Impact of pH on Iron Redox Transformations in Simulated Freshwaters

Oct 26, 2018 - The impact of the pH of natural waters on the various pathways contributing to the formation and decay of Fe(II) in the presence of Suw...
1 downloads 0 Views 658KB Size
Subscriber access provided by UNIV OF WESTERN ONTARIO

Environmental Processes

Impact of pH on iron redox transformations in simulated freshwaters containing natural organic matter Shikha Garg, Chao Jiang, and T. David Waite Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b03855 • Publication Date (Web): 26 Oct 2018 Downloaded from http://pubs.acs.org on October 28, 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 44

Environmental Science & Technology

1 2

Impact of pH on iron redox transformations in simulated freshwaters

3

containing natural organic matter

4 Shikha Garg, Chao Jiang and T. David Waite*

5 6 7

School of Civil and Environmental Engineering, The University of New South Wales, Sydney,

8

NSW 2052, Australia

9 10 11 12

Environmental Science & Technology

13

Revised and resubmitted

14

October 2018

15 16 17 18

*Corresponding

author: Phone +61-2-9385 5060; Email [email protected] 1 ACS Paragon Plus Environment

Environmental Science & Technology

19 20

Abstract

21

The impact of the pH of natural waters on the various pathways contributing to the formation and

22

decay of Fe(II) in the presence of Suwannee River Fulvic Acid (SRFA) is investigated in this

23

study. Our results show that thermal Fe(III) reduction occurs as a result of the presence of

24

hydroquinone-like moieties in SRFA with the rate of Fe(III) reduction by these entities relatively

25

invariant with change in pH in the range 6.8-8.7. The Fe(II) oxidation rate in the dark is

26

controlled by its interaction with O2 and increases with increase in pH with the overall outcome

27

that the steady-state Fe(II) concentration in the dark is strongly affected by solution pH. On

28

irradiation, a portion of the hydroquinone-like moieties present are oxidized to form

29

semiquinones that are capable of reducing Fe(III) and/or oxidizing Fe(II) under circumneutral pH

30

conditions. The extent of photo-generation of semiquinones on irradiation of SRFA and the

31

persistence of these radicals increases significantly with decrease in pH. Due to the higher

32

concentration and longevity of these organic moieties under low pH conditions, the impact of pH

33

on steady-state Fe(II) concentration is less pronounced in previously irradiated SRFA solution

34

compared to that observed in dark SRFA solution. Under irradiated conditions, the rates of Fe

35

transformation (including both Fe(II) oxidation and Fe(III) reduction) is nearly independent of

36

pH. While ligand to metal charge transfer (LMCT) is the dominant pathway for photochemical

37

Fe(III) reduction, Fe(II) oxidation under irradiated conditions mainly occurs as a result

38

interaction with O2, semiquinones and other short-lived oxidants. Overall, our data supports the

39

conclusion that, as a result of the contribution from photo-generated organic moieties to Fe redox

40

transformations, the steady-state Fe(II) concentration in irradiated surface waters containing

41

natural organic matter may not be impacted significantly by changes in pH. 2 ACS Paragon Plus Environment

Page 2 of 44

Page 3 of 44

Environmental Science & Technology

42

3 ACS Paragon Plus Environment

Environmental Science & Technology

43

TOC Art

44

4 ACS Paragon Plus Environment

Page 4 of 44

Page 5 of 44

Environmental Science & Technology

45

1. INTRODUCTION

46

Iron is a critical nutrient for all living organisms but its availability to organisms is limited even

47

though it is one of the most abundant elements in the earth’s crust. In natural waters, iron exists

48

principally in two oxidation states, namely, as ferric (Fe(III) and/or ferrous (Fe(II) iron. The rates

49

of transformation of Fe between these oxidation states are generally considered to be strongly

50

dependent on pH. With decrease in pH, the inorganic Fe(III) speciation changes with associated

51

increase in the solubility of iron oxyhydroxides. Change in inorganic Fe(III) speciation also

52

impacts the reactivity (including reducibility) of these species in natural environments. In

53

addition, change in pH affects the Fe(II) oxidation kinetics with the persistence of Fe(II) in

54

natural waters increasing with decrease in pH.1-4 While it is apparent that lowering the pH of

55

natural waters will increase Fe bioavailability when iron is present in inorganic form, the impact

56

of pH changes under conditions where natural organic matter (NOM)-Fe interactions are

57

important and where redox transformations of both Fe and NOM are operative remains unclear.

58

The variation in pH of natural sunlit environments where humic or fulvic acid-type NOM exists

59

may impact iron binding by NOM.5 The variation in the Fe binding capacity of NOM may result

60

in changes in Fe solubility and may also have a significant impact on the Fe(II) oxidation

61

kinetics which are recognized to vary considerably in the presence of various organic

62

compounds.6-12 Furthermore, changes in the rate of oxidation of organically-complexed Fe(II)

63

with pH will also affect the overall pH-dependence of the rate of Fe(II) formation and decay.8, 11,

64

13

65

a result of changes in (i) the reduction potential of the redox-active organic and inorganic

66

moieties (such as reactive oxygen species; ROS) involved in Fe transformations, (ii) the extent of

67

photo-generation of various organic moieties and ROS involved in Fe transformations and/or (iii)

The redox transformations of Fe in the presence of NOM are also expected to vary with pH as

5 ACS Paragon Plus Environment

Environmental Science & Technology

68

the lifetime of various organic moieties and ROS. Due to the complex nature of natural organics

69

and the interplay between these various factors, it is difficult to clearly predict the effect of pH

70

on Fe redox cycling in natural waters.

