Electrochemical Transformations of Perfluoroalkyl Acid (PFAA

Subscriber access provided by Karolinska Institutet, University Library

Remediation and Control Technologies

Electrochemical Transformations of Perfluoroalkyl Acid (PFAA) Precursors and PFAAs in Groundwater Impacted with Aqueous Film Forming Foams Charles E. Schaefer, Sarah Choyke, P. Lee Ferguson, Christina Andaya, Aniela Burant, Andrew Chapin Maizel, Timothy J. Strathmann, and Christopher P. Higgins Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b02726 • Publication Date (Web): 22 Aug 2018 Downloaded from http://pubs.acs.org on August 22, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 33

Environmental Science & Technology

Electrochemical Transformations of Perfluoroalkyl Acid (PFAA) Precursors and PFAAs in Groundwater Impacted with Aqueous Film Forming Foams

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

Charles E. Schaefer1,*, Sarah Choyke2, P. Lee Ferguson2, Christina Andaya3, Aniela Burant4, Andrew Maizel4, Timothy J. Strathmann4, Christopher P. Higgins4 CDM Smith, 110 Fieldcrest Avenue, #8, 6th Floor, Edison, NJ 08837

1

2

Nicholas School of the Environment, Duke University, Durham, North Carolina 27708, United States 3

APTIM, 17 Princess Road, Lawrenceville, NJ 08648

4

Department of Civil and Environmental Engineering, Colorado School of Mines, Golden, CO 80401

*

CORRESPONDING AUTHOR: Mailing address: CDM Smith, 110 Fieldcrest Avenue, #8, 6th Floor, Edison, NJ 088837. (732)-590-4633. E-mail: [email protected]

Submitted to Environmental Science & Technology

Key Words: PFOS, PFOA, electrochemical, boron-doped diamond, AFFF

ACS Paragon Plus Environment


Page 2 of 33

42

Abstract

43

While oxidative technologies have been proposed for treatment of waters impacted by

44

aqueous film forming foams (AFFFs), information is lacking regarding the

45

transformation pathways for the chemical precursors to the perfluoroalkyl acids (PFAAs)

46

typically present in such waters. This study examined the oxidative electrochemical

47

treatment of poly- and perfluoroalkyl substances (PFASs) for two AFFF-impacted

48

groundwaters. The bulk pseudo first order rate constant for PFOA removal was 0.23 L h-1

49

A-1; for PFOS, this value ranged from 0.084 to 0.23 L h-1 A-1. Results from the first

50

groundwater studied suggested a transformation pathway where sulfonamide-based

51

PFASs transformed to primarily perfluorinated sulfonamides and perfluorinated

52

carboxylic acids (PFCAs), with subsequent defluorination of the PFCAs. Transient

53

increases in the perfluorinated sulfonamides and PFCAs were observed. For the second

54

groundwater studied, no transient increases in PFAAs were measured, despite the

55

presence of similarly structured suspected PFAA precursors and substantial

56

defluorination. For both waters, suspected precursors were the primary sources of the

57

generated fluoride. Assessment of precursor compound transformation noted the

58

formation of keto-perfluoroalkane sulfonates only in the second groundwater. These

59

results confirm that oxidation and defluorination of suspected PFAA precursors in the

60

second groundwater underwent transformation via a pathway different than that of the

61

first groundwater.

62 63 64 65


2

Page 3 of 33


66 67

Introduction Groundwater impacted with poly- and perfluoroalkyl substances (PFASs) originating

68

from fire fighting activities where aqueous film forming foams (AFFFs) were used has

69

become a major environmental concern and challenge. Several studies have noted the

70

impacts to groundwater that likely occurred as a result of these fire fighting activities (1,

71

2). Perfluoroalkyl acids (PFAAs) are among the most troublesome compounds observed

72

in these impacted groundwaters, as PFAAs are recalcitrant to natural biotic and abiotic

73

transformation processes. Of these PFAAs, perfluorooctanoic acid (PFOA) and

74

perfluorooctane sulfonate (PFOS) have a health advisory level prescribed by the United

75

States Environmental Protection Agency (USEPA) of 0.07 µg/L, both individually and

76

combined (3). PFOS and PFOA have been detected in groundwater at concentrations that

77

are orders of magnitude above this health advisory level (2, 4, 5).

