Contents lists available at ScienceDirect

Journal of Hazardous Materials

journal homepage: www.elsevier.com/locate/jhazmat

Stabilization and solidification remediation of soil contaminated with poly-and perfluoroalkyl substances (PFASs)Mattias Sörengårda,⁎, Dan B. Klejab, Lutz Ahrensaa Department of Aquatic Sciences and Assessment, Swedish University of Agricultural Sciences (SLU), P.O. Box 7050, 75007, Uppsala, Swedenb Swedish Geotechnical Institute, Kornhamnstorg 61, 111 27, Stockholm, Sweden

G R A P H I C A L A B S T R A C T

A R T I C L E I N F O

Keywords:PFASPFOSStabilzationSolidificationRemediation

A B S T R A C T

Remediation methods for soils contaminated with poly- and perfluoroalkyl substances (PFASs) are urgently neededto protect the surrounding environment and drinking water source areas from pollution. In this study, the stabi-lization and solidification (S/S) technique was tested on aged PFAS-contaminated soil that were artificially spikedwith 14 PFAS. To further reduce leaching of PFASs in S/S-treated soil, seven different additives were tested at 2%concentration: powdered activated carbon (PAC), Rembind®, pulverized zeolite, chitosan, hydrotalcite, bentonite,and calcium chloride. Standardized leaching tests on S/S-treated soil revealed that leaching of 13 out of 14 targetPFASs (excluding perfluorobutane sulfonate (PFBA)) was reduced by, on average, 70% and 94% by adding PAC andRembind®. Longer-chained PFASs such as perfluorooctane sulfonate (PFOS), which is considered persistent,bioaccumulative and toxic, were stabilized by 99.9% in all S/S treatments when PAC or Rembind® was used as anadditive. The S/S stabilization efficiency depended on PFAS perfluorocarbon chain length and functional group,e.g., it increased on average by 11–15 % per CF3-moeity and was on average 49% higher for the perfluorosulfonates(PFSAs) than the perfluorocarboxylates (PFCAs). Overall, the S/S treatment with active carbon-based additivesshowed excellent performance in reducing leaching of PFASs, without marked loss of physical matrix stability.

https://doi.org/10.1016/j.jhazmat.2019.01.005Received 3 September 2018; Received in revised form 31 December 2018; Accepted 2 January 2019

⁎ Corresponding author.E-mail address: mattias.sorengard@slu.se (M. Sörengård).

Journal of Hazardous Materials 367 (2019) 639–646

Available online 03 January 20190304-3894/ © 2019 Elsevier B.V. All rights reserved.

T

1. Introduction

Poly- and perfluoroalkyl substances (PFASs) are an emerging anddiversified group of synthetic chemicals with unique physicochemicalproperties [1,2]. Due to these properties, PFASs are used in numerousindustrial and consumer products, such as electronics, clothing, cook-ware, lubricants, and aqueous firefighting foam (AFFF) [2]. Typicalpoint sources of PFAS emissions are wastewater treatment plants,landfill leachates, biosolid fertilizers applied to agricultural land, andAFFF spills at fire-training facilities, while typical non-point sources areatmospheric deposition, indoor air, and street runoff [3]. Use of PFAS-containing AFFF at fire-training facilities is a particular concern be-cause large volumes of AFFFs are released into adjacent soil during ashort period [4], and then transported from the soil to the surroundingenvironment. Because of the soil contamination, firefighter trainingfacilities have been shown to be sources of long-term release of PFASs,even after the introduction of restrictions on use of AFFF [5–7]. Nu-merous studies have also linked PFAS-contaminated soils at firefightertraining facilities to elevated PFAS concentrations in drinking waterwells, resulting in human exposure [8,9].

The PFASs are characterized by their extreme persistency and un-ique physicochemical properties, such as simultaneous hydrophobicityand oleophobicity, and thus are a challenge for conventional soil re-mediation techniques [10]. Different treatment techniques for re-moving PFASs from drinking water and wastewater have been devel-oped [11,12], but water treatment techniques cannot be used directlyfor soil treatment purposes [13,14]. In addition, currently available soilremediation techniques are either too costly to be used (e.g., soil in-cineration, pump and treat systems) [15,16]. An alternative treatmentfor PFAS-contaminated soil and groundwater is phytoremediation,which uses plants for in situ uptake of PFASs and subsequent in-cineration, but this treatment is time-consuming [17]. Another poten-tial remediation method is stabilization of PFASs in soil by addition ofactivated carbon (AC) and modified clays, which has been shown toconsiderably reduce PFAS leaching from various soils 14,18,19].However, previous studies have tested only a few PFASs and a limitednumber of sorbents, and have not assessed the long-term treatmentefficiency.

Stabilization and solidification (S/S) is a well-used remediationtechnique for soils and sediments but, to our knowledge, has not beentested previously for PFAS-contaminated soil. It is a practicable, cost-efficient remediation method and has been used for various othercontaminants and matrices such as soil, sediment, industrial sludge[20], and waste [21]. In-situ S/S remediation is considered more eco-nomically and environmentally sustainable compared to e.g. excavateand landfilling [22]. The S/S method involves mixing cementitiousbinder and additives into the contaminated matrix, either in situ or exsitu, and aims to immobilize contaminants by i) physical protectionthrough matrix solidification, thereby lowering the hydraulic con-ductivity and reducing contaminant exposure to leaching, and ii) che-mical protection through contaminant stabilization by reducing thewater solubility of contaminants through increased precipitation,change of oxidation state, and adsorption, thereby reducing their che-mical solubility [23]. Modeling studies have shown that S/S treatmentcan extend the leaching process by over 100 [24] to 1000 [25] years formetals and polyaromatic hydrocarbons (PAHs). The mechanicalstrength of the treated soil is a key variable for S/S remediation, be-cause porosity and permeability generally decrease with increasingstrength, which affecting contaminant leaching over a longer timeperiod [26]. Mechanical strength development of concrete in S/Streatment applications is complex, because of large variations in che-mical and physical properties of soils and sediments. Therefore, opti-mization of binder ingredients, binder to soil ratio, and water content isrecommended to achieve higher strength and lower leachability foreach compound of interest [27].

The leaching behavior of PFASs is mainly influenced by adsorption

processes, predominantly by both hydrophobic and electrostatic inter-actions [28]. The hydrophobic interactions of PFASs are dependent onchain length and electrostatic interactions among functional groups[29]. The binding interactions can be influenced by adsorbent char-acteristics such as surface area, pore size, surface charge, and organiccontent, but also by solution chemistry parameters such as pH andconcentration of inorganic ions and dissolved organic carbon (DOC)[11,[28]. The S/S treatment uses a cementitious matrix, but PFASs havebeen shown to leach from this kind of matrix, e.g., contaminated con-crete slabs have been identified as environmental PFAS point sources[30]. A successful way to enhance the stabilization of organic micro-pollutants in S/S-treatment is to add a suitable adsorption-enhancingadditive (e.g., AC for PAHs) [31]. Accurate identification of suitableadditives for PFAS-contaminated soil is therefore necessary for thesuccess of S/S remediation.

