Absorption of short-chain to long-chain perfluoroalkyl substances using swellable organically modified silica

Eva K. Stebel , Kyndal A. Pike , Huan Nguyen , Heather A. Hartmann , Mattaeus J. Klonowski , Michaela G. Lawrence , Rachel M. Collins , Claire E. Hefner and Paul L. Edmiston *
Department of Chemistry, The College of Wooster, Wooster, OH 44691, USA. E-mail: pedmiston@wooster.edu

Received 30th April 2019 , Accepted 11th July 2019

First published on 11th July 2019


Organosilica synthesis chemistry via the sol–gel method was used to create porous organosilica adsorbents designed to remove PFAS compounds from water. A set of adsorbents was created by incorporating a fluorophilic amide or fluoroalkyl function group along with quaternary groups for ion exchange within a continuous porous matrix. A distinguishing aspect of the adsorbents was the ability to volumetrically swell 2.5× (>6 mL g−1) in the presence of organic liquids. Swelling results from a flexible pore architecture and was used to create expanded mesopores which were hypothesized to yield greater adsorption capacity for PFAS compounds. Adsorption isotherms and adsorption kinetics experiments were conducted for a set of 12 PFAS analytes in deionized water and 50 mM NaCl in order to understand the effect of adsorbate characteristics and ionic strength on adsorption. PFAS of all chain lengths ranging from C4–C10 were bound to adsorbents possessing both fluorophilic and cationic groups. In comparison, only long-chain PFAS >C6 are bound when the adsorbents are exclusively hydrophobic or fluorophilic. An adsorbent was created where a cationic imidazolium containing polymer was entrapped within the swellable pores. The polymer-infused organosilica demonstrated the highest capacity across the suite of 12 short-chain and long-chain PFAS. In column experiments, the polymer organosilica yielded 47 mg g−1 capacity for perfluorooctanoic acid (PFOA) prior to breakthrough with an influent concentration of 200 μg L−1, approximately an order a magnitude greater than GAC. The high adsorption capacity is attributed to expandable pores which can accommodate polymer and PFAS adsorbates while maintaining sufficient access for mass transport.



Water impact

Perfluoroalkyl substances (PFAS) are an emerging class of environmental contaminants. This research involves the design and characterization of specialized adsorbents to remove PFAS from water. These results should aid practitioners and researchers interested in PFAS remediation.

Introduction

Poly- and perfluoroalkyl substances (PFAS) are a class of emerging environmental contaminants1,2 resulting from use in fire suppressants and other products such as coatings.3–5 The presence of PFAS in the aquatic environment is a concern since toxicological data6 suggest a linkage to reproductive effects,7,8 neuro-development issues in children,9 and changes in immune response.10,11 As a result, methods to remediate PFAS in water are being developed. PFAS compounds are chemically inert and resistant to biodegradation due to the strength of the C–F bond. Thus, removal of PFAS from water has focused on the use of adsorbents such as granular activated carbon (GAC),12–15 ion exchange (IEx) resins,16–20 and other adsorbents21 (see recent reviews).22,23 The primary PFAS solutes studied have been perfluorooctane sulfonate (PFOS) and perfluorooctanoic acid (PFOA), which are the most commonly regulated PFAS globally24 as exemplified by the U.S. EPA establishing a health advisory concentration of 70 ng L−1 PFOS and PFOA combined. Both GAC and IEx resins have shown to be effective in adsorbing PFOA and PFOS. However, PFAS are a diverse class of compounds comprising anionic, cationic, and neutral species each with a diversity of fluoroalkyl chain lengths.1,25 Short chain compounds such as perfluorobutanoic acid (PFBA), which are more soluble in water, have been found to be particularly difficult to remove using traditional sorbents.26 Adsorption of PFAS is also adversely affected by the presence of other solutes such as natural organic matter and salts.27

Alternative adsorbents for PFAS have been studied including organic frameworks,28 amine-modified cellulose,29 graphene,30 and nanomaterials.31 One goal has been to improve PFAS affinity by using functional groups that can make specific interactions with adsorbates using fluoroalkyl-modified hydrogels32 or cyclodextrins.33 There is a continuous search for higher capacity, low cost granular adsorbents to improve the economics of treatment by extending the time between replacement of adsorbent beds. As a result, a variety of natural materials have explored as adsorbents.34–36

Here, we used the sol–gel process37 to create organosilica adsorbents with pore structures and chemical functionality to study the adsorption of a wide suite of PFAS from water. Adsorbents were based on swellable organically modified silica (SOMS, commercial name: Osorb®) which is a hydrophobic and continuously porous resin that reversibly swells 2.5× in the presence of organic adsorbates.38 SOMS is synthesized by the polycondensation of bis(trimethoxysilylethyl)benzene, BTEB (1, Fig. 1) under controlled conditions. Methoxysilane groups react to form Si–O–Si linkages forming a network where self-assembly directed by the bridging aryl group yields a porous and mechanically flexible matrix. Alternative organosilane precursors can be co-polymerized with BTEB to alter the surface chemistry while preserving the flexible pore structure.39 A fluoroalkyl silane (2, Fig. 1) and quaternary amine silane (3) were explored as functional groups that would facilitate the adsorption of anionic PFAS compounds through fluorophilic and ionic interactions, respectively. The swelling of SOMS can also be used to alter the chemical functionality of the pores by physically entrapping polymers. A solution of commercially available cationic polymer poly[(3-methyl-1-vinylimidazolium chloride)-co-(1-vinylpyrrolidone)] (4) was thus added to the pores via swelling. After removing the solvent by drying, the pores shrink in size physically immobilizing the absorbed polymer embedded in hydrophobic pores.


image file: c9ew00364a-f1.tif
Fig. 1 Structures of alkoxysilane precursors used for organosilica synthesis (1–3). Polymer (4) encapsulated in SOMS.

Organosilica adsorbents in the form of SOMS were hypothesized to be a useful tool in studying PFAS adsorption in order to optimize affinity and capacity. First, the pore size can be expanded by swelling SOMS with a water miscible organic liquid prior to use. Several studies suggest that PFAS adsorbates aggregate to form micelles and hemi-micelles upon adsorption.40–42 Wider pores would help to facilitate transport and accommodate PFAS aggregates.43 Second, the surface chemistry of the organosilica was varied by polymerization of different silane monomer mixtures. Systematic addition of a fluoroalkyl group and a cationic (ion exchange) group to the pre-existing hydrophobic (aryl rich) pores was used to determine the effect on PFAS adsorption. In the absence of optimal silane precursors, 4 was entrapped which has cationic groups for ion exchange and amide groups. Amides have been shown to make an attractive interaction with organofluorine,44 not previously explored for PFAS adsorption.