71

In this study, we investigate the impact of pH on Fe(III) reduction kinetics and Fe(II) oxidation

72

kinetics over the pH range 6.8-8.7 in Suwannee River Fulvic Acid (SRFA) solutions. Fe

73

transformations in SRFA solutions are investigated under three conditions in this study, namely,

74

non-irradiated, previously irradiated and continuously irradiated. The non-irradiated condition

75

may be considered representative of deep waters to which light does not penetrate while

76

previously irradiated solutions may be considered to represent surface waters at night that has

77

been irradiated during the prior daytime.

78

considered representative of surface waters during daytime. In order to obtain a comprehensive

79

understanding of iron speciation in natural waters, it is imperative to understand the nature and

80

rates of iron transformations under each of these conditions. Note that the term “surface waters”

81

used here refers to the top few centimeters of the water body where photochemical processes will

82

be important. In deeper waters, thermal processes will control Fe transformation since

83

penetration of light will be unimportant due to complete absorption of light by NOM present in

84

the surface waters. The results of this study extend our understanding from earlier studies of iron

85

redox transformations at acidic pH (3-5) to pH conditions more typical of natural waters.14-17

86

Readers are referred to our previous studies under acidic conditions to compare and contrast with

87

the observations presented here. The results obtained here also have direct relevance to

88

atmospheric aerosols and rainwater where organic complexation of Fe is recognized to maintain

89

Fe in dissolved form and photochemical reduction of organically-complexed Fe(III) is

90

considered to be an important process. 18-22

The continuously irradiated condition may be

6 ACS Paragon Plus Environment

Page 6 of 44

Page 7 of 44

Environmental Science & Technology

91

While the focus of this study is on the impact of pH on Fe(II) and Fe(III) transformations under

92

dark and irradiated conditions, additional attention should be given in future studies to the effects

93

of other parameters such as ionic strength, dissolved oxygen concentration and temperature on

94

SRFA-mediated iron redox transformations.

95

2. EXPERIMENTAL METHODS

96

2.1.

Reagents

97

All solutions were prepared using 18 MΩ.cm resistivity Milli-Q (MQ) water (TOC < 0.1 mg.L-1)

98

unless stated otherwise. All experiments were performed at a controlled room temperature of

99

22 °C. No changes in the temperature of the solution were observed in the dark studies however

100

the temperature of the experimental solution increased slightly (by ~2°C) during irradiation for

101

10 min. All chemicals were analytical grade and were purchased from Sigma-Aldrich unless

102

stated otherwise. All glass and plastic ware were soaked in 3% HCl for at least 24 h prior to use.

103

All stock solutions were stored at 4 °C in the dark when not in use.

104

All pH measurements were undertaken using a Hanna 210 pH meter with pH adjustments made

105

using high purity HCl and NaOH. A maximum pH variation of ± 0.1 unit was allowed during

106

experiments. The pH electrode was calibrated on the NBS scale using NIST-traceable buffer

107

solutions (pH 4.01, 7.01 and 10.01). For experiments at pH 8.3, a 2 mM NaHCO3 solution in

108

equilibrium with CO2 in the atmosphere was used. For experiments at pH 6.8, 7.3 and 8.7, 2 mM

109

NaHCO3 solution in equilibrium with synthetic air containing 15000, 6000 and 200 ppm CO2

110

(HiQ® certified calibration standards; BOC) respectively was used. To allow equilibrium of CO2

111

between the solution and the gas phase, solutions were sparged in Dreschel bottle for two hours

112

prior to experiments. All buffer solutions contained 2 mM NaHCO3 and 10 mM NaCl. Addition

7 ACS Paragon Plus Environment

Environmental Science & Technology

113

of SRFA to the buffer solution resulted in a slight decrease in pH (by 0.1-0.2 units) at pH 8.3 and

114

8.7 but was adjusted back to the original pH by addition of a small amount of NaOH.

115

Stock solutions of 2 g.L-1 standard SRFA (International Humic Substances Society; batch

116

number 2S101F), 0.1 M ferrozine (3-(2-pyridyl)-5,6-diphenyl-1,2,4-triazine-4′,4″-disulfonic acid

117

sodium salt; abbreviated as FZ), 100 µM Amplex Red (AR; Invitrogen), 50 kU.L-1 horseradish

118

peroxidase (HRP), 4 mM Fe(II) chloride in 0.2 M HCl, 2 mM Fe(III) chloride in 0.2 M HCl

119

and

120

containing Cu and Zn) were prepared as described in our earlier work. 16

121

To prevent Fe(III) precipitation, the ratio of iron: SRFA concentrations was maintained at 10

122

nmoles Fe/mg-SRFA (0.056 w/w%). The final Fe and SRFA concentrations used in all

123

experiments was 100 nM and 10 mg.L-1, respectively. The ratio of iron: SRFA concentrations

124

used here is consistent with the ratio used in our earlier studies under acidic conditions thereby

125

allowing comparison of iron transformations under these different pH conditions.14-16 The ratio

126

of [Fe]:[SRFA] employed here (0.056 w/w%) can be generalized to natural waters that are

127

reported to have [Fe]:[DOC] ratios ranging between 0.0002−15 w/w%.23

128

2.2 Experimental setup

129

The experimental setup for the photochemistry experiments was the same as that described in our

