Breakdown Products from Perfluorinated Alkyl Substances (PFAS

Subscriber access provided by MIDWESTERN UNIVERSITY

Remediation and Control Technologies

Breakdown products from perfluorinated alkyl substances (PFAS) degradation in a plasma-based water treatment process Raj Kamal Singh, Sujan Fernando, Sadjad Fakouri Baygi, Nicholas Multari, Selma Mededovic Thagard, and Thomas M. Holsen Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b07031 • Publication Date (Web): 15 Feb 2019 Downloaded from http://pubs.acs.org on February 17, 2019

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 30

Environmental Science & Technology

1

Breakdown products from perfluorinated alkyl substances (PFAS)

2

degradation in a plasma-based water treatment process

3 4

Raj Kamal Singh †, Sujan Fernando ‡, Sadjad Fakouri Baygi ‡, Nicholas Multari †, Selma

5

Mededovic Thagard †, Thomas M. Holsen *‡

6

† Plasma Research Laboratory, Department of Chemical and Biomolecular Engineering, Clarkson

7

University, Potsdam, New York 13699, United States

8

‡ Department of Civil and Environmental Engineering, Clarkson University, Potsdam, New York

9

13699, United States

10 11

* Corresponding author; E-mail: [email protected]; # +1-315-268-3851

12 13 14 15 16 17 18

KEYWORDS: BYPRODUCTS; CYCLIC PERFLUOROALKANES; PERFLUOROALKYL

19

SUBSTANCES; PFOA; PFOS; PLASMA

1 ACS Paragon Plus Environment


20

ABSTRACT

21

Byproducts produced when treating perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate

22

(PFOS) in water using a plasma treatment process intentionally operated to treat these compounds

23

slowly to allow for byproduct accumulation were quantified. Several linear chain perfluoroalkyl

24

carboxylic acids (PFCAs) (C4 to C7) were identified as byproducts of both PFOA and PFOS

25

treatment. PFOA, perfluorohexane sulfonate (PFHxS) and perfluorobutane sulfonate (PFBS) were

26

also found to be byproducts from PFOS degradation. Significant concentrations of fluoride ions,

27

inorganic carbon and smaller organic acids (trifluoroacetic acid, acetic acid and formic acid) were

28

also identified. In addition to PFCAs, PFHxS and PFBS, trace amounts of 43 PFOA-related and

29

35 PFOS-related byproducts were also identified using a screening and search-based algorithm.

30

Minor concentrations of gas-phase byproducts were also identified (< 2.5% of the F originally

31

associated with the parent molecules) some of which are reported for the first time in

32

perfluoroalkyl substance degradation experiments including cyclic perfluoroalkanes (C4F8, C5F10,

33

C6F12, C7F14 and C8F16). The short chain PFCAs detected suggest the occurrence of a step-wise

34

reduction of the parent perfluoroalkyl substances (PFAS) molecule, followed by oxidation of

35

intermediates, perfluoroalkyl radicals and perfluoro alcohols/ketones. Using a fluorine mass

36

balance, 77% of the fluorine associated with the parent PFOA and 58% of the fluorine associated

37

with the parent PFOS were identified. The bulk of the remaining fluorine was determined to be

38

sorbed to reactor walls and tubing using sorption experiments in which plasma was not generated.

39 40 41


Page 2 of 30

Page 3 of 30


42

1. INTRODUCTION

43

Perfluoroalkyl substances (PFAS) are carbon-chain based organo-fluorine compounds containing

44

stable C-F bonds. The most commonly encountered PFAS are perfluorooctanoic acid (PFOA) and

45

perfluorooctane sulfonate (PFOS), which were widely used in many commercial and industrial

46

applications. Disposal of PFAS-containing products, the discharge of industrial and municipal

47

wastewater and the use of aqueous film forming foams in firefighting has resulted in widespread

48

PFAS contamination of surface water and groundwater including many drinking water supplies.1–4

49

PFAS have recently received considerable attention due to their ubiquitous presence, recalcitrance

50

in the environment, and toxic properties.1,5

51

The most commonly employed water treatment technologies for the removal of PFAS from water

52

are activated carbon and ion-exchange. However, relatively short breakthrough times and

53

generation of waste including saturated adsorbent and concentrated brine solution (from ion-

54

exchange resin regeneration), which require further treatment or disposal make the search for

55

effective treatment technologies important.6 Previous studies have consistently demonstrated the

56

difficulty of breaking the stable C-F bonds by hydroxyl radicals generated as primary oxidants

57

from conventional advanced oxidation processes (AOPs) such as UV/H2O2/O3.7 However, other

58

AOPs

59

microwave/persulfate,12 and ionizing radiation (electron beam,13 γ-irradiation14) including

60

plasma15 have been reported to be effective for treatment of some PFAS.

