratory-scale studies have observed that the persulfate radical was capable of perfluorooctanoic acid (PFOA) degradation but was not effective for perfluorooctane sulfonate (PFOS) degradation.12 A recent pilot-scale field study showed moderate (∼40%) PFAS reductions in an aquifer after treatment with persulfate, ozone and hydrogen peroxide.13

Conventional adsorbents, including activated carbon, ion exchange polymers and resins, have been studied extensively for ex situ treatment of PFAS-impacted waters. These sorbents are effective at removing PFAS from water over a range of concentrations.4,10 Despite their effectiveness on longer chain length PFAS, these sorbents are expensive and are difficult to implement in situ, and do not retain shorter chain length PFAS to the same extent as they retain the longer chain length PFAS.14 Organic carbon (OC) content, electrostatic interaction and pH have long been understood as factors influencing PFAS adsorption.15–17 A previous study by Xiao et al.18 showed that PFOA and PFOS are susceptible to removal by adsorption to flocs formed by Alum coagulation. Previous work from our group has shown the addition of coagulants to be effective on Ottawa sand, increasing sorption capacities by factors ranging from 4–45.19 This paper presents a new in situ treatment methodology that involves the injection of soluble cationic polymers that act to sequester PFAS in the aquifer, thereby retarding further migration of the plume. Cationic drinking water coagulants were selected to span a range of molecular weights. They included polyaluminum chloride, polyamine (a co-polymer of epichlorohydrine and dimethyl amine), and polydiallyldimethyl ammonium chloride (polyDADMAC). The latter two are polymers varying in size with polyamine averaging 750 kDa and polyDADMAC being larger averaging 2–3 MDa. Earlier work with polyaluminum chloride did not result in increases in adsorption (unpublished), indicating that the size of the coagulant is important. The central hypothesis of this paper is that amending aquifer material with cationic polymer coagulants will enhance PFAS sorption to the aquifer material by increasing the OC content and positive charge of solid phase. Previous work on Ottawa sand indicated positive results, but this paper represents a more realistic scenario for remediation of an actual site. This hypothesis was tested on 6 PFAS from the USEPA's Unregulated Contaminant Monitoring Rule 3 (UCMR3) list: three carboxylates (perfluoro-heptanoate [PFHpA]; PFOA; perfluoro-nonanoate acid [PFNA]) and three sulfonates (perfluoro-butane sulfonate [PFBS]; perfluoro-hexane sulfonate [PFHxS]; PFOS). A combination of batch sorption tests and one-dimensional column studies were undertaken to determine the ability of polyDADMAC and polyamine to enhance PFAS adsorption. Long-term efficacy of this method was tested by flushing the treated columns with PFAS-free simulated groundwater and evaluating PFAS desorption.

Materials and methods

Chemicals

Polyamine and polyDADMAC were purchased from Accepta (Manchester, UK). PFOS (>98%) and PFOA (>96%) and isotopically labeled standards; 13C8-PFOA (99%), 13C8-PFOS (99%),18O2-PFHxS (99%), 13C3-PFHxS (99%), 13C4-PFOS (99%) and 13C4- PFHpA (99%) were purchased from Cambridge Isotope Laboratories (Andover, MA). PFBS (99%), PFHxS (99%), PFHpA (>98%), and PFNA (>98%) and isotopically labeled standards 13C5-PFNA (99%) and 13C9-PFNA (99%) were purchased from Wellington Laboratories (Ontario, CA). Stock solutions of all PFAS and their labeled standards were prepared in Optima Grade Methanol (Fisher Scientific, Waltham, MA) at a concentration of 0.5 and 5 mg L−1 respectively. Glass auto-sampler vials (2 mL) equipped with Viton septa (ChromTech, Apple Valley, MN) were used for PFAS analysis to minimize potential contamination.