130

earlier work.16 Briefly, irradiation of SRFA and/or Fe-SRFA solutions were performed in a 1 cm

131

path length quartz cuvette (volume ~ 3.5 mL) using a 150 W Xe lamp equipped with air mass

132

filters (ThermoOriel). The spectral irradiance of the lamp and the absorbed photon irradiance (in

133

μEinstein.m-2 .s-1) of 10 mg·L-1 SRFA as a function of wavelength was reported in our previous

134

work and is independent of pH.14

3 kU.mL-1 stock solution of superoxide dismutase (SOD from bovine erythrocytes

8 ACS Paragon Plus Environment

Page 8 of 44

Page 9 of 44

Environmental Science & Technology

135

To avoid the influence of rapid Fe(II) oxidation, 1 mM FZ was introduced in the experiments in

136

which Fe(III) reduction was investigated with this approach similar to the ‘FZ trapping’

137

technique that has been used in other studies.24, 25 The presence of 1 mM FZ outcompetes Fe(II)

138

oxidation completely as confirmed by the observation that no decrease in Fe(II) concentration

139

occurs within 1 h in the presence of 1 mM FZ at all pHs investigated here. Under these

140

conditions, the measured Fe(II) formation rates on Fe(III) reduction are the same as the Fe(III)

141

reduction rates. Note that under the experimental conditions used here, FZ-mediated Fe(III)

142

reduction was unimportant since the rate of Fe(III) reduction was the same in the presence of 0.5,

143

1, and 2 mM FZ (p > 0.05 calculated using one-way ANOVA; Figure S1).

144

For measurement of Fe transformation rates in non-irradiated SRFA solutions, appropriate

145

volumes of Fe(III) or Fe(II) stock were added to 30 mL of buffer solution containing 10 mg.L-1

146

SRFA and samples withdrawn at regular time intervals to measure either the concentration of

147

Fe(II)FZ3 formed or the concentration of Fe(II) remaining, respectively, using the FZ method.

148

For measurement of Fe transformation rates under previously-irradiated conditions, appropriate

149

volumes of Fe(III) or Fe(II) stock were added to 3 mL of pre-irradiated 10 mg.L-1 SRFA solution

150

in a 1 cm quartz cuvette and the concentration of Fe(II)FZ3 formed or concentration of Fe(II)

151

remaining, respectively, after a certain time was measured using the FZ method. For

152

measurement of Fe transformation rates in continuously-irradiated conditions, appropriate

153

volumes of Fe(III) or Fe(II) stock were added to 3 mL of 10 mg.L-1 SRFA solution in a 1 cm

154

quartz cuvette and the solution irradiated for 1, 2, 5 or 10 minutes. The concentration of

155

Fe(II)FZ3 formed or concentration of Fe(II) remaining after irradiation was measured using the

156

FZ method. More detailed description of the experimental setup used for non-irradiated,

9 ACS Paragon Plus Environment

Environmental Science & Technology

157

previously irradiated and continuously irradiated SRFA solutions is provided in the Supporting

158

Information (S1).

159

Oxygen removal experiments at pH 6.8, 7.3, 8.3 and 8.7 (where pH was controlled by CO2

160

equilibrium with 2 mM NaHCO3 solution) were carried out by sparging 2 mM NaHCO3

161

solutions with argon gas containing 15000, 6000, 300 and 200 ppm CO2 (HiQ® certified

162

calibration standards, BOC), respectively, in a Dreschel bottle for 2 hours prior to experiments.

163

The efficiency of oxygen removal was estimated to be ~ 95% based on the measurement of the

164

rate of oxygenation of inorganic Fe(II) in solutions that were treated in the same manner. In all

165

oxygen removal experiments, sparging was continued throughout the experiments in order to

166

ensure that sufficiently low oxygen levels were maintained at all times.

167

2.3 Measurement of Fe(II)

168

Concentrations of total Fe(II) were measured using the FZ method26 as described in our earlier

169

work.16 FZ traps both inorganic Fe(II) and SRFA-complexed Fe(II) forming the Fe(II)FZ3

170

complex, the concentration of which was measured using Visible Spectroscopy employing a 1 m

171

pathlength cell coupled with an Ocean Optics Spectrophotometry system. Detailed description of

172

the method is provided in Supporting Information (SI).

173

2.4 Modelling and statistical analysis approaches

174

Kinetic modelling was performed using the software package Kintecus.27 Statistical analysis was

175

performed using one-way analysis of variance (ANOVA) at the 5% significance level. It should

176

be noted that the modeling undertaken in this study is not a “fitting exercise”. Rather, a model is

177

developed based on a proposed set of chemical reactions and their associated rate equations.

178

Good agreement between model output and experimentally determined results over a range 10 ACS Paragon Plus Environment

Page 10 of 44

Page 11 of 44

Environmental Science & Technology

179

of conditions (as shown later) confirms that the chosen reaction set is a reasonable

180

description of the underlying processes operating under the experimental conditions used.