The overall aim of the present study was to assess whether S/Streatment can be used for remediation of PFAS-contaminated soil. Thiswas done by testing the performance of S/S with and without additivesin terms of the leaching behavior and the mechanical strength of thesolidification. Specific objectives were to i) identify and evaluate con-ventional and novel S/S additives for S/S remediation of PFAS-con-taminated soil, ii) assess the leaching behavior of a wide range of PFASs(n= 14) with varying chain lengths and functional groups following S/S treatment of contaminated soil, and iii) evaluate the mechanicalstability (strength) of the soil after S/S treatment, in order to validatethe applicability and efficiency of this treatment for PFAS-contaminatedsoil.

2. Material and methods

2.1. Analytical standards

The target PFASs comprised C3-C10 perfluoroalkyl carboxylates(PFCAs) (PFBA, PFPeA, PFHxA, PFHpA, PFOA, PFNA, PFDA, PFUnDA),C4, C6, C8 perfluoroalkane sulfonates (PFSAs) (PFBS, PFHxS, PFOS),perfluorooctanesulfonamide (FOSA), and 6:2 and 8:2 fluorotelomersulfonic acids (FTSAs). In total, nine isotopically labeled internal stan-dards (ISs) were used: [13]C4-PFBA, [13]C2-PFHxA, [13]C4-PFOA,[13]C5-PFNA, [13]C2-PFDA, [13]C2-PFUnDA, [18]O2-PFHxS, [13]C4-PFOS and [13]C8-FOSA (purity > 99%, Wellington Laboratories,Guelph, ON). Abbreviation, supplier, and purity of the native and iso-topically labeled PFAS standards are listed in Table S1 in SupportingInformation (SI).

2.2. Sampling, soil characteristics and spiking

Soil was taken using a stainless steel shovel from a depth of 0.45-0.55 m at a site in Högåsa, Sweden (15°27´E, 58°30´N) in July 2016.The soil is characterized as a loamy sand (sand 82%, silt 15% and clay3%). It consists of quartz, K-feldspar, and plagioclase as the main mi-neral constituents and has a low carbon content (0.36%) [32]. It wasselected for this study because it has high leaching potential due to itscoarse texture and low organic carbon content [32]. For the experi-ments, a set of spiked soil samples was prepared. A 1.0 kg portion of drysoil was first spiked with 114 mL of a 0.10 mg mL−1 PFAS standardmixture in methanol (c= 11.4 μg g−1 dry weight (dw) for each in-dividual PFAS) and then 1 L Millipore water was added to adjust thewater content to a sludge consistency for ease of homogenization. Themixture was homogenized in a blender (Tefal Blender Classic) for 0.5 hand a 2.0 kg portion of the spiked mass was combined with 18 kg ofunspiked soil (resulting in c= 0.60 μg g−1 dw for each individualPFAS). This mixture of spiked and unspiked soil was homogenized in apolypropylene (PP) bucket (60 cm length, 30 cm width, and 20 cmheight) in an end-over-end shaker at 30 rpm for 24 h, air-dried at roomtemperature for two weeks in a fume hood, and then further homo-genized in the same end-over-end shaker at 30 rpm for an additional

M. Sörengård et al. Journal of Hazardous Materials 367 (2019) 639–646

640

120 h. Finally, the soil was aged for 4 months in darkness at roomtemperature (20 °C) before S/S treatment in the laboratory.

2.3. Stabilization-solidification treatment mixtures and preparation

The S/S treatment was performed by mixing together a binder, theaged PFAS-spiked soil, different types of additives, and water (Table 1).The binder consisted of a constant combination of Portland Cement(PC) and fly ash (FA) (Cement BAS, Cementa, Sweden) and groundgranulated blast-furnace base slag (GGBS) (Merit 5000, Merox,Sweden) with a PC:FA:GGBS ratio of 1:1:2 (w/w). The soil and binderwere mixed with a ratio of 9:1 (reference samples which did not includesoil and did not include binder were also prepared). In total, sevencommercially available additives were tested at a concentration of 2%(w/w) of the binder weight (corresponding to 0.2% (w/w) of the dryPFAS-contaminated soil): powdered activated carbon (PAC, pulverizedFiltrasorb 400, Calgon Carbon, Sweden), Rembind® (AC-based andcontaining amorphous aluminum hydroxide and kaolin clay, Ziltek,Australia), pulverized zeolite (Clinoptilolite, Ecoconceptearth,Sweden), chitosan (600,000–800,000 Da, Acros Organics, Sigma Al-drich, Sweden), hydrotalcite (Mg6Al2CO3(OH)16·4(H2O), Kisuma Che-micals, Japan), bentonite (montmorillonite, Swedish Nuclear Fuel andWaste Management Company, Sweden), and calcium chloride (purity93%, Sigma Aldrich, Sweden) (Table 1). The water content was opti-mized by strength tests (see Figure S1 in SI) and was kept constant at70% (previous S/S studies have used a water content ranging between60 [33] and 100 [34] %), however a too low water content can generateviscosity problems in field scale mixing [35]

The S/S treatments (each of treatment 1–8 and the reference sam-ples) were prepared separately for the leaching tests and the strengthtest, because the latter needed to be performed on intact cylinders. Alltreatment mixtures were prepared separately in a mixer (Electrolux BE21 A, Sweden) at 200 rpm for 5 min until homogeneity, added to fi-berglass tubes (5 cm diameter; 5 cm height for the leaching tests and5 cm diameter; 30 cm height for the strength tests), and cured at 4 °C for28 days. The homogeneity was confirmed using visually validation byx-ray tomography (Koestel and Larsbo, 2014) (Figure S2 in SI).

2.4. PFAS leaching tests and PFAS analysis

Tests on PFAS leaching were performed in triplicate in batch modefor each individual treatment mixture (1–8 in Table 1). The test typewas similar to the European standard compliance leaching test (EN12457-1 [44]), which is often used for long-term S/S material evalua-tion. After curing, the different S/S-treated soils were crushed sepa-rately and sieved to a 0.1–2.0 mm fraction. Then 15 g portions of each

soil were mixed with 30 mL of Millipore water in 50 mL PP-tubes(Corning®, nonpyrogenic) (liquid/solid (L/S) ratio of 2) and shaken for24 h at 200 rpm in an end-over-end shaker (Reax 2, Heidolph). Ashaking time of 24 h may be not enough to reach equilibrium (e.g.,[29]), but the conditions ensure comparability with the Europeanstandard for leaching tests (EN 12457-1). Reference liquid sampleswere prepared in the same way, but without S/S-treated soil (n= 3).

After equilibration, the suspensions were centrifuged at 3000 rpmfor 15 min and 1000 μL of the aqueous phase (cw) were transferred toEppendorf tubes and directly spiked with 100 μL of the IS mixture(c = 10 ng mL−1) together with 900 μL HLPC grade methanol. TheEppendorf tubes were vortexed and centrifuged and their contents werefiltered through recycled cellulose syringe filters (0.45 μm, Sartorius)into 2.0 mL auto-injector glass vials (Eppendorf, Germany).

