Alternate Reductants with VB12 to Transform C8 and C6

Subscriber access provided by READING UNIV

Article

Alternate Reductants with VB12 to Transform C8 and C6 Perfluoroalkyl Sulfonates: Limitations and Insights into IsomerSpecific Transformation Rates, Products and Pathways Saerom Park, Chloe De Perre, and Linda S Lee Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.7b03744 • Publication Date (Web): 12 Nov 2017 Downloaded from http://pubs.acs.org on November 17, 2017

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 free 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 accessible to all readers and 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.

Environmental Science & Technology 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 26

Environmental Science & Technology

1

Alternate Reductants with VB12 to Transform C8 and C6 Perfluoroalkyl Sulfonates:

2

Limitations and Insights into Isomer-Specific Transformation Rates, Products and

3

Pathways

4 5

Saerom Park1,2, Chloe de Perre1, and Linda S. Lee1,2,*

6

Purdue University, 1Department of Agronomy, 2Ecological Science and Engineering, West

7

Lafayette, IN 47907-2054

8 9

*

Corresponding author at: Department of Agronomy, Purdue University, West Lafayette, IN

10

47907, USA. Tel.: +1 765 494 8612; fax: +1 765 496 2926.

11

E-mail address: [email protected] (L.S. Lee).

12 13

Revision Prepared for Environmental Science and Technology

14

Revised October 29, 2017

15 16 17 18 19 20 21 22

ACS Paragon Plus Environment


23

Graphical Art

24 25


Page 2 of 26

Page 3 of 26

26 27


ABSTRACT Previous studies evaluating VB12 with Ti(III)-citrate for potential use in in-situ

28

remediation of perfluorooctane sulfonate (PFOS) found that linear (L)-PFOS was unaltered. We

29

explored if alternate reductants could overcome this limitation with a primary focus on nanoscale

30

zero valent zinc (nZn0). Transformation over time with VB12-nZn0 was quantified at 22, 70 and

31

90°C for PFOS, at 70°C for perfluorohexane sulfonate (PFHxS), and VB12-nFe0 and VB12-

32

Pd0/nFe0 at 70°C for PFOS. Only branched (br-) isomers were transformed generating F‒ (no

33

SO42‒) and polyfluoroalkyl intermediates/products. The absence of L-PFOS transformation by

34

VB12 appears to be due to the inability of L-perfluoroalkyl sulfonates to complex with VB12

35

and not an activation energy issue that can be overcome by stronger reductants/catalysts. At

36

90°C, 95% of br‒PFOS isomers were transformed within 5 d. Isomer-specific removal rates were

37

positively correlated to the br-CF3’s proximity to the terminal CF3. Br-PFHxS transformation

38

was approximately two times slower with less defluorination than br-PFOS. C8/C7 and C6/C5

39

polyfluorinated sulfonates from br‒PFOS and br‒PFHxS, respectively, were identified as both

40

intermediates and apparent dead-end products. Pathways included 4 F replaced by 2 H and a

41

C=C bond, and serial F replacement by H with up to 12 F atoms removed from br-PFOS.

42 43



44

INTRODUCTION

45

Perfluorooctane sulfonate (PFOS) has received much attention due to its widespread distribution

46

in the environment, environmental persistence, biological and chemical recalcitrance, potential

47

toxicity, and bioaccumulative property.1-5 Recently, several other perfluoroalkyl acids (PFAAs)

48

including perfluorohexane sulfonate (PFHxS) have been added to the US EPA Unregulated

49

Contaminant Monitoring Regulation (UCMR 3) list due to their frequent occurrence and

50

persistence.6 Perfluoroalkyl sulfonic acids (PFSAs) including PFOS and PFHxS have been

51

commonly used as components of aqueous-film forming foams (AFFFs).7,8 PFSAs are

52

formulated by electrochemical fluorination yielding a mixture of linear (L‒) and branched (br‒)

53

isomers (~70/30% L‒/br‒).9,10 Due to their superior effectiveness for extinguishing hydrocarbons

54

fuel fires, AFFFs have been commonly and repeatedly used at military bases and airports for

55

emergency personnel training for more than the past three decades. This has led to PFOS and

56

PFHxS being detected in the low ppb to low ppm range at these sites,7,8,11-14 which is well above

57

the recent US EPA announced Provisional Health Advisory value for drinking water of 70 ng/L

