Silicate Binding and Precipitation on Iron Oxyhydroxides

Jan 5, 2018 - These species are the attachment sites from which oligomerization and polymerization reactions occur, starting at loadings exceeding ∼...
0 downloads 10 Views 1MB Size
Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE

Article

Silicate Binding and Precipitation on Iron Oxyhydroxides Masakazu Kanematsu, Glenn Waychunas, and Jean-François Boily Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.7b04098 • Publication Date (Web): 05 Jan 2018 Downloaded from http://pubs.acs.org on January 5, 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 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 28

Environmental Science & Technology

1

Silicate Binding and Precipitation on Iron Oxyhydroxides

2

Masakazu Kanematsu 1, 2, Glenn A. Waychunas 1, Jean-François Boily 2,*

3 4

1

5

Berkeley, CA 94720, U.S.A.

6

2

Energy Geosciences Division, Lawrence Berkeley National Laboratory, One Cyclotron Road,

Department of Chemistry, Umeå University, SE-901 87 Umeå, Sweden

7 8

*corresponding author. + 46 73 833 2678; [email protected]

9 10 11 12

Submitted to Environmental Science & Technology

13 14 15 16 17 18 19 20 21

August 9, 2017.

22

In revised form on October 26, 2017

23

In 2nd revised form, on December 28nd, 2017

ACS Paragon Plus Environment

1

Environmental Science & Technology

24

Page 2 of 28

Abstract

25

Silica-bearing waters in nature often alter the reactivity of mineral surfaces via deposition

26

of Si complexes and solids. In this work, Fourier transform infrared (FTIR) spectroscopy was

27

used to identify hydroxo groups at goethite (α-FeOOH) and lepidocrocite (γ-FeOOH) surfaces

28

that are targeted by ligand exchange reactions with monomeric silicate species. Measurements of

29

samples first reacted in aqueous solutions then dried under N2(g) enabled resolution of the

30

signature O-H stretching bands of singly (−OH), doubly- (µ-OH) and triply-coordinated (µ3-OH).

31

Samples reacted with Si for 3 and 30 d at pH 4 and 7 revealed that –OH groups were

32

preferentially exchanged by silicate, and that µ-OH and µ3-OH groups were not exchanged.

33

Based on knowledge of the disposition of –OH groups on the major crystallographic faces of

34

goethite and lepidocrocite, and the response of these groups to ligand exchange prior

35

oligomerization, our work points to the predominance of rows of mononuclear monodentate

36

silicate species, each separated by at least one –OH group. These species are the attachment sites

37

from which oligomerization and polymerization reactions occur, starting at loadings exceeding

38

~1 Si/nm2. Only above such loadings can reaction products grow away from rows of –OH groups

39

and form hydrogen bonds with non-exchangeable µ-OH and µ3-OH groups. This can occur in

40

systems with soluble Si concentrations as low as ~0.7 mM after 30-d reaction time. These

41

findings have important repercussions for our understanding of the fate of waterborne silicate

42

ions exposed to minerals.

43 44

. Keywords: silicate; iron oxyhydroxide; adsorption; oligomerization; polymerization.

45 46

ACS Paragon Plus Environment

2

Page 3 of 28

Environmental Science & Technology

47 48

1. Introduction

49

Iron (oxyhydr)oxides are of common occurrence in nature, and play key roles in

50

environmental processes, primarily through metal or anion adsorption and interfacial electron

51

transfer processes.1, 2 Adsorption of silicate ions is particularly relevant to aquatic and terrestrial

52

environments, where (submillimolar) dissolved silica concentrations can vary with pH and

53

mineral composition.3, 4 Silicates form mineral-bound molecular complexes and polymeric gels

54

that profoundly affect interfacial reactions involved in the biogeochemical cycling of elements in

55

nature.5-7,8,

56

understanding of the mechanism through which silicate ions bind and precipitate at mineral

57

surfaces.

9

Accounting for the impact that Si plays in this context has long called for an

58 59

Previous studies9-12 dedicated to this important issue have helped constrain ideas on

60

reaction mechanisms and timescales of mineral-silicate interactions.7, 8, 13-16 Swedlund et al.8, 17

61

and Waychunas et al.16 have, for instance, addressed the potential roles that steric constraints

62

play on Si binding at crystallographically oriented surfaces. The appeal for this approach lies in

63

rationalizing Si binding in terms of the known two-dimensional disposition of OH groups

64

available for ligand exchange. This approach becomes particularly powerful when working with

65

surfaces of contrasting orientations, and distinct populations of (hydr)oxo groups.