When evaluating adsorbents for PFAS adsorption it is helpful to consider parameters relevant to remediation. Recent criticism of previous work characterizing adsorbents31 pointed out that adsorption performance of PFOA and PFOS have been routinely measured at concentrations much higher than levels expected in the environment (mg L−1vs. ng L−1), thus limiting their relevance. The work reported here attempts to bridge this gap by working with equilibrium concentrations, Ce, in the ppb to ppt range. Further critique in the literature notes that a wider range of PFAS compounds should be tested in addition to PFOA and PFOS. Mixtures of PFAS are often present in the environment including those in the EPA's UCMR 3 list, some of which may emerge as having adverse health effects. PFAS solutes included in this study (Table 1) included 7 perfluoro-carboxylates, 3 perfluorosulfonates, 1 neutral PFAS, and 1 cationic compound to provide a fuller picture of performance. Finally, co-solutes can impact the adsorption performance. Here, elevated ionic strength was studied due to its possible effect on ionic interactions involving PFAS adsorbates.45

Table 1 PFAS compounds used for adsorbent characterization
PFAS name Acronym Formula CAS
Perfluorodecanoic acid PFDA C10HF19O2 335-76-2
Perfluorononanoic acid PFNA C9HF17O2 375-95-1
Perfluorooctanoic acid PFOA C8HF15O2 335-67-1
Perfluoroheptanoic acid PFHpA C7HF13O2 375-85-9
Perfluorohexanoic acid PFHxA C6HF11O2 307-24-4
Perfluoropentanoic acid PFPeA C5HF9O2 2706-90-3
Perfluorobutanoic acid PFBA C4HF7O2 375-22-4
Perfluorooctanesulfonic acid PFOS C8HF17O3S 1763-23-1
Perfluorohexanesulfonic acid PFHxS C6HF13O3S 355-46-4
Perfluorobutanesulfonic acid PFBS C4HF9O3S 375-73-5
Perfluorooctanesulfonamide PFOSA C8H2F17NO2S 754-91-6
Perfluorooctane sulfonamide quaternary ammonium PFOSaAm C13H13F17N2O2S 13417-01-1


Experimental

Materials

Silane precursors 1, BTEB; 2, (tridecafluoro-1,1,2,2-tetrahydrooctyl)trimethoxysilane; and 3, N-trimethoxysilylpropyl-N,N,N-trimethylammonium chloride (50% in methanol) were purchased from Gelest. Tetrabutyl ammonium fluoride (TBAF, 1.0 M solution in THF); 4, Luviquat™ FC 370 polymer, hexamethyldisilazane (HMDS), Norit® 1240 W activated carbon, neat PFAS compounds, and all solvents were obtained from Aldrich. PFAS standards were obtained from Wellington Laboratories. Oasis® HLB (200 mg, 6 mL) solid phase extraction cartridges were obtained from Waters.

Adsorbent synthesis

SOMS exclusively comprised of 1 was prepared as described previously.39 Fluoroalkyl-modified organosilica, “F-SOMS”, was prepared by adding 3.0 g of 1 and 0.55 g of 2 to 14 mL acetone. Polymerization was catalyzed by the addition of 580 μL 0.225 M TBAF solution (90 μL 1 M TBAF, 490 μL water) resulting in gelation. After aging for 6 days at 25 °C, the gel was crushed into ∼5–10 mm pieces, rinsed twice with acetone, and immersed in 20 mL of 5% v/v HMDS in acetone for 48 h at 25 °C. The gel pieces were thoroughly rinsed with acetone in using Soxhlet extraction, and dried at room temperature. The solid material was ground using a ball mill and sieved. Quaternary amine modified organosilica, “QA-SOMS”, was prepared in the same manner as F-SOMS except 3.0 g of 1 and 800 μL of the vendor provided solution of 3 were combined to create the gel. SOMS with entrapped polymer, “poly-SOMS”, was prepared by applying 10.0 mL of a 100 mg mL−1 of 4 in methanol to 5.0 g of SOMS which fully absorbed the solution. The methanol was evaporated at 25 °C to dryness.

Characterization

Surface area and pore volume were measured by N2 adsorption at 77 °K using a Beckman Coulter SA-3100 instrument. Samples were outgassed at 120 °C for 120 min prior to measurement. Surface area was calculated using the BET method46 and the pore size distribution was calculated using the BJH method.47 Swelling capacity (mL of liquid absorbed per g SOMS) was measured by gravimetry after titrating the sorbent to the fully swollen state with neat acetone. The change in volume was measured by visible light microscopy using image analysis software (Image J). Scanning electron microscopy was conducted using Hitachi S-4700 after coating the samples with Pt. Imaging was done at 10.0 kV. Reproducibility was tested by repeating measurements using the same sample and also by measuring adsorption by separate batches of material synthesized by the same procedure. FT-IR spectra of poly-SOMS were taking using a Perkin-Elmer FT-IR 2000 using KBr pellets whereas spectra for other materials were measured using a Nicolet 6700 FT-IR using a diamond-ATR accessory. Fluorine elemental analysis was conducted by Galbraith Laboratories.

Adsorption kinetics and isotherm measurements

Adsorption kinetics was measured by adding 200 mg of adsorbent to 1 L high density polyethylene (HDPE) bottles. SOMS, F-SOMS, and QA-SOMS were pre-wetted by applying a minimal volume of ethanol (∼500 μL) to the adsorbent. Next, 500 mL of 2000 μg L−1 solution of individual PFAS in DI water or 50 mM NaCl was added to the bottles which were shaken for 14 hours at 170 rpm with 2 mL aliquots taken at 0.25, 0.5, 1, 2, and 14 h intervals. Sample aliquots were filtered with 0.45 μm cellulose acetate syringe filters, except for PFOSA and PFOSaAm, which were centrifuged to remove any adsorbent particles. Cellulose acetate filters were shown not to remove PFAS compounds except for PFOSA and PFOSaAm in separate experiments. Samples were immediately diluted with 50[thin space (1/6-em)]:[thin space (1/6-em)]50 with 100% methanol or 50[thin space (1/6-em)]:[thin space (1/6-em)]50 with 95% methanol/5% ammonium acetate if the compound was a perfluorinated carboxylate. An additional control experiment was done by adding the polymer 4 to water containing PFAS solutes and measuring that the polymer did not lead to bias in the measurement, which was not observed. The concentration of PFAS was measured by LC-MS for each aliquot and the initial solution which was compared to a control which was a sample of filtered or centrifuged untreated solution. Internal standards and perfluoro-n-[1,2-13C2]octanoic acid (M2PFOA) and sodium perfluoro-1-[1,2,3,4-13C4]octanesulfonate (MPFOS) were added prior to measurement to selected samples. Data was fit to an integrated a pseudo-second order kinetic model.48
 
image file: c9ew00364a-t1.tif(1)
where q(t) is the amount adsorbed at time t, k2 is the rate constant, and qe = q(t → ∞), the equilibrium capacity. RStudio (Version 1.1.456) was used to fit isotherms and pseudo second order kinetics.

Adsorption isotherms were conducted for each individual PFAS at C0 concentrations of 250, 500, 1000, 1500, and 2000 μg L−1 in either DI water or 50 mM NaCl using a 40 mg L−1 dosage of adsorbent. SOMS and F-SOMS were pre-swollen with ethanol prior to addition of aqueous PFAS solutions. Bottles were shaken for 14 h and the PFAS concentration measured compared to control samples of the initial solution using the same procedures above. Data was fit to the Freundlich isotherm in the form:

 
image file: c9ew00364a-t2.tif(2)
by linear regression using RStudio to determine 1/n and the Freundlich constant, KF. All adsorption isotherms were run in duplicate.

Column experiments

Adsorbents were pre-wetted either by vigorous agitation in water for GAC and poly-SOMS or swelling with ethanol for SOMS and F-SOMS. Continuous flow column experiments were performed by placing 500 mg of sorbent in 16 mm Pharmacia XK16 glass/polyamide-polyethylene column creating a bed height of approximately 1 cm. Polypropylene tubing was used for all fluid handling. Adsorbents were initially rinsed on the column with 250 mL of DI water at 2 mL min−1. PFAS solutions were applied at 1.0 mL min−1 with a contact time of 2 min. Treatment columns were run with influent concentrations of either 2.0 or 200 μg L−1. Effluent samples were collected periodically in conjunction with matched influent samples. Aliquots were immediately diluted 50[thin space (1/6-em)]:[thin space (1/6-em)]50 with either 100% methanol or 95% methanol/5% ammonium acetate solution and PFAS concentration measured by LC-MS. Internal standards were added to select samples for quality control. When analyzing effluent from 2.0 μg L−1 applied PFAS concentration 500 mL aliquots were collected and concentrated using solid-phase extraction as previously described.49,50 Surrogates sodium perfluoro-[13C8]-octanesulfonate (M8PFOS) and perfluoro-n-[13C8]octanoic acid (M8PFOA) were used to measure recovery.