58

for combined perfluorooctanoic acid and PFOS.15 Other states in the USA are enforcing even

59

lower PFOS levels, e.g., 11 ng/L for PFOS in Michigan.16 Given the frequent occurrence of

60

PFSAs above regulatory limits in groundwater especially at AFFF-impacted sites,14,17 and the

61

fact that PFSAs are unable to be further degraded by microbes,18,19 remediation technologies

62

amenable for in-situ use are needed. There are some technologies that show promise for ex-situ

63

remediation of groundwater, e.g., pump and treat, which are primarily sorptive filtration

64

processes or destruction technologies that require high energy or extreme conditions.20 However,

65

there are only a few that explore PFOS transformation using technologies with in-situ potential

66

and none for PFHxS transformation. Park et al. (2016)21 explored heat-activated persulfate (60.5


Page 4 of 26

Page 5 of 26


67

mM) at 85°C; 84 mM at 90°C for 100 h), but observed no transformation of PFOS (0.92 µM).

68

Likewise, palladium (Pd)-coated nano-sized zero valent iron (Pd0nFe0) particles for potential use

69

in a permeable reactive barrier did not yield any PFOS transformation.22 In permanganate

70

systems, Liu et al. (2012)23 reported ~50% PFOS removal in 18 days (d) at 65ºC and pH=4.2, but

71

with only 5% defluorination and < 40% desulfonation. No isomer delineation was done so it is

72

unclear if L-PFOS was degraded and only ~10% PFOS removal was observed at neutral pH.

73

Vitamin B12 (VB12) is a well- known efficient electron mediator produced naturally by

74

anaerobic soil microorganisms.24 Ochoa-Herrera et al. (2008)25 evaluated PFOS transformation

75

by VB12 at 70°C and pH 9 with Ti(III)-citrate (36 mM) as a reductant which showed 71%

76

defluorination of br‒PFOS isomers in 5 d, but no removal of L-PFOS. Although the % L-PFOS

77

relative to total PFOS will vary over time due to variations in sources and differential

78

partitioning,12, 26-28 typically close to half or more of the total PFOS present will be L-PFOS, thus

79

an in-situ technology that cannot attack L-PFOS has limited use. Furthermore, the effectiveness

80

of VB12 in the systems evaluated by Ochoa-Herrera et al. (2008)25 required high concentrations

81

of Ti(III)-citrate. Therefore, we explored alternate reductants for use in VB12 systems with a

82

primary focus on nano-sized zero valent zinc (nZn0), which has a higher reduction potential than

83

Ti(III),29 and to a lesser extent (one temperature and time), nFe0 and Pd0nFe0. Transformation

84

over time with VB12-nZn0 was investigated at 22, 70 and 90°C for PFOS and at 70°C for PFHxS,

85

with F- generation and isomer-specific PFSA removal rates quantified. L-PFOS did not degrade.

86

For br-isomers, organic intermediates were identified at each sampling time in the 70°C

87

experiments, and feasible reaction pathways were proposed. In VB12 systems with nFe0 or

88

Pd0nFe0, which were evaluated for only PFOS at 70°C after a 5-d reaction, no L-PFOS

89

degradation was observed.



90

MATERIAL AND METHODS

91

Chemicals. PFOS isomer structures and nomenclature are provided in Fig. S1 in the Supporting

92

Information (SI). Technical heptadecafluorooctane sulfonate potassium salt (PFOS, C7F17SO3K,

93

≥ 98%), technical tridecafluorohexane-1-sulfonate potassium salt (PFHxS, C6F13SO3K, ≥ 98%),

94

and Vitamin B12 (VB12, Cyanocobalamin, C63H88CoN14O14P, ≥ 98%) were purchased from

95

Sigma-Aldrich (St. Louis, MO, USA). Palladium acetate (Pd(C2H3O2)2, 99.98% metal basis) was

96

obtained from Alfa Aesar (Ward Hill, MA). Mass-labelled sodium perfluoro-1-[13C8] octane

97

sulfonate (M8PFOS, 13C8F17SO3Na, > 99%) and sodium perfluoro-1-[1,2,3-13C3] hexane

98

sulfonate (M3PFHxS, 13C312C3SO3Na, ≥ 99%) were obtained from Wellington Laboratories