In order to determine liquid/solid partitioning coefficients (kd), thespiked initial soil (cs,0) was analyzed in triplicate in accordance to amethod described elsewhere [[45]]. In brief, 3.0 g of freeze-dried (over7 days) solid sample were spiked with 100 μL of an IS mixture(c = 10 ng mL−1) and extracted with solid-liquid extraction using30 mL methanol (LiChrosolv, Merck, Germany). The extract was con-centrated under a nitrogen gas stream to 1 mL, and cleaned up with25 mg ENVI-Carb (120/400, Supelco, USA) and 50 μL glacial acetic acid(Merck, Germany). After centrifuging at 3000 rpm for 15 min, 0.5 mL ofthe supernatant was transferred to a PP tube (Eppendorf, Germany)together with 0.5 mL Millipore water, vortexed for 10 min, centrifugedat 15,000 rpm for 15 min, and filtered through a 0.45 μm cellulosesyringe filter (Sartorius, Germany) into a 2.0 mL auto-injector glass vial.

The PFAS analysis of all samples was performed using ultra highperformance liquid chromatograph coupled to tandem mass spectro-scopy (UHPLC-MS/MS) (Quantiva TSQ; Thermo Fisher) using a BEH-C18 column (1.7 μm, 50 mm, Waters) and an injection volume of 10 μL.The eluent gradient was set to 12 min, and the mobile phases wereMillipore water and Acetonitrile plus 5 mM ammonium acetate. Theisotope dilution method was used for quantification, using an eight-point calibration curve (0.01–100 ng mL−1), and the data were eval-uated using TraceFinder™ software (Thermo Fisher). The pH was mea-sured in the sample leachates (n = 1) for all tested materials (691 pHMeter, Metrohm, Switzerland).

2.5. Mechanical strength test

An unconfined compressive strength (UCS) test was performed induplicate for all treatment mixtures (Table 1) using a standardizedmethod (ISO/TS 17892-7:2005, Swedish Standards Institute, 2005[46]), where the shear strength [kPa] was measured. The UCS value (q ,ukPa) was then calculated as:

Table 1The different stabilization and solidification (S/S) treatments tested. In all treatments, the soil consisted of aged spiked sandy loam soila. The binder consisted of acombination of Portland Cement (PC), fly ash (FA), and ground granulated blast-furnace base slag (GGBS) at PC:FA:GGBS ratio of 1:1:2 (v/v). The water content was70% (w/w).

TreatmentNo.

Soila:Binder Additive(2 % w/w)

Median effective additive particlesize

Expected interaction effect of additive References to PFAS sorptionstudies

1 9:1 No additive N/Ab N/A2 9:1 PAC 12 μmc Increased surface [36]3 9:1 Rembind® 500 μmc Increased surface [37]4 9:1 Zeolite 58 μmc Increased surface [36]5 9:1 Chitosan 600,000-800,000 Da Increased hydrophobic/ cationic sorption [38]6 9:1 Hydrotalcite 66 μmc Increased sorption to cationic mineral [39]7 9:1 Bentonite 306 Da [40] Increased sorption and mechanical strength [41]8 9:1 Calcium chloride N/A Improved mechanical strength and sorption through

divalent ion bridge[42,43]

ReferenceB 0:1 No additive N/A N/AReferenceS 1:0 No additive N/A N/A

a Soil spiked with 0.10 mg mL−1 PFAS standard mixture in methanol and then mixed with unspiked soil to give c = 0.6 μg g−1 dw soil for each individual PFAS.bN/A = Not applicable. cMeasured by laser-scattering particle size distribution analyzer (LA-950, Horiba, Japan) (see Figure S6 in SI).

M. Sörengård et al. Journal of Hazardous Materials 367 (2019) 639–646

641

=q 2u f (1)

where f is the shear strength [kPa].

2.6. Quality control and quality assurance

The PFAS concentrations in the blanks ranged from non-detected to0.00029 μg g−1 dw (6:2 FTSA)) for the solid samples (n= 3) and non-detected to 0.00063 μg mL−1 dw (6:2 FTSA) for the aqueous phase(n= 3) (Tables S4 and S7 in SI). The method detection limit (MDL) wascalculated using average blank and three times of the standard devia-tion or, if no PFAS was detected in the blanks, the lowest calibrationpoint was used. The MDL ranged between 0.00001 μg g−1 and 0.00089μg g−1 for the solid phase, and between 0.00001 μg mL-1 and 0.00063μg mL-1 for the aqueous phase (Tables S4 and S7 in SI). The recovery ofthe method was calculated for all samples, based on the loss of IS duringsample preparation and matrix effects, and was compared against thecalibration curve. The method recovery was on average 76% ± 11% forthe solid phase and 85% ± 10% for the aqueous phase for individualPFASs (Tables S3 and S6 in SI). The average regression coefficient ( r2)of the calibration curves was 0.9970 ± 0.0025 and the measurement(replicate) error in the leaching experiment (n = 3) was on average11% ± 7.5% for the solid phase and 14% ± 4.4% for the aqueous phasefor individual PFASs (Tables S2 and S5 in SI).

2.7. Calculations

To evaluate the effect of the S/S treatments on the leaching beha-vior of the different PFAS, the retention (Re) was assessed by comparingthe individual PFAS concentrations in the water phase (cw [μg mL−1])of treated S/S soil c( )w to that in non-treated S/S soil c( )w ref, as:

=Re CC

w

w ref, (2)

where Re < 1 indicates increased PFAS retention.The solid-water partitioning coefficient (kd) [kg L−1] was calculated

as:

=k CCd

s

w (3)

where cs [μg g soil−1 dw] is the concentration of an individual PFAS inthe soil solid phase and cw [μg mL−1] is the measured concentration ofthat individual PFAS in the aqueous phase (Table S10 in SI). cs wascalculated as:

= =c c c orcc w c V

ws s w ss s i w w i

s i,0

,0 , ,

, (4)

where cs,0 is the measured initial concentration of the PFAS in the solidsoil sample [μg g−1 dw] (Table S8 in SI), ws i, is the solid sample weight[g], and Vw i, is the aqueous volume [mL].

