Atmospheric Processing of Volcanic Glass: Effects on Iron Solubility

Subscriber access provided by Brown University Library

Article

Atmospheric processing of volcanic glass: effects on iron solubility and redox speciation Elena Charlene Maters, Pierre Delmelle, and Steeve Bonneville Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.5b06281 • Publication Date (Web): 29 Apr 2016 Downloaded from http://pubs.acs.org on May 4, 2016

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 30

Environmental Science & Technology

TOC/Abstract Art 273x152mm (150 x 150 DPI)

ACS Paragon Plus Environment


Atmospheric processing of volcanic glass: effects on iron solubility and redox speciation Elena C. Maters1,*, Pierre Delmelle1, Steeve Bonneville2 1

Earth and Life Institute, Environmental Sciences, Université catholique de Louvain, Croix du Sud 2, bte L7.05.10, B-1348 Louvain-la-Neuve, Belgium

2

Biogéochimie et Modélisation du Système Terre, Département Géosciences, Environnement et Société, Université Libre de Bruxelles, Avenue Franklin Roosevelt 50, CP160/02, B-1050 Brussels, Belgium

*Corresponding author Email: [email protected] Tel.: +32 (0)10 47 36 38 Fax: +32 (0)10 47 45 25

ACS Paragon Plus Environment

Page 2 of 30

Page 3 of 30


1

ABSTRACT

2

Volcanic ash from explosive eruptions can provide iron (Fe) to oceanic regions where this

3

micronutrient limits primary production. Controls on the soluble Fe fraction in ash remain poorly

4

understood but Fe solubility is likely influenced during atmospheric transport by condensation-

5

evaporation cycles which induce large pH fluctuations. Using glass powder as surrogate for ash,

6

we experimentally simulate its atmospheric processing via cycles of pH 2 and 5 exposure. Glass

7

fractional Fe solubility (maximum 0.4%) is governed by the pH 2 exposure duration rather than

8

by the pH fluctuations, however; pH 5 exposure induces precipitation of Fe-bearing

9

nanoparticles which (re-)dissolve at pH 2. Glass leaching/dissolution release Fe(II) and Fe(III)

10

which are differentially affected by changes in pH; the average dissolved Fe(II)/Fetot ratio is

11

~0.09 at pH 2 versus ~0.18 at pH 5. Iron release at pH 2 from glass with a relatively high bulk

12

Fe(II)/Fetot ratio (0.5), limited aqueous Fe(II) oxidation at pH 5, and possibly glass-mediated

13

aqueous Fe(III) reduction may render atmospherically-processed ash a significant source of

14

Fe(II) for phytoplankton. By providing new insight into the form(s) of Fe associated with ash as

15

wet aerosol versus cloud droplet, we improve knowledge of atmospheric controls on

16

volcanogenic Fe delivery to the ocean.

2 ACS Paragon Plus Environment


Page 4 of 30

17

INTRODUCTION

18

In vast areas of the world’s oceans (>30%), marine primary production (MPP) is limited by an

19

insufficient supply of bioavailable iron (Fe),1,2,3 an essential micronutrient for processes

20

including photosynthesis, respiration and nitrogen fixation.4 Continental dusts such as soil

21

particles, glacial flour, and fly ash play a key role in alleviating the Fe deficiency in these waters

22

upon atmospheric deposition.5,6 Since MPP represents a control on carbon dioxide exchange

23

between the atmosphere and the ocean, and thereby contributes to climate regulation over

24

millennial time scales,7,8 considerable effort has been dedicated to quantifying Fe input to

25

seawater by continental dust deposition.5,9,10

26

Volcanic ash is increasingly recognized as an Fe source to the surface ocean.11 The millenial Fe

27

flux to the Pacific Ocean from these aluminosilicate particles produced by explosive eruptions is

28

estimated to be comparable to that from mineral dust from arid and semi-arid regions,12 with

29

millimeter- to meter-scale ash layers in deep ocean sediment evidencing ash input to the ocean