99

(Ontario, Canada) for use as internal standards (IS). For isomer identification and composition

100

quantification in technical PFOS and technical PFHxS, potassium perfluoro-1-octanesulfonate

101

(L‒PFOS), sodium perfluoro-1-hexanesulfonate (L‒PFHxS), potassium perfluorooctane

102

sulfonate (Technical grade), sodium perfluoro-1-methylheptane sulfonate (P1MHpS; 1‒PFOS),

103

sodium perfluoro-3-methylheptane sulfonate (P3MHpS; 3‒PFOS), sodium perfluoro-4-

104

methylheptane sulfonate (P4MHpS, 4‒PFOS), sodium perfluoro-5-methylheptane sulfonate

105

(P5MHpS; 5‒PFOS), and sodium perfluoro-5-methylheptane sulfonate (P6MHpS; 6‒PFOS)

106

were obtained from Wellington Laboratories (Ontario, Canada). nZn0 (99.7%, 40 ~ 60 nm) and

107

nFe0 (99.9%, ~25 nm) were obtained from SkySpring Nanomaterials, Inc. (Houston, TX).

108

Sources of other chemicals and solvents used are provided in SI.

109 110

Batch experiments. Batch experiments were conducted to assess if 0.2 g nZn0 in 10-mL 0.4 mM

111

VB12 can transform technical PFOS (8 µM) at 22, 70, and 90°C and at 70°C using 0.2 g nFe0

112

and Pd0nFe0 (1.0% of Pd weight basis to 0.2 g nFe0) at an initial pH (pHi) = 10.4 in unbuffered


Page 6 of 26

Page 7 of 26


113

solutions. Synthesis of Pd0nFe0 particles was described in Park et al. (2017).22 L‒PFOS

114

transformation in the absence of br-PFOS isomers and transformation of technical PFHxS (8 µM)

115

were also evaluated in VB12-nZn0 systems (pHi = 10.4) but only at 70°C. We chose an initial

116

high pH condition to assess the role of alternate reductants based on previous work by Ochoa-

117

Herrera et al. (2008)25 for PFOS (pH 6.4 to 8.9) using Ti(III) as the reductant and Amir and Lee

118

(2011)30 for tetrachloroethene (pH 5 to 9) using nZn0 as a reductant that showed removal in

119

VB12 systems increases with increasing pH. We did not want to use buffers; therefore, pHi was

120

set at 10.4 by adjusting with 0.001 M NaOH such that the pH would hopefully stay in the

121

alkaline range during the reaction to maintain VB12 reactivity.

122

All sample preparation and treatments were conducted in an anaerobic chamber (> 95%

123

N2, < 5% H2). High density 120-mL polyethylene (HDPE) crimp serum bottles were used;

124

HDPE was previously shown to have negligible adsorption of PFOS and PFHxS.31 To each

125

bottle, nZn0 or other alternate reductants (0.2 g) was added followed by 5 mL each of VB12 (0.8

126

mM) and PFSA (16 µM) solutions. Bottles were capped with rubber stoppers sealed with

127

aluminum crimp caps and wrapped with aluminum foil to prevent light exposure. Samples were

128

placed in a preheated oven (70 & 90°C) or kept at room temperature (22 ± 0.5°C) under static

129

conditions. Final pH (pHf) was measured in one sample at the last sampling time. The pH

130

measured at the end of each sampling time (1, 2, 3, 5, 7, 11, and 21 d for 22 and 70°C for VB12-

131

nZn0; 1, 2, 3, and 5 d for 90°C for VB12-nZn0; and 5-d for nFe0 and Pd0nFe0) is referred to as

132

pHf (final pH for that sample time). Heated samples were quickly cooled to room temperature

133

(22 ± 0.5°C) by placing them in ice followed by centrifugation (1 h at 3,300 rpm; 2,042 g). The

134

aqueous phase was transferred to a 50-mL polypropylene (PP) tube and the remaining solid

135

particles were extracted 5 times successively with 15-mL of acidified methanol (10:90, v/v, 1%



136

acetic acid/MeOH). No residual PFSAs were detected in the 5th extraction. When discussing

137

PFSA removal in this study, PFSAs removal is defined as PFSAs not recovered after exhaustive

138

extraction of the solid reductants (nZn0, nFe0, or Pd0nFe0). Therefore, % PFSA was quantified by

139

comparing the combined PFSAs mass in the aqueous phase and extracts of the solid particles to

140

the initial PFSA mass measured in the applied PFSA solution × 100.