3. Results and discussion

3.1. Solid-liquid partitioning coefficient of PFASs from the leaching tests

The starting concentration of PFASs in the solid soil sample (Cs,0)(n= 3) ranged from 0.20 to 1.5 μg g−1 dw for individual PFASs(mean = 0.49 μg g−1 dw). In the S/S treatment of the PFAS-spiked soilwith no binder and no additive (References in Table 1), the kd rangedbetween 1.8 L kg-1 (PFHxA) and 81 L kg−1(FOSA) (Fig. 1), which iswithin the range reported for PFAS soil adsorption in other studies (e.g.,[47]). Compared with this value for ReferenceS, in the S/S treatmentsinvolving an additive (2–8 in Table 1) the kd value was significantlyhigher for 13 of the 14 target PFASs (all except PFBA) (two sided t-test,MATLAB, p < 0.05, see Table S11 in SI). Based on the additive used,the median kd value in the leaching experiments increased in the fol-lowing general order: chitosan bentonite no additive calciumchloride < zeolite < hydrotalcite < PAC < Rembind®. For mostPFASs, addition of chitosan, bentonite, calcium chloride, zeolite orhydrotalcite had no significant effect compared with the treatment in-volving a binder but without an additive (Treatment 1 in Table 1) (forstatistical data, see Table S12 in SI). In contrast, previous studies haveshown good sorption of PFASs to these materials or have found that, ingeneral, the presence of divalent cations such as Ca2+ can enhancesorption of PFASs [38,39,41,43]. The highest kd values were achievedwith Rembind®, ranging from 1.9 L kg−1 for PFBA to 22 000 L kg-1 forPFOS (mean kd 5700 L kg−1 for all 14 PFAS), and with PAC, rangingfrom 7.1 L kg−1 for PFBA to 3500 L kg−1 for PFOS (mean kd of 920 Lkg−1 for all 16 PFAS). Comparable kd values have been reported forPFAS sorption to pristine AC, with kd values ranging from 26 000 to 61000 L kg-1 for PFOS after 48 h of shaking [36], and ranging from 240 Lkg-1 and 120 000 L kg-1 for PFOS after 10 min of shaking with pristinegranulated AC and pristine PAC, respectively [48]. With each addi-tional CF2 moiety in compound chain length, the kd value increased byon average (including all experiments) 0.33 log units for PFCAs (R2 =0.91, p < 0.001, n= 48, ANOVA,Matlab) and 0.23 log units for PFSAs(R2 = 0.52, p = 0.49, n= 18, ANOVA,Matlab) (Fig. 1, Figure S3 in SI).The significant linear correlation of log kd with increasing

Fig. 1. Partitioning coefficient kd [L kg soil−1] for poly- and perfluoroalkyl substance (PFASs) in spiked untreated soil (no binder, no additives) and in S/S treatmentswith and without an additive (see Table 1). A) perfluorosulfonates (PFSAs), 6:2 and 8:2 fluorotelomer sulfonic acids (FTSAs), and perfluorooctanesulfonamide(FOSA), and B) perfluorocarboxylates (PFCAs).

M. Sörengård et al. Journal of Hazardous Materials 367 (2019) 639–646

642

perfluorocarbon chain length indicates that the adsorption mechanismis strongly driven by hydrophobic interaction. Similarly, previous stu-dies identified the importance of hydrophobic sorption mechanism forPFASs and have found kd value increases of 0.50-0.60 log units and0.52-0.75 log units per CF2 moiety for suspended particulate matter(SPM) and sediment, respectively 2947. In addition, comparing theeffect of the functional group of PFASs with the same perfluorocarbonchain length, the kd value for FOSA in the non S/S-treated spiked soilwas, on average, 1.4 log units, 0.97 log units, and 0.52 log units higherthan for the corresponding PFCAs, FTSAs, and PFSAs, respectively(Fig. 1 and Figure S4 in SI). In contrast, in the S/S-treated soil, themeasured kd values were similar for FOSA and PFSAs, but were onaverage 1.1 log units and 0.46 log units higher than for the PFCAs andFTSAs, respectively. Similar results have been reported in previousstudies [29,47]. The sorption difference between the functional groupscan be explained by electrostatic interaction mechanisms [29,47].

3.2. pH effect on the solid-liquid partitioning coefficient in S/S matrix

PFASs sorption to solid phases is dependent on pH and the presencesof divalent cations in the water phase indicating the importance ofelectrostatic sorption mechanisms [28,29,49,50]. The anionic PFASssorption strength is often observed to be inversely correlated with pH,whereby the abundant hydroxide ions (OH−) are competing with theanionic PFASs to the solid phase adsorption sites 28. The high pH as-sociated with S/S (average pH = 10 ± 0.8) compared to the untreatedsoil (pH = 6.4) does therefore not generate the optimal electrostaticsolution chemistry for sorption of anionic PFASs. Though S/S treatedsoil had a high pH, most PFASs (except the PFCAs with a per-fluorocarbon chain length < C5: PFBA and PFPeA) were successfullystabilized with S/S due to hydrophobic sorption (Fig. 2 and Fig. 3). Thehydrophobic sorption interactions for PFBA and PFPeA is small becauseof the short perfluorocarbon chain, and therefore electrostatic interac-tion is a more important sorption mechanism for PFBA and PFPeAcompared to the longer chained PFASs. PFBA showed stronger parti-tioning to the solid phase in the untreated reference soil compared tothe S/S treated soil (Fig. 1 and Fig. 2), which can be explained by thehigher pH in the untreated soil compared to the S/S treated soil andhence reduced electrostatic sorption. The pH can also have an effect onthe solubility of PFASs depending on their protonation state. FOSA hasan acid dissociation constant (pKa) of 6.2–6.5 [51], while the PFCAs,PFSAs and FTSAs have generally a pKa< 2 meaning that they areprotonated anions at environmentally relevant pH [52]. For example,FOSA (Kd = 25 L kg-1) had approximately a three time higher sorptionstrength compared to PFOS (Kd = 81 L kg-1) in the reference soil, whichmay be explained by the fact that the leachate pH (pH = 6.4) was closeto the pKa of FOSA (6.2–6.5), which means that FOSA was not fullyprotonated in the reference soil. On the other hand, the sorptionstrength of FOSA and PFOS was comparable in the S/S treatment(average Kd = 328 L kg-1 and 331 L kg-1 for FOSA and PFOS, respec-tively) possibly due to the high pH in the S/S treatment (10 ± 0.8),which is higher than the pKa of FOSA and PFOS. Thus, FOSA exhibitedmainly hydrophobic interactions with the reference soil since it existedmainly as neutral species, whereas in the S/S treated soil FOSA ionicinteractions played also an important role. The effect of PFASs’ proto-nation state on sorption behavior is often not regarded in PFAS sorptionstudies [28,50]. More studies are needed in the future in particular onnew identified PFASs [53], with possible anionic, cationic, neutral andzwitterionic functional groups [54].

3.3. S/S treatment efficiency on PFAS contaminated soil

The leaching tests revealed that the S/S treatment changed theleaching behavior for all PFASs compared with the spiked untreatedsoil, and that the reduction in PFAS in the water phase was highly in-fluenced by their characteristics (Fig. 2). For the non-AC based S/S

treatments (i.e., no additive, calcium chloride, chitosan, zeolite, hy-drotalcite and bentonite), leachability (Cw/Cw,ref) decreased linearly, by15% per C–F moiety (R2 = 0.94, p = 4 × 10− [29], n = 48, ANOVA,Matlab), for C3-C10 PFCAs. For the PFSAs, leachability decreased by11% per C–F moiety (R2 = 0.90, p = 7 × 10-12, n= 18, ANOVA, Ma-tlab). An exponential decrease in the leachability of C3-C10 PFCAs wasobserved for S/S treatments using AC as the additive (i.e., PAC andRembind®), by Cw/Cw,ref = 97.7x-4.36 for PAC (R2 = 0.94, n = 48) andCw/Cw,ref = 3344.3x-6.6 for Rembind® (R2 = 0.95, n = 48), where x isthe increasing number of C–F moieties (see Figure S5 in SI). Previousstudies have reported a similar trend of increasing adsorption of PFASsto the solid phase with increasing perfluorocarbon chain length andhave attributed this to increasing hydrophobic bonding strength withincreasing chain length [28,29]. Another important factor influencingthe leachability was the functional group, which showed the followingtrend for the non-AC based S/S-treatments:

= > > > >PFSAs Re Eq FTSAs FOSA PFCAs( 0.08 ( . 2)) (0.10) (0.28)(0.33)

(Fig. 2 and Figure S4 in SI). This indicates that the compounds with asulfonate functional group (i.e., PFSAs and PTSAs) leached less in acementitious matrix compared with the neutral sulfonamide (FOSA)and the carboxylates (i.e., PFCAs). For the S/S-treatment with AC-basedadditives, the differences in leachability were not as pronounced, whichcan be explained by the strong sorption effect of AC, and hence lowerconcentrations close to MDL, which were associated with higher un-certainty (Fig. 1, Figure S4 in SI).