181 182 183

3. RESULTS AND DISCUSSION 3.1 Fe(III) reduction and Fe(II) oxidation kinetics in non-irradiated SRFA solutions in the

184

pH range 6.8-8.7

185

3.1.1 Fe(III) reduction rate

186

As shown in Figure 1, the Fe(II) concentration increased as a result of thermal reduction of Fe(III)

187

with the Fe(III) reduction rate slightly increasing with decrease in pH. For all pH conditions

188

examined, the Fe(III) reduction rate is rapid in the first 10 min then slows down in the later

189

stages. The decrease in the Fe(III) reduction rate is not due to oxidation of Fe(II) since any Fe(II)

190

formed is trapped by FZ thereby preventing Fe(II) oxidation by dioxygen and other oxidants

191

present in our experimental matrix. Also, the Fe(II) concentration generated on Fe(III) reduction

192

in the later stages does not reach a plateau (Figure 1) with this result excluding the possibility of

193

depletion of a single Fe(III) reductant in this solution. It thus appears that the Fe(II) generation

194

profile can be attributed to:

195

(i)

the presence of two or more Fe(III) reductants with the decrease in the initial rapid Fe(III)

196

reduction rate in the first 10 min due to the depletion of a stronger Fe(III) reductant (i.e.

197

Rs; eq.1) with a relatively weaker Fe(III) reductant (Rw; eq.2) present in SRFA solution

198

responsible for the slower Fe(III) reduction in the later stages; and/or

199 200

(ii)

the presence of two or more different Fe(III) species (which we will denote as Fe(III)L1 and Fe(III)L2) with the ferric iron present in Fe(III)L1 form readily reducible (eq. 3) while

11 ACS Paragon Plus Environment

Environmental Science & Technology

Page 12 of 44

201

the ferric iron present as Fe(III)L2 is reduced relatively slowly (eq. 4) by Fe(III) reducing

202

organic moieties (R) that are intrinsically present.

203

k1 R s  Fe(III)   R s  Fe(II)

(1)

204

k2 R w  Fe(III)   R w  Fe(II)

205

L R + Fe(III)L1   R   Fe(II)L1 1

(3)

206

L R + Fe(III)L2  R   Fe(II)L2

(4)

k

k

(2)

2

207

The presence of two ligand classes in SRFA is consistent with the results of an earlier study on

208

Fe(II) oxidation kinetics in SRFA solution which showed formation of a weaker Fe(II)-SRFA

209

complex in the initial stages followed by the formation of a second type of Fe(II)–fulvic acid

210

complex that is resistant to oxidation by both O2 and H2O2 in the later stages.28 Voelker and

211

Sulzberger also showed slow formation of a less reducible Fe(III)-SRFA complex.9 Additionally,

212

the difference in the Fe(II) oxidation kinetics observed in the presence of organic ligands

213

released by Synechococcus PCC7002 under Fe-limiting and high-Fe conditions has been

214

accounted for by considering the presence of two types of organic ligands exhibiting differing Fe

215

chelation properties.29 It should be noted however that the possible existence of more than two

216

classes of redox-active organic groups of differing reactivity towards Fe(III) could not be

217

entirely excluded at this stage. These possibilities are discussed further in later sections.

218

3.1.2 Fe(II) oxidation rate

219

To fully understand iron redox transformations in non-irradiated SRFA solutions, we examined

220

the Fe(II) oxidation kinetics under dark conditions. At all pHs examined, the Fe(II) concentration 12 ACS Paragon Plus Environment

Page 13 of 44

Environmental Science & Technology

221

decreases almost linearly in the initial stages due to rapid Fe(II) oxidation while a relatively

222

slower Fe(II) oxidation rate is observed in the later stages (Figure 1b). As shown in Figure 1b,

223

the initial rate of Fe(II) oxidation is strongly dependent on pH. For example, the half-life of Fe(II)

224

is ~ 5 min at pH 6.8 and < 10 s at pH 8.7. The pH dependence of Fe(II) oxidation has been

225

reported in many studies with this dependence mainly attributed to the change in Fe(II)

226

speciation with change in pH.1, 2 The half-lifes of Fe(II) measured here in SRFA solutions of

227

various pH are shorter than those reported for inorganic Fe(II) at comparable pHs (Table S1)1, 2

228

with this result confirming previous findings that the complexation of Fe(II) by SRFA increases

229

the Fe(II) oxidation rate.13,

230

resulted in a significant decrease in the Fe(II) oxidation rate at all pH values investigated with

231

this result supporting the conclusion that Fe(II) oxidation mostly occurs as a result of its

232

interaction with dioxygen in non-irradiated SRFA solution.

233

The slower Fe(II) oxidation rate observed in the later stages (Figure 1b) is presumably a result of

234

increase in the rate of Fe(III) reduction when sufficient Fe(III) is generated on Fe(II) oxidation

235

(Figure 1a) and/or due to the formation of an Fe(II)-SRFA complex that is resistant to oxidation

236

by dioxygen in the later stages as suggested by Miller and co-workers.28 Based on the measured

237

extent of Fe(III) reduction as a function of time (Figure 1a), the pseudo-first order rate constant

238

of Fe(III) reduction is estimated to be in the range ~ 2.2×10-3 – 2.5×10-5 s-1 while the initial

239

pseudo-first order Fe(II) oxidation rate constant is around 0.05±0.02 s-1 at pH 8.3 with these

240

results supporting the conclusion that Fe(III) reduction cannot explain the slowing of Fe(II)

241

oxidation kinetics in the later stages of the reaction. The inability of a reduction process to

242

sufficiently describe the data provides support for the presence of two types of Fe–SRFA

243

complexes in accord with the findings of earlier work. 9, 28

28, 30

As shown in Figure 1c, partial (~95%) removal of dioxygen

13 ACS Paragon Plus Environment

Environmental Science & Technology

244 245

3.2 Fe(III) reduction and Fe(II) oxidation kinetics in previously irradiated SRFA solution

246

in the pH range 6.8-8.7

247

3.2.1 Fe(III) reduction rate

248

When Fe(III) was added to SRFA solution that had been previously irradiated for 10 minutes