141 142

PFSA and isomer analysis. Aqueous phase samples were diluted with MeOH (1:1 v/v) and

143

solvent extracts were diluted with water so all samples prior to injection were 1:1 v/v

144

MeOH/water to reduce adsorption to high-performance liquid chromatography (HPLC) vials,

145

minimize matrix effects, and optimize sensitivity. All samples were analyzed using a Shimadzu

146

(Nexera x2) ultra-HPLC (uPLC) with an AB Sciex Quadrupole Time of Flight (QTOF) 5600

147

mass spectrometer (MS) (detailed in SI). Immediately prior to analysis, 30 µL of IS was added to

148

each HPLC vial and used for quantification. PFSA isomer analysis was done using an Ascentis

149

Express F5 PFP column (2.1 × 100 mm, 2.7 µm, 90 Å, Sigma-Aldrich) with a Phenomenex

150

column pre-filter (detailed in SI). Retention times of the PFSA isomers were confirmed using

151

commercially-available individual PFSA isomer standards. The isomeric composition of

152

technical PFOS from Sigma-Aldrich used in the batch reactions was quantified using individual

153

isomer-specific calibration curves with M8PFOS IS correction (linear, r2 > 0.99). Percent isomer

154

compositions were calculated by the ratio of each isomer-specific concentration to total PFOS

155

concentration, which was determined by integration of all isomer peaks. The 3- and 4-PFOS

156

isomers were quantified together due to co-elution (Fig. S2A). Although we had a 1-PFOS

157

standard, the 1-PFOS peak was not well resolved from the dm-PFOS and 2-PFOS isomers with

158

peaks being of low intensity and broad. Therefore, the difference between total isomers in


Page 8 of 26

Page 9 of 26


159

technical PFOS minus the isomers that were quantified independently provided estimates for the

160

sum of the other unspecified isomers. For PFHxS isomer analysis, only br‒ versus L‒ was

161

quantified in technical PFHxS using L‒PFHxS standard calibration curves with M3PFHxS IS

162

correction (linear, r2 > 0.99). The difference between total PFHxS concentration and L‒PFHxS

163

quantified independently was used as an estimated of the total br‒PFHxS isomers. A

164

chromatogram of br- and L-PFHxS peaks is exemplified in Fig. S2B. Estimated percent isomeric

165

compositions of technical PFOS and PFHxS, isomer-specific limits of detection (LOD) and

166

limits of quantification (LOQ) are reported in Table S1.

167 168

Identification of organic intermediates/products. Initial screening of any volatile organic

169

intermediates or final products (herein intermediates and final products will also be generically

170

referred to as products) was performed by pulling a headspace sample through the septa with a

171

stainless steel needle attached to a 5-mL disposable plastic syringe and directly injecting onto a

172

Shimadzu 17A gas chromatograph (GC) with an electron capture detector (ECD). For organic

173

products in all aqueous samples and solvent extracts of the nZn0 particles, undiluted samples

174

were analyzed using an EVO C18 column (2.1 x 100 mm, 5 µm, 100 Å) with a Phenomenex

175

column pre-filter coupled to a uPLC/QTOF MS using three different methods for each

176

corresponding purpose (detailed in SI including Table S2). Peaks were identified as potential

177

organic product candidates only if peak intensities of parent masses (m/z) in the samples were at

178

least 10 times higher than those in the PFSAs stock solution and matrix controls. The most

179

feasible theoretical chemical formulas of the parent masses observed were initially explored

180

using PeakView/MasterView and then proposed using Chemdraw (ver.15, Perkin Elmer). To

181

assess the validity of the proposed theoretical chemical formulas, three steps were followed. First,



182

the parent mass of candidates and the proposed theoretical chemical formulas were compared by

183

calculating the relative difference between their masses referred to as a difference error (DE in

184

ppm) as follows: (|candidate mass ‒ theoretical mass)/(candidate mass)) × 1,000,000 ppm|).