In general, the additives bentonite, calcium chloride, hydrotalcite,chitosan and zeolite did not significantly reduce the leachability ofindividual PFASs compared with no additive (p < 0.05, two-sided t-test, MATLAB) (Fig. 3). In contrast, the additives containing AC (i.e.,PAC and Rembind®) significantly reduced the leachability for in-dividual PFASs (i.e., PFPeA, PFHxA, PFHpA, PFOA, PFBS, PFHxS, andPFOS) (p < 0.05, two-sided t-test, MATLAB) (Table S12 in SI). ForPFHxA, PFHpA, and PFBS, the leachability was reduced by a factor of10 (> 90%), whereas for PFOA, PFNA, PFDA, PFUnDA, PFHxS, PFOS,6:2 FTSA, 8:2 FTSA, and FOSA the leachability was reduced by a factorof 100 to close to 1000 (> 99.0-99.9%) (Fig. 3). The lower reduction inleachability of the shorter-chain PFASs (i.e., PFHxA, PFHpA, and PFBS)can be explained by their high water solubility and small molecularweight, leading to lower hydrophobic bond strength and faster diffusionto the water phase [29,47,28]. To the best of our knowledge, this is thefirst study to examine the effects of S/S treatment on leaching of PFASs,but previous studies on other organic micropollutants have reported areduction in leachability of 91–99.9 % for 2-chloroaniline [55], ni-trobenzene [56], dioxide, furan [31], and polyaromatic hydrocarbons(PAHs) [31] (Table S13 in SI). As found in the present study for the lowmolecular weight PFASs (i.e., PFBA (213 Da) and PFPeA (263 Da)

Fig. 2. Average poly- and perfluoroalkyl substance (PFAS) concentration in thewater phase after stabilization and solidification (S/S) treatments involvingnon-activated carbon based additives (Cw

) or no S/S treatment (Cw,ref) as afunction of perfluorocarbon chain length for: perfluorosulfonates (PFCAs,n = 48), perfluorosulfonates (PFSAs, n = 18), 6:2 and 8:2 fluorotelomer sul-fonic acid (FTSAs, n = 12) and perfluorooctanesulfonamide (FOSA, n = 6).

M. Sörengård et al. Journal of Hazardous Materials 367 (2019) 639–646

643

(Figs. 2 and 3), a previous study found that S/S-treatment of wastematerials using an unspecified adsorbent as an additive was dependenton the contaminant molecular size of the PAHs, with the lowest mo-lecular weight compound (naphthalene) being least stabilized of 10PAHs investigated [57]. In another study investigating addition of 0–10% AC for phenol-contaminated soil, leachability was reduced by 40%with an optimum of 2% added AC, without altering the solidificationperformance [58]. This was explained by an increase in specific surfacearea of the material from 5 m2 g−1 with 0% AC to 18 m2 g−1 with 2%AC due to the high surface area of the AC. It has also been shown thatAC by itself can stabilize PFOS in various soils, with an increase in kd of1–3 log units depending on the fraction of soil organic carbon [14].However, concerns about the long-term stabilization properties of AChave been raised, as old AC shows low sorption capacity for PFASs andeven tends to desorb certain PFASs (i.e., PFBA, PFPeA) over time [12].Other studies have argued that during crystallization the cement canencapsulate, thereby blocking access by AC particles, suggesting thatpre-mixing of the additive before adding the binder can enhance ad-ditive performance in S/S-treatment [59].

3.4. Mechanical strength tests

Besides stabilization, the strength of the solidified soil is a key

parameter in S/S treatment, since a stronger concrete generally haslower permeability and porosity. The results from the UCS tests (Fig. 4)indicated that, in general, the treatment with no additive had thehighest strength, on average 4600 kPa (except hydrotalcite, average4800 kPa), while the treatments with additives had a lower mean UCSranging from 2900 kPa (Rembind®) to 4200 kPa (calcium chloride). Thehighest UCS was obtained when 100% binder was used (on average,9900 kPa). Although this is an unrealistic situation, this result indicatesthat a higher soil to binder ratio could increase the mechanical strengthperformance. Rembind® had the lowest UCS (2900 kPa), i.e., thestrength was 38% lower than in the treatment with no additive. An-other study which tested a variety of conventional and novel S/S bindercombinations found a mean UCS of < 500 kPa, while top-performingbinders had a UCS of > 2000 to 3500 kPa [60]. All treatment mixturestested in the present study met the UCS standard for the U.S. EPA (USA,1996), which is based on ex-situ S/S treatment and disposal of con-taminated risk waste at landfills (350 kPa or hydraulic conductivity <1. 10−9 ms-1) [61], and the standard for the Netherlands and France(1000 kPa) [51]. However, Rembind® (2900 kPa), chitosan (3300 kPa),and bentonite (3100 kPa) did not meet the UCS standard for Canada(3500 kPa), Environment Canada, 1991) [62]. Overall, however, thestandard UCS value depends on individual cases and can be as low as140 kPa for in situ burial purposes [63] and up to 7000 kPa for use as a

Fig. 3. Retention of poly- and perfluoroalkyl substance(PFAS) concentrations after stabilization and solidifi-cation (S/S) treatment with an additive (PAC,Rembind®, hydrotalcite, zeolite, chitosan, bentonite,and calcium chloride) and without an additive.Significant differences (two-sided t-test p < 0.05,n = 3) are indicated by filled dots. The dotted linerepresents no difference between with and withoutadditives.

Fig. 4. Unconfined compressive strength (UCS) of the stabilization and solidification (S/S) treatment with and without an additive, and with 100% binder (no soil).The vertical lines represent S/S performance guideline values (n = 2).