30

throughout Earth’s history.13 Further, geochemical analyses of ocean sediment and ice sheet drill

31

cores point to a relationship between periods of intense volcanism and global cooling at several

32

points in time possibly driven by increases in MPP induced by ash fallout.14,15 Importantly,

33

volcanic impacts on climate via perturbations to the carbon cycle, including by ocean Fe

34

fertilization by ash, are increasingly evoked14-16 alongside the more established effect on climate

35

of radiative forcing by volcanogenic sulfate aerosols.17 Recent field and laboratory results

36

confirm that ash deposition can modify seawater biogeochemistry by releasing Fe.11,18-22

37

A wide range of Fe release values from volcanic ash in (sea)water (18-37 000 nmol Fe g-1 ash in

38

1 h) has previously been reported,12,18,23 and controls on the fraction of Fe in ash that can be


Page 5 of 30


39

supplied to the ocean remain poorly understood.11 Olgun et al.12 found no correlation between

40

ash Fe content (~1 to 8 at.%) and Fe solubility for over forty samples from different eruptions.

41

Ayris and Delmelle24 highlighted the possibility of various volcanic and atmospheric controls on

42

ash Fe solubility but these have yet to be fully elucidated; the former have only recently been

43

explored by thermodynamic modeling of ash-gas interactions at high temperature25 while the

44

latter still await investigation. Ash is likely subjected to physicochemical processes during long

45

range transport similar to those known to enhance Fe solubility in airborne mineral dust.26-28 In

46

particular, exposure to water condensation-evaporation cycles can significantly modify Fe

47

partitioning between dissolved and particulate phases, with large pH fluctuations in the solution

48

surrounding solid particles (i.e., a highly acidic film in the ‘wet aerosol’ phase outside of clouds

49

versus a less acidic droplet in the ‘cloud droplet’ phase within clouds29) suggested to be a key

50

aspect of atmospheric processing.27,29-31

51

Condensation-evaporation cycles probably affect volcanic ash which is co-emitted with acidic

52

gases and condensates (e.g., H2SO4, HCl, possibly HF) during eruption and whose hygroscopic

53

nature promotes water adsorption.32,33 Atmospheric processing of ash has not previously been

54

studied, yet knowledge of what governs Fe solubility in ash during its lifecycle from magma

55

source to ocean sink is essential for assessing its capacity to deliver bioavailable Fe to the surface

56

ocean.11 Consideration of Fe redox speciation is also important given that ferrous (II) Fe is much

57

more soluble than ferric (III) Fe and thus may be regarded as the form most readily bioavailable

58

in seawater.2,34-36 Dissolved Fe redox speciation has seldom been reported in previous

59

atmospheric processing studies on continental dusts30,37,38 and has never been measured in

60

volcanic ash leachates.



Page 6 of 30

61

Here we investigate experimentally for the first time the influence of pH variations on Fe(II) and

62

Fe(III) mobilization from a powdered glass as a proxy for volcanic ash transported long distances

63

to the ocean. Specifically, a time series of dissolved Fe(II) and Fe(III) concentrations are

64

measured in batch reactors containing glass suspensions in H2SO4 solution either cycled between

65

pH 2 and pH 5 or kept constant at pH 2 as a point of comparison. In addition, we apply bulk and

66

surface analytical and geochemical modeling techniques to elucidate changes in Fe speciation

67

within the solid and aqueous phases induced by the simulated atmospheric processing.