Soil characterization

Soil was obtained from a Department of Defense Facility located in the South Central United States, located within the Central Redbed Plains of the Central Lowland Physiographic Province. The soil is silty sand with some clay derived from the Permian-aged (roughly 250 million years old) Hennessey Group,20 and is colored red by ferric anhydride. These and similar soils in the region are rich in iron and often referred to as “red beds”.21 Soil was dried at 100 °C overnight then sieved between 40–50 mesh. Polypropylene centrifuge tubes (50 mL) were used for batch tests while 2.5 × 10 cm borosilicate glass columns from Kimble Chase (Rockwood, TN) were used in column tests. Pre-existing PFAS contamination of the soil was evaluated by mixing 5 g of soil with 20 mL of Optima grade methanol for 72 hours in a centrifuge tube. After centrifugation for 15 minutes, the methanol supernatant was decanted and analyzed via high pressure liquid chromatography/mass spectrometry (HPLC/MS) for PFAS content. The PFAS concentrations were all below detection. Soil pH, ammonium acetate (NH4OAc), cation-exchange-capacity (CEC), calcium carbonate (CaCO3), nitrate as nitrogen (NO3-N), ammonium as nitrogen (NH4-N), total organic carbon (TOC), and cation concentrations (Ca2+, K+, Mg2+, Na+, Al3+) were determined by the Research Analytical Laboratory (University of Minnesota, St. Paul MN) and provided in ESI (Table S.1). Cation exchange capacity was determined by both the direct and summation method. TOC was determined by combustion analysis of a sub-sample fumigated with HCl to remove carbonates. To determine point of zero charge (PZC) of the soil, the drift method was employed. Mixtures containing 5 g of soil and 20 mL of water (1, 0.1, or 0.01 M KCl background electrolyte) were modified to the desired initial pH (2, 4, 6, 8, and 10) using HCl or NaOH. Samples were placed on a shaker table for 48 hours, and final pH values were measured with a pH probe. The final pH versus the change in pH was plotted. A quick yet reliable estimate of the soil's PZC was obtained by selecting the final pH at which the line crosses a change in pH of 0 (ESI Fig. S1).

Enhancer isotherms

To determine initial polyDADMAC and polyamine dosages to be used in batch and column tests, adsorption isotherms of the polymers on test soil were determined by adding 5 g of soil, 25 mL of simulated groundwater (10 mM NaHCO3 buffered at pH 7). The initial polyDADMAC or polyamine concentration ranged from 0 to 5000 mg L−1. After mixing for 24 hours on a wrist action shaker, the contents of the tubes were separated by centrifugation and the final polyDADMAC or polyamine concentration in the supernatant was determined using TOC analysis.

PFAS adsorption isotherms

Batch sorption tests were performed in triplicate for each of the six individual PFAS selected by adding 5 g of soil and 20 mL of simulated groundwater (10 mM NaHCO3 buffered at pH 7) into 50 mL polypropylene centrifuge tubes. PFAS were spiked into each tube to reach initial concentrations of 10, 20, 40, 50, 75, 100 μg L−1. Two additional batches were prepared by repeating the previous steps and adding polyDADMAC or polyamine at 5000 mg L−1 (based on enhancer isotherms). After 24 hours of mixing and centrifugation at 2000 rpm for 15 minutes, 1 mL of supernatant was decanted and placed into a 2 mL glass HPLC auto-sampler vial along with 100 ng of isotopically labeled internal standard for subsequent HPLC-MS analysis. To gain a further understanding of sorption behavior, additional batch experiments were performed at pH of 5, 8, and 9 for PFOS and PFOA, individually. Adsorption to the centrifuge walls was tested and determined to be negligible, confirming an earlier study with the same centrifuge tubes.19

PFAS columns

A series of one-dimensional column tests was performed to measure PFAS transport through test soil in the presence and absence of either polyDADMAC or polyamine. Glass borosilicate columns (2.5 × 10 cm, Kimble Chase, Rockwood, TN) were packed with dry soil in one cm increments and capped with glass wool and endcaps to contain the soil. The bulk density of the packed soil column was determined from the amount of solid phase added and the internal volume of the column. To achieve complete water saturation, the dry-packed columns were then flushed with CO2 gas followed by at least 3 pore volumes of 10 mM NaHCO3 solution buffered to pH 7 at a flow rate of 0.12 mL min−1. The porosity (0.43) was determined by weighing a dry packed soil column, saturating the system with water (3 pore volumes), and the weighing the column again. The difference in weight was used to determine the porosity and water-saturated pore volume of the column. The flow rate gives a linear velocity of 0.8 m per day, which simulates a fast-flowing groundwater with a relatively low retention time in the column and thus a worst-case scenario for a sorptive treatment. To amend soil with polyDADMAC or polyamine, a 5000 mg L−1 solution of was pumped through the column and effluent was continuously monitored for TOC. Once the relative effluent concentration (C/C0) reached unity, indicating saturation of the soil with the polymers, the influent reservoir was switched to a solution containing 100 μg L−1 (ppb) of an individual PFAS. Approximately 7.2 mL of effluent were collected in 10 mL glass test tubes every hour on a rotating fraction collector. After 100% breakthrough of PFAS was recorded, the influent reservoir was switched back to simulated groundwater to examine the reversibility of this process, and effluent samples were collected for an additional 30 pore volumes.