249

(and then returned to darkness), Fe(II) was generated due to thermal Fe(III) reduction at all pH

250

values (Figure 2a) in a manner similar to that observed in non-irradiated SRFA solution (Figure

251

1). However, the Fe(III) reduction in previously irradiated SRFA solution was pH dependent

252

with the Fe(II) concentrations generated on Fe(III) reduction at pH 6.8 and 7.3 substantially

253

higher than those at pH 8.3 and 8.7 (Figure 2a). In addition, the concentration of Fe(II) generated

254

in previously irradiated SRFA solution was higher than the Fe(II) generated in non-irradiated

255

SRFA solution, especially at pH 6.8 and pH 7.3. This observation suggests that an Fe(III)

256

reductant was generated on irradiation of SRFA in the pH range 6.8-8.7 with this reductant

257

playing a more significant role at lower pH. The increased Fe(III) reduction could also be due to

258

changes in Fe(III) reactivity as a result of changes in Fe(III) speciation however this possibility

259

appears unlikely since the solution conditions (i.e. pH, ionic strength, etc) were exactly the same

260

as those used in non-irradiated SRFA solution. Note that the changes in Fe-binding ability of

261

SRFA during irradiation, which were hypothesized to account for the changes in Fe(II) oxidation

262

kinetics in dark and irradiated conditions in a recent study containing exudates from

263

Synechococcus PCC702,29 were found to be unimportant here since similar Fe binding capacities

264

and Fe binding constants have been obtained for non-irradiated and previously-irradiated SRFA

265

solutions (see Figure S2 and Table S2) using the competitive ligand method described by Fujii

266

and co-workers.

5

This finding seems reasonable given that excess SRFA concentrations 14 ACS Paragon Plus Environment

Page 14 of 44

Page 15 of 44

Environmental Science & Technology

267

(compared to Fe concentrations) and a short irradiation time (10 min) were used in this study.

268

Thus, we conclude that the increased Fe(III) reduction in previously-irradiated SRFA solution is

269

as a result of generation of an Fe(III) reductant.

270

3.2.2 Fe(II) oxidation rate

271

As shown in Figure 2, Fe(II) oxidation rates when Fe(II) is added to SRFA solution that had been

272

previously irradiated for 10 min (and then returned to darkness) are higher than those measured

273

in non-irradiated SRFA solutions (Figure 1), especially at pH 6.8 and 7.3. The effect of the

274

increased Fe(II) oxidation rate is more dramatic in the initial stages when Fe(II) exists at

275

relatively higher concentration with this effect becoming insignificant in the later stages when

276

Fe(II) was mostly oxidized. Furthermore, the Fe(II) oxidation rate is largely independent of pH,

277

especially in the initial stages in previously-irradiated SRFA solution. This is an interesting

278

observation and suggests that decrease in pH may not impact Fe(II) oxidation kinetics as

279

significantly as previously envisaged, at least in irradiated surface waters.

280

The increased Fe(II) oxidation rate in previously irradiated SRFA solution suggests that an Fe(II)

281

oxidant was generated on irradiation of SRFA with the rate of Fe(II) oxidation by this oxidant

282

comparable with the rate of Fe(II) oxidation by dioxygen. This observation is in agreement with

283

our earlier work14 under acidic conditions in which we reported that semiquinone-like moieties

284

(A−) were generated on irradiation and were capable of oxidizing Fe(II) in acidic environments

285

where Fe(II) oxygenation is negligible. At circumneutral pH, it is likely that the generation of

286

semiquinone-like moieties on irradiation is responsible for the increase in the rate of Fe(II)

287

oxidation in previously irradiated solution.

15 ACS Paragon Plus Environment

Environmental Science & Technology

288

3.2.3

289

Lifetime of the additional Fe(II) oxidant and Fe(III) reductant generated on irradiation of SRFA

290

In order to obtain further insight into the identity and nature of the Fe(III) reductant and Fe(II)

291

oxidant generated on irradiation of SRFA, we determined the lifetime of the Fe(III) reductant

292

and Fe(II) oxidant experimentally. As shown in Figure 3a, when Fe(III) was added to SRFA

293

solution that had been irradiated for 10 min and stored in the dark prior to Fe(III) addition, Fe(II)

294

generation decreased as the duration for which the SRFA solution was stored in the dark

295

increased (the complete Fe(II) generation profile for various storage durations is shown in Figure

296

S3). As shown, in previously irradiated SRFA solution that had been stored in the dark for 2.5

297

hours, the Fe(II) generation is close to the Fe(II) generation in non-irradiated SRFA solution at

298

pH 8.3 and 8.7 (Figure 3a) with this result suggesting that the lifetime of the Fe(III) reductant

299

generated during irradiation is ≤ 2.5 hours. The Fe(III) reductant at pH 6.8 and 7.3 is longer-lived

300

since the Fe(III) reduction rates in previously-irradiated SRFA solution that had been stored in

301

the dark for 2.5 hours prior to Fe(III) addition are still higher than those observed in non-

302

irradiated SRFA solutions at pH 6.8 and 7.3. Such longevity eliminates the possibility of

303

31 playing a role in Fe(III) reduction in previouslysuperoxide ( O 2 ; t1/2 ~ 100 s at pH 8.3)

304

irradiated SRFA solution. Indeed, Fe(III) reduction rates in the presence and absence of 25 kU.L-

305

1 SOD

306

0.05, calculated using one-way ANOVA; Figure S4) with this result supporting the conclusion

307

that O 2 does not reduce Fe(III) in previously irradiated SRFA solution.