68

MATERIALS AND METHODS

69

Volcanic Glass Sample

70

A glass powder of andesitic composition (SiAl0.34Fe0.09Mg0.13Ca0.13Na0.13K0.03Ti0.01O3) was used

71

in the present study as a proxy for the primary constituent of volcanic ash.39 Moreover, ash

72

delivered to the open ocean following long range transport is likely to be enriched in glassy

73

fragments due to earlier gravitational settling of crystalline particles.40 In addition, the largest

74

and most explosive eruptions correspond to violent caldera-forming ignimbrite events which

75

generate ash clouds dominated by the glassy component and which are most susceptible to lead

76

to ash deposition far from source.15,41 Details on synthesis and characterization of the glass are

77

provided in the Supporting Information (SI). The particle size distribution is broadly comparable

78

to that reported for natural ash from various explosive eruptions,42 with particles spanning 100 µm in diameter capable of being transported 100s to 1000s of km from the volcano

80

before gravitational settling.42-44


Page 7 of 30


81

Batch Dissolution Experiments

82

The atmospheric processing experiment involved exposing the glass powder to a H2SO4 solution

83

subjected to changing acidity via three cycles of pH 2 and pH 5 over a 72 h period. This

84

approach has previously been applied to simulate atmospheric processing of various Fe-bearing

85

dusts.27,29-31,45 The preference for using H2SO4 over other atmospheric acidic compounds (e.g.,

86

HNO3, organic acids) that may also interact with ash particles is justified on the basis that H2SO4

87

is readily produced by oxidation of volcanic SO2 and typically dominates the acid aerosol load in

88

an atmospheric ash cloud.17 Twelve hours of exposure at each pH value, mimicking the wet

89

aerosol (pH 2) and cloud droplet (pH 5) phases, was chosen based on the estimated lifetime of

90

aerosols and clouds in the atmosphere.46 Here the terms ‘wet aerosol’ and ‘cloud droplet’ are

91

used to designate two potential pH conditions and not to encompass all physicochemical

92

conditions corresponding to these phases in the atmosphere. The solution pH was raised by

93

addition of 6 M and 1 M (dropwise > pH 3) NH4OH and lowered by addition of 3.6 M H2SO4. A

94

control experiment exposing the glass powder to a H2SO4 solution at constant pH 2 for 36 h was

95

also performed as a point of comparison. Details of the experimental protocol are given in the SI.

96

Briefly, triplicate experiments were conducted in polypropylene batch reactors covered with

97

parafilm at 25 °C in the dark under constant gentle stirring at a solid-to-solution ratio of 1 g L-1.

98

This value is intermediate within the very wide range of particle loadings estimated in

99

atmospheric aerosols and clouds.47 During the experiments, sub-samples of the batch solution

100

were collected at various time intervals (minute to hour scale), filtered through 0.2 µm cellulose

101

acetate membrane filters, and stored in capped plastic tubes in the dark at ~4 °C until dissolved

102

Fe analyses (within two days). Sub-samples at pH 5 were acidified (to ~pH 2) with 1.8 M H2SO4

103

immediately after filtration to preserve the dissolved Fe and Fe(II)/Fe(III) ratio. At the end of the 6 ACS Paragon Plus Environment


Page 8 of 30

104

experiments, the glass remaining in solution was recovered by vacuum filtration, rinsed with

105

ultrapure water and dried in air for subsequent spectroscopic and/or microscopic analyses.

106

Dissolved Iron Analyses

107

Total Fe (Fetot = Fe(II) + Fe(III)) and Fe(II) concentrations in solution sub-samples were

108

determined colorimetrically by the Ferrozine method48 using a Genesys 10S UV-Vis

109

spectrophotometer and a 1 cm cell path length. The Ferrozine, buffer and reducing reagents were

110

prepared as described by Viollier et al.49 Standard solutions ranging from 0 to 5000 ppb of Fe(II)

111

were prepared from (NH4)2Fe(SO4)26H2O dissolved in pH 2 H2SO4. The detection limit was 2.5

112

ppb (~0.05 µM). Measurements were performed within minutes of Ferrozine and buffer reagent

113

addition and under inactinic illumination to minimize the potential for (photo)oxidation of

114

Fe(II).50 The Fe(III) concentration was calculated from the difference between measured Fetot and

115

Fe(II) concentrations.