66 67

Recent work in our group developed possibilities for addressing this question by

68

monitoring populations of hydroxo groups on goethite (GT; α-FeOOH), rod-shaped

69

lepidocrocite (RL; γ-FeOOH)) and lath-shaped lepidocrocite (LL; γ-FeOOH)) nanoparticles of

ACS Paragon Plus Environment

3

Environmental Science & Technology

Page 4 of 28

70

contrasting structures and crystal habits (Figs. 1, S1, S2). These groups, which can be singly (–

71

OH), doubly (µ-OH), and triply (µ3-OH) coordinated with underlying Fe atoms,18,

72

distinct distributions on the dominant crystallographic faces of GT, RL and LL particles (Fig. 1,

73

Table S1). Our approach, which tracks the vibration spectroscopic response of these OH groups,

74

is used to identify reactive binding sites, and in many cases, even the crystallographic faces on

75

which reactions occur. This approach therefore offers new ideas for understanding Si binding

76

and polymerization on minerals, adding to previous work in this area.7, 8, 14, 15

19

have

77 78

To this end, this work was focused on GT, RL and LL particles, for which the vibrational

79

spectra of surface hydroxo groups are well-understood (Figs. 1, S1, S2 and Table S1).19-22

80

Following an approach we have previously developed, we first reacted LL, RL and GT with

81

solutions of silicate in aqueous solutions.18-20 We thereafter tracked the reacted sites in dried

82

samples. Drying was necessary to remove water which otherwise overwhelms the vibration

83

spectroscopic responses of surface hydroxo groups. This strategy consequently opens the

84

possibility for identifying Si surface complexes in water-unsaturated media, such as in the

85

vadose zones of sediments, Earth’s atmosphere, planetary surfaces, and even in technological

86

settings. In this study, we show that –OH groups are the sole ligand exchange sites for silica at

87

iron oxyhydroxide surfaces, and therefore the primary attachment sites from which

88

oligomerization and polymerization reactions occur. We also suggest that monomeric

89

monodendate Si complexes are predominantly organized along rows of –OH groups on the

90

dominant crystallographic faces of the nanoparticles under study.

91 92

ACS Paragon Plus Environment

4

Page 5 of 28

Environmental Science & Technology

93 94

2. Methods

95

2.1 Batch adsorption experiments

96

Details on the methods of synthesis, preparation and characterization of LL, RL and GT

97

can be found in the Supplementary Information (SI; Figs. S1, S2, Table S1) section, as well as in

98

previous studies from our group.23-25 Stock solution of silicic acid (0.2, 0.4, 0.6, 0.8, 1.0, 1.5, 2.0,

99

4.0 and 6.0 mM) were freshly prepared by dissolution of trimethyl orthosilicate (TMOS) in

100

vigorously stirred Milli-Q water at pH 2.26 This strategy was beneficial for minimizing risk of

101

oligomerization/polymerization in the aqueous solutions, and therefore chiefly limiting this study

102

to the case of the adsorption of monomeric silicate species. The absence of solution polymer was

103

confirmed by the lack of Si-O stretching bands indicative of oligomerization and polymerization,

104

evaluated by Fourier transform infrared (FTIR) spectra of the stock solutions. Only the solutions

105

with Si concentrations exceeding ~2.0 mM may possibly have formed polymers during the

106

course of adsorption experiments.

107 108

All Si adsorption experiments were conducted in the absence of background electrolyte

109

ions, and under an atmosphere of humidified N2(g). This minimized any competitive adsorption

110

reactions that would occur under dry conditions. Still, because counterions were present in the

111

form of unadsorbed Si and of titrants (HCl, NaOH), adsorption reactions were conducted under

112

conditions of low but positively-charged surfaces (iep=7.7 for RL and LL; iep=9.4 for G),

113

namely at pH 4 and 7. Reactions were initiated by mixing 5 mL suspensions of FeOOH (400

114

m2/L) to 5 mL of Si stock solutions, and by adjusting pH values to 4.0 and 7.0 by addition of

115

small volumes of 1.0 M HCl or NaOH. The suspensions were then equilibrated on an end-to-end

ACS Paragon Plus Environment

5

Environmental Science & Technology

Page 6 of 28

116

rotator for up to 30 d in a room kept at 298 ± 1 K during which samples were collected to

117

monitor Si uptake kinetics. Suspensions were then monitored for pH, centrifuged (~2000 g for

118

30 min) and the supernatants analyzed for residual silicic acid concentrations. The centrifuged

119

mineral pastes were analyzed by FTIR as described in Section 2.3.

120 121

Determination of silicic acid content in the supernatants was carried out with the

122

molybdenum blue spectrophotometric method for the kinetic adsorption experiments, and by

123

Inductively Coupled Plasma/Optical Emission Spectrometry (ICP-OES) for adsorption isotherm

124

experiments. In the spectrophotometric method, reagents (Si-bearing standards and samples

125

mixed with ammonium molybdate, hydrochloric acid, oxalic acid, ascorbic acid solution and

126

acetone) were stirred, then reacted for 10 min prior to absorbance measurements with a uv-vis

127

spectrophotometer (Shimadzu UV-2100). Spectra over the entire 190-800 nm region were

128

monitored but the absorbances at 650 nm ± 10 nm were taken to evaluate total Si concentrations.

129

The measurement error of this method was no more than 3%. Total dissolved Si concentrations

130

in a 1% HNO3 matrix were determined by ICP-OES (Perkin Elmer, Optima 5300 DV), with a

131

detection limit of ~12 ppb.