Analytical measurements

PFAS concentrations were analyzed by LC-MS using an Agilent 1200/6410 HPLC-MS (QqQ) with using Infinity Lab C18 Poroshell 120 21 × 100 mm column, particle size 2.7 μm and PEEK tubing is used for solvent flow. Mobile phases were A: 5 mM ammonium acetate in water B: 95% methanol + 5 mM ammonium acetate with a flow rate of 0.300 mL min−1 at temperature of 35 °C. Injection volumes were 5 μL. Single PFAS analytes with expected concentrations >2 μg L−1 were measured using direct injection with isocratic elution with MRM detection utilizing both a quantitative and qualitative transitions of analytes and internal standards (see ESI Table S1 for parameters). SPE extracts were of single component samples were run in the same procedure. Mixtures of PFAS compounds were run using gradient elution (ESI Table S2). A laboratory reagent blank and laboratory control sample was run before each set of samples and several times during each daily worklist. A laboratory control was run with each sample sub-set or experimental method or set of samples. The limit of quantitation was 0.65–6.5 μg L−1 in direct injection mode (ESI, Table S3) with replicate precision <3%.

Results and discussion

Adsorbent characterization

Organosilica adsorbents were characterized by scanning electron microscopy (Fig. 2 and ESI Fig. S5), FT-IR, and porosimetry (Table 2). F-SOMS, which introduced 0.13 mol mol−1 of 2 relative to 1 which was used exclusively in previous work, had little impact on the microstructure (Fig. 2 and Table 2). Both SOMS and F-SOMS swelled upon addition of acetone: 6.8 vs. 6.5 mL acetone per g adsorbent, respectively resulting in a 2.5× change in volume. In contrast, the surface area and pore volume of QA-SOMS was reduced (Table 2). QA-SOMS required over 5 min to reach a gel state during polymerization compared to <1 min for SOMS. QA-SOMS was more friable and became a powder during the synthesis process, whereas SOMS and F-SOMS remained a monolithic material until being ground to granular form prior to testing. The reduction in gel time and reduced hardness may be due to charge–charge repulsions during polymerization that reduce crosslinking. Spectroscopic evidence for reduced crosslinking of QA-SOMS is correlated with the appearance of a band at 890 cm−1 characteristic of Si–O with a non-bridging oxygen51 and reduction in the shoulder at 1080 cm−1 of the primary Si–O–Si band (see ESI, Fig. S1).
image file: c9ew00364a-f2.tif
Fig. 2 SEM micrographs of A, SOMS; B, F-SOMS; and C, poly-SOMS in the unswollen state.
Table 2 Physical properties of adsorbents
Adsorbent Description (precursor) Particle size (μm) BET surface area (m2 g−1) Pore volume (mL g−1) Pore diameter distribution (%) Swella (mL g−1)
<6 nm 6–80 nm >80 nm
a Volume of liquid acetone absorbed per gram of sorbent.
SOMS Swellable modified silica (1) 250–450 650 1.03 29 69 2 6.1
F-SOMS Fluoroalkyl modified (1, 2) 250–450 660 1.04 30 69 1 6.6
QA-SOMS Quaternary amine modified (1, 3) 50–150 500 0.60 71 26 3 6.3
Poly-SOMS Entrapped cationic polymer (4) 250–450 155 0.21 35 63 1 6.1
GAC Norit® 1240 W 450–600 975 0.52 55 39 6


FT-IR spectroscopy indicated that the Si–O–Si region from 950–1200 cm−1 were similar for SOMS and F-SOMS. F-SOMS showed an increase in peak area at 1250 and 1100 cm−1 due the introduction of C–F stretching modes.52 Determining the extent of incorporation of the fluoroalkyl precursor (2) in the final material was difficult to quantitate using FT-IR due to the overlap of the C–F bands with the Si–O–Si region from 1100–1250 cm−1. Elemental analysis found that F-SOMS was composed of 13.6% fluorine by weight representing 80% theoretical incorporation of 2 in the final material, assuming complete polycondensation. Full condensation is unlikely, thus, there was substantial incorporation of the fluoroalkyl group.

The ability of SOMS to swell allows two distinct forms of the adsorbent: either an unwetted-unswollen form or material that is pre-swollen to expand the pores. Adsorption of PFAS by either unswollen or swollen SOMS was evaluated through batch equilibrium experiments starting with a concentration of C0 = 2000 μg L−1. It was found that when unswollen the adsorption of PFAS compounds >C4 was significantly reduced. For example, the extent of adsorption of PFOA and PFOS after 14 h using swollen-SOMS was 91%, respectively, compared to 29% and 8% if SOMS was unswollen (see ESI Fig. S2 for complete data). In contrast, PFBA adsorption was higher when using unswollen SOMS: 46% removal vs. 0% removal when SOMS was swollen. The ability of unswollen SOMS to adsorb PFBA may be due to the adsorbate's smaller size which allows entry into the collapsed air-filled pores of the hydrophobic SOMS matrix. Unswollen SOMS has previously shown good capacity to adsorb volatile organic solutes, including short chain carboxylate acids, from water.53 Combining both unswollen and swollen SOMS to treat short-chain and long-chain substances, respectively, may be a useful treatment strategy. However, for consistency, in all subsequent experiments, SOMS and F-SOMS were pre-swollen prior to measurement of adsorption by adding a volume of ethanol just sufficient to open the pores and wet the interior surfaces. Absorbed ethanol was displaced by water by dilution when the material added to aqueous samples. After ethanol displacement the pores remain open. Pore collapse only occurs when capillary forces by entrained solvent constrict SOMS during evaporation.

The ability of the cationic polymer 4 to remain encapsulated in poly-SOMS upon rinsing with water was tested. A sample of poly-SOMS was rinsed with 500 volumes of DI water over 20 min collecting the flow through. After drying, the mass of the poly-SOMS decreased 14 ± 3% post water rinse. Confirmation via mass balance was achieved by evaporating the water to dryness and measuring the residue for polymer by gravimetry and FT-IR. The non-volatile residue was determined to be polymer equivalent to the mass lost upon rinsing. Poly-SOMS was hydrophilic prior to rinsing, however, the granular material became hydrophobic after rinsing. The data suggest that a minor fraction of the polymer remained on the surface during the swell-dry encapsulation process which was readily removed by water rinse. Remaining polymer appeared to be fully encapsulated and not removed by water rinse. Encapsulation was also verified by measuring the amount of polymer in poly-SOMS before and after use in column experiments using FT-IR (see below).