61

Plasma-based water treatment is a technology that, using only electricity, converts water into a

62

mixture of highly reactive species including •OH, O, H•, HO2•, O2•‒, H2, O2, H2O2 and aqueous

63

electrons (e-aq) called a plasma.16,17 When applied for water treatment, plasmas are generally

64

formed by means of an electrical discharge between two electrodes-one high voltage and one

such

as

photocatalysis,8

electrochemical

oxidation,9


persulfate,10

sonolysis,11


65

grounded-within or contacting the contaminated water. To date, a variety of electrical discharge

66

plasma reactor types featuring different electrode arrangements have been used to treat a wide

67

range of organic and inorganic contaminants including pharmaceuticals18,19 and VOCs20, among

68

other compounds.21

69

Recently, we have developed an enhanced contact plasma reactor for the destruction of PFAS that

70

was able to remove PFOA and PFOS to below regulatory limits with removal efficiencies greater

71

than those of leading alternative technologies.22 The reactor features a high voltage electrode in

72

the gas, just above the liquid surface and a grounded ring electrode submerged just beneath the

73

liquid surface to achieve contact between plasma streamers and the entire reactor volume. Plasma

74

is formed by applying a sufficiently high electrical potential between the high voltage and

75

grounded electrodes via an external plasma-generating network. Argon gas is pumped through

76

submerged diffusers to produce bubbles and form a layer of foam on the liquid surface. This foam

77

concentrates surfactant-like contaminants (e.g., PFAS) and enhances the contact between the

78

liquid and the plasma, exposing the contaminant at the gas-liquid interface to reactive oxidative

79

and reductive species in the plasma.21 Unlike other AOPs which are solely based on the production

80

of OH radicals, plasma treatment involves generation of reductive species (hot and aqueous

81

electrons, hydrogen radicals) and ions (e.g., argon ions) without requiring any chemical inputs.

82

Aqueous electrons have been shown to be directly, and argon ions indirectly, responsible for the

83

degradation of PFAS during a plasma treatment.22

84

Most AOP studies report short-chain (C2-C7) perfluoroalkyl carboxylic acids (PFCAs), fluoride

85

ions, and carbon dioxide as byproducts of PFOA and PFOS degradation. However the initial

86

reactions involved can be distinct, involving different types of reactive species: electron transfer

87

between Fe3+ ions and PFOA in an UV-Fenton process,23 abstraction of an electron from PFOA 4 ACS Paragon Plus Environment

Page 4 of 30

Page 5 of 30


88

by holes generated in a photocatalytic process,24 direct electron transfer from –COOH group of

89

PFOA to the anode in an electrochemical process,9 and attack of sulfate radical causing –COOH

90

cleavage of PFAS in persulfate oxidation.10 The main mechanism involved in sonochemical

91

degradation of PFOS is based on pyrolysis, where the -SO3- group can cleave at temperatures

92

around 1000 K close to the bubble-water interface. In addition, ultrasonic pressure waves which

93

can generate internal bubble temperatures of up to 4000 K lead to an instantaneous mineralization

94

of PFOA/PFOS to F-, SO42-, CO, and CO2 without generating short chain PFCAs.25 In ionizing

95

radiation, attack of aqueous electrons, hydrogen atoms and other reducing species may be the

96

predominant initiating reaction mechanism of PFOA degradation.13,14

97

There is very limited information on the fluorine mass balance in previous studies which suggests

98

that many byproducts may not have been identified. For example, there is little information on the

99

gaseous byproducts that may be produced. In addition, most AOP degradation studies have

100

reported short chain (C4-C7) PFCAs as the main byproducts from PFOS treatment but not shorter

101

chain perfluoroalkane sulfonates (PFSAs).