132 133

2.3 FTIR Spectroscopy

134

Aliquots (10-20 µL) centrifuged wet pastes of the mineral suspensions were applied onto

135

an Attenuated Total Reflectance (ATR) cell using a spatula (Golden Gate, single-bounce

136

diamond cell), then dried under N2(g) in the analytical chamber of a FTIR spectrometer (Bruker

137

Vertex 70/V FTIR) in a room kept at 298 ± 1 K. Measurements were carried out in the 600-4500

138

cm-1 range at a resolution of 4.0 cm-1 and a forward/reverse scanning rate of 10 Hz, resulting in

ACS Paragon Plus Environment

6

Page 7 of 28

Environmental Science & Technology

139

1000 co-added spectra for each sample. A DLaTGS detector was used for these measurements,

140

and the Blackman-Harris 3-term apodization function was used to correct phase resolution.

141

Spectra were collected every 0.5 h during water evaporation until all O-H stretching and bending

142

modes of free water disappeared. The spectra of the resulting dried samples were then used to

143

identify the surface OH groups exchanged by silicate binding.

144 145

2.4 Molecular Dynamics

146

Simulations of the (110) face of goethite (Pbnm spacegroup) and the (100) face of

147

lepidocrocite were carried out to study the impact of Si binding on the hydrogen bonding

148

environments of surface hydroxo groups. Classical molecular dynamics (MD) simulations of this

149

face were carried out with the CLAYFF 27 force field for bulk oxygen and hydroxyls, the revised

150

CLAYFF parameters

151

details on the simulation approach, and Table S2 for parameters). The simulation cell was

152

generated using program METADISE30 with the crystallographic parameters of Wyckoff 31, and

153

chosen based on the terminations with the smallest energies and dipole moments. Dangling

154

surface bonds were then saturated with surface oxygens (Os) or protons, where applicable.

155

Charge-neutral

156

Fe792O660(OH)660(OsH)264(Os)66 for goethite and Fe960O840(OH)840(OsH)180 for lepidocrocite, with

157

slabs separated by a 8 nm thick void. Simulations were limited to Fe-bound H3SiO41- molecules

158

each replacing a –OH group in the form of a mononuclear monodentate complex. This choice

159

was guided from the experimental data of this study suggesting that up to ~2/3 of –OH groups on

160

rows of –OH groups can be ligand-exchanged at loadings of 2 Si/nm2. More details of the

161

simulations strategy, and analyses of the production runs, can be found in the SI.

28

slabs

for bulk Fe3+, and the Heinz et al.

were

∼3

nm

thick

and

29

parameters for H3SiO41- (cf. SI for

initially

had

the

composition

ACS Paragon Plus Environment

7

Environmental Science & Technology

Page 8 of 28

162 163

3. Results & Discussions

164

Time-resolved adsorption experiments of Si on LL, RL and GT (Fig. S3) revealed rapid

165

uptake of Si over the first 3-5 d of reaction time. This was followed by substantially slower

166

binding rates over ~25 d without full attainment of thermodynamic equilibrium. Because the

167

experiments involved freshly-made Si solutions below ~2 mM from TMOS we exclude the

168

possibility that polymerized solution species are present, and hence that polymer adsorption and

169

considerations for entropically-driven32 adsorption are not required to explain these results.

170

Adsorption isotherms at 3 and 30 d reaction time (Fig. 2; cf. Fig. S4 for data and model in log-

171

log plot) are all of IUPAC33 type IV, and consist of (i) an adsorption/oligomerization regime

172

reaching a Langmuirian maximum mostly at ~2 sites·nm-2, and of (ii) a more variable

173

polymerization/precipitation regime favored by increased reaction time. The results also show

174

that greater loadings achieved at pH 7 are driven by polymerization/precipitation.

175

The concentration fields of these two regimes were tentatively resolved by the following

176 177

model: ΓSi =Γads

178

൫K ads ∙ൣSiaq ൧൯

m

1+൫K ads ∙ൣSiaq ൧൯

m

+K poly ൫ൣSiaq ൧൯

n

(1)

179

In this equation, soluble Si concentrations ([Siaq]) were used to predict Si surface loadings (ΓSi)

180

in terms of (i) adsorption reactions (left-hand term: Langmuir-Freunlich isotherm with total site

181

density

182

polymerization/precipitation reactions (right-hand term: Freundlich isotherm with constant Kpoly

183

and non-ideality factor n). The resulting model (Table S3) suggests that no significant

184

polymerization/precipitation occurs below at least ~1.5-2.0 Si/nm2, with soluble Si levels as low

of

Γads,

adsorption

constant

Kads

and

non-ideality

factor

m),

and

(ii)

ACS Paragon Plus Environment

8

Page 9 of 28

Environmental Science & Technology

185

as ~1 mM for 3 d and ~0.7 mM for 30 d reaction time. The adsorption regime can be adequately

186

predicted using representative crystallographic site densities of –OH groups on all faces of the

187

particles under study (Table S1): namely 3.30 –OH/nm2 in GT, 2.99 –OH/nm2 in RL and 2.65 –

188

OH/nm2 in LL. Only the data for LL particles at pH 7 required larger values because of

189

polymerization/precipitations reactions; here we used a proxy density value of ~6 –OH/nm2. All

190

Si binding data and the accompanying model provide the macroscopic basis upon which the

191

nature of Si binding can be described using FTIR spectroscopy.