78

Electrochemical treatment of PFOA and PFOS, as well as longer and shorter-chained

79

PFAAs, has shown promise. Electrochemical studies on the treatment of PFAAs typically

80

have focused on using either mixed metal oxide (MMO) (6-8) or boron-doped diamond

81

(BDD) anodes (9-12), where oxidative treatment in electrolyte solutions amended with

82

PFOA, PFOS, or a mixture of PFAAs has been demonstrated. Transformation products

83

including shorter-chained perfluorinated carboxylic acids (PFCAs) and fluoride have

84

been observed, and detailed transformation mechanisms regarding the PFAAs involving

85

an “unzipping” process have been proposed (9, 10, 13).

86

Only a relatively few electrochemical studies involving PFAAs in natural

87

groundwater systems, and in the presence of the full range of PFASs typically associated

88

with AFFF-impacted waters, have been performed. One study compared the


3


89

defluorination kinetics of PFOS and PFOA during electrical treatment using a BDD

90

anode in both electrolyte and natural groundwater matrices (12). Results of this study

91

showed that natural groundwater constituents had only minimal impacts on PFOA and

92

PFOS treatment. However, AFFF-impacted waters typically contain a wide range of

93

polyfluorinated compounds that are susceptible to oxidative transformation to PFAAs

94

(14); electrochemical treatment of these PFAA precursors was not performed in this

95

previous study.

96

Page 4 of 33

Only a very limited assessment of PFAA precursors associated with AFFF-impacted

97

groundwater has been performed during electrochemical treatment. Electrochemical

98

treatment of AFFF-impacted groundwater using a BDD anode was recently performed

99

(15), but only PFAAs and 6:2 fluorotelomer sulfonate were evaluated, and reaction

100

kinetics were not assessed in the groundwater. In another study, electrochemical

101

treatment of PFAAs, along with PFAA precursors 6:2 fluorotelomer sulfonamide alkyl

102

betaine, 6:2 fluorotelomer sulfonamide propyl N,N dimethylamine, and 6:2 fluorotelomer

103

sulfonate, were evaluated in effluent collected from a wastewater treatment plant (16).

104

Electrochemical oxidation of the precursors resulted in transient formation of PFCAs.

105

While this study provided useful insight into the electrochemical oxidation pathways, the

106

PFASs in the wastewater treatment plant effluent were significantly different from those

107

encountered in AFFF-impacted groundwater. Perfluorinated sulfonates, perfluoroalkyl

108

sulfonamide amines, and perfluoroalkyl sulfonamide amino carboxylates, which (in

109

addition to PFCAs) are present in AFFF formulations manufactured by 3M (4), were not

110

part of the study. Thus, the impacts of PFAA precursors associated with AFFF on

111

electrochemical treatment remain unclear, and the potential transformation pathways of


4

Page 5 of 33


112

these precursor compounds have not been reported. Proper assessment of electrochemical

113

approaches for treatment of AFFF-impacted waters will require a more comprehensive

114

assessment of PFAA precursors, as these precursors may be present in greater quantities

115

than the PFAAs (14).

116

The overall goal of this study was to assess the transformation of PFAA precursors

117

present in AFFF during electrochemical treatment. Specifically, potential PFAA

118

precursor transformation mechanisms and rates were determined, and the impacts of the

119

precursors on the overall treatment of PFAAs (including PFOS and PFOA) were assessed.

120

Two AFFF-impacted waters also were assessed to examine the potential impacts of

121

“fresh” versus “aged” AFFF constituents. Findings from this work highlight the

122

importance of considering PFAA precursor fate when designing electrochemical

123

treatment for AFFF-impacted waters.

124 125

Experimental

126

Materials

127

PFOA (96% purity) was purchased from Sigma Aldrich. Two natural groundwaters,

128

designated W1 and W2, were used for all the electrochemical experiments. W1 was

129

collected from a facility with no known AFFF impacts, while W2 was collected from a

130

US Department of Defense facility in the vicinity of a fire training area where AFFF was

131

used; 3M AFFF was likely one of the products used at this location, as suggested by an

132

empty drum of this solution identified at the site. Basic water quality parameters and

133

dominant PFAA levels (in the case of W1, after spiking with 3M AFFF) are provided in


5


134

Table 1. The AFFF solution used for spiking W1 was manufactured by 3M (2001), and

135

was previously characterized and provided by Dr. Jennifer Field as (4).

Page 6 of 33

136 137

Electrochemical System

138

Electrochemical experiments were performed similarly to those described previously (12)

139

using a single compartment Microflow Cell (ElectroCell North America, Inc.). The

140

cathode material was stainless steel, and the anode material was boron-doped diamond on

141

niobium support (Condias, GmbH, Germany). The active surface area of each electrode

142

was 10 cm2. The distance between electrodes was 4 mm.