Analytical methods

Polymer concentrations were determined with a Shimadzu TOC-L analyzer. Potassium hydrogen phthalate (KHP) was used as a calibration standard for non-purgeable OC analysis. Aqueous samples were injected into a 680° catalytic oven in the presence of a platinum catalyst, oxidized to CO2, and measured by non-dispersive infrared detection. Aqueous PFAS concentrations were measured using a Hewlett-Packard series 1050 HPLC paired to a series 1100 single quadrapole MS. A Betasil C18 (50 mm × 2.1 mm × 10 μm) analytical column (Thermo-Scientific, Waltham MA) was used to achieve analyte separation. The mobile phases, delivered at a flow rate of 0.20 mL min−1, consisted of A: 2.0 mM ammonium acetate in a mixture of 90% water (from a reverse osmosis system) and 10% methanol (Optima Grade, Fisher Scientific); and B: 2.0 mM ammonium acetate in methanol (Optima Grade, Fisher Scientific). The A/B ratio was ramped linearly from 78/22 to 33/67 in the first 3 minutes, maintained for 1 minute, then changed linearly from 33/67 to 0/100 over 4 minutes, and held for an additional 7 minutes. Mass to charge ratios (m/z) for UCMR3 analytes are; PFBS: 299, PFHpA: 363, PFHxS: 399, PFOA: 413, PFNA: 463, PFOS: 499. Masses were calculated using as single point relative response factor (RRF) method and isotopically labeled internal standards at (100 μL of ∼1 μg mL−1 PFAS). The single point calibration22 and use of a single quadrapole LC/MS is justified based on the results of earlier published studies in this laboratory.19,22,23 Instrumental limit of detection (LOD) and limit of quantification (LOQ) were calculated based on a peak signal to noise (S/N) ratios of 3[thin space (1/6-em)]:[thin space (1/6-em)]1 and 10[thin space (1/6-em)]:[thin space (1/6-em)]1 respectively. LODs were determined to be 5 μg L−1 for each PFAS, while this is higher than the established health guidelines it is sufficient to determine equilibria in batch experiments and breakthrough in column experiments. Glass auto-sampler vials (2 mL) equipped with Viton septa (ChromTech, Apple Valley, MN) were used for PFAS analysis to minimize potential contamination. Procedural blanks were conducted using simulated groundwater, soil but no added coagulant or PFAS. These procedural blanks did not produce PFAS above the detection limits.

Statistical analysis

To calculate 95% confidence intervals (CI) for fits of isotherms and breakthrough curves, statistical analysis was performed using Microsoft Excel. The LINEST function was applied to obtain the standard error for each fit. The 95% CI was calculated from the standard error. Regressed partition coefficients were compared within each batch with a t-test, with a significance level set to the widely accepted value of α = 0.05. Batches and columns (except polymer isotherms) were all run in triplicate. The means of these repeated experiments were also compared, using a t-test with a significance level of α = 0.05. Statistical analysis was preformed using Microsoft Excel.

Results and discussion

Enhancer isotherms

Isotherms of the polymers (polyDADMAC and polyamine) were constructed on 40–50 mesh size fraction of test soil. Mass of enhancer sorbed per mass of soil was plotted against equilibrium dissolved concentrations (Fig. 1). The data were fit to the Langmuir isotherm where q is mass of enhancer sorbed per mass of soil,
 
image file: c9ew00426b-t1.tif(1)

image file: c9ew00426b-f1.tif
Fig. 1 Sorption isotherms of polyDADMAC (a) and polyamine (b) on test soil, fit to the Langmuir isotherm model. The root mean squared differences between the data and Langmuir isotherm curves were 2.8 and 0.48 for polyDADMAC and polyamine, respectively.

K L is the Langmuir constant, CE is the equilibrium concentration of the enhancer and Qm is the maximum sorption capacity. In order to ensure that the soil was saturated with polymer a value exceeding Qm (i.e. 5000 mg L−1 polyDADMAC or polyamine) was selected as the initial dosage to be applied to the PFAS experiments.