M. Sörengård et al. Journal of Hazardous Materials 367 (2019) 639–646

644

load-bearing concrete construction material [64]. Ultimately, aPC:FA:GGBS ratio of 1:1:2 at a water:cement ratio of 0.7 for a loamysand fulfils the different UCS standards. Comparing the two AC ad-ditives with the best performance in reducing the leachability of PFASs(i.e., Rembind® and PAC), PAC had a lower negative impact on the USC,with an average decrease of 14% compared to Rembind® with anaverage decrease of 38% (Fig. 4). The higher adsorption by Rembindcompared to PAC (Fig. 2) can be explained by higher uncertainties atlow concentrations (Cw), while another hypothetical explanation is theincreased air-water-material interface at lower concrete strength. Lowerstrength of concrete is primarily associated with increased formation ofair-bubbles on the micro-scale within the concrete [65], and specificunsaturated air-water-material interface have shown to induced an in-creased adsorption of PFASs due to their surfactant properties [66]. Apossible air-water-material interface effect on PFASs adsorption in ce-mentious matrix needs further validation.

It is important to note that a carbon poor sandy silt soil was used inthis study, while it is known that soils with either high organic carboncontent and clay minerals have shown to increase PFAS partitioning tosoil [28]. However, high organic carbon content and clay minerals canhave disturbing and retarding effects on the curing of the S/S process,and thus producing a material with less physical performance [67,68].Practitioners [27,35] suggest that before S/S treatment in the field,laboratory studies should be performed optimizing S/S binder recipesfor the specific contaminated soil due to the complexity of binding soilto aggregates. Either a higher soil to binder ratio can be used to increasethe physical performance [69], or a high carbon content can be handledby using, for example, magnesium oxide (MgO)-bearing binder incombination with Portland cement (PC-cement) [70] or addition ofsodium silicate [56,67], and a high clay content can be handled byadding lime to the binder mixture 68]. PFAS-contaminated soils areoften also co-contaminated with non-aqueous phase liquid hydro-carbons and non-fluorinated surfactants, originating from firefightingfoams, effecting the PFAS partitioning in the soil and in the cementiousS/S matrix [71–73]. Complex soil and contaminant mixtures effects onS/S treatment was not included in the scope of the study, although theauthors encourage future studies on this topic.

4. Conclusions

The promising results of the leaching and strength tests in thisstudy, combined with the predicted long-term treatment efficienciesand the commercial availability of the technology, indicate that S/Streatment of contaminated soils could meet the urgent demand toprevent PFASs leaching from polluted soil hotspots into the environ-ment and drinking water sources. The S/S treatment with activecarbon-based additives showed excellent performance in reducingleaching of PFASs, without losing considerable physical stability of thematrix. However, the tests performed in this study do not reflect theconjoined effects of both solidification and stabilization and the resultsare scale-dependent, so S/S treatment should be tested on field scale infuture studies.

Acknowledgement

This work was supported by the project PFAS-PURE from VINNOVA(2015-03561).

Appendix A. Supplementary data

Supplementary material related to this article can be found, in theonline version, at doi:https://doi.org/10.1016/j.jhazmat.2019.01.005.

References

[1] A.B. Lindstrom, M.J. Strynar, E.L. Libelo, Polyfluorinated compounds: past, present,

and future, Environ. Sci. Technol. 45 (2011) 7954–7961.[2] R.C. Buck, J. Franklin, U. Berger, J.M. Conder, I.T. Cousins, P.D. Voogt, A.A. Jensen,

K. Kannan, S.A. Mabury, L. van, Perfluoroalkyl and polyfluoroalkyl substances inthe environment: terminology, classification, and origins, Integr. Environ. Assess.Manag. 7 (2011) 513–541.

[3] L. Ahrens, Polyfluoroalkyl compounds in the aquatic environment: a review of theiroccurrence and fate, J. Environ. Monit. 13 (2011) 20–31.

[4] B.J. Place, J.A. Field, Identification of novel fluorochemicals in aqueous film-forming foams used by the US military, Environ. Sci. Technol. 46 (2012)7120–7127.

[5] M. Filipovic, A. Woldegiorgis, K. Norström, M. Bibi, M. Lindberg, A.-H. Österås,Historical usage of aqueous film forming foam: a case study of the widespreaddistribution of perfluoroalkyl acids from a military airport to groundwater, lakes,soils and fish, Chemosphere 129 (2015) 39–45.

[6] L. Ahrens, K. Norström, T. Viktor, A.P. Cousins, S. Josefsson, Stockholm ArlandaAirport as a source of per- and polyfluoroalkyl substances to water, sediment andfish, Chemosphere 129 (2015) 33–38.

[7] C. Baduel, C.J. Paxman, J.F. Mueller, Perfluoroalkyl substances in a firefightingtraining ground (FTG), distribution and potential future release, J. Hazard. Mater.296 (2015) 46–53.

[8] I. Gyllenhammar, U. Berger, M. Sundström, P. McCleaf, K. Eurén, S. Eriksson,S. Ahlgren, S. Lignell, M. Aune, N. Kotova, A. Glynn, Influence of contaminateddrinking water on perfluoroalkyl acid levels in human serum - a case study fromUppsala, Sweden, Environ. Res. 140 (2015) 673–683.

[9] J. Bräunig, C. Baduel, A. Heffernan, A. Rotander, E. Donaldson, J.F. Mueller, Fateand redistribution of perfluoroalkyl acids through AFFF-impacted groundwater, Sci.Total Environ. (2017) 596–597 360–368.

[10] I. Ross, J. McDonough, J. Miles, P. Storch, P. Thelakkat Kochunarayanan, E. Kalve,J. Hurst, S. Soumitri Dasgupta, J. Burdick, A review of emerging technologies forremediation of PFASs, Remediat. J. 28 (2018) 101–126.

[11] E. Arias, M. Mallavarapu, R. Naidu, Treatment technologies for aqueous per-fluorooctanesulfonate (PFOS) and perfluorooctanoate (PFOA): a critical reviewwith an emphasis on field testing, Environ. Technol. Innov. 4 (2015) 168–181.

[12] P. McCleaf, S. Englund, A. Östlund, K. Lindegren, K. Wiberg, L. Ahrens, Removalefficiency of multiple poly- and perfluoroalkyl substances (PFASs) in drinking waterusing granular activated carbon (GAC) and anion exchange (AE) column tests,Water Res. 120 (2017) 77–87.

[13] Y. Yao, T.U. Sack, K. Volchek, C.E. Brown, PFC-contaminated Soil and ItsRemediation Strategies: a Review. In, (2015), pp. 314–339.

[14] D. Kupryianchyk, S.E. Hale, G.D. Breedveld, G. Cornelissen, Treatment of sitescontaminated with perfluorinated compounds using biochar amendment,Chemosphere 142 (2016) 35–40.

[15] L. Paterson, T. Siemens Kennedy, D. Sweeney, SLR Consulting (Canada) Ltd.Remediation of Perfluorinated Chemicals at a Former Fire Fighting Training Area.in, (2008).

[16] C. Eschauzier, P. Voogt, H.-J. de Brauch, F.T. Lange, in Polyfluorinated Chem.Transform. Prod, Springer, Berlin, Heidelberg, 2012, pp. 73–102, https://doi.org/10.1007/978-3-642-21872-9_5.

[17] L. Gobelius, J. Lewis, L. Ahrens, Plant uptake of per- and polyfluoroalkyl substancesat a contaminated fire training facility to evaluate the phytoremediation potentialof various plant species, Environ. Sci. Technol. 51 (2017) 12602–12610.