143

All experiments were performed in batch mode, where a polypropylene vessel

144

served as the groundwater reservoir (Figure S1). Groundwater (0.25 L) was recirculated

145

through the electrochemical cell at 0.10 L/min using a peristaltic pump. Flow rates were

146

verified using a flowmeter. All experiments were performed under constant current

147

conditions, while monitoring voltage. Power was supplied using an E3633A 200W power

148

supply (Agilent). Current densities of 0 (no current controls) and 25 mA/cm2 were used;

149

one additional test at a current density of 200 mA/cm2 also was used. All experiments

150

were performed at room temperature (approximately 25 degrees C).

151

For experiments performed using groundwater W1, the groundwater was amended

152

with Na2SO4 so that the sulfate concentration in the groundwater increased by 500 mg/L

153

sulfate. This sulfate addition was performed to increase the conductivity of the water so

154

that the desired current density could be attained at an applied voltage similar to that in

155

the W2 electrochemical experiments. In addition, as mentioned above, W1 was amended

156

with 3M AFFF solution (0.02 mL AFFF solution to 250 mL of W1 groundwater). Thus,


6

Page 7 of 33


157

the AFFF-spiked W1 groundwater served as the “fresh” AFFF-impacted groundwater,

158

while the W2 groundwater served as the “aged” AFFF-impacted groundwater. All

159

groundwater was passed through a 20 µm filter prior to initiating the electrochemical

160

experiments to prevent any particulates from entering the electrochemical cell.

161

The groundwater solution in each experiment was monitored as a function of time

162

throughout the duration of each experiment, which typically lasted 8 hours. Samples were

163

collected for determination of pH, anions, and PFASs. Temperature of the recirculated

164

groundwater also was monitored.

165

Initially, duplicate samples were collected at select timepoints and immediately

166

quenched by mixing with 20 µl of a sterile 1.5 g/L sodium thiosulfate solution to

167

scavenge any residual oxidant species remaining in the sample (17). Preliminary tests

168

(data not shown) indicated that addition of the quenching agent did not impact the levels

169

of PFASs detected in the electrochemically treated samples, so this preservation step was

170

discontinued in later experiments. Control experiments were also performed without

171

applied current to account for any PFAS losses, such as sorption or volatilization, not

172

attributable to electrochemical treatment.

173

An additional experiment was performed in duplicate using PFOA (initial

174

concentration of 20 mg/L) in 150 cm3 of 1480 mg/L sodium sulfate; the current density

175

was 25 mA/cm2. This PFOA experiment was used to serve as a comparison to PFOA

176

transformation rates in the more complex W1 and W2 groundwater systems, which

177

contained a mixture of PFASs.

178 179

Analytical Methods


7


180

Page 8 of 33

An Oakton probe (Part no. WD-35634-14) was used to measure sample pH. Anions

181

were analyzed via ion chromatography using EPA Method 300.0, and perchlorate was

182

analyzed via ion chromatography using EPA Method 314.2. The detection limit for

183

anions (except perchlorate) was 200 µg/L; the detection limit for perchlorate was 0.25

184

µg/L. Descriptions of the quantitative analyses for PFAAs and the semi-quantitative

185

analyses for potential PFAA precursors are provided in the Supporting Information. Total

186

oxidizable precursor analysis, based on the previously developed methods (14), were

187

performed on W1 and W2 by SGS Axys Analytical Services Ltd. (BC, Canada).

188 189

Results and Discussion

190

PFAS Composition of W1 and W2

191

Levels of PFAAs in W1 and W2 are provided in Table 1. The remaining dominant

192

fluorinated compounds (>106 area counts) present in W1 and W2 prior to electrochemical

193

treatment (t=0 timepoint in the batch experiments), based on high resolution mass

194

spectrometry (HRMS), are summarized in Tables S1 and S2, respectively. Tables S1 and

195

S2 also include compounds that showed transient increases during electrochemical

196

treatment; confidence levels and similarity scores for these compounds are provided in

197

Tables S3 and S4. For W1, consistent with analysis of 3M AFFF performed previously

198

(4), several classes of sulfonamido compounds with perfluorinated tails were detected in

199

the AFFF-spiked groundwater. Perfluorinated chain lengths of n= 4 through 6 typically

200

were the dominant species. However, perfluorooctane sulfonamide (n=8), a potential

201

precursor of PFOA and/or PFOS (14,18), was identified. The presence of both


8

Page 9 of 33


202

perfluoroalkane sulfonamides (FASAs) and perfluoroalkane sulfinates (PFASis) indicate

203

that previously identified PFAA precursors are present in the AFFF-spiked W1 (19, 20).