Adsorption kinetics

The rate of adsorption of individual PFAS solutes was measured by batch depletion experiments at pH 6.6. The adsorption kinetics for PFOA (Fig. 3) were typical where SOMS materials had a faster rate of adsorption compared to GAC. Adsorption to QA-SOMS was the most rapid, however, the particle size of the material was much smaller and cannot be directly compared. Most of the kinetic data fit well to a pseudo-second order rate equation to calculate k2. Overall, the rate of adsorption to SOMS and F-SOMS were 3–10× faster compared to poly-SOMS or GAC (Table 3, see ESI Tables S4–S6 and Fig. S3 for complete data sets). Faster rates of adsorption to SOMS is attributed to the open swollen pore structure. In contrast, adsorption to poly-SOMS was slower, presumably since it was a porous matrix collapsed around polymer. The rate of adsorption to poly-SOMS was generally constant across the range of PFAS compounds regardless of chain length suggesting that diffusion into the polymer filled pores does not depend strongly on PFAS size. For SOMS and F-SOMS, a reduction in adsorption rate for longer-chain compounds was observed with a 10-fold reduction in adsorption rate for PFNA relative to PFOA.
image file: c9ew00364a-f3.tif
Fig. 3 Adsorption kinetics of PFOA uptake using SOMS (○); F-SOMS (image file: c9ew00364a-u1.tif); QA-SOMS (image file: c9ew00364a-u2.tif); poly-SOMS (image file: c9ew00364a-u3.tif); and GAC (●). Dosage 200 mg L−1, temperature 25 °C, constant agitation, DI water.
Table 3 Pseudo-second order rate constants and equilibrium adsorption capacity for PFAS solutes
PFAS k 2 (g adsorbent per mg PFAS h) @ ionic strength = 0 mM/50 mM q e (mg g−1)b @ ionic strength = 0 mM/50 mM
SOMS F-SOMS QA-SOMS Poly-SOMS GAC SOMS F-SOMS QA-SOMS Poly-SOMS GAC
a All fits to eqn (1). Yielded r2 values >0.98 unless indicated by *. Error and r2 values are reported in ESI. k2 > 20 indicates system came to equilibrium in less than 15 min. Conditions: C0 = 2000 ppb, 200 mg L−1 dosage, modified with 50 mM NaCl. b Value of qe = 10.0 μg g−1 indicates greater than 99.9% removal of PFAS analyte at equilibrium. No ads = minimal adsorption of this compound. n/m = not measured.
PFDA 5.6/>20 1.9/19.6 0.7/5.3 1.1/1.2 0.8/1.4 5.0/9.9 5.8/10.0 6.0/6.2 10.0/9.9 7.5/9.9
PFNA 1.0/8.2 0.7/4.0 4.9/10.5 1.2*/1.4 0.9/1.4 9.2/9.9 8.9/9.9 8.3/6.3 9.1/9.9 6.5/9.5
PFOA 10.4/6.9 6.9/>20 2.5/1.9 1.1/0.6 1.6/0.7 9.9/10.0 9.8/10.0 4.5/8.9 10.0/7.0 9.9/9.5
PFHpA 6.3/3.2 3.2/8.8 0.4*/2.4 1.0/0.7 0.2/0.4 8.0/3.0 7.6/3.1 8.7/2.9 10.0/7.1 10.0/9.9
PFHxA 1.7/3.1 2.4/3.6 0.3*/2.1 0.7/1.2 1.0/0.4 2.9/4.9 2.6/4.9 7.3/2.9 10.0/7.5 3.0/1.9
PFPeA 5.0/7.2 No ads >20*/0.1* 0.9/1.3 0.4/0.6 2.0/1.5 No ads 3.9/5.6 10.0/5.8 9.9/7.2
PFBA No ads No ads >20/0.7 1.0/2.2 0.7/0.8 No ads No ads 3.8/2.5 8.9/6.8 6.8/6.0
PFOS 7.8/14 5.9/26.3 >20/>20 1.9/3.1 0.2/22 4.5/9.9 7.8/9.9 9.9/9.9 10.0/9.9 9.8/9.9
PFHxS 4.0/>20 2.4/8.7 >20/16 0.7/0.9 0.1/0.4 1.1/3.9 2.1/9.7 9.9/9.0 10.0/7.9 10.0/9.2
PFBS 11.4/>20 No ads n/m 0.1*/0.7 0.1/0.6 0.4/1.2 No ads n/m 8.4/2.1 6.7/3.4
PFOSA 5.2/3.9 2.6/6.3 n/m 5.6/2.3 0.6*/0.4* 9.4/8.0 9.0/8.3 n/m 9.4/6.7 8.8/8.4
PFOSaAm 1.9/>20 0.2/0.3 >20/12 1.1*/1.6 0.3/1.4 8.8/9.9 9.9/10.0 9.9/9.3 6.7/5.9 9.3/8.0


Increasing the ionic strength led to an approximately 2-fold increase in the rate of adsorption of PFAS solutes to SOMS and F-SOMS (Table 3). Ionic strength had little impact on the rate of adsorption to poly-SOMS or GAC, with the exception of PFOS adsorption which was more rapid in the presence of 50 mM NaCl. Overall, the kinetic data indicate that there may be distinct differences in the barriers to mass transport and/or adsorption mechanisms depending on adsorbent type. Adsorption to SOMS and F-SOMS appears to be facilitated by the open pore system via hydrophobic interactions since minimal adsorption of short-chain PFAS was observed.

The affinity of PFAS adsorption, as measured by the adsorption capacity at equilibrium (qe), varied depending on the adsorbent. Hydrophobic adsorbents SOMS and F-SOMS showed negligible adsorption of short-chain PFAS solutes including PFBS and PFBA suggesting that the degree of affinity between the fluoroalkyl groups and the adsorbent is too weak. The presence of fluoroalkyl groups in F-SOMS did not lead to a substantial increase in affinity of other PFAS solutes compared to SOMS putting into question the hypothesized improvement of affinity due to specific interactions with fluoroalkyl groups.

GAC and poly-SOMS showed broad effectiveness to adsorb PFAS compounds regardless of chain length and polar group. Poly-SOMS had the highest affinity demonstrating near complete removal of all PFAS solutes except for the cationic PFOSaAm compound which was presumably hindered due to like-charge repulsion. Elevated ionic strength had a deleterious impact on the adsorption of shorter chain compounds to poly-SOMS including PFBA, PFBS, and PFPeA. Overall, GAC, and in particular poly-SOMS, show consistent performance in term of adsorption rate and capacity for a wide range of PFAS solutes.

Ionic interactions may play an important role in the ability of poly-SOMS to adsorb short-chain PFAS such as PFBA. The modifying polymer is 30% imidazolium and 2.0 meq g−1 of cationic groups (resulting in 0.16 meq g−1 for poly-SOMS). Thus, an ionic interaction is possible between the imidazolium and the carboxylate or sulfonates of PFAS. Decreases in the affinities of PFBA, PFBS, and PFPeA upon addition of 50 mM NaCl as measured by a reduction in qe also suggest that ionic interactions play a role in binding short chain PFAS.

Adsorbent isotherms

Individual adsorption isotherms for all 12 PFAS compounds binding to SOMS, F-SOMS, poly-SOMS, and GAC were measured at ionic strengths of 0 mM and 50 mM NaCl, pH 6.6. Poly-SOMS was found to have on average the highest capacity across the set of PFAS solutes tested (Fig. 4 for representative compounds PFBA, PFBS, PFOS, PFOS; see ESI Fig. S4 for all isotherms). Adsorption isotherms for poly-SOMS typically did not reach saturation under the conditions tested. The lack of saturation in the adsorption to poly-SOMS was unusual given that the pore volume of 0.21 mL g−1 was the lowest for all the adsorbents (Table 2). It is hypothesized that swelling of poly-SOMS induced by adsorption of PFAS provides a mechanism to generate additional pore volume.
image file: c9ew00364a-f4.tif
Fig. 4 Adsorption isotherms for PFBA, PFBS, PFOA, and PFOS to SOMS (○); F-SOMS (image file: c9ew00364a-u4.tif); poly-SOMS (image file: c9ew00364a-u5.tif); and GAC (●). Dosage: 40 mg L−1, 14 h contact time, constant agitation, DI water.

PFBA yielded isotherms that differed from other PFAS solutes (Fig. 4). First, adsorption to SOMS and F-SOMS was minimal, as also observed in the kinetics measurements. Second, adsorption of PFBA to poly-SOMS and GAC from DI water reaches a maximum and then decreases at higher concentrations. The unusual shape of the isotherm may be due to the substantial fraction of PFBA in solution altering the equilibrium for adsorption. Such behaviour may be important in understanding the challenges in removing PFBA from water. A standard isotherm for PFBA was observed at elevated ionic strength (see ESI).