PFAS batch tests

The use of both coagulants significantly increased the sorption of PFAS to the test soil (Fig. 2). The PFAS sorption data were best fit using the linear adsorption model
 
q = CEKD(2)
where q is the solid phase concentration of PFAS (ng PFAS per g soil) and plotted against equilibrium dissolved PFAS concentrations (CE: μg PFAS per L). The soil-water distribution coefficient (KD) was calculated from the slope of each linear isotherm. An important consideration in the sorption mechanism are hydrophobic interactions between PFAS and organic matter associated with soil particles. This was determined by normalizing KD values to the fraction of organic carbon (foc) in each system, to obtain organic carbon-water partitioning coefficients (KOC). Test soil has an foc of 0.015, and Qm of polyDADMAC and polyamine are 6.99 and 8.11 mg OC per g soil, respectively. The addition of polyDADMAC or polyamine to a solution containing suspended test soil and PFAS significantly increased PFAS sorption as measured by both the KD and KOC values (Table 1). For the polyDADMAC-amended soil, the resulting sorption coefficients (KD) increased by factors of 3.1 (PFHpA) to 6.1 (PFOS). Similarly, the addition of polyamine resulted in an increase in KD by a factor of 2.0 (PFOS) to 4.1 (PFHxS). All KD and KOC values determined after polymer treatment were significantly greater than the untreated soil (Table S.2 ESI).

image file: c9ew00426b-f2.tif
Fig. 2 Sorption isotherms on test soil fit to a linear model. Closed circles represent control conditions, open circles represent the addition of polyamine and closed triangles represent the addition of polyDADMAC. Error bars represent standard deviation of triplicate measurements (most are smaller than the symbol).
Table 1 Linear soil-water distribution coefficients (KD) and organic carbon-water partition coefficients (KOC) obtained from batch tests PFAS on untreated test soil (control) and test soil amended with polyDADMAC or polyamine
K D/KOC(L/kg) ± 95% CIa PFBS PFHxS PFOS PFHpA PFOA PFNA
a 95% CI of slope.
Control
K D 0.62 ± 0.12 0.45 ± 0.08 1.44 ± 0.35 0.43 ± 0.04 1.62 ± 0.24 0.72 ± 0.13
K OC 41.3 ± 7.7 30.0 ± 5.4 96.0 ± 2.3 28.7 ± 2.5 108.0 ± 15.9 48.0 ± 8.8
Polyamine
K D 1.53 ± 0.24 1.84 ± 0.29 2.94 ± 0.42 1.12 ± 0.34 10.7 ± 0.70 1.3 ± 0.47
K OC 66.2 ± 10.5 79.6 ± 12.6 127.2 ± 8.2 48.5 ± 14.7 463.0 ± 30.0 56.2 ± 10.2
PolyDADMAC
K D 1.92 ± 0.25 2.3 ± 0.19 8.75 ± 1.36 1.56 ± 0.36 7.12 ± 0.98 2.93 ± 0.93
K OC 87.3 ± 11.4 104.6 ± 8.7 397.9 ± 61.8 93.7 ± 16.3 322.9 ± 44.4 133.3 ± 4.24


Consistent with the general paradigm that PFAS with longer carbon chains exhibit greater sorption,15 this trend was observed within the sulfonate group especially after the addition of polyDADMAC and polyamine, and was especially apparent for the sulfonated species. Carboxylates did not follow this rule as consistently as the sulfonates, with PFOA exhibiting more sorption than its longer-chained analog, PFNA. Another trend observed was that sulfonates exhibited greater sorption than carboxylates of comparable fluorinated carbon chain lengths (i.e. PFHpA vs. PFHxS). This is likely due to the more hydrophobic nature of the sulfonate head group compared to carboxylates.15 An exception to this appears to be PFOA, which exhibited similar or greater sorption than PFOS (depending on polymer added). This PFOA result for both chain length and functional group is atypical of most studies where PFOA is less sorbtive than PFNA and PFOS. This anomaly is not observed when looking at the column data (see below). In comparing our KD values, our values are lower than those reported in the literature (2–120 L kg−1) which may be the result of differences in organic carbon, pH, and ionic strength.23–26

By adding a coagulant to the system, we are increasing the organic carbon content of the soil. Many researchers have shown that PFAS often partitions more strongly to higher organic carbon soils.15,27 If the increase in sorption observed in this study was due solely to the additional organic carbon content of the sorbed coagulants, then KOC values should remain constant for each PFAS, regardless of experimental condition. Upon calculating KOC values for each system, the KOC values increased with similar magnitudes as KD (Table 1). This is indicative of a sorption process that is not solely driven by the increase of organic carbon and hydrophobic interactions from the addition of polyDADMAC or polyamine. One possible mechanism contributing to the enhancement of PFAS sorption could be the increase in positive charge resulting from the adsorption of the cationic polymers. It is interesting to note that polyDADMAC is the larger polymer with a greater charge density, and for most cases has a greater impact on the sorption of PFAS. An earlier study indicated that polyDADMAC has a lower equilibrium constant for PFOS than polyamine.28 Furthermore, the soil seems to have a higher capacity for polyamine than polyDADMAC (Fig. 1). Both of these results suggest that polyamine would have a greater effect on increased sorption of PFAS. Perhaps, the charge density being greater for polyDADMAC overcomes these other factors, or a combination of increased organic carbon and charge density is responsible.