308

As shown in Figure 3b, with increase in the duration for which the SRFA solution was stored in

309

the dark prior to Fe(II) addition, the rate of Fe(II) oxidation decreased, especially at pH 6.8 and

(an enzyme that catalyzes the decay of O 2 to O2 and H2O2) were essentially identical (p >

16 ACS Paragon Plus Environment

Page 16 of 44

Page 17 of 44

Environmental Science & Technology

310

7.3. (Note that since the impact of photo-generated Fe(II) oxidant is more significant in the initial

311

stages, we compare the Fe(II) concentration remaining after 30 s rather than steady-state Fe(II)

312

concentration here; the complete Fe(II) oxidation kinetics profile is shown in Figure S5). We can

313

confirm that this additional Fe(II) oxidant generated on SRFA irradiation was not H2O2 since no

314

increase in the Fe(II) oxidation rate was observed on addition of 400 nM H2O2 (which is similar

315

to the H2O2 concentration generated on irradiation of 10 mg.L-1 SRFA; Figure S6) to 10 mg.L-1

316

SRFA solution under all pH conditions examined here (Figure S7).

317

One interesting observation is that, at any given pH, the lifetimes of the additional Fe(III)

318

reductant and the Fe(II) oxidant present in previously-irradiated SRFA solution were similar

319

(Table S3; see Supporting Information S2 for details of this calculation) with this result

320

supporting the hypothesis that the additional Fe(II) oxidant and Fe(III) reductant generated on

321

irradiation of SRFA are the same entity. It is likely that one entity (most likely semiquinone-like

322

radicals) are generated during irradiation of SRFA and is responsible for both Fe(II) oxidation as

323

well as Fe(III) reduction. A purported role of semiquinone moieties in Fe(III) reduction and

324

Fe(II) oxidation has been suggested previously.32-36 Furthermore, the presence of semiquinone

325

moieties in SRFA and other natural organics is widely reported by other researchers.37-40 As such,

326

the involvement of semiquinone-like moieties present in SRFA in Fe redox transformations

327

appears likely.

328

3.3 Fe(III) reduction and Fe(II) oxidation kinetics in continuously irradiated SRFA

329

solution in the pH range 6.8-8.7

330

3.3.1 Fe(III) reduction rate

17 ACS Paragon Plus Environment

Environmental Science & Technology

Page 18 of 44

331

When SRFA solutions containing Fe(III) were irradiated continuously, the Fe(II) concentration

332

increased as a result of Fe(III) reduction (Figure 4). The rates of Fe(II) generation in

333

continuously irradiated solutions are significantly higher than those observed in non-irradiated

334

(Figure 1) and previously irradiated SRFA solutions (Figure 2) at all pH values examined with

335

these results suggesting that a reduction pathway other than those discussed in earlier sections

336

plays an important role in continuously irradiated SRFA solution. This observation also suggests

337

that photochemically-mediated Fe(III) reduction will be dominant in sunlit surface waters while

338

thermal Fe(III) reduction will play a role in deep waters or in surface waters during the night

339

time.

340

There are two potential pathways for rapid photochemical Fe(III) reduction, namely, ligand to

341

metal charge transfer (LMCT) within the Fe(III)SRFA complex and/or inorganic Fe(III)24, 41-45

342

44-46 and/or reduction by photo-generated O 2 (i.e., superoxide-mediated iron reduction (SMIR)).

343

Note that the presence of any short-lived photo-excited organic moieties may also explain the

344

rapid photochemical Fe(III) reduction observed however electron transfer from these moieties to

345

quinones and/or oxygen would appear far more likely. Furthermore, due to the strong evidence

346

in the literature supporting an LMCT pathway for Fe(III) reduction,

347

consideration is given to the possible role of short-lived photo-excited organic moieties in

348

photochemical Fe(III) reduction. To probe the role of O 2 , the Fe(III) reduction rate was

349

measured in the presence of 25 kU.L-1 SOD. As shown in Figure S8, only a small decrease

350

(≤10%) in Fe(III) reduction rates is observed on SOD addition suggesting that O 2 plays a

351

minor role in photochemical Fe(III) reduction, at least under the experimental conditions

352

investigated here. This observation is consistent with recent reports on the role of O 2 in Fe(III)

353

reduction in open oceans49-52 which also suggested that Fe present as soluble or colloidal organic 18 ACS Paragon Plus Environment

24, 41-45, 47, 48

no further

Page 19 of 44

Environmental Science & Technology

354

complexes is not reduced by O 2 . Thus, we conclude that Fe(III) reduction by LMCT is likely

355

to be the dominant pathway of Fe reduction in irradiated surface waters, at least under conditions

356

where Fe is bound to humic or fulvic acids.