192 193

Building upon these results, the O-H stretching region was used to identify hydroxo

194

groups targeted by Si binding. Sites resolvable by FTIR are those that are isolated, weakly

195

hydrogen-bonded or involved in a homogeneous network of hydrogen bonds (Figs. 3 and S5-S7;

196

Table S1). On GT, these include –OH sites (3661 cm-1) of the (110) face, representing ~82% of

197

all –OH sites, and present along rows (O-O distance of ~3 Å) in which ~50% form hydrogen

198

bonds with one another (Fig. 1).19 In contrast, –OH sites of the (021) face are not resolvable as

199

distinct bands as they are involved in a complex network of hydrogen bonds (Fig. S1).19

200

Resolvable sites of GT also include µ3II-OH (3490 cm-1) of the (110) face, which form a

201

hydrogen bond with one overlying –OH group.21 On RL and LL, resolvable –OH groups (3667

202

cm-1) are those of the (001) face, and represent ~62 % of these groups. As on the (110) face of

203

GT, these groups are also exposed along rows but in which only ~15% form hydrogen bonds

204

with one another.19 This difference arises from the lack of a hydrogen bond with a neighboring

205

µ3-OH group. The remaining –OH groups of RL and LL cannot be resolved, and are present as

206

chemisorbed waters (η–OH2) on the (100) face, and as defects (~0.9 site/nm2) on the (010)

ACS Paragon Plus Environment

9

Environmental Science & Technology

Page 10 of 28

207

face.34 Additional resolvable groups on RL and LL include µ-OH of the (010) face (3626 cm-1),

208

and µ3-OH groups (3534 and 3552 cm-1) of the (100) face (Figs. 1, S1).

209

Band intensities and frequencies of resolvable groups were systematically affected by

210

variations in Si loadings (Fig. 3 for pH 7 data for 3 d reaction time; cf. Figs. S5-S7 for all data).

211

Bands of –OH in LL and RL (3667 cm-1) and GT (3661 cm-1) all underwent considerably

212

stronger changes than those of the less exchangeable µ-OH and µ3-OH groups. Most notably, Si

213

binding in the first Si/nm2 ligand-exchanged ~40-60% of the detectable –OH sites by FTIR,

214

while binding at ~1-2 Si/nm2 involved considerably less exchange (Fig. 4a). These data can

215

consequently be used to suggest that:

216

(i)

ligand exchange predominates at < 1 Si/nm2;

217

(ii)

Si binding at ~1-2 Si/nm2 is driven by oligomerization onto pre-existing monomeric Si

218

species, followed by polymerization/precipitation reactions requiring no or little

219

consumption of –OH groups;

220 221

(iii)

Growth of the polymer at > ~2 Si/nm2 consumes the remaining –OH groups and/or alters their hydrogen bonding environments.

222

In accordance with these observations, the appearance of silanols (e.g. Fe-O-Si-(OH)3) — seen

223

in O-H stretching bands at 3720 (hydrogen-bonded) and 3740 (isolated) cm-1 (Figs. 3, 4b)35 —

224

correlates with the steep losses in intensities of –OH bands ~1 Si/nm2 (Fig. 4a). Above these

225

loadings, the appearance of silanols has a weaker dependence on Si loadings (Fig. 4b), an

226

observation that is consistent with the concept of increased amounts of Si-O-Si linkages. The

227

appearance of the latter can be appreciated in the Si-O stretching region (Fig. S8) in the form of

228

polymeric species (~1062 and 1123-1130 cm-1)7, 8 at Si loadings exceeding 1.8 Si/nm2 at 3d and

229

1.1 Si/nm2 at 30 d. Less clear from the Si-O stretching region is the possible concomitant

ACS Paragon Plus Environment

10

Page 11 of 28

Environmental Science & Technology

230

appearance of oligomeric species (~985-993cm-1). Finally, monomeric ( 2 Si/nm2 at 30 d reaction time in Figs. S6 & S7), Si

249

complexes may have possibly begun forming hydrogen bonds with µ3-OH sites in a similar

250

manner described for the (010) face of LL and RL. Additionally, we stress that a blue-shift of

251

the 3490 cm-1 band of GT to ~3560 cm-1 was not the result of ligand exchange of µ3,II-OH but

252

rather of the rupture or weakening of the original intersite µ3,II-OH ⋅⋅⋅OH− hydrogen bond (Fig.

ACS Paragon Plus Environment

11

Environmental Science & Technology

Page 12 of 28

253

1) resulting from −OH/silicate exchange. The sensitivity of this band to protonation and ligand

254

exchange was demonstrated in previous studies from our group.19-20,25

255 256

Si binding mechanisms

257

Combining the information gained from the adsorption isotherms and vibration

258

spectroscopy, we draw the following suggestions on Si binding and precipitation mechanisms at

259

iron oxyhydroxide surfaces. First, the O-H stretching region provides evidence that (i) Si

260

binding targets only –OH groups, and that (ii) ~40-60% of the –OH groups that can be detected

261

by vibration spectroscopy remain unexchanged at the Langmuirian maxima (~2 sites/nm2 in Fig.