Batch sorption studies were repeated for PFOA and PFOS at pH of 5, 8 and 9 to gain understanding as to how solution chemistry affects PFAS sorption and the enhancement capabilities of polyamine and polyDADMAC (ESI Fig. S.2). Consistent with previous findings that pH influences PFAS sorption by way of electrostatic interaction,15,16,29 the adsorption of PFOS and PFOA in single compound equilibration experiments decreased with increasing pH. This finding is explained by the surface charge of soils becoming more positive with more acidic conditions, thus electrostatically attracting anionic charges. At pH 8, the adsorption of PFOS or PFOA was lower than at either higher or lower pH. This coincides with the PZC of test soil, which was determined to be 8.1 (ESI Fig. S.1). At pH 9, PFOS and PFOA adsorption is virtually non-existent when polymer is not present. At this pH, the surface of the soil particles is basic, possessing a net negative charge. Therefore, in the absence of polymer and in the presence of net negative charge, one would not expect PFAS to sorb. However, in the presence of polyamine and polyDADMAC, adsorption of PFOS and PFOA increased at pH values above the PZC of the test soil. The polymers, being cationic, increase in sorption at this high pH. The presence of the polymers is able to overcome any electrostatic repulsion between the PFAS and the negatively charged soil surface.

PFAS column tests

To investigate the effect that polymer addition will have in dynamic systems, column studies were performed with untreated soil and soil pre-loaded with either polyDADMAC or polyamine. To determine the volume of 5000 mg L−1 polymer solution required to treat the soil, effluent breakthrough curves for polyDADMAC and polyamine were obtained (see ESI Fig. S.3). Complete breakthrough (i.e. effluent concentration equal to influent concentration) of the polymers occurred after flushing with 10–15 pore volumes of the 5000 mg L−1 solution. Therefore, columns were treated with 16 pore volumes (approximately 0.48 L) prior to introduction of the PFAS solutions. Based on breakthrough curves obtained for PFAS in test soil amended with either polyDADMAC or polyamine, retention of PFAS on the columns was significantly increased (Fig. 3), with breakthrough observed at 8 to 20 pore volumes later than the control. Each breakthrough curve was fit to a one-dimensional form of the advective-dispersive reactive (ADR) transport equation using least squares fit in excel:
 
image file: c9ew00426b-t2.tif(3)
where R is the retardation factor, and D is the diffusion coefficient. The model was then confirmed using CXTFIT. After R for each breakthrough curve was determined, KD was back calculated using the equation:
 
image file: c9ew00426b-t3.tif(4)
where ρB is the soil bulk density (1.56 g cm−3) and θW is the volumetric water content (0.43).

image file: c9ew00426b-f3.tif
Fig. 3 Breakthrough curves of PFAS through untreated test soil (closed circles), preloaded with polyDADMAC (black triangles) or polyamine (open circles). Data are presented as effluent concentration normalized to influent concentration versus number of pore volumes (1 PV is ∼30 mL). Error bars representing one standard deviation of the triplicate samples are smaller than symbols. Solid lines represent fitted curves.

The results of the column tests showed similar results to those of the batch tests in that sorption, or in this case retention, of PFAS onto test soil significantly increased in the presence of polyDADMAC or polyamine (Table S.3). For example, initial breakthrough of the shortest chain sulfonate (PFBS) on a polyDADMAC amended soil column occurred at ∼9 pore volumes (PV). Whereas PFOS breakthrough on a polyDADMAC amended soil column occurred at ∼20 PV's. Similar results were observed in that longer chained sulfonates and carboxylates exhibited more sorption than their shorter analogues. While in batch tests, PFOA displayed more sorption than its heavier analog (PFNA), this anomaly was not observed in the column tests.