204

The compounds identified in W2 were similar in structure to those identified in W1.

205

However, compounds containing the sulfonated end groups of the non-fluorinated

206

branches (e.g., S-OHPrAmPr-FQASA-OHPrS) detected in W2 were not detected in W1.

207

Perfluorinated chain lengths of n=4 through 6 were the dominant species for these

208

sulfonated sulfonamides. As with W1, FASAs (known PFAA precursors (14, 18)), were

209

also present in W2. For n=5 and 6, FASA levels (based on integrated area counts) were

210

10- to 100-times less than in W1. For n=4 and 8, FASA levels were similar (within a

211

factor of 2) for both W1 and W2. These n=8 FASA levels, as well as the presence of n=8

212

compounds for both N-SPAmP-FASA and PFASA-PDA, suggest that W2 has similar or

213

greater potential for formation of PFOA or PFOS from oxidative transformation of

214

precursors as does W1. W2 did not contain any detectable PFASi’s, another known class

215

of PFAA precursors (19).

216

To further assess the potential for precursor transformation to PFAAs, Figure S2

217

shows the baseline (prior to electrochemical treatment) fluorine content based on

218

integrated area response for all the precursors (n=4 through 8) present in W1 and W2.

219

Figure S2 shows that the fluorine content in potential precursor compounds in W2 were

220

approximately 5-times greater than in W1, which again suggests that W2 has equal or

221

greater likelihood of forming PFAAs upon electrochemical oxidation. However, the

222

response factors for the potential precursors in W1 and W2 may vary considerably for

223

each compound, thus the information provided in Figure S2 may not be an appropriate

224

indicator of potential PFAA formation via oxidation (although, considering the structural


9


225

similarity of the compounds, the data in Figure S2 is expected to provide an order of

226

magnitude type estimate).

227

Page 10 of 33

Oxidative transformation of PFAA precursors also was assessed using the total

228

oxidizable precursor (TOP) assay (14). Results are provided for both W1 and W2 in

229

Figure S3. Despite the apparent abundance of potential precursor compounds present in

230

W2 relative to W1, the TOP assay indicates that the PFAA precursors present in W2 are

231

negligible compared to W1, as PFCAs increased approximately 200-times in W1 during

232

the TOP assay. These results suggest that PFAA formation via oxidation likely originates

233

from CEtAmPr-FASA-PrAs, AmPr-FASA-PrAs, and FASAs that are present only in W1

234

(or, present in much greater quantities in W1 than in W2). It is plausible that these

235

compounds were originally present in the AFFF source materials associated with W2, but

236

were transformed in situ prior to collecting the W2 groundwater.

237

Interestingly, AmPr-FASA and the other potential precursor compounds present in

238

W2 (Figure S2) do not appear to substantially contribute to PFAA formation via chemical

239

oxidation. This could be due to their oxidation pathway, or due to the fact that their

240

concentrations are too low to measurably contribute to the PFAA mass already present in

241

W2. This will be further explored in subsequent sections as PFAA formation and the

242

fluoride balance are assessed during electrochemical treatment.

243 244

PFAS Transformations during Electrochemical Treatment – W1

245

Electrochemical treatment at 25 mA/cm2 required an applied voltage of

246

approximately 13 V and 16 V for W1 and W2, respectively. A small increase in


10

Page 11 of 33


247

temperature from approximately 25 to 30 degrees C occurred during treatment. The pH

248

for both waters remained circumneutral.

249

The PFASs shown in Tables S1 and S2 for W1 and W2 were generally removed

250

during electrochemical treatment. Figure S4 shows the decreases in CEtAMPr-FASA-

251

PrAs, AmPr-FASA-PrAs, and AmPr-FASAs for W1. Figure S5 shows the transient

252

increases in OAmPr-FASAs, MeFASAAs, FASAs, and PFASi’s for W1; these

253

compounds all show an increase followed by a decrease. It is important to note that

254

potential precursor levels analyzed via HRMS were determined without stable isotope

255

internal standards, so results should be interpreted with caution.