357

3.3.2 Fe(II) oxidation rate

358

As shown in Figure 4, the Fe(II) oxidation rates in continuously irradiated SRFA solutions are

359

largely independent of pH and were lower than those observed in previously irradiated SRFA

360

solutions at all pH values investigated here (see Figure S9 for comparison of Fe(II) oxidation

361

rates in continuously and previously irradiated SRFA solutions) with this result possibly due to

362

the rapid reduction of any Fe(III) formed via Fe(II) oxidation in continuously irradiated SRFA

363

solutions (Figure 4a). In our earlier work under acidic conditions,16 we had shown that a short-

364

lived Fe(II) oxidant (possibly peroxyl-like radicals) are generated on irradiation of SRFA

365

however it is difficult to determine the role of these radicals in Fe(II) oxidation since Fe(II)

366

oxygenation is very rapid under these conditions. The half-life of Fe(II) at circumneutral pH

367

determined here is only few seconds (Figure 4) which is much shorter than the half-life of Fe(II)

368

in continuously irradiated SRFA solution under acidic conditions (~ 2 minutes) where peroxyl-

369

like radicals are the main oxidant.16 These results suggest that peroxyl-like radicals, if formed,

370

will play a minor role in Fe(II) oxidation under circumneutral pH conditions. Furthermore, as

371

discussed earlier, the role of photo-generated H2O2 is also not important under continuously

372

irradiated conditions since the Fe(II) oxidation rates in SRFA solutions in the presence and

373

absence of 400 nM H2O2 are essentially identical (Figure S7).

374

3.3 Kinetics and mechanism of Fe redox transformations

19 ACS Paragon Plus Environment

Environmental Science & Technology

375

Based on the results and discussion presented in earlier sections, we draw the following

376

conclusions regarding the mechanism of Fe redox transformations in the pH range 6.8-8.7:

377 378

1) The rate of Fe(III) reduction by organic moieties intrinsically present in SRFA in the dark is not strongly dependent on pH in the pH range examined here;

379

2) Either a new class of redox-active moieties is generated or an increase in the

380

concentration of an existing redox-active moiety occurs on irradiation of SRFA with

381

these changes resulting in an increase in the Fe(III) reduction rate in previously-irradiated

382

SRFA solution;

383 384 385 386

3) Fe(III) reduction in continuously irradiated conditions occurs via an LMCT pathway under all pH conditions investigated here; 4) Fe(II) oxidation in non-irradiated SRFA solution occurs for the most part via oxygenation with the rate of Fe(II) oxygenation increasing with increase in pH;

387

5) The rate of Fe(II) oxidation decreases after a few minutes of exposure of Fe(II) to oxic

388

SRFA solutions as result of formation of a second class of Fe(II)-SRFA complex that is

389

resistant to oxidation, at least at pH 8.3 and 8.8. This second-class of Fe(II)-SRFA

390

complex may be formed from re-arrangement of the original Fe(II)-SRFA complex to a

391

more stable conformation or may involve a second type of binding site;

392

6) A relatively long-lived Fe(II) oxidant is generated on irradiation of SRFA and plays an

393

important role in Fe(II) oxidation at lower pHs (i.e. pH 7.3 and 6.8). As a result of Fe(II)

394

oxidation by this additional Fe(II) oxidant at lower pH, the overall rate of Fe(II) oxidation

395

in previously-irradiated SRFA solution is essentially independent of pH;

20 ACS Paragon Plus Environment

Page 20 of 44

Page 21 of 44

Environmental Science & Technology

396

7) The lifetime of the additional Fe(III) reductant and Fe(II) oxidant generated on irradiation

397

of SRFA is similar with this result supporting the hypothesis that the same light-

398

generated entity is capable of both reducing Fe(III) and oxidizing Fe(II).

399

While the exact nature of the Fe(III) reducing group present intrinsically in SRFA and/or

400

generated on irradiation of SRFA cannot be determined based on our experimental results alone,

401

their behavior is consistent with that of quinone-like moieties. Hydroquinone-like moieties (A2−),

402

which were reported to exist in SRFA solution under acidic conditions,14 are likely to play a role

403

in Fe(III) reduction at the pH conditions investigated here. The Fe(III) reducing or Fe(II)

404

oxidizing entity generated on irradiation of SRFA possibly includes semiquinone-like moieties

405

(A−) that have been reported to be generated on irradiation of SRFA solution under acidic

406

conditions (eq. 5). 14

407

h  A 2 O   A

(5)

2

408

It should be noted that direct verification of the presence of these entities is not possible due to

409

their low concentration (shown to be only a few µmoles.g-1 SRFA) compared to that of bulk

410

organics.

411

transformation differ from the bulk semiquinone moieties that are present and, as is evident from

412

the long lifetime of these moieties, do not undergo rapid reaction with O2. The exact means of

413

stabilization of semiquinone-like entities in SRFA is not clear from our work but could be

414

attributed to electron delocalization and/or the presence of divalent metal ions such as Ca and

415

Mg which have been reported to stabilize semiquinone radicals within humic structures by

416

creating bridging interactions and intramolecular aggregation. 37, 39 Alternatively complexation of

53

Furthermore, we would like to highlight that the entities involved in Fe

21 ACS Paragon Plus Environment

Environmental Science & Technology

417

semiquinone radicals by divalent cations may also stabilize these radicals as has been reported to

418

be the case for quinoid compounds. 54-56

419

As discussed earlier, the Fe(III) reduction kinetics in non-irradiated SRFA solution suggests the

420

presence of (i) two (or more) redox-active organic moieties (i.e. a “one ligand class” model) or

421

(ii) two or more Fe(III) species with differing reactivity (i.e. a “two ligand class” model). Both

422

of these models of Fe(III) reduction in non-irradiated SRFA solution are shown in the reaction

423

schematic in Figure 5 and discussed in detail below.