116

RESULTS

117

Total Fe Trends

118

Dissolved Fetot concentrations (in µmol g-1 of glass) in H2SO4 solution cycled between pH 2 and

119

pH 5 over 72 h, and the corresponding measured pH values (± 0.1 pH unit), are shown in Figure

120

1. During the first pH 2 phase, dissolved Fetot concentrations increased rapidly within 1 h and

121

attained a maximum value of 5.1 ± 0.2 µmol g-1 after 12 h. During the first pH 5 phase, dissolved

122

Fetot concentrations declined steeply within 1 min and attained a minimum value of 1.1 ± 0.2

123

µmol g-1 after 12 h (at 24 h total). This overall pattern was reproduced during the subsequent two

124

cycles of pH change. Lowering the solution pH for the second and third pH 2 phases induced a


Page 9 of 30


125

steep increase in dissolved Fetot concentrations to 3.2 ± 0.1 and 3.4 ± 0.1 µmol g-1 after 1 min,

126

respectively, exceeding that measured after 1 min of exposure for the first pH 2 phase (1.0 ± 0.1

127

µmol g-1). In addition, dissolved Fetot concentrations increased overall during the second and

128

third pH 2 phases, reaching values of 5.8 ± 0.2 and 6.4 ± 0.1 µmol g-1 after each 12 h period (at

129

36 and 60 h total), respectively. In contrast, raising the solution pH for the second and third pH 5

130

phases induced an almost immediate decrease in dissolved Fetot concentrations, reaching values

131

of 0.8 ± 0.1 and 0.7 ± 0.1 µmol g-1 after each 12 h period (at 48 and 72 h total), respectively.

132

Dissolved Fetot concentrations (in µmol g-1 of glass) in H2SO4 solution at pH 2 over 36 h,

133

representing the control experiment, are shown in Figure 2a. Also incorporated in this plot for

134

comparison are the dissolved Fetot concentrations from the pH 2 phases of the cycling experiment

135

merged together, i.e., with the pH 5 phases removed. The pattern of dissolved Fetot

136

concentrations during the control and cycling experiments are remarkably similar with a rapid

137

initial increase within the first hour and a more gradual increase thereafter to final values of 6.2 ±

138

0.1 and 6.4 ± 0.1 µmol g-1 after 36 h, respectively.

139

Fe(II) and Fe(III) Trends

140

Dissolved Fetot, Fe(II) and Fe(III) concentrations (in µmol g-1 of glass) in H2SO4 solution cycled

141

between pH 2 and pH 5 over 72 h are shown in Figure 1. Dissolved Fe(III) concentrations

142

followed a cyclic pattern as noted above for dissolved Fetot concentrations. Dissolved Fe(II)

143

concentrations increased to 1.5 ± 0.2 µmol g-1 after 12 h during the first pH 2 phase, dropping to

144

< 0.1 µmol g-1 1 min after NH4OH addition, but increasing continuously to 0.3 ± 0.2 µmol g-1

145

over the following 12 h of the first pH 5 phase (at 24 h total). This general trend recurred during

146

the second and third cycles, with Fe(II) concentrations reaching 0.6 ± 0.3 and 0.8 ± 0.1 µmol g-1



Page 10 of 30

147

at the end of each successive pH 2 phase (at 36 and 60 h total) and 0.2 ± 0.1 and 0.4 ± 0.1 µmol

148

g-1 at the end of each successive pH 5 phase (at 48 and 72 h total). Overall, during the pH 2

149

phases, dissolved Fetot concentrations are comparatively high and mostly consist of dissolved

150

Fe(III) (average Fe(II)/Fetot = 0.09). During the pH 5 phases, dissolved Fetot concentrations are

151

comparatively low and are also mostly accounted for by dissolved Fe(III) although with a higher

152

relative contribution of dissolved Fe(II) (average Fe(II)/Fetot = 0.18) which increased from the

153

beginning to the end of each pH 5 phase (0.04 to 0.33, 0.03 to 0.31, and 0.07 to 0.49,

154

respectively).