102

The primary objective of this work was to determine the species produced during PFAS

103

degradation in a plasma treatment process. The reactor used was intentionally operated to treat the

104

PFAS slowly to allow for byproduct accumulation and subsequent identification. Both liquid and

105

gas phase byproducts were quantified, their distribution across the phases determined and a

106

fluorine mass balance constructed. Unconventional liquid phase byproducts were identified and

107

their concentrations indirectly estimated using a computer-based algorithm with the data acquired

108

from ultra-performance liquid chromatograph–quadrupole time of flight–high resolution mass

109

spectrometer (UPLC-QToF-HRMS) analysis. Using the byproducts identified, a degradation

110

mechanism for PFOA and PFOS in the plasma process was proposed. 5 ACS Paragon Plus Environment


111

2. MATERIALS AND METHODS

112

Chemicals

113

Analytical grade PFOA and PFOS (linear chain only, no branched isomers) were purchased from

114

Sigma Aldrich (St. Louis, MO). The standards of linear perfluoroalkyl carboxylic acids, sulfonates

115

and labeled internal standards (refer to supplementary information (SI) Table S1) were purchased

116

from Wellington Laboratories (Guelph, ON). Methanol and acetonitrile (LC-MS grade) were

117

purchased from Thermo Fisher Scientific (Waltham, MA).

118

Electrical circuit and plasma reactor

119

A custom built high-voltage power supply was used for the generation of plasma. Pulses of 40 Hz

120

frequency were generated by charging a 2 nF capacitor, and discharging it through a rotating spark

121

gap. A negative voltage of 30 kV was used for all the experiments. A schematic for the electrical

122

circuit is shown in Figure S1 (SI) and discussed elsewhere.22

123

The plasma reactor was a glass vessel of diameter 17.3 cm and height 19 cm (total reactor volume

124

3.8 L), which consisted of a sharpened nickel-chromium rod (diameter = 2.2 mm) as a high voltage

125

(HV) electrode, and an aluminum ring (outer diameter = 9.8 cm, inner diameter = 6 cm) as a ground

126

electrode. The HV electrode was placed in the headspace region and the ground electrode was

127

fitted as a ring around the diffuser circumference submerged in the liquid. The spacing between

128

the HV and ground electrode was 2.7 cm (1.5 cm in liquid, 1.2 cm in gas). The top opening of the

129

glass vessel was sealed with an airtight polymer cap which was adapted to allow for integration of

130

electrodes, extraction of liquid and gaseous samples, and gas recirculation. Argon gas was

131

introduced through the submerged diffuser and recirculated through the headspace with a flow rate

132

of 4 L/min. The schematic of the overall experimental setup in shown in Figure 1. To allow for


Page 6 of 30

Page 7 of 30


133

byproduct accumulation, the discharge energy was much lower than was used in the high rate

134

plasma reactor in our previous study (0.24 vs 0.63 J/pulse).22

135 136

Figure 1. Schematic of the experimental setup.

137

Experimental procedure

138

At the start of the experiments the reactor was filled with 1.5 L of either 8.3 mg/L of PFOA or

139

PFOS solution and then treated for time intervals of 30, 60, and 120 min. The initial solution pH

140

was around 4.6 and the electrical conductivity was adjusted to 300 µS/cm using 0.1 M NaCl. Gas

141

and liquid samples were collected using air tight gas canisters (1 L) and syringes, respectively. To

142

measure sorption of PFAS onto reactor components, sorption experiments were performed,

143

maintaining the same experimental conditions but without plasma being generated.

144



145

Analytical procedures

146

Liquid byproducts

147

Methods. A Waters Acuity UPLC coupled to a Q-TOF (Xevo G2-XS) and HR-MS was used for

148

the analysis of PFAS and their liquid byproducts. Separation was performed with a Waters Acquity

149

HSS T3 column (2.1 mm x 100 mm, 1.8 µm) and samples were analyzed by electrospray in

150

negative ion mode. Samples were diluted with methanol (1:3 ratio) and then sonicated and

151

centrifuged prior to injection (20 µL). A detailed description of analytical method is provided in

152

supplementary information (Text S1 and Table S2).