As with the batch sorption studies, comparison of KD to KOC's was carried out to gain more insight as to the mechanism by which sorption is enhanced. Again, the KOC values did not remain constant with polymer loading, suggesting that the sorption mechanism is not limited to an increase in organic content resulting from the addition of polyDADMAC and polyamine. Therefore, electrostatic interactions between PFAS and the polymers is thought to dominate the sorption mechanism.

To investigate the longer-term viability of the method, desorption experiments were performed immediately after column tests by flushing the column with PFAS free simulated groundwater at the same rate (0.12 mL min−1) for an additional 30 pore volumes (∼60 h). Results for PFOS in the absence of polymer indicates that it desorbs and is flushed from the column within 8 pore volumes (Fig. 5). By calculating the area under the curve of the flushed breakthrough curves in Fig. 5, one can estimate the amount of PFAS recovered. All control columns (no coagulant added) showed near 100% mass recovery within 8 pore volumes, indicating that sorption of PFAS onto this soil was reversible (Fig. 5) This result is consistent with previous studies that show PFAS to be mobile in aquifer materials.30,31 Desorption from soil pre-treated with both polyDADMAC and polyamine show greatly decreased recoveries, indicating that the association of PFAS, polymers and soil is quite strong (Fig. 4 and 5), with less than or equal to 15% recovery in all cases.


image file: c9ew00426b-f4.tif
Fig. 4 Desorption of PFOS from soil columns when flushed with PFAS free simulated groundwater. Dotted lines indicate when PFAS addition was stopped.

image file: c9ew00426b-f5.tif
Fig. 5 Recovery percentages of PFAS from untreated test soil columns (black bars) and soil columns pre-treated polyamine and polyDADMAC (gray and hollow bars).

Summary

Batch reactor tests showed a significant increase in PFAS KD values in the presence of either polyamine or polyDADMAC, in some cases by a factor of 6 or greater. Normalizing the adsorption coefficient to organic carbon content did not explain the increased adsorption on the polymer-treated soil. This finding indicated mechanisms in addition to hydrophobic interactions between PFAS and the organic carbon content increase associated with the addition of polymers. Because the intent of this research was to develop a new in situ treatment method, column studies were subsequently performed to assess performance of the polymer treatment under dynamic conditions. Retention of PFAS in the soil columns was significantly increased, resulting in substantially delayed breakthrough of PFAS following either polyamine or polyDADMAC treatment. In general, the batch and column results follow the general trend that longer-chain PFAS exhibited greater sorption and retention than shorter-chained analogs. This is also true for differences accounting for functional group, where sulfonates exhibited more sorption than carboxylates. polyDADMAC resulted in greater enhanced PFAS sorption than polyamine in batch tests for all PFAS but PFOA. The column results showed that polyDADMAC significantly out-performed polyamine for all PFAS with the exception of PFOA and PFNA. Desorption studies suggest that PFAS sorption using this method is not readily reversible, especially in comparison to untreated soil, which is a desirable quality for long term in situ sequestration of PFAS and containment of groundwater plumes. The results of this work indicate that these cationic polymers could be added to an aquifer to act as a permeable adsorptive barrier to PFAS in a groundwater plume. Knowledge gaps that remain include the longevity of such a barrier, the effectiveness for other PFAS, including short chain carboxylates and compounds that are precursors to the perfluorochemicals tested herein, the degradability of the polymers in the environment in the presence and absence of PFAS, and the precise mechanism governing the increased adsorption.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

Support for this research was provided by the Strategic Environmental Research and Development Program (SERDP) under contract W912HQ-14-C-0042 for Project ER-2425, “Development of a Novel Approach for In Situ Remediation of PFC Contaminated Groundwater Systems”. This work has not been subject to SERDP review, and no official endorsement should be inferred. The authors would also like to thank the engineers at CH2M (now Jacobs) for providing the excavated soil.