Adsorption isotherms were fit to the Freundlich isotherm (for complete data set see ESI Table S7). Based on the fitted equations to the Freundlich model, capacity at Ce = 200 μg L−1 was calculated (Fig. 5). The Ce value was selected because the concentration is near environmental relevance and was within the measured equilibrium concentration range of all data sets. Several notable results were observed. Both poly-SOMS and GAC showed broad adsorption performance across the entire set of PFAS. Poly-SOMS generally had 2–5× the capacity compared to GAC at equilibrium and was the best performing adsorbent in DI water reaching a maximum adsorption capacity of 22.8 ± 0.6 mg g−1 for PFHxA. In contrast, SOMS and F-SOMS were less effective in adsorbing short-chain PFAS compounds from DI water indicative that interactions with the fluoroalkyl group are insufficient to provide removal when the chain length is smaller. Thus, the hypothesis that the introduction of fluoroalkyl groups would increase affinity of PFAS adsorption is not supported by the data. In general, F-SOMS and SOMS have similar capacity for each individual PFAS compound. Hydrophobic surfaces appear to be sufficient to promote adsorption of long-chain PFAS compounds regardless of surface composition assuming the fluoroalkyl groups added by co-polymerization are accessible to make adsorbate interactions. These results differ from previous reports where organofluorine modified adsorbents33 have demonstrated better performance to the adsorption of PFOA and PFOS. What is unknown is whether modification with fluoroalkyl groups leads to greater selectivity in adsorption of PFAS compounds in mixtures with non-fluorinated organic solutes. Selectivity will be tested in the future to fully understand whether there are benefits to creating fluoroalkyl-modified adsorbents.


image file: c9ew00364a-f5.tif
Fig. 5 Adsorption capacities for anionic PFAS by organosilica compounds and GAC in DI water (left) and 50 mM NaCl (right) at Ce= 200 μg L−1.

Addition of fluoroalkyl groups to SOMS was further explored by varying the amount of fluoroalkyl content by varying the relative amounts of precursor 1 to 2 from 0–0.47 mol mol−12. The BET surface area, pore volume, and capacity to adsorb PFOS were measured (see ESI Table S8). Results show that the affinity of PFOS remains constant for mole fractions of 2 between 0–0.13 mol mol−1. As the fraction of fluoroalkyl precursor increases beyond 0.13 mol mol−1, PFOS capacity decreases with a concomitant decrease in surface area and pore volume of the adsorbent. These data suggest that a more open pore structure provides for additional capacity. The material prepared with 0.13 mol mol−12 was chosen as F-SOMS since the surface area, pore volume, and swelling was equivalent to SOMS while possessing the maximum amount of fluoroalkyl groups.

The adsorption capacity of neutral and cationic PFAS solutes was also evaluated. GAC and poly-SOMS yielded capacities equivalent to short chain anionic PFAS (Table 4). Adsorption capacity of poly-SOMS for the cationic PFOSaAm was measurable, but lower than any other PFAS including PFBA. Adsorption of PFOSA and PFOSaAm by SOMS and F-SOMS was substantial. F-SOMS had adsorption capacity at Ce = 200 μg L−1 of >30 mg g−1 for both compounds. F-SOMS showed a 2-fold increase in capacity for PFOSA over SOMS (35 vs. 17 mg PFOSA per g adsorbent). As a result, there may be improved affinity upon the addition of fluoroalkyl groups in the absence of adsorbate charge.

Table 4 Adsorption capacity for neural and cationic PFAS
PFAS Adsorbent Capacity (mg g−1)
C e = 200 μg L−1a
a Error ± 5%. Dosage: 40 mg L−1, 14 h contact time.
PFOSA (neutral) SOMS 17.7
F-SOMS 34.9
Poly-SOMS 8.8
GAC 5.1
PFOSaAm (cationic) SOMS 35.5
F-SOMS 36.7
Poly-SOMS 2.0
GAC 4.9


Ionic strength was varied to determine the effect on adsorption. Under conditions of elevated ionic strength (50 mM NaCl) SOMS and F-SOMS had enhanced adsorption capacity for PFOA and especially PFOS (Fig. 5). For example, adsorption capacity of PFOS was 67 ± 3 mg g−1 adsorbed to F-SOMS at Ce 200 = μg L−1, which is 4× what was observed for GAC under the same conditions. Enhancement of adsorption of PFAS to hydrophobic SOMS in the presence of salt occurred for only longer-chain compounds ≥C7 and thus may be attributable to the “salting out effect” where the partitioning onto the adsorbent is enhanced in presence of dissolved ions because the water is more ordered and compressible and the cavity volume available to accommodate an in situ fluoroalkyl group is reduced.54 Higher capacities observed for longer chain PFAS may also be due to enhancement of adsorbate–adsorbate interactions, with one explanation being that higher ionic strength screens ion–ion repulsions of the PFAS headgroups resulting in micelle formation once the concentration in the pores becomes high enough. When examining the effect of elevated ionic strength on the performance of GAC, addition of salt generally improves adsorption of longer-chain PFAS, but negatively impacts adsorption of short chain compounds such as PFBA. Reduction in adsorption of short-chain compounds suggests that the ability of GAC to adsorb PFAS made be due in part to ion exchange. The effect of ionic strength on the adsorption of poly-SOMS shows a strong dependence on the length of the PFAS chain length (Fig. 6). Adsorption isotherms for longer chain compounds such as PFDA, PFNA, and PFOA show no dependence on ionic strength, suggesting that adsorption of poly-SOMS by these compounds is driven primarily by interactions of the fluoroalkyl groups with the sorbent and themselves via adsorbate–adsorbate interactions. In contrast, the addition of salt reduced affinity of short chain PFAS compounds to poly-SOMS, suggesting that ionic interactions with polymer quaternary amine groups may be the predominant interaction. Overall, the data suggest that successful adsorption of a wide range of PFAS compounds requires multiple types of interactions: ionic interactions to capture short chain compounds and hydrophobic interactions to promote association of long-chain compounds. Regardless of PFAS compound, poly-SOMS yielded higher capacity than carbon at either ionic strength condition.


image file: c9ew00364a-f6.tif
Fig. 6 Adsorption isotherms of PFAS solutes to poly-SOMS in DI water (image file: c9ew00364a-u6.tif) and 50 mM NaCl (□).

Modification of SOMS by the addition of fluoroalkyl groups or quaternary amine groups was useful in that the effects due to differences of surface chemistry could be evaluated independently of morphology by using an equivalent organosilica scaffold with similar pore structure. The pores can be expanded by swelling which presumably aids mass transport by further reducing differences in adsorption to changes in pore size. Longer-chain PFAS compounds were adsorbed to hydrophobic SOMS. Adsorption of the full range of anionic PFAS of all chain lengths appears to require adsorbents that are hydrophobic and possess cationic groups. GAC was generally effective across the full range of PFAS compounds suggesting that the surface is also heterogeneous (hydrophobic, ionic) and/or that smaller pores assist in adsorption of short-chain PFAS.