153

Quality Assurance. For quality assurance and control, all samples were spiked with 2 ng of

154

labeled internal standards. Based on the analysis of method blanks, the limit of detection (LOD)

155

was determined to be 0.2 ng/L, however, the lower limit of quantification (LOQ) was set to be 1

156

ng/L with signal to noise ratio of 10:1. Six point calibration in the range of 10 and 5000 ng/L was

157

used for the quantification of samples using C-13 isotopic dilution or internal standard methods

158

(details are provided in Table S1 (SI)). Quantification was performed with MassLynksTM 4.1

159

(Waters Corporation) using regression fit of r2 > 0.98 and deviation < 30%. Calibration standards

160

were reinjected in the sample sequence to validate the time dependent response from the

161

instrument. Analyzed data were quantified if surrogate recovery was between 70 and 120%.

162

Concentrations of PFAS in the samples were normalized with respect to surrogate recovery.

163

Simulation for identification of unconventional liquid byproducts. The identification and

164

confirmation of unconventional byproducts is challenging due to the lack of commercial standards.

165

However, with the aid of high-resolution mass spectrometry it was possible to identify a number

166

of potential byproducts based on mass accuracy (within 5 ppm) of the measured molecular ion [M-


Page 8 of 30

Page 9 of 30


167

H]-. For this work, a screen and search algorithm which uses this approach developed previously

168

by Baygi et al. was used.26 The algorithm has four basic steps: (1) identification of candidate

169

compounds with elemental composition compatible with PFOA or PFOS, (2) generation of

170

candidate compound spectra matrix using their theoretical isotopic distribution, (3) candidate list

171

screening by matching theoretical m/z with mass spectral data, and (4) retention time

172

determination of targeted m/z using isotopic profile determination. The output was the elemental

173

composition for each m/z feature in the chromatogram which have a formula that resembles a

174

PFAS.

175

Gaseous byproducts

176

The gas-phase PFAS decomposition products were analyzed using a Markes Unity and CIA

177

thermal desorption system coupled to a Thermo Trace GC Ultra gas chromatograph and a Thermo

178

DSQ II quadrupole mass spectrometer. The thermal desorption system was used to sample 50 mL

179

of the headspace gas from the canisters which was trapped and focused on a cold trap at 0ºC and

180

then injected (split 20:1) into the GC by heating the cold trap rapidly (20ºC/sec) to 300ºC.

181

Separation was carried out on a Rt-Alumina BOND/CFC column (30 m length x 0.53 mm i.d x 10

182

µm film thickness) at constant pressure of 5 psi. Oven temperature was initially set at 100ºC for

183

one minute, then ramped at a rate of 5ºC/min to 200ºC and held for 10 minutes. The transfer line

184

to the MS was set to 200ºC. The MS was operated in negative ion chemical ionization (NICI)

185

mode, where methane was used as a reagent gas with the ion source temperature set to 150ºC. Data

186

were collected in the mass range of 50 and 500 amu.

187 188



189

Fluoride quantification

190

Fluoride ions (F-) were measured as per EPA 9214 method with an Accumet Excel XL60 meter

191

kit (Fisher Scientific) with a combination electrode (Accumet) and total ionic strength adjustment

192

buffer (TISAB, VWR Chemicals).

193

Ion chromatography

194

Sulfate (SO42-), formate (HCOO-), acetate (CH3COO-) and trifluoroacetate (CF3COO-) ion

195

concentrations were measured using ion chromatography (Dionex Integrion HPIC) equipped with

196

an AS11 (4 mm i.d.) column and a conductivity detector. Sodium hydroxide solution (NaOH, 23

197

mM) was used as the eluent with a constant flow rate of 1 mL/min.

198

Inorganic carbon analysis

199

Inorganic carbon was measured by difference between total carbon before and after acid sparging

200

using a TOC analyzer (Shimadzu Model OC-VCPH).

201

RESULTS AND DISCUSSION

202

Quantification of liquid byproducts

203

Plasma-based water treatment degrades PFOA and PFOS into various short-chain PFCAs and

204

PFSAs. In this study, 90% of the PFOA was removed in 60 min, a much slower rate than in the

205

high rate plasma reactor used in our earlier study, where a similar percentage was removed in 30

206

min.22 Perfluoroheptanoic acid (PFHpA), perfluorohexanoic acid (PFHxA), perfluoropentanoic

207

acid (PFPeA) and perfluorobutanoic acid (PFBA) are common byproducts of both PFOA and

208

PFOS degradation, with two additional byproducts (perfluorohexane sulfonate (PFHxS) and

209

perfluorobutane sulfonate (PFBS)) detected for PFOS only. The concentrations of all the 10 ACS Paragon Plus Environment

Page 10 of 30

Page 11 of 30


210

byproducts increased in the first 60 minutes of treatment and then decreased by the end of the

211

experiment (Figure 2a and b). The structures of all byproducts analyzed in this study are shown in

212

Table S3. For PFOS, the peak concentrations of PFOA (C8) and PFHpA (C7) were at 30 and 60

213

minutes of treatment time respectively, and the trend of decreasing peak concentrations of PFOA

214

(C8) > PFHpA (C7) > PFHxA (C6) > PFPeA (C5) > PFBA (C4) suggests a step-wise chain

215

propagation degradation of PFOS with PFOA as the first byproduct.