References

  1. J. Gaillard, B. Veyrand, M. Thomas, X. Dauchy, V. Boiteux, P. Marchand, B. Le Bizec, D. Banas and C. Feidt, Tissue Uptake, Distribution, and Elimination of Perfluoroalkyl Substances in Juvenile Perch through Perfluorooctane Sulfonamidoethanol Based Phosphate Diester Dietary Exposure, Environ. Sci. Technol., 2017, 51, 7658–7666 CrossRef CAS PubMed .
  2. E. F. Houtz, C. P. Higgins, J. A. Field and D. L. Sedlak, Persistence of perfluoroalkyl acid precursors in AFFF-impacted groundwater and soil, Environ. Sci. Technol., 2013, 47, 8187–8195 CrossRef CAS .
  3. Z. Liu, Y. Lu, P. Wang, T. Wang, S. Liu, A. C. Johnson, A. J. Sweetman and Y. Baninla, Pollution pathways and release estimation of perfluorooctane sulfonate (PFOS) and perfluorooctanoic acid (PFOA) in central and eastern China, Sci. Total Environ., 2017, 580, 1247–1256 CrossRef CAS .
  4. N. Merino, Y. Qu, R. A. Deeb, E. L. Hawley, M. R. Hoffmann and S. Mahendra, Degradation and Removal Methods for Perfluoroalkyl and Polyfluoroalkyl Substances in Water, Environ. Eng. Sci., 2016, 33, 615–649 CrossRef CAS .
  5. K. A. Barzen-Hanson, S. E. Davis, M. Kleber and J. A. Field, Sorption of Fluorotelomer Sulfonates, Fluorotelomer Sulfonamido Betaines, and a Fluorotelomer Sulfonamido Amine in National Foam Aqueous Film-Forming Foam to Soil, Environ. Sci. Technol., 2017, 51, 12394–12404 CrossRef CAS .
  6. C. A. Moody, J. Field, S. Strauss, G. Hebert and M. Odom, Occurrence and distribution of perfluorinated surfactants in groundwater impacted by fire-fighting activity, Book of Abstracts, 219th ACS National Meeting, San Francisco, CA, March 26–30, 2000, 2000, ENVR-027 Search PubMed .
  7. C. A. Moody, G. N. Hebert, S. H. Strauss and J. A. Field, Occurrence and persistence of perfluorooctanesulfonate and other perfluorinated surfactants in groundwater at a fire-training area at Wurtsmith Air Force Base, Michigan, USA, J. Environ. Monit., 2003, 5, 341–345 RSC .
  8. M. Filipovic, A. Woldegiorgis, K. Norstroem, M. Bibi, M. Lindberg and A.-H. Oesteraas, Historical usage of aqueous film forming foam: A case study of the widespread distribution of perfluoroalkyl acids from a military airport to groundwater, lakes, soils and fish, Chemosphere, 2015, 129, 39–45 CrossRef CAS .
  9. J. L. Guelfo and C. P. Higgins, Subsurface Transport Potential of Perfluoroalkyl Acids at Aqueous Film-Forming Foam (AFFF)-Impacted Sites, Environ. Sci. Technol., 2013, 47, 4164–4171 CrossRef CAS .
  10. V. A. A. Espana, M. Mallavarapu and R. Naidu, Treatment technologies for aqueous perfluorooctanesulfonate (PFOS) and perfluorooctanoate (PFOA): A critical review with an emphasis on field testing, Environmental Technology & Innovation, 2015, 4, 168–181 Search PubMed .
  11. E. R. McKenzie, R. L. Siegrist, J. E. McCray and C. P. Higgins, Effects of Chemical Oxidants on Perfluoroalkyl Acid Transport in One-Dimensional Porous Media Columns, Environ. Sci. Technol., 2015, 49, 1681–1689 CrossRef CAS .
  12. S. Park, L. S. Lee, V. F. Medina, A. Zull and S. Waisner, Heat-activated persulfate oxidation of PFOA, 6:2 fluorotelomer sulfonate, and PFOS under conditions suitable for in-situ groundwater remediation, Chemosphere, 2016, 145, 376–383 CrossRef CAS .
  13. D. Eberle, R. Ball and T. B. Boving, Impact of ISCO Treatment on PFAA Co-Contaminants at a Former Fire Training Area, Environ. Sci. Technol., 2017, 51, 5127–5136 CrossRef CAS .
  14. P. McCleaf, S. Englund, A. Oestlund, K. Lindegren, K. Wiberg and L. Ahrens, Removal efficiency of multiple poly- and perfluoroalkyl substances (PFASs) in drinking water using granular activated carbon (GAC) and anion exchange (AE) column tests, Water Res., 2017, 120, 77–87 CrossRef CAS .
  15. C. P. Higgins and R. G. Luthy, Sorption of Perfluorinated Surfactants on Sediments, Environ. Sci. Technol., 2006, 40, 7251–7256 CrossRef CAS .