262

4). Thus, even if Si would have been selectively bound by all the (FTIR-silent) sites of the

263

terminations of the particles (e.g. 0.6/nm2 on (021) of GT, and 0.5-1.0/nm2 on (100) of RL; Table

264

S1) we evaluate that a minimum of 1.0-1.5 Si/nm2 must be bound to ~3 –OH/nm2 of the

265

dominant faces. Because bidendate (e.g. (≡Fe-O)2-SiO2 or ≡Fe-O2-SiO2) complexation would

266

require ligand exchange with the majority of these sites, it cannot explain the levels of

267

unexchanged –OH densities at the Si Langmuirian maxima (Fig. 4). Even a density of 1.0 Si/nm2

268

on these faces would ideally leave one intervening –OH group between each bound Si (e.g. [-OH

269

–OSiO3Hp –OH]) if a monomeric monodentate species prevailed. This view also falls in line

270

with related accounts on arsenate36 and phosphate18 binding on the same type of rows of –OH

271

groups, and where steric constraints have been argued in favor for monomeric monodendate

272

species. We must also highlight that other studies37,

273

monodentate and bidentate species, the relative occurrences of which are affected by surface

274

loadings and pH.

38

have also pointed to coexisting

275

ACS Paragon Plus Environment

12

Page 13 of 28

Environmental Science & Technology

276

Further evidence supporting the prevalence of mononuclear monodendate Si in the

277

systems considered in this work can be observed through the impact of such complexes on the

278

hydrogen bonding network developed at mineral surfaces. In GT, this can be observed through

279

the splitting of the original 3661 cm-1 band to a triplet at 3651, 3661 and 3670 cm-1 to loadings

280

up to ~2 Si/nm2 (Figs. 3, 5, S9). MD simulations (Fig. S10, (110) face of GT with 0.6 Si/nm2)

281

suggest that this triplet originates from the breakdown of a hydrogen bonding network initially

282

involving 50% of –OH groups (–OH···–OH) to populations as low as 10 %. In contrast to GT, Si

283

binding on the RL/LL induces a simpler response, with its original 3667 cm-1 band shifting to

284

~3655 cm-1. Consistent with this observation, MD simulations (Fig. S10) suggest that this

285

simpler response can be explained by the smaller populations of intersite hydrogen bonding (Fig.

286

1), lying at ~10-20 % irrespective of Si loading. Fig. 6 shows a tentative band assignment for

287

each of these contrasting hydrogen bond strengths. It also includes plausible hydrogen bonding

288

environments for isolated (3740 cm-1) and hydrogen bonding (3720 cm-1) silanols. MD suggests

289

that the latter are responsible for 2-2.5 H-bonds/Si on GT (Fig. S10).

290 291

Once the prevailing monodentate complex has achieved loadings exceeding ~1-2 Si/nm2,

292

oligomerization/precipitation of additional Si species can proceed via the formation of new Si-O-

293

Si linkages with Fe-bound Si. The appearance of these linkages can be seen through the

294

appearance of bands at ~1062 and 1123-1130 cm-1 (Fig. S8). These additional species tend to

295

appear at conditions corresponding to the precipitation edges of the adsorption isotherms (Fig. 2).

296

Only at this stage are oligomerization/polymerization reactions no longer directional (Fig. 1),

297

thus allowing reaction products to expand laterally from rows of the type [-OH –OSiO3Hp –OH –

298

OSiO3Hp] over areas populated by non-exchangeable µ-OH and µ3-OH sites. Only then would

ACS Paragon Plus Environment

13

Environmental Science & Technology

Page 14 of 28

299

band intensities of these two latter groups be attenuated via hydrogen bonding with the newly-

300

grown oligomeric/polymeric species. This concept could, as such, represent an extension of

301

previous ideas by Swedlund et al.17, 39, except that oligomeric/polymeric units first grow along

302

rows of –OH groups then laterally over the entire surface.

303 304

Based on these inferences, we expect that silicate monomers are likely to sorb differently

305

at mineral surfaces with dissimilar arrangements of exchangeable –OH groups. For example, if

306

exchangeable –OH groups are relatively sparse, the only way for much silicate to bind to the

307

surface will be via formation of silicate polymers attached by single surface Fe-O-Si bonds. At

308

the other extreme, where there are high densities of exchangeable hydroxyls, the precise distance

309

between such hydroxyls may favor bidentate attachment over monodentate attachment initially.

310

Examples on iron oxyhydroxides may possibly include particle terminations exposing pairs of

311

OH groups at equatorial edges of Fe, or at defects. Differences in the density and disposition of –

312

OH groups would also directly impact the onset of oligomerization/polymerization reactions.

313

These possibilities would need to be explored by further analysis, notably by X-ray absorption

314

spectroscopy, and by detailed exploration of sorption clusters with varying protonation states by

315

molecular modelling.