256

Electrochemical treatment of W1 also showed transient increases in PFCAs, but not

257

corresponding increases in the perfluorinated sulfonic acids (PFSAs), as shown in Figure

258

1. These results are consistent with the TOP assay (Figure S3), as well as the data of

259

Houtz and Sedlak (14). Finally, fluoride generation was observed (Figure S6).

260

Collectively, these data suggest that oxidative electrochemical treatment of the PFASs in

261

W1 (including many of the precursors included in the 3M AFFF) proceeds through initial

262

oxidation steps (Figure 2) that include a combination of oxidation of the terminal amine,

263

dealkylation of the sulfonamide, and defluorination of the carbon chain. The pathway

264

shown in Figure 2 assumes that the branched sulfonamide structures (CEtAMPr-FASA-

265

PrAs and AmPr-FASA-PrAs,) oxidatively transform yielding OAmPr-FASAs, which are

266

then rapidly oxidized as shown. It is speculated that a currently unidentified precursor(s)

267

results in the formation of MeFASAAs. All the intermediate species shown in Figure 2

268

showed transient increases during electrochemical treatment.


11


269

Page 12 of 33

Previously proposed pathways for the aerobic biotransformation of

270

ethylperﬂuorooctane sulfonamide have indicated that formation of OAmPr-FASAs

271

preceeds the formation of FASAs (18, 19), thus the formation of OAmPr-FASAs

272

observed herein is consistent with the oxidative formation of FASAs. Mejia-Avendaño et

273

al. (20) have shown the formation of FASA from AmPr-FASAs. The formation of FASAi

274

from FASAs also has been observed during aerobic biotransformation processes (19).

275

However, the formation of PFCAs from sulfonamido precursors has only been observed

276

through abiotic pathways (14, 21), and not biotic pathways (18, 19). Houtz and Sedlak

277

(14) also have shown that abiotic oxidation of both MeFASAAs and FASAs results in the

278

formation of PFCAs, which is consistent with the oxidation pathway shown in Figure 2.

279

Electrochemical oxidation of non-fluorine containing sulfonamides, with cleavage of the

280

S-N bond, has been previously demonstrated (22). The unzipping and defluorination of

281

PFAAs via electrochemical approaches have been well documented (23).

282

Electrochemical dealkylation and amine oxidation for non-fluorine containing

283

compounds also have been well documented (24, 25).

284

In W1, the generation of the n=7 and n=8 PFCAs is much less than that of the

285

shorter-chained PFCAs (Figure 1). This is likely due to the relative abundance of n=4 to

286

6 precursors present and/or the transformation of the longer (7 and 8 chain) PFCAs to

287

shorter chain PFCAs. The only n=8 precursor identified in W1 was perfluorooctane

288

sulfonamide (FOSA), and no n=7 precursors were identified. The generation of PFHpA

289

likely was due to the electrochemical oxidation of PFOA (6). These results are consistent

290

with those observed by Houtz et al. (26), who observed substantial increases in n=4

291

through 6 PFCAs due to chemical oxidation of 3M AFFF, but no reported increases in


12

Page 13 of 33


292

n=7 or 8 PFCAs. It is also possible that the small increase in PFHpA observed herein was

293

due to an unidentified precursor present in the dissolved AFFF solution.

294 295 296

PFAS Transformations during Electrochemical Treatment – W2 Figure S7 shows the decreases in S-OHPrAmPr-FASA-OHPrS, SPrAmPr-

297

FASAPrS,S-OHPrAmPr-FASAA, SPrAmPr-FASAA, SPrAmPr-FASA, SPr-FASA, and

298

AmPr-FASA for W2. Figure S8 shows the transient increases in FASAs (n=4 and n=6)

299

and K-PFASs (n=2,3,4,6) for W2. The transiently generated FASA levels for n=4 and 6

300

were 10- to 100-times less than those for W1, and K-PFAS area counts were generally 3-

301

to 10-times less than FASA area counts. Unlike W1, W2 showed no transient increases in

302

OAmPr-FASA, MeFASAAs, or PFASi’s; W2 also showed no transient increases in

303

PFCAs or PFSAs (Figure 3), but did yield fluoride in quantity similar to that observed for

304

W1 (Figure S6). An additional experiment was performed at a lower current density (15

305

mA/cm2), and with sampling at 1 and 2 hours to ensure that a large transient increase in

306

PFAAs did not occur at early timepoints (t

Electrochemical Transformations of Perfluoroalkyl Acid (PFAA

Recommend Documents