[18] P. Das, E. Arias, V. Kambala, M. Mallavarapu, R. Naidu, Remediation of per-fluorooctane sulfonate in contaminated soils by modified clay adsorbent - A risk-based approach topical collection on remediation of site contamination, Water AirSoil Pollut. 224 (2013).

[19] S.E. Hale, H.P.H. Arp, G.A. Slinde, E.J. Wade, K. Bjørseth, G.D. Breedveld,B.F. Straith, K.G. Moe, M. Jartun, Å. Høisæter, Sorbent amendment as a remediationstrategy to reduce PFAS mobility and leaching in a contaminated sandy soil from aNorwegian firefighting training facility, Chemosphere 171 (2017) 9–18.

[20] O. Malliou, M. Katsioti, A. Georgiadis, A. Katsiri, Properties of stabilized/solidifiedadmixtures of cement and sewage sludge, Cem. Concr. Compos. 29 (2007) 55–61.

[21] S. Paria, P.K. Yuet, Solidification-stabilization of organic and inorganic con-taminants using portland cement: a literature review, Environ. Rev. 14 (2006)217–255.

[22] M.J. Harbottle, A. Al-Tabbaa, C.W. Evans, A comparison of the technical sustain-ability of in situ stabilisation/solidification with disposal to landfill, J. Hazard.Mater. 141 (2007) 430–440.

[23] J. Kumpiene, A. Lagerkvist, C. Maurice, Stabilization of As, Cr, Cu, Pb and Zn in soilusing amendments – a review, Waste Manag. 28 (2008) 215–225.

[24] C. Ludwig, C.A. Johnson, M. Käppeli, A. Ulrich, S. Riediker, Hydrological andgeochemical factors controlling the leaching of cemented MSWI air pollution con-trol residues: a lysimeter field study, J. Contam. Hydrol. 42 (2000) 253–272.

[25] S. Bozkurt, L. Moreno, I. Neretnieks, Long-term processes in waste deposits, Sci.Total Environ. 250 (2000) 101–121.

[26] B.D. Bone, L.H. Barnard, D.I. Boardman, P. Carey, C.D. Hills, H.M. Jones,C. MacLeod, M. Tyrer, L.H. Barnard, D.I. Boardman, P.J. Carey, C.D. Hills,H.M. Jones, C.L. MacLeod, M. Tyrer (Eds.), Review of Scientific Literature on theUse of stabilisation/Solidification for the Treatment of Contaminated Soil, SolidWaste and Sludges, 2004 UK Environment Agency.

[27] C. Hills, P. Gunning, E. Bates, Stabilisation-Solidification of Contaminated Soil andWaste, Hygge Media, 2015.

[28] Z. Du, S. Deng, Y. Bei, Q. Huang, B. Wang, J. Huang, G. Yu, Adsorption behaviorand mechanism of perfluorinated compounds on various adsorbents—a review, J.Hazard. Mater. 274 (2014) 443–454.

[29] C.P. Higgins, R.G. Luthy, Sorption of perfluorinated surfactants on sediments,Environ. Sci. Technol. 40 (2006) 7251–7256.

M. Sörengård et al. Journal of Hazardous Materials 367 (2019) 639–646

645

[30] C. Baduel, J.F. Mueller, A. Rotander, J. Corfield, M.-J. Gomez-Ramos, Discovery ofnovel per- and polyfluoroalkyl substances (PFASs) at a fire fighting training groundand preliminary investigation of their fate and mobility, Chemosphere 185 (2017)1030–1038.

[31] E.R. Bates, E. Sahle-Demessie, D.W. Grosse, Solidification/stabilization for re-mediation of wood preserving sites: treatment for dioxins, PCP, creosote, and me-tals, Remediat. J. 10 (2000) 51–65.

[32] H. Kirchmann, S. Snäll, J. Eriksson, L. Mattsson, Properties and classification of soilsof the Swedish long-term fertility experiments: V. Sites at Vreta Kloster and Högåsa,Acta Agric. Scand. Sect. B Soil Plant Sci. 55 (2005) 98–110.

[33] S. Firouzbakht, S. Gitipour, A.F.H. Rooz, Cement-based and AC-BasedSolidification/Stabilization processes of aliphatic hydrocarbons contaminated soilsin Kharg Island oil terminal, Int. J. Environ. Res. 11 (2017) 439–448.

[34] F. Wang, H. Wang, A. Al-Tabbaa, Leachability and heavy metal speciation of 17-year old stabilised/solidified contaminated site soils, J. Hazard. Mater. 278 (2014)144–151.

[35] J.R. Conner, S.L. Hoeffner, A critical review of stabilization/solidification tech-nology, Crit. Rev. Environ. Sci. Technol. 28 (1998) 397–462.

[36] V. Ochoa-Herrera, R. Sierra-Alvarez, Removal of perfluorinated surfactants bysorption onto granular activated carbon, zeolite and sludge, Chemosphere 72(2008) 1588–1593.

[37] J. Szabo, J. Hall, M. Magnuson, Treatment of Perfluorinated Alkyl Substances inWash Water Using Granular Activated Carbon and Mixed Media, at < U.S.Environmental Protection Agency, 2017 > https://www.epa.gov/homeland-security-research.

[38] Q. Zhang, S. Deng, G. Yu, J. Huang, Removal of perfluorooctane sulfonate fromaqueous solution by crosslinked chitosan beads: sorption kinetics and uptake me-chanism, Bioresour. Technol. 102 (2011) 2265–2271.

[39] R. Rattanaoudom, C. Visvanathan, Removal of PFOA by hybrid membrane filtrationusing PAC and hydrotalcite, Desalin. Water Treat. 32 (2011) 262–270.

[40] 40.BELBaR. Stability of Clay Colloids Sorption/desorption of Pollutants on ColloidsErosion of Clay Modelling of Clay Colloids, (2016) at < http://www.skb.se/belbar/wp-content/uploads/2016/04/D6.14-Proceedings_BELBaR_2016_with_EndUserReviewBoardEvaluation.pdf > .

[41] R. Zhang, W. Yan, C. Jing, Mechanistic study of PFOS adsorption on kaolinite andmontmorillonite, Colloids Surf. Physicochem. Eng. Asp. 462 (2014) 252–258.

[42] C.Y. Tang, Fu Shiang Q, D. Gao, C.S. Criddle, J.O. Leckie, Effect of solutionchemistry on the adsorption of perfluorooctane sulfonate onto mineral surfaces,Water Res. 44 (2010) 2654–2662.

[43] F. Wang, K. Shih, Adsorption of perfluorooctanesulfonate (PFOS) and per-fluorooctanoate (PFOA) on alumina: influence of solution pH and cations, WaterRes. 45 (2011) 2925–2930.

[44] CEN. Characterisation of Waste - Leaching - Compliance Test for Leaching ofGranular Waste Materials and Sludges - Part 1: One Stage Batch Test at a Liquid toSolid Ratio of 2 l/kg for Materials With High Solid Content and With Particle SizeBelow 4 Mm (without or With Size Reduction), (2002).