  16. F. Wang and K. Shih, Adsorption of perfluorooctanesulfonate (PFOS) and perfluorooctanoate (PFOA) on alumina: Influence of solution pH and cations, Water Res., 2011, 45, 2925–2930 CrossRef CAS .
  17. Q. Yu, R. Zhang, S. Deng, J. Huang and G. Yu, Sorption of perfluorooctane sulfonate and perfluorooctanoate on activated carbons and resin: Kinetic and isotherm study, Water Res., 2009, 43, 1150–1158 CrossRef CAS .
  18. F. Xiao, M. F. Simcik and J. S. Gulliver, Mechanisms for removal of perfluorooctane sulfonate (PFOS) and perfluorooctanoate (PFOA) from drinking water by conventional and enhanced coagulation, Water Res., 2013, 47, 49–56 CrossRef CAS .
  19. Y. H. Aly, C. Liu, D. P. McInnis, B. A. Lyon, J. Hatton, M. McCarty, W. A. Arnold, K. D. Pennell and M. F. Simcik, In situ remediation method for enhanced sorption of perfluoro-alkyl substances onto Ottawa sand, J. Environ. Eng., 2018, 144, 04018086 CrossRef .
  20. N. E. J. Curtis and W. E. Ham, in Geology and Earth Resources of Oklahoma. An Atlas of Maps and Cross Sections, ed. K. S. Johnson, C. C. Branson, N. E. J. Curtis, W. E. Ham, W. E. Harrison, M. V. Marcher and J. F. Roberts, Oklahoma Geological Survey, Norman, Oklahoma, 1972 Search PubMed .
  21. A. G. Institute, Dictionary of Geological Terms, Anchor Press/Doubleday, Garden City, New York, 1976 Search PubMed .
  22. N. J. M. Fitzgerald, A. Wargenau, C. Sorenson, J. Pedersen, N. Tufenkji, P. J. Novak and M. F. Simcik, Partitioning and Accumulation of Perfluoroalkyl Substances in Model Lipid Bilayers and Bacteria, Environ. Sci. Technol., 2018, 52, 10433–10440 CrossRef CAS .
  23. R. L. Johnson, A. J. Anschutz, J. M. Smolen, M. F. Simcik and R. L. Penn, The adsorption of perfluorooctane sulfonate onto sand, clay, and iron oxide surfaces, J. Chem. Eng. Data, 2007, 52, 1165–1170 CrossRef CAS .
  24. F. Xiao, X. Zhang, L. Penn, J. S. Gulliver and M. F. Simcik, Effects of Monovalent Cations on the Competitive Adsorption of Perfluoroalkyl Acids by Kaolinite: Experimental Studies and Modeling, Environ. Sci. Technol., 2011, 45, 10028–10035 CrossRef CAS .
  25. F. Li, X. Fang, Z. Zhou, X. Liao, J. Zou, B. Yuan and W. Sun, Adsorption of perfluorinated acids onto soils: Kinetics, isotherms, and influences of soil properties, Sci. Total Environ., 2019, 649, 504–514 CrossRef CAS .
  26. J. Milinovic, S. Lacorte, M. Vidal and A. Rigol, Sorption behaviour of perfluoroalkyl substances in soils, Sci. Total Environ., 2015, 511, 63–71 CrossRef CAS .
  27. R. L. Johnson, A. J. Anschutz, J. M. Smolen, M. F. Simcik and R. L. Penn, The Adsorption of Perfluorooctane Sulfonate onto Sand, Clay, and Iron Oxide Surfaces, J. Chem. Eng. Data, 2007, 52, 1165–1170 CrossRef CAS .
  28. E. L. Anderson, M. P. S. Mousavi, Y. H. Aly, X. V. Chen, M. F. Simcik and P. Buhlmann, Remediation of Perfluorooctylsulfonate Contamination by in Situ Sequestration: Direct Monitoring of PFOS Binding to Polyquaternium Polymers, ACS Omega, 2019, 4, 1068–1076 CrossRef CAS .
  29. J. Tang and Y.-l. Zhang, Advance in studies of PFOA and PFOS adsorption and desorption in soils/sediments, Turang (Beijing, China), 2014, vol. 46, pp. 599–606 Search PubMed .
  30. J. P. Giesy and K. Kannan, Perfluorochemical surfactants in the environment, Environ. Sci. Technol., 2002, 36, 146A–152A CrossRef CAS .
  31. F. Xiao, M. F. Simcik, T. R. Halbach and J. S. Gulliver, Perfluorooctane sulfonate (PFOS) and perfluorooctanoate (PFOA) in soils and groundwater of a US metropolitan area: Migration and implications for human exposure, Water Res., 2015, 72, 64–74 CrossRef CAS .

Footnote

Electronic supplementary information (ESI) available. See DOI: 10.1039/c9ew00426b

This journal is © The Royal Society of Chemistry 2019