Column experiments

Granular adsorbents were packed into beds and used to treat water containing PFAS compounds to evaluate treatment performance. QA-SOMS yielded material too fine to create a granular bed of adsorbent, so was not tested. Single component solutions were applied to F-SOMS, poly-SOMS, and GAC at an effluent concentration of C0 = 200 μg L−1. Single PFAS compounds were tested to evaluate how equilibrium adsorption isotherms match data obtained from column experiments where the contact time is significantly shorter. For all PFAS compounds tested (PFOS, PFOA, and PFBA) SOMS, F-SOMS, and GAC demonstrated a capacity at breakthrough that was less than the equilibrium capacity at Ce = 200 μg L−1 (Table 5). For example, F-SOMS had a capacity of 0.7 mg PFOS g−1 at column breakthrough Ceffluent = 40 μg L−1vs. 8.5 mg PFOS per g at Ce = 200 μg L−1 as measured using an adsorption isotherm. The comparatively lower capacity at breakthrough is expected since the breakthrough point was determined when the effluent is 5× lower than 200 μg L−1 (i.e. pre-equilibrium). In addition, and the contact time on the column was less compared to batch experiments used to measure the isotherm (2 min vs. 14 h). Interestingly, the capacity values at column breakthrough for poly-SOMS were much higher and exceeded the capacity as measured for Ce = 200 μg L−1 using the adsorption isotherm. For example, 47 mg PFOA per g was adsorbed at the breakthrough concentration of C/C0 < 0.2 (40 μg L−1) as compared to 12 ± 1 mg g−1 at Ce = 200 μg L−1 on the isotherm. A possible explanation for the higher adsorption capacity of poly-SOMS during continuous long-term flow is that 14 h was insufficient for equilibrium to be established in batch experiments. It is hypothesized that during the time of column flow, PFOA was strongly adsorbed to poly-SOMS and able to diffuse into the polymer filled pores. Due to the process of inward diffusion, the PFAS adsorbates may become deeply sequestered. Swelling may also contribute to the increased capacity in continuous flow experiments where adsorption may have triggered matrix expansion and addition sites to adsorb PFOA (see below). Further work is ongoing to explore whether the increased capacity by poly-SOMS is observed for other PFAS solutes during continuous flow. The increase in capacity for PFAS removal under continuous flow is fortuitous since using a granular media bed is the practical way such material would be used for treatment.
Table 5 Adsorption capacity at breakthrough for PFOA, PFOS, and PFBA
PFAS Capacity at breakthrough (mg g−1)
Matrix: DI water Matrix: 50 mM NaCl
F-SOMS Poly-SOMS GAC F-SOMS GAC
a n/m = not measured.
PFOS 0.7 n/ma 2.2 14.2 0.5
PFOA 0.9 >47 1.5 9.9 0.5
PFBA 0 10.9 0.1 n/m n/m


A mixture of all 12 PFAS compounds (C0 = 200 μg L−1 each) was applied to a poly-SOMS column in order to test the removal of in the presence of co-solutes. Breakthrough for each PFAS compound was distinct for each substance with shorter chain compounds (ex. PBFA) eluting first (Fig. 7) similar to results seem previously for GAC and ion exchange resins.13 Interestingly, the poly-SOMS showed a temporary improvement in treatment performance after 15[thin space (1/6-em)]000 applied bed volumes resulting in a dip in the breakthrough curve. The cause of the temporary increase in PFAS affinity is unknown, but may a manifestation of adsorbate-induced swelling (note: backpressure increased 5–10% during the course of the experiment which may be a manifestation of loss of intra-particle bed volume). After 65[thin space (1/6-em)]000 applied bed volumes 11/12 PFAS compounds had reached breakthrough. Neutral PFOSA was still being fully removed and long-chain PFDA and PFOS compounds also showed strong removal at the 65[thin space (1/6-em)]000 bed-volume mark. Displacement of PFBA and PFPeA began to occur at the 50[thin space (1/6-em)]000-bed volume mark at the time PFOA reached breakthrough. The total amount of adsorbed PFAS was 80 mg g−1 at 65[thin space (1/6-em)]000 bed-volumes.


image file: c9ew00364a-f7.tif
Fig. 7 Column breakthrough experiment for poly-SOMS. Influent was 200 μg L−1 of each PFAS solute in DI water pH 6.6. Contact time, 2 min.

FT-IR spectrometry taken before use and after recovery of poly-SOMS from the column, determined that >85% of the initial polymer, 4, remained after 65[thin space (1/6-em)]000 bed volumes (see ESI Fig. S5). The distinct amide band at 1650 cm−1 of 4 was used for quantitation. Loss in the polymer was likely due to rinsing off residues from the surface that was unentrapped during preparation (see above).

Scanning electron microscopy was used to examine poly-SOMS as synthesized compared to poly-SOMS recovered after application of 65[thin space (1/6-em)]000 bed volumes of 200 μg mL−1 each PFAS as a mixture (Fig. 8). Prior to treatment the surface of poly-SOMS generally resembled SOMS with a general morphology of interconnected organosilica colloid particles (Fig. 8A). Polymer comprising 8% w/w of poly-SOMS may be evident by some of the pores being less distinct than SOMS (Fig. 2A). After PFAS adsorption the surface of the poly-SOMS particles was much smoother with >1 μm globular structures adhering to various places on the surface (Fig. 8B). It is presumed that these structures are comprised of accumulated PFAS as there were no other solutes in the water. Similarly, macro-aggregates of PFAS have been observed by transmission electron microscopy in spent ion exchange resins.17 Poly-SOMS particles were fractured by short duration grinding to expose the interior pore structures (Fig. 8C and D). Electron microscopy revealed a generally similar pore morphology between as synthesized poly-SOMS versus resin removed from the column, although after PFAS adsorption appearance of residues in some regions was observed (Fig. 8D). Based on the microscopy, there is evidence that PFAS accumulates at or near the surface of poly-SOMS in multi-layer or macro-aggregate structures. The hypothesis that swelling was involved in enhancing capacity was not supported by electron microscopy as the pore sizes are generally equivalent between fresh and used resin. Instead, capacity may be influenced by PFAS interacting with co-adsorbates. The displacement of short-chain PFBA during the mixed PFAS column experiment may be due to replacement of PFBA with a long-chain PFAS compound in regions of accumulated PFAS, although there is no direct evidence for this hypothesis and further testing would be required in future work.


image file: c9ew00364a-f8.tif
Fig. 8 SEM micrographs of poly-SOMS as synthesized (A & C) vs. after recovery from column mixed PFAS post-breakthrough (B & D). A &B are images of the surface of the adsorbent particles. C & D are the interior of the adsorbents accessed via grinding.

Measurement of treatment performance at low PFAS concentrations was used to determine if SOMS-based adsorbent had sufficient affinity to achieve effluent <70 ng L−1 given an environmentally relevant input concentration of 2.0 μg L−1. The breakthrough point (C/C0 = 0.2) for 2.0 μg L−1 PFOS on a SOMS column with an ionic strength of 50 mM occurred at 12[thin space (1/6-em)]000 column volumes (see ESI Table S9). The PFOS effluent concentration was 72 ng L−1 for SOMS and 37 ng L−1 for F-SOMS averaged over the first 4000 column volumes indicating that the fluoroalkyl-modified material may yield higher affinity at low concentrations. Overall, initial results show that SOMS materials have the capability to treat water to meet guidelines. Future work will require exploration of the effectives of treatment when compared to other adsorbents in addition to the treatment effectiveness in natural waters including ground water and surface water.

Conclusions

The ability of modified porous organosilica materials to adsorb a range of PFAS compounds was explored relative to GAC. The organosilica adsorbents (SOMS) had the ability to swell and could be modified during synthesis to alter the pore surface chemical composition. Hydrophobic SOMS and fluoroalkyl-modified F-SOMS demonstrated the ability to adsorb long-chain PFAS solutes, especially under conditions of high ionic strength. Poly-SOMS modified to encapsulate a commercially available cationic polymer had the ability to adsorb the full range of PFAS solutes with adsorption capacities higher than GAC. Poly-SOMS showed enhanced adsorption capacity during column experiments over capacities measured in batch equilibrium experiments. Data showed that porous hydrophobic adsorbents removed long-chain PFAS solutes from water. Cationic groups were necessary to adsorb short-chain PFAS anions presumably by ionic interaction in the absence of interactions provided by longer fluoroalkyl groups. Cationic groups on the surface may assist in maintaining charge neutrality, especially if PFAS adsorbates aggregate. Adsorbate–adsorbate interactions may play an important role in increasing adsorption capacity.