155

Dissolved Fetot, Fe(II) and Fe(III) concentrations (in µmol g-1 of glass) in H2SO4 solution at pH 2

156

over 36 h are shown in Figure 2b. Both dissolved Fe(II) and dissolved Fe(III) contribute

157

significantly, though the former somewhat less than the latter, to the dissolved Fetot

158

concentrations. The relative proportion of Fe(II) in the control pH 2 solution (average Fe(II)/Fetot

159

= 0.35) exceeds that measured in the cycling pH 2 solution (average Fe(II)/Fetot = 0.09) which

160

was intermittently subjected to pH 5 conditions (Figure 2c).

161

DISCUSSION

162

Our results provide first insight into Fe mobilization from volcanic ash under pH fluctuations

163

associated with cloud condensation-evaporation cycles, demonstrating that fractional Fe

164

solubility in the glass is highest (up to 0.4%) under pH 2 conditions and lowest (up to 0.1%)

165

under pH 5 conditions (Figure 1). These observations are consistent with the pH dependence of

166

both Fe solubility and aluminosilicate glass dissolution51,52 and imply that Fe mobilization from

167

glassy ash may dominate while in wet aerosol form compared to in cloud droplet form during

168

atmospheric transport. Shi et al.29 similarly concluded that the low pH phase outside of clouds


Page 11 of 30


169

drives Fe dissolution in mineral dust. Figure 1 also shows that Fe solubility increases with each

170

successive cycle, with dissolved Fetot reaching 5.1 µmol g-1 and 6.4 µmol g-1 at the end of the

171

first and third pH 2 phases, respectively. However, the similar Fetot release during the pH 2

172

control experiment and the pH 2 phases of the cycling experiment (Figure 2a) suggests that the

173

pH fluctuation itself is not necessary to enhance Fe solubility in volcanic glass; rather it is the

174

duration of low pH exposure that dictates the total Fe mobilized from the material. These

175

observations are consistent with findings of Shi et al.29 on Fe solubility in mineral dust, although

176

as these authors note, cloud condensation-evaporation cycles may significantly influence the

177

form of Fe delivered to the surface ocean, i.e., predominantly as nanoparticulate Fe in droplet

178

deposition versus as dissolved Fe in aerosol deposition.

179

During cloud processing of mineral dust, Shi et al.31 proposed that ferric oxyhydroxide

180

precipitates as the pH increases to 5-6 (Fe(III)(aq) + 3OH-(aq) Fe(OH)3(s)) and re-dissolves at pH

181

2. This process was held responsible for the neoformation of ferrihydrite nanoparticles on cloud-

182

processed dust in the laboratory and the occurrence of ferrihydrite nanoparticles on wet-

183

deposited dust collected far from source in the field. Iron-rich nanoparticles have also been

184

generated by Kadar et al.53,54 from suspensions of continental Fe-bearing materials, including

185

crushed lapilli from Etna volcano (Italy), in pH 2 solutions subsequently raised to pH 6. In the

186

present study with volcanic glass, secondary ferric oxyhydroxides also likely formed at pH 5.

187

Equilibrium calculations performed with the Visual MINTEQ geochemical program predict

188

undersaturation with respect to ferrous oxyhydroxide (Fe(OH)2) and supersaturation with respect

189

to ferric oxyhydroxide (Fe(OH)3) in the pH 5 solution based on Fe(II) and Fe(III) concentrations

190

of 1.5 and 3.6 µmol L-1, respectively (i.e., measured at the end of the first pH 2 phase). In

191

addition, analysis of the glass material recovered at the end of the pH cycling experiment by 10 ACS Paragon Plus Environment


Page 12 of 30

192

transmission electron microscopy coupled with energy dispersive X-ray spectroscopy (TEM-

193

EDX) revealed the presence of Fe-bearing nanoparticle aggregates (Figure S2 in the SI).