[45] L. Ahrens, S. Felizeter, R. Sturm, Z. Xie, R. Ebinghaus, Polyfluorinated compoundsin waste water treatment plant effluents and surface waters along the River Elbe,Germany. Mar. Pollut. Bull. 58 (2009) 1326–1333.

[46] Swedish Standards Institute. Geotechnical Investigation and Testing - LaboratoryTesting of Soil - Part 7: Unconfined Compression Test on Fine-grained Soil (ISO/TS17892-7:2004, (2005).

[47] L. Ahrens, S. Taniyasu, L.W.Y. Yeung, N. Yamashita, P.K.S. Lam, R. Ebinghaus,Distribution of polyfluoroalkyl compounds in water, suspended particulate matterand sediment from Tokyo Bay, Japan, Chemosphere 79 (2010) 266–272.

[48] M.C. Hansen, M.H. Børresen, M. Schlabach, G. Cornelissen, Sorption of per-fluorinated compounds from contaminated water to activated carbon, J. SoilsSediments 10 (2010) 179–185.

[49] H. Campos Pereira, M. Ullberg, D.B. Kleja, J.P. Gustafsson, L. Ahrens, Sorption ofperfluoroalkyl substances (PFASs) to an organic soil horizon – effect of cationcomposition and pH, Chemosphere 207 (2018) 183–191.

[50] Y. Li, D.P. Oliver, R.S. Kookana, A critical analysis of published data to discern therole of soil and sediment properties in determining sorption of per and poly-fluoroalkyl substances (PFASs), Sci. Total Environ. (2018) 628–629 110–120.

[51] S. Rayne, K. Forest, A new class of perfluorinated acid contaminants: primary and

secondary substituted perfluoroalkyl sulfonamides are acidic at environmentallyand toxicologically relevant pH values, J. Environ. Sci. Health A Tox. Hazard. Subst.Environ. Eng. 44 (2009) 1388–1399.

[52] S. Rayne, K. Forest, K.J. Friesen, Computational approaches may underestimate pKavalues of longer-chain perfluorinated carboxylic acids: implications for assessingenvironmental and biological effects, J. Environ. Sci. Health - ToxicHazardousSubst. Environ. Eng. 44 (2009) 317–326.

[53] L.W.Y. Yeung, S.A. Mabury, Are humans exposed to increasing amounts of uni-dentified organofluorine? Environ. Chem. 13 (2016) 102–110.

[54] G. Munoz, S.V. Duy, P. Labadie, F. Botta, H. Budzinski, F. Lestremau, J. Liu,S. Sauvé, Analysis of zwitterionic, cationic, and anionic poly- and perfluoroalkylsurfactants in sediments by liquid chromatography polarity-switching electrosprayionization coupled to high resolution mass spectrometry, Talanta 152 (2016)447–456.

[55] S. Gallo, L. Zampori, G. Dotelli, P. Meloni, I.N. Sora, R. Pelosato, Use of admixturesin organic-contaminated cement-clay pastes, J. Hazard. Mater. 161 (2009)862–870.

[56] J. Liu, X. Nie, X. Zeng, Z. Su, Cement-based solidification/stabilization of con-taminated soils by nitrobenzene, Front. Environ. Sci. Eng. China 6 (2012) 437–443.

[57] E. Mulder, J.P. Brouwer, J. Blaakmeer, J.W. Frénay, Immobilisation of PAH in wastematerials, Waste Manag. 21 (2001) 247–253.

[58] J. Liu, X. Nie, X. Zeng, Z. Su, Long-term leaching behavior of phenol in cement/activated-carbon solidified/stabilized hazardous waste, J. Environ. Manage. 115(2013) 265–269.

[59] R.E. Crane, D.P. Cassidy, V.J. Srivastava, Activated carbon preconditioning to re-duce contaminant leaching in cement-based stabilization of soils, J. Environ. Eng.U. S. 140 (2014).

[60] F. Wang, H. Wang, F. Jin, A. Al-Tabbaa, The performance of blended conventionaland novel binders in the in-situ stabilisation/solidification of a contaminated sitesoil, J. Hazard. Mater. 285 (2015) 46–52.

[61] United States environmental protection agency, Prohibition on the Disposal of BulkLiquid Hazardous Waste in Landfills – Statutory Interpretive Guidance, (1996).

[62] Environment Canada, Proposed Evaluation Protocol For Cement-Based SolidifiedWastes, Environment Canada, 1991.

[63] J.N. Meegoda, A.S. Ezeldin, H.-Y. Fang, H.I. Inyang, Waste immobilization tech-nologies, Pract. Period. Hazard. Toxic Radioact. Waste Manag. 7 (2003) 46–58.

[64] A.A. Tabbaa, J.A. Stegemann, Stabilisation/Solidification Treatment andRemediation: Proceedings of the International Conference on Stabilisation/Solidification Treatment and Remediation, CRC Press, Cambridge, UK, 2005, pp.12–13. April 2005.

[65] C. Lian, Y. Zhuge, S. Beecham, The relationship between porosity and strength forporous concrete, Constr. Build. Mater. 25 (2011) 4294–4298.

[66] Y. Lyu, M.L. Brusseau, W. Chen, N. Yan, X. Fu, X. Lin, Adsorption of PFOA at theair–Water interface during transport in unsaturated porous media, Environ. Sci.Technol. 52 (2018) 7745–7753.

[67] A. Coz, A. Andrés, S. Soriano, J.R. Viguri, M.C. Ruiz, J.A. Irabien, Influence ofcommercial and residual sorbents and silicates as additives on the stabilisation/solidification of organic and inorganic industrial waste, J. Hazard. Mater. 164(2009) 755–761.

[68] E.I. Stavridakis, Influence of liquid limit and slaking on cement stabilized clayeyadmixtures, Geotech. Geol. Eng. 17 (1999) 145–154.

[69] R. Malviya, R. Chaudhary, Factors affecting hazardous waste solidification/stabi-lization: a review, J. Hazard. Mater. 137 (2006) 267–276.

[70] F. Jin, F. Wang, A. Al-Tabbaa, Three-year performance of in-situ solidified/stabi-lised soil using novel MgO-bearing binders, Chemosphere 144 (2016) 681–688.

[71] J.L. Guelfo, C.P. Higgins, Subsurface Transport Potential of Perfluoroalkyl Acids atAqueous Film-Forming Foam (AFFF)-Impacted Sites, (2013), https://doi.org/10.1021/es3048043.

[72] Y.-M. Li, F.-S. Zhang, Characterization of a cetyltrimethyl ammonium bromide-modified sorbent for removal of perfluorooctane sulphonate from water, Environ.Technol. 35 (2014) 2556–2568.

[73] G. Pan, C. Jia, D. Zhao, C. You, H. Chen, G. Jiang, Effect of cationic and anionicsurfactants on the sorption and desorption of perfluorooctane sulfonate (PFOS) onnatural sediments, Environ. Pollut. 157 (2009) 325–330.

M. Sörengård et al. Journal of Hazardous Materials 367 (2019) 639–646

646