424

3.4.1. One ligand class model

425

In this model, we hypothesize that semiquinone-like moieties, A−, are the strong Fe(III) reductant

426

(i.e. Rs) and are responsible for the rapid generation of Fe(II) in the initial stages in non-

427

irradiated and previously irradiated SRFA solutions. Concomitantly, hydroquinone-like moieties,

428

A2−, intrinsically present in SRFA, reduce Fe(III) relatively slowly as reflected in the slower

429

Fe(II) generation rate in the later stages of the reaction. On irradiation, the oxidation of A2− is

430

further facilitated as a result of O and/or singlet oxygen generation, thereby resulting in 2

431

increase in the concentration of A−, facilitating further oxidation of Fe(II) and reduction of Fe(III)

432

in previously-irradiated SRFA solution.

433

3.4.2. Two ligand class model

434

In this model, two different classes of ligand (denoted as L1 and L2) are assumed to exist in

435

SRFA and to form Fe-SRFA complexes of differing reactivity. Note that the second class of

436

Fe(III)SRFA and Fe(II)SRFA complexes (i.e. Fe(III)L2 and Fe(II)L2) may be formed via

437

rearrangement of the weaker Fe(III)L1 complex to a more stable conformation and/or interaction

438

of Fe(III)' with L2 as indicated in the reaction schematic. Note that both mechanisms will result 22 ACS Paragon Plus Environment

Page 22 of 44

Page 23 of 44

Environmental Science & Technology

439

in kinetically equivalent outcomes as long as the ligand concentrations (i.e., L1 and L2

440

concentrations) are significantly in excess of the Fe concentration present. The ferric iron

441

present in the Fe(III)L1 complex is reduced much more rapidly than the Fe(III) present in the

442

Fe(III)L2 complex by both thermal and photochemical reduction pathways. It is hypothesized

443

that the Fe(II) present in the Fe(II)L1 complex undergoes rapid oxidation by oxygen in the dark

444

and by oxygen and semiquinone-radicals in previously-irradiated SRFA solutions. In comparison,

445

the Fe(II) present in Fe(II)L2 is hypothesized to be relatively inert to oxidation.

446

We have attempted to describe our experimental results using both models (see Table S4 and 1)

447

however the one ligand class model (shown in Table S4) does not provide a good description of

448

Fe(II) transformation kinetics under irradiated conditions (see Figures S11-S13). In comparison,

449

the two ligand class model (Table 1) provides an excellent description of our experimental

450

results (for both Fe(III) reduction and Fe(II) oxidation; see Figures 1, 2 and 4) under all

451

conditions investigated here. Detailed description of the key reactions and justification of the rate

452

constants used is provided in the Supporting Information (S3). The mechanism of iron redox

453

transformation presented here is similar to that reported under acidic conditions16 but with two

454

main differences; namely (i) the presence of two types of iron binding ligands in SRFA, and (ii)

455

reduction of Fe(III) by semiquinone radicals. The observation that semiquinone radicals reduce

456

Fe(III) and/or oxidize Fe(II) under alkaline conditions but act as Fe(II) oxidant only under acidic

457

conditions16 is consistent with the fact that the reduction potential of the quinone/semiquinone

458

redox couples decreases with increase in pH (for example, pe0 = 5.42 for pH < 4.1 and pe0 = 1.32

459

at pH > 4.1 for the benzoquinone/semiquinone redox couple),35 rendering semiquinone radicals

460

better reductants at higher pHs. The key assumptions and parameters used in developing the two-

461

ligand class kinetic model are:

23 ACS Paragon Plus Environment

Environmental Science & Technology

Page 24 of 44

462

1) The reactions for ROS and organic radical generation used in the kinetic model (i.e.

463

reactions 1-14; Table 1) are obtained from our earlier work

464

made to rate constants for some of the reactions due to changes in the experimental

465

conditions (such as pH and reactor volume) from those used previously. The changes in the

466

reactor volume alters the fluence value and hence the rate of photon absorption by SRFA.

467

2) The initial concentration of hydroquinone-like moieties (i.e. Fe(III) reductant that is

468

intrinsically present) is the same as that determined under acidic conditions (i.e. 35.4

469

µmoles.g-1SRFA) in our earlier study.14 As mentioned in our earlier study,14 the

470

concentration of the moieties responsible for Fe(III) reduction is much lower than the

471

reported electron donating capacity of SRFA57, 58 with this capacity attributed, for the most

472

part, to the presence of phenolic moieties in SRFA. As such it is reasonable to conclude that

473

the entities responsible for Fe(III) reduction are distinct from bulk phenolic moieties.

474

3) The initial concentration of semiquinone-like moieties (i.e., the photo-generated Fe(III)

475

reductant and Fe(II) oxidant) in non-irradiated SRFA solution was assumed to be zero as was

476

reported to be the case under acidic conditions.14 The concentration of these moieties in

477

previously-irradiated SRFA solution was determined based on best-fit to the Fe(III) reduction

478

and Fe(II) oxidation data obtained under these conditions. The total concentration of the

479

quinone-like moieties involved in Fe transformations is much lower than the bulk quinone

480

concentration in SRFA58 with this fact supporting the conclusion that these entities are

481

distinct from the bulk quinone moieties in SRFA.

482

4) Due to the excess SRFA concentration used here, the concentration of inorganic Fe(III)

483

(i.e. Fe(III)’) in equilibrium with the SRFA-complexed Fe(III) is fairly low ( 0.9) presented here with the discrepancy in model prediction and

490

experimental data well within the variability in experimental data (