185

Second, the empirical candidate and theoretical isotopes masses were also compared using the

186

same DE calculation. If the DE of both the parent candidate masses and their isotopes were less

187

than 15 ppm, the samples were re-analyzed using a product ion mode to trigger MS/MS spectra

188

of the parent candidate mass using a higher collision energy (CE of -30 or -50 kV) than normal

189

CE (CE -10 kV). Third, product ions of the candidate parent masses were compared with those

190

of the theoretical fragments using DE. If calculated DE of all product ions were also below 15

191

ppm, the candidate mass and proposed theoretical chemical formulas was deemed a reasonable

192

fit.

193 194

Inorganic products analysis. The generation of F‒ and/or SO32‒ was expected as products of

195

PFSA transformation. SO32– is not stable and rapidly oxidizes to SO42– when exposed to air22,32

196

therefore, solutions were intentionally exposed to air to allow complete conversion to SO42– prior

197

to analysis of SO42–. F‒ and SO42‒ in aqueous samples along with PFSA stock solutions and

198

matrix controls were analyzed following EPA Method 300.033 on an Agilent 1100 Ion

199

Chromatograph coupled to an Alltech electrical conductivity detector and using a NaOH mobile

200

phase. External standard curves and peak heights were used for quantification. Background

201

levels if any were subtracted from samples.

202 203

Statistical analysis. Rates of PFSA isomer removal using all individual replicates were

204

determined using a linear regression model in Excel 2016 (p-value = 0.05). One-way analyses of


Page 10 of 26

Page 11 of 26


205

variance were used for all statistical analyses to determine the significant difference among the

206

results at a p-value of 0.05.

207 208

RESULTS AND DISCUSSION

209

PFOS removal and defluorination with VB12-nZn0. Only br-PFOS isomers were removed

210

and transformed by VB12-nZn0 similar to what was found with VB12 and Ti-citrate.25 Even

211

when only the L‒PFOS isomer was present, no removal was observed (101 ± 7 % recovered, p =

212

0.2318, Fig. S3); therefore, lack of L-PFOS removal is not due to a competitive effect in the

213

multiple isomer system (technical PFOS). Fluoride generation accompanied br-PFOS removal,

214

but no sulfate generation was observed. Natural log plots of br-PFOS isomer removal relative to

215

br-PFOS isomers at t=0 and % defluorination for reactions at 22, 70 and 90°C are summarized in

216

Fig. 1. Defluorination was calculated based on the moles of F- generated relative the total moles

217

of F- available from br-PFOS (17 moles of F- per mole of initial PFOS). As temperature

218

increased, both br‒PFOS isomer removal and defluorination increased. By 5 d at 90°C, 95% of

219

br-PFOS was removed with primarily 3&4-PFOS remaining (Fig. 2). The decrease in pH was

220

also greater with increasing temperature: pHf values of 10.2, 8.7, and 7.9 for reactions with

221

PFOS at 22°C (21 d), 70°C (21 d) and 90°C (5 d), respectively, and 8.6 (21 d) for PFHxS at

222

70°C. Cobalt (Co) has a unique oxidation-dependent color with Co(III) being red, Co(II) amber,

223

and Co(I) light blue to colorless.34,35 Sample solutions were red, amber, and very light yellow to

224

colorless for 22, 70 and 90°C, respectively at the last sampling date, thus evident of enhanced Co

225

reduction with increasing temperature. The reduction to super reduced Co(I) at 90°C may have

226

been further facilitated by temperature-enhanced nZn0 corrosion to Zn2+.36 At 90°C, the degree

227

of defluorination of br-PFOS between 3-d and 5-d samples were not statistically different (p=



Page 12 of 26

228

0.6343) and reflect 100% defluorination. However, several closely and partially co-eluting

229

unidentified peaks that increased with reaction time and temperature challenged quantifying

230

accurately F- concentrations especially at later times at 90°C (exemplified in Fig. S4). Regardless,

231

the trend of increasing F- concentration over time was consistent and clear.

232 233

PFOS removal and defluorination with nFe0 and Pd0nFe0. When nFe0 was applied as an

234

alternate reductant in the high pH (pHi = 10.4) VB12 system, PFOS removal at 70ºC was not

235

statistically different (p= 0.0227) than observed with nZn0 as the reductant with ~20% technical

236

PFOS removed in 5 d (Fig. S5) and no significant removal of L-PFOS was observed. Addition of

237

Pd0 to nFe0 system as a catalyst reduced the degree of PFOS removal (

Alternate Reductants with VB12 to Transform C8 and C6

Recommend Documents