194

To further elucidate factors governing Fe dissolution during the pH cycling experiment, a first-

195

order kinetic model29,55 was fitted to the time evolution of dissolved Fetot concentrations

196

measured during the pH 2 phases, assuming simultaneous dissolution of two Fe pools within the

197

glass characterized by different dissolution kinetics:

198

Mt = Ʃ(M0-M0 x e-kt)

199

Where Mt and M0 are the Fe concentrations (µmol g-1) in solution at time t and initially in a

200

particular pool within the glass, respectively, and k is the dissolution rate constant (h-1). The

201

choice of two pools was dictated by the occurrence of Fe in at least two positions in the

202

aluminosilicate glass (i.e., as glass network modifier or former) as described below. Values of

203

parameters representing the initial Fe content (M0) and the rate constant (k) for each pool, given

204

in Table S2 in the SI, were optimized using the Excel Solver function to obtain the best fit to the

205

experimental data during each of the three pH 2 phases (Figure 3a). These parameters have been

206

constrained mathematically and so cannot be said to correspond to particular Fe speciations

207

within the glass, but rather provide a model representation of changes in Fe mobilization from

208

more than one pool within the material. The model predicts a change in M0 values across

209

successive pH 2 phases indicative of a progressive increase in Fe contained within the faster

210

dissolving pool and decrease in Fe contained within the slower dissolving pool (Figure 3b),

211

consistent with the enhancement of fractional Fe solubility measured over the first, second and

212

third pH 2 phases (Figure 1). The increase in dissolved Fetot concentration measured after the

213

first minute of each of the pH 2 phases (1.0 ± 0.1, 3.2 ± 0.1, 3.4 ± 0.1 µmol g-1, respectively)

214

suggests that the Fe originally sourced from the glass was converted into a more readily soluble 11 ACS Paragon Plus Environment

Page 13 of 30


215

form over time, e.g., into a secondary Fe oxyhydroxide. In other words, Fetot concentrations

216

measured during the second and third pH 2 phases likely reflect initial immediate dissolution of

217

Fe-bearing nanoparticles formed during the preceding pH 5 phases as well as continuous gradual

218

dissolution of the glass material.

219

The Fetot mobilization trend in acid solution, namely a rapid initial Fe release transitioning to a

220

slower and steadier Fe release (Figure 2a), is consistent with glass leaching and dissolution.

221

Volcanic glass exhibits a complex Fe(II)-Fe(III) distribution reflecting properties such as melt

222

composition, temperature, pressure and oxygen fugacity.56,57 Generally, Fe(II) occurs in six-fold

223

coordination as a glass modifier while Fe(III) exists in four-fold coordination as a glass

224

former.58,59 In our glass, with a bulk Fe(II)/Fetot ratio of 0.5 (Table S1 in the SI), both Fe(II) and

225

Fe(III) may be present as glass modifiers while Fe(III) probably also occurs as a glass

226

former.58,59 Acidic leaching initiates breakdown of the aluminosilicate via the ready removal of

227

glass modifying elements, including Fe(II) and possibly Fe(III), by exchange with protons in

228

solution.60 Dissolution concurrently liberates the residual glass modifying elements, including Si,

229

Al and possibly Fe(III), by proton mediated hydrolysis, with breakage of the Si-O bond being the

230

rate limiting and final step in dissolution.61 These processes are supported by X-ray

231

photoelectron spectroscopy (XPS) analysis of the treated glass which show surface cation

232

depletion and features of proton-mediated exchange/hydrolysis (Figures S3 and S4 in the SI).

233

Further, the rapid initial (within 1 h) Fetot release from our glass in pH 2 solution may result in

234

part from leaching, with the transition (from ~3 to 6 h) to a slower Fetot release (Figure 2a)

235

reflecting the longer time taken for Fe to diffuse from greater depths within the glass surface.

236

Near-instantaneous leaching and dissolution of ultrafine particles (

Atmospheric Processing of Volcanic Glass: Effects on Iron Solubility

Recommend Documents