Conflicts of interest

The process to prepare the organosilica materials described within have been patented by The College of Wooster and licensed to ABS Materials, Inc. The corresponding author has a financial interest in the commercialization of the patents.

Acknowledgements

The research was funded in part by the Strategic Environmental Research and Development Program (grant ER18-1300) through the U.S. Department of Defense. Authors wish to thank the Molecular and Cellular Imaging Center at The Ohio State University for assistance with SEM.

Notes and references

  1. K. A. Barzen-Hanson, S. C. Roberts, S. Choyke, K. Oetjen, A. McAlees, N. Riddell, R. McCrindle, P. L. Ferguson, C. P. Higgins and J. A. Field, Discovery of 40 Classes of Per- and Polyfluoroalkyl Substances in Historical Aqueous Film-Forming Foams (AFFFs) and AFFF-Impacted Groundwater, Environ. Sci. Technol., 2017, 51, 2047–2057 CrossRef CAS.
  2. M. M. Schultz, D. F. Barofsky and J. A. Field, Fluorinated Alkyl Surfactants, Environ. Eng. Sci., 2003, 20, 487–501 CrossRef CAS.
  3. A. G. Paul, K. C. Jones and A. J. Sweetman, A First Global Production, Emission, And Environmental Inventory For Perfluorooctane Sulfonate, Environ. Sci. Technol., 2009, 43, 386–392 CrossRef CAS PubMed.
  4. S. Posner, in Polyfluorinated Chemicals and Transformation Products, ed. T. P. Knepper and F. T. Lange, Springer Berlin Heidelberg, Berlin, Heidelberg, 2012, pp. 25–39 Search PubMed.
  5. M. Kotthoff and M. Bücking, Four Chemical Trends Will Shape the Next Decade's Directions in Perfluoroalkyl and Polyfluoroalkyl Substances Research, Front. Chem., 2018, 6, 103 CrossRef.
  6. Z. Zeng, B. Song, R. Xiao, G. Zeng, J. Gong, M. Chen, P. Xu, P. Zhang, M. Shen and H. Yi, Assessing the human health risks of perfluorooctane sulfonate by in vivo and in vitro studies, Environ. Int., 2019, 126, 598–610 CrossRef CAS.
  7. J. S. Kang, T.-G. Ahn and J.-W. Park, Perfluorooctanoic acid (PFOA) and perfluooctane sulfonate (PFOS) induce different modes of action in reproduction to Japanese medaka (Oryzias latipes), J. Hazard. Mater., 2019, 368, 97–103 CrossRef CAS.
  8. P. Song, D. Li, X. Wang and X. Zhong, Effects of perfluorooctanoic acid exposure during pregnancy on the reproduction and development of male offspring mice, Andrologia, 2018, 50, e13059 CrossRef.
  9. B. B. Høyer, J. P. Bonde, S. S. Tøttenborg, C. H. Ramlau-Hansen, C. Lindh, H. S. Pedersen and G. Toft, Exposure to perfluoroalkyl substances during pregnancy and child behaviour at 5 to 9years of age, Horm. Behav., 2018, 101, 105–112 CrossRef.
  10. L. Zheng, G.-H. Dong, Y.-H. Jin and Q.-C. He, Immunotoxic changes associated with a 7-day oral exposure to perfluorooctanesulfonate (PFOS) in adult male C57BL/6 mice, Arch. Toxicol., 2009, 83, 679–689 CrossRef CAS.
  11. J. C. DeWitt, M. M. Peden-Adams, J. M. Keller and D. R. Germolec, Immunotoxicity of Perfluorinated Compounds: Recent Developments, Toxicol. Pathol., 2012, 40, 300–311 CrossRef CAS.
  12. J. D. McNamara, R. Franco, R. Mimna and L. Zappa, Comparison of Activated Carbons for Removal of Perfluorinated Compounds From Drinking Water: Comparison of Activated Carbons for Removal of Perfluorinated Compounds From Drinking Water, J. - Am. Water Works Assoc., 2018, 110, E2–E14 CrossRef CAS.
  13. P. McCleaf, S. Englund, A. Östlund, 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.
  14. P. Westreich, R. Mimna, J. Brewer and F. Forrester, The removal of short-chain and long-chain perfluoroalkyl acids and sulfonates via granular activated carbons: A comparative column study, Remediation, 2018, 29, 19–26 CrossRef.
  15. D. Zhang, Q. Luo, B. Gao, S.-Y. D. Chiang, D. Woodward and Q. Huang, Sorption of perfluorooctanoic acid, perfluorooctane sulfonate and perfluoroheptanoic acid on granular activated carbon, Chemosphere, 2016, 144, 2336–2342 CrossRef CAS PubMed.
  16. S. Deng, Q. Yu, J. Huang and G. Yu, Removal of perfluorooctane sulfonate from wastewater by anion exchange resins: Effects of resin properties and solution chemistry, Water Res., 2010, 44, 5188–5195 CrossRef CAS.
  17. A. Zaggia, L. Conte, L. Falletti, M. Fant and A. Chiorboli, Use of strong anion exchange resins for the removal of perfluoroalkylated substances from contaminated drinking water in batch and continuous pilot plants, Water Res., 2016, 91, 137–146 CrossRef CAS.
  18. K. E. Carter and J. Farrell, Removal of Perfluorooctane and Perfluorobutane Sulfonate from Water via Carbon Adsorption and Ion Exchange, Sep. Sci. Technol., 2010, 45, 762–767 CrossRef CAS.
  19. S. Woodard, J. Berry and B. Newman, Ion exchange resin for PFAS removal and pilot test comparison to GAC, Remediation, 2017, 27, 19–27 CrossRef.
  20. A. Zaggia, L. Conte, L. Falletti, M. Fant and A. Chiorboli, Use of strong anion exchange resins for the removal of perfluoroalkylated substances from contaminated drinking water in batch and continuous pilot plants, Water Res., 2016, 91, 137–146 CrossRef CAS.
  21. S. T. M. L. D. Senevirathna, S. Tanaka, S. Fujii, C. Kunacheva, H. Harada, B. R. Shivakoti and R. Okamoto, A comparative study of adsorption of perfluorooctane sulfonate (PFOS) onto granular activated carbon, ion-exchange polymers and non-ion-exchange polymers, Chemosphere, 2010, 80, 647–651 CrossRef CAS.
  22. I. Ross, J. McDonough, J. Miles, P. Storch, P. T. Kochunarayanan, E. Kalve, J. Hurst, S. S. Dasgupta and J. Burdick, A review of emerging technologies for remediation of PFASs, Remediation, 2018, 28, 101–126 CrossRef.
  23. Z. Du, S. Deng, Y. Bei, Q. Huang, B. Wang, J. Huang and G. Yu, Adsorption behavior and mechanism of perfluorinated compounds on various adsorbents—A review, J. Hazard. Mater., 2014, 274, 443–454 CrossRef CAS.
  24. C. L. Wilder, Community exposures to per-and polyfluoroalkyl substances in drinking water: A national issue, J. Environ. Health, 2017, 80, 38 Search PubMed.
  25. M. Clara, S. Scharf, S. Weiss, O. Gans and C. Scheffknecht, Emissions of perfluorinated alkylated substances (PFAS) from point sources—identification of relevant branches, Water Sci. Technol., 2008, 58, 59–66 CrossRef CAS PubMed.
  26. M. Ateia, A. Maroli, N. Tharayil and T. Karanfil, The overlooked short- and ultrashort-chain poly- and perfluorinated substances: A review, Chemosphere, 2019, 220, 866–882 CrossRef CAS.