316 317

Molecular resolution of adsorption and oligomerization/polymerization reactions made in

318

this study has several implications in our appreciation of Si complex formation in soils and

319

aquatic environments. These findings necessarily imply that drying of soils would produce

320

silicate attachment sites facilitating polymerization at mineral surfaces. Long-term exposure of Si

321

to surfaces effectively produces substantial coverage of reactive surface sites with silicate

ACS Paragon Plus Environment

14

Page 15 of 28

Environmental Science & Technology

322

monomer and/or polymers. This can occur even at soluble Si concentrations as low as ~0.7 mM,

323

and therefore below the solubility limit of amorphous SiO2 under acidic to circumneutral

324

conditions. As polymerized silicate complexation would be difficult to desorb or dissolve, there

325

would necessarily be a large alteration in reactivity, for example with respect to contaminant and

326

nutrient transformation and transport in aquatic environments. Having noted these implications,

327

this study consequently calls for renewed efforts by the research community to improve our

328

understanding of the interfacial phenomena and structure driving the reactivity of Si-coated

329

mineral surfaces.

330 331

Acknowledgements

332

This research was supported by the Swedish Research Council (2016-03808) to J.-F. B,

333

and by a JSPS Postdoctoral Fellowships for Research Abroad to M.K. G.A.W. acknowledges

334

support from the Office of Science, Office of Basic Energy Sciences of the U.S. Department of

335

Energy (BES-DOE) under Contract No. DE-AC02-05CH11231. We appreciate the technical

336

support from Kenichi Shimizu, Xiaowei Song, and Solomon Tesfalidet during the course of this

337

work.

338 339

Supporting information available

340

Description of mineral synthesis and characterization as well as MD simulations; Si binding

341

modeling constants; Physical properties of FeOOH particles, including surface structure and

342

imaging; Si adsorption kinetics; ATR-FTIR spectra of the Si-O stretching region in Si-reacted

343

GT; Narrow O-H stretching region; MD-derived hydrogen bonding populations.

344

ACS Paragon Plus Environment

15

Environmental Science & Technology

Page 16 of 28

345 346 347 348

References

349 350 351

1. Cornell, R. M.; Schwertmann, U., The iron oxides : Structure, properties, reactions, occurrences, and uses. 2nd, completely revised and extended ed.; Wiley-VCH: Weinheim, 2003; p xxxix, 664.

352 353 354 355

2. Brown, G. E.; Henrich, V. E.; Casey, W. H.; Clark, D. L.; Eggleston, C.; Felmy, A.; Goodman, D. W.; Gratzel, M.; Maciel, G.; McCarthy, M. I.; Nealson, K. H.; Sverjensky, D. A.; Toney, M. F.; Zachara, J. M., Metal oxide surfaces and their interactions with aqueous solutions and microbial organisms. Chem. Rev. 1999, 99, (1), 77-174.

356 357

3. McKeague, J. A.; Cline, M. G., Silica in soil solutions. I. The form and concentration of dissolved silica in aqueous extracts of some soils. Can. J. Soil. Sci. 1963, 43, (1), 70-82.

358 359

4. Alvarez, R.; Sparks, D. L., Polymerization of silicate anions in solutions at low concentrations. Nature 1985, 318, (6047), 649-651.

360 361

5. Vempati, R. K.; Loeppert, R. H., Influence of structural and adsorbed Si on the transformation of synthetic ferrihydrite. Clays Clay Min. 1989, 37, (3), 273-279.

362 363 364

6. Cismasu, A. C.; Michel, F. M.; Tcaciuc, A. P.; Brown Jr, G. E., Properties of impuritybearing ferrihydrite III. Effects of Si on the structure of 2-line ferrihydrite. Geochim. Cosmochim. Acta 2014, 133, 168-185.

365 366 367

7. Swedlund, P. J.; Miskelly, G. M.; McQuillan, A. J., An attenuated total reflectance IR study of silicic acid adsorbed onto a ferric oxyhydroxide surface. Geochim. Cosmochim. Acta 2009, 73, (14), 4199-4214.

368 369 370

8. Swedlund, P. J.; Miskelly, G. M.; McQuillan, A. J., Silicic acid adsorption and oligomerization at the ferrihydrite-water interface: interpretation of ATR-IR spectra based on a model surface structure. Langmuir 2010, 26, (5), 3394-401.

371 372

9. Waltham, C. A.; Eick, M. J., Kinetics of arsenic adsorption on goethite in the presence of sorbed silicic acid. Soil Sci. Soc. Am. J. 2002, 66, (3), 818-825.

373 374

10. Meng, X. G.; Bang, S.; Korfiatis, G. P., Effects of silicate, sulfate, and carbonate on arsenic removal by ferric chloride. Water Res. 2000, 34, (4), 1255-1261.

375 376

11. Davis, C. C.; Chen, H. W.; Edwards, M., Modeling silica sorption to iron hydroxide. Environ. Sci. Technol. 2002, 36, (4), 582-587.