216



217 218

Figure 2. Concentration profiles of (a) PFOA and (b) PFOS and their byproducts with treatment

219

time; experiments were performed in triplicate, and the average with standard deviations are

220

displayed in the plots. The reactor was intentionally operated to treat the PFAS slowly to allow for

221

byproduct accumulation and subsequent identification. Molar concentration profiles of these

222

byproducts are shown in Figure S2.

223 224

These byproducts were further mineralized to CF3COOH, CH3COOH, HCOOH, F- and SO42- (the

225

latter only in the case of PFOS) and the concentrations of all these compounds increased with

226

treatment time (Figure 3). Inorganic carbon concentrations also increased by 0.18 and 0.20 mg/L

227

after a 120 min treatment from PFOA and PFOS degradation, respectively. Significant

228

concentrations of sulfate ions were detected after 120 min of treatment of PFOS, which

229

corresponds to 85% of the theoretical sulfate that could be produced from the oxidation of sulfur


Page 12 of 30

Page 13 of 30


230

present in 8.3 mg/L (initial concentration) of PFOS. It should be noted that the concentration of

231

gaseous CO2 was not measured. These results collectively indicate significant mineralization of

232

the parent and their byproduct compounds.

233

234



235 236

Figure 3. Concentrations of smaller organic acids, fluoride and sulfate produced during the

237

degradation of (a) PFOA and (b) PFOS.

238 239

As mentioned, plasma electrons, aqueous electrons and argon ions are the main species responsible

240

for PFAS degradation. As shown in Figure 4, the attack of these species on the –COOH functional

241

group of the PFOA molecule may result in the formation of unstable perfluoroalkyl radicals such

242

of •C7F15 moieties, which in subsequent radical recombination with •OH may lead to the formation

243

of perfluoro alcohols (C7F15OH).7 Thermally unstable enol (-OH) could be transformed into the

244

more stable keto (=O) form (C6F13COF) by HF elimination caused by the thermal transfer of e-

245

aq.

246

molecule. Chain propagation reactions involving reductive and oxidative species and subsequent

247

hydrolysis yield short chain PFCAs. The degradation pathway of PFOS appears to be similar to

13

Further, the hydrolysis of C6F13COF yields C6F13COOH with the loss of one more HF


Page 14 of 30

Page 15 of 30


248

that of PFOA with the inclusion of a chain initiation reaction involving the attack of electrons or

249

argon ions on the C-S bond resulting in SO3- group cleavage from the terminal carbon and

250

formation of •C8F17 radicals. The chain propagation reactions of •C8F17 result in the formation of

251

short-chain PFCAs (Figure 4). Short chain PFSAs such as PFHxS and PFBS may be formed by

252

the reactions of SO3•- with •C6F13 and •C4F9, respectively. Please note that perfluoroheptane

253

sulfonate (PFHpS) or perfluoropentane sulfonate (PFPeS) were not measured due to a lack of

254

analytical standards, but it is likely that they would also have formed following the proposed

255

degradation pathway.

256 257

Figure 4. Proposed degradation pathway for PFOA and PFOS in plasma treatment. Note that of

258

the PFAS shown only perfluoropropanoic acid (PFPA) was not quantified.

259



260

Chemical reactions of plasma reactive species with PFOA, PFOS and their byproducts can lead to

261

the formation of a large number of transient and stable compounds. A number of novel byproducts

262

were identified based on accurate mass measurements and isotopic profile (section 2.4.1). Using

263

this approach, 43 and 35 novel byproducts of PFOA and PFOS, respectively were identified (Table

264

S4). Based on the response (peak area), these byproducts were further subdivided into three classes

265

as low (area

Breakdown Products from Perfluorinated Alkyl Substances (PFAS

Recommend Documents