  27. T. D. Appleman, C. P. Higgins, O. Quiñones, B. J. Vanderford, C. Kolstad, J. C. Zeigler-Holady and E. R. V. Dickenson, Treatment of poly- and perfluoroalkyl substances in U.S. full-scale water treatment systems, Water Res., 2014, 51, 246–255 CrossRef CAS.
  28. W. Ji, L. Xiao, Y. Ling, C. Ching, M. Matsumoto, R. P. Bisbey, D. E. Helbling and W. R. Dichtel, Removal of GenX and Perfluorinated Alkyl Substances from Water by Amine-Functionalized Covalent Organic Frameworks, J. Am. Chem. Soc., 2018, 140, 12677–12681 CrossRef CAS PubMed.
  29. M. Ateia, M. F. Attia, A. Maroli, N. Tharayil, F. Alexis, D. C. Whitehead and T. Karanfil, Rapid Removal of Poly- and Perfluorinated Alkyl Substances by Poly(ethylenimine)-Functionalized Cellulose Microcrystals at Environmentally Relevant Conditions, Environ. Sci. Technol. Lett., 2018, 5, 764–769 CrossRef CAS.
  30. S. Lath, D. A. Navarro, D. Losic, A. Kumar and M. J. McLaughlin, Sorptive remediation of perfluorooctanoic acid (PFOA) using mixed mineral and graphene/carbon-based materials, Environ. Chem., 2018, 15, 472–480 CrossRef CAS.
  31. N. B. Saleh, A. Khalid, Y. Tian, C. Ayres, I. V. Sabaraya, J. Pietari, D. Hanigan, I. Chowdhury and O. G. Apul, Removal of poly- and per-fluoroalkyl substances from aqueous systems by nano-enabled water treatment strategies, Environ. Sci.: Water Res. Technol., 2019, 5, 198–208 RSC.
  32. P.-J. Huang, M. Hwangbo, Z. Chen, Y. Liu, J. Kameoka and K.-H. Chu, Reusable Functionalized Hydrogel Sorbents for Removing Long- and Short-Chain Perfluoroalkyl Acids (PFAAs) and GenX from Aqueous Solution, ACS Omega, 2018, 3, 17447–17455 CrossRef CAS.
  33. L. Xiao, Y. Ling, A. Alsbaiee, C. Li, D. E. Helbling and W. R. Dichtel, β-Cyclodextrin Polymer Network Sequesters Perfluorooctanoic Acid at Environmentally Relevant Concentrations, J. Am. Chem. Soc., 2017, 139, 7689–7692 CrossRef CAS.
  34. A. Rostvall, W. Zhang, W. Dürig, G. Renman, K. Wiberg, L. Ahrens and P. Gago-Ferrero, Removal of pharmaceuticals, perfluoroalkyl substances and other micropollutants from wastewater using lignite, Xylit, sand, granular activated carbon (GAC) and GAC+Polonite® in column tests – Role of physicochemical properties, Water Res., 2018, 137, 97–106 CrossRef CAS.
  35. V. Ochoa-Herrera and R. Sierra-Alvarez, Removal of perfluorinated surfactants by sorption onto granular activated carbon, zeolite and sludge, Chemosphere, 2008, 72, 1588–1593 CrossRef CAS.
  36. Q. Zhang, S. Deng, G. Yu and J. Huang, Removal of perfluorooctane sulfonate from aqueous solution by crosslinked chitosan beads: Sorption kinetics and uptake mechanism, Bioresour. Technol., 2011, 102, 2265–2271 CrossRef CAS PubMed.
  37. L. L. Hench and J. K. West, The sol-gel process, Chem. Rev., 1990, 90, 33–72 CrossRef CAS.
  38. C. M. Burkett and P. L. Edmiston, Highly swellable sol–gels prepared by chemical modification of silanol groups prior to drying, J. Non-Cryst. Solids, 2005, 351, 3174–3178 CrossRef CAS.
  39. C. M. Burkett, L. A. Underwood, R. S. Volzer, J. A. Baughman and P. L. Edmiston, Organic–Inorganic Hybrid Materials that Rapidly Swell in Non-Polar Liquids: Nanoscale Morphology and Swelling Mechanism, Chem. Mater., 2008, 20, 1312–1321 CrossRef CAS.
  40. W. Chen, X. Zhang, M. Mamadiev and Z. Wang, Sorption of perfluorooctane sulfonate and perfluorooctanoate on polyacrylonitrile fiber-derived activated carbon fibers: in comparison with activated carbon, RSC Adv., 2017, 7, 927–938 RSC.
  41. Á. Piñeiro, G. Prieto, J. M. Ruso, P. V. Verdes and F. Sarmiento, Surface films of short fluorocarbon–hydrocarbon diblocks studied by molecular dynamics simulations: Spontaneous formation of elongated hemimicelles, J. Colloid Interface Sci., 2009, 329, 351–356 CrossRef.
  42. 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.
  43. M. Nassi, E. Sarti, L. Pasti, A. Martucci, N. Marchetti, A. Cavazzini, F. Di Renzo and A. Galarneau, Removal of perfluorooctanoic acid from water by adsorption on high surface area mesoporous materials, J. Porous Mater., 2014, 21, 423–432 CAS.
  44. F. Hof, D. M. Scofield, W. B. Schweizer and F. Diederich, A Weak Attractive Interaction between Organic Fluorine and an Amide Group, Angew. Chem., Int. Ed., 2004, 43, 5056–5059 CrossRef CAS PubMed.
  45. 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.
  46. S. Brunauer, P. H. Emmett and E. Teller, Adsorption of Gases in Multimolecular Layers, J. Am. Chem. Soc., 1938, 60, 309–319 CrossRef CAS.
  47. E. P. Barrett, L. G. Joyner and P. P. Halenda, The Determination of Pore Volume and Area Distributions in Porous Substances. I. Computations from Nitrogen Isotherms, J. Am. Chem. Soc., 1951, 73, 373–380 CrossRef CAS.
  48. Y.-S. Ho, Review of second-order models for adsorption systems, J. Hazard. Mater., 2006, 136, 681–689 CrossRef CAS.
  49. S. P. J. van Leeuwen and J. de Boer, Extraction and clean-up strategies for the analysis of poly- and perfluoroalkyl substances in environmental and human matrices, J. Chromatogr. A, 2007, 1153, 172–185 CrossRef CAS.
  50. N. Yamashita, K. Kannan, S. Taniyasu, Y. Horii, T. Okazawa, G. Petrick and T. Gamo, Analysis of Perfluorinated Acids at Parts-Per-Quadrillion Levels in Seawater Using Liquid Chromatography-Tandem Mass Spectrometry, Environ. Sci. Technol., 2004, 38, 5522–5528 CrossRef CAS.
  51. H. Aguiar, J. Serra, P. González and B. León, Structural study of sol–gel silicate glasses by IR and Raman spectroscopies, J. Non-Cryst. Solids, 2009, 355, 475–480 CrossRef CAS.
  52. S. Park, S. Nam, L. Kim, M. Park, J. Kim, T. K. An, W. M. Yun, J. Jang, J. Hwang and C. E. Park, Synthesis and characterization of a fluorinated oligosiloxane-containing encapsulation material for organic field-effect transistors, prepared via a non-hydrolytic sol–gel process, Org. Electron., 2012, 13, 2786–2792 CrossRef CAS.
  53. P. L. Edmiston, A. R. Gilbert, Z. Harvey and N. Mellor, Adsorption of short chain carboxylic acids from aqueous solution by swellable organically modified silica materials, Adsorption, 2018, 24, 53–63 CrossRef CAS.
  54. S. Endo, A. Pfennigsdorff and K.-U. Goss, Salting-Out Effect in Aqueous NaCl Solutions: Trends with Size and Polarity of Solute Molecules, Environ. Sci. Technol., 2012, 46, 1496–1503 CrossRef CAS.

Footnote

Electronic supplementary information (ESI) available. See DOI: 10.1039/c9ew00364a

This journal is © The Royal Society of Chemistry 2019