ACS Paragon Plus Environment

16

Page 17 of 28

Environmental Science & Technology

377 378 379

12. Christl, I.; Brechbuehl, Y.; Graf, M.; Kretzschmar, R., Polymerization of Silicate on Hematite Surfaces and Its Influence on Arsenic Sorption. Environ. Sci. Technol. 2012, 46, (24), 13235-13243.

380 381 382

13. Pokrovski, G. S.; Schott, J.; Garges, F.; Hazemann, J. L., Iron (III)-silica interactions in aqueous solution: Insights from X-ray absorption fine structure spectroscopy. Geochim. Cosmochim. Acta 2003, 67, (19), 3559-3573.

383 384 385 386

14. Song, Y. T.; Swedlund, P. J.; McIntosh, G. J.; Cowie, B. C. C.; Waterhouse, G. I. N.; Metson, J. B., The Influence of Surface Structure on H4SiO4 Oligomerization on Rutile and Amorphous TiO2 Surfaces: An ATR-IR and Synchrotron XPS Study. Langmuir 2012, 28, (49), 16890-16899.

387 388

15. Ferras, Y.; Robertson, J.; Swedlund, P. J., The Application of Raman Spectroscopy to Probe the Association of H4SiO4 with Iron Oxides. Aquatic Geochemistry 2017, 23, (1), 21-31.

389 390 391

16. Waychunas, G.; Jun, Y.-S.; Eng, S.; Trainor, T. P., Anion sorption topology on hematite: Comparison of arsente and silicate. . In Adsorption of Metals by Geomedia II, Barnett, M. O.; Kent, D., Eds. 2008; pp 31-66.

392 393 394

17. Swedlund, P. J.; Hamid, R. D.; Miskelly, G. M., Insights into H4SiO4 surface chemistry on ferrihydrite suspensions from ATR-IR, Diffuse Layer Modeling and the adsorption enhancing effects of carbonate. J. Colloid Interface Sci. 2010, 352, (1), 149-157.

395 396

18. Ding, X.; Song, X.; Boily, J.-F., Identification of fluoride and phosphate binding sites at FeOOH surfaces. J. Phys. Chem. C 2012, 116, (41), 21939-21947.

397 398

19. Song, X. W.; Boily, J. F., Structural controls on OH site availability and reactivity at iron oxyhydroxide particle surfaces. Phys. Chem. Chem. Phys. 2012, 14, (8), 2579-2586.

399 400

20. Song, X.; Boily, J.-F., Surface Hydroxyl Identity and Reactivity in Akaganeite. Journal of Physical Chemistry C 2011, 115, (34), 17036-17045.

401 402

21. Rustad, J. R.; Boily, J. F., Density functional calculation of the infrared spectrum of surface hydroxyl groups on goethite (alpha-FeOOH). Am. Mineral. 2010, 95, (2-3), 414-417.

403 404

22. Ding, X.; Song, X.; Boily, J.-F., Identification of fluoride and phosphate binding sites at FeOOH surfaces. Journal of Physical Chemistry C 2012, 116, (41), 21939-21947.

405 406 407

23. Lewis, D. G.; Farmer, V. C., Infrared adsorption of surface hydroxyl-groups and lattice vibrations in lepidocrocite (g-FeOOH) and boehmite (g-AlOOH). Clay Min. 1986, 21, (1), 93100.

408 409 410

24. Schwertmann, U.; Cornell, R. M., Iron oxides in the laboratory : Preparation and characterization. 2nd completely rev. and extended ed.; Wiley-VCH: Weinheim ; New York, 2000; p xviii, 188.

ACS Paragon Plus Environment

17

Environmental Science & Technology

Page 18 of 28

411 412 413

25. Boily, J.-F.; Felmy, A. R., On the protonation of oxo- and hydroxo-groups of the goethite (α-FeOOH) surface: A FTIR spectroscopic investigation of surface O-H stretching vibrations. Geochim. Cosmochim. Acta 2008, 72, (14), 3338-3357.

414 415

26. Brinker, C. J., Hydrolysis and condensation of silicates: Effects on structure. J. Non-Cryst. Sol. 1988, 100, 31-50.

416 417 418

27. Cygan, R. T.; Liang, J. J.; Kalinichev, A. G., Molecular models of hydroxide, oxyhydroxide, and clay phases and the development of a general force field. J. Phys. Chem. B 2004, 108, (4), 1255-1266.

419 420

28. Kerisit, S., Water structure at hematite–water interfaces. Geochim. Cosmochim. Acta 2011, 75, (8), 2043-2061.

421 422 423

29. Heinz, H.; Lin, T. J.; Mishra, R. K.; Emami, F. S., Thermodynamically Consistent Force Fields for the Assembly of Inorganic, Organic, and Biological Nanostructures: The INTERFACE Force Field. Langmuir 2013, 29, (6), 1754-1765.

424 425 426

30. Watson, G. W.; Kelsey, E. T.; deLeeuw, N. H.; Harris, D. J.; Parker, S. C., Atomistic simulation of dislocations, surfaces and interfaces in MgO. Journal of the Chemical SocietyFaraday Transactions 1996, 92, (3), 433-438.

427 428

31. Wyckoff, R. W. G., Crystal Structures 1. 1963 ed.; Interscience Publishers: New York, 1963.

429 430 431 432

32. Sparks, D. J.; Romero-González, M. E.; El-Taboni, E.; Freeman, C. L.; Hall, S. A.; Kakonyi, G.; Swanson, L.; Banwart, S. A.; Harding, J. H., Adsorption of poly acrylic acid onto the surface of calcite: an experimental and simulatino study. Phys Chem Chem Phys 2017, 17, 27375.

433 434 435 436

33. Sing, K. S. W.; Everett, D. H.; Haul, R. A. W.; Moscou, L.; Pierotti, R. A.; Rouquerol, J.; Siemieniewska, T., Reporting physisorption data for gas solid systems with special reference to the determination of surface area and porosity (Recommendations 1984). Pure and Applied Chemistry 1985, 57, (4), 603-619.

437 438

34. Boily, J.-F.; Kozin, P. A., Particle morphological and roughness controls on mineral surface charge development. Geochim. Cosmochim. Acta 2014, 141, (Supplement C), 567-578.

439 440 441

35. Yu, P.; Kirkpatrick, R. J.; Poe, B.; McMillan, P. F.; Cong, X., Structure of Calcium Silicate Hydrate (C-S-H): Near-, Mid-, and Far-Infrared Spectroscopy. Journal of the American Ceramic Society 1999, 82, (3), 742-748.

442 443

36. Loring, J. S.; Sandstrom, M. H.; Noren, K.; Persson, P., Rethinking arsenate coordination at the surface of goethite. Chemistry 2009, 15, (20), 5063-72.

444 445

37. Kubicki, J. D.; Paul, K. W.; Kabalan, L.; Zhu, Q.; Mrozik, M. K.; Aryanpour, M.; PierreLouis, A. M.; Strongin, D. R., ATR-FTIR and Density Functional Theory Study of the Structures,

ACS Paragon Plus Environment

18

Page 19 of 28

Environmental Science & Technology

446 447

Energetics, and Vibrational Spectra of Phosphate Adsorbed onto Goethite. Langmuir 2012, 28, (41), 14573-14587.

448 449 450

38. Kwon, K. D.; Kubicki, J. D., Molecular orbital theory study on surface complex structures of phosphates to iron hydroxides: Calculation of vibrational frequencies and adsorption energies. Langmuir 2004, 20, (21), 9249-9254.

451 452 453

39. Swedlund, P. J.; Sivaloganathan, S.; Miskelly, G. M.; Waterhouse, G. I. N., Assessing the role of silicate polymerization on metal oxyhydroxide surfaces using X-ray photoelectron spectroscopy. Chemical Geology 2011, 285, (1), 62-69.

454

2012, 337, 1200-1203.

455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471

ACS Paragon Plus Environment

19

Environmental Science & Technology

Page 20 of 28

472 473 474

Figure Captions

475

Figure 1. Schematic representation of crystal habits of rod lepidocrocite (RL), lath lepidocrocite

476

(LL) and goethite (GT) particles under study. See Fig. S1 for transmission electron images.

477

Disposition of surface hydroxo groups on dominant crystallographic faces, taken from results of

478

Song and Boily.19 LL and RL particles expose neutrally-charged µ-OH groups on the (010) face,

479

and ~0.9/nm2 of –OH (~10% of the µ-OH population) from imperfections. The (100) face of LL

480

and GT and the (110) and (100) faces of GT consist of rows of −OH, µ–OH, and µ3–OH groups.

481

About 15% of –OH groups in LL and RL and about 50% of these groups in GT donate a

482

hydrogen bond to a neighboring –OH. The greater population in GT is cause by a permanent

483

hydrogen bond involving and underlying µ3–OH group of structural origin. Terminations of LL

484

and RL consist of (001)-like faces consisting of µ3–OH and bare Fe (Lewis acid) sites (not

485

shown). In GT, the (021) face exposes, on the other hand, both –OH and µ–OH groups, that are

486

strongly hydrogen bonded. (see Fig. S1 and Song and Boily19 for further details).

487 488

Figure 2. Adsorption isotherms of Si on GT, RL and LL particle surface (298 K, no background

489

electrolyte), determined by ICP-OES. Curves were generated using Eq. 1 with the modeling

490

parameters of Table S3. Horizontal grey dashes lined and shaded area indicate the

491

crystallographic density (ρ -OH) of singly-coordinated groups.

492 493

Figure 3. ATR-FTIR spectra of O-H stretching region of N2-dry (298 K) GT, RL and LL

494

particles under study . (a) No silicate present. (b-d) Exposed to aqueous solutions of silicic acid

ACS Paragon Plus Environment

20

Page 21 of 28

Environmental Science & Technology

495

at pH 7 for 3 d, then dried under N2(g) at 298 K. See Figs. S5-S7 for full data sets of pH (4 and

496

7) and reaction time (3 and 30 d). We note that bands of –OH groups in samples prepared at pH

497

4 (Figs. S5-S7) are always of lower intensities that those prepared at pH 7 because protonation of

498

–OH to –OH2 alters the hydrogen bonding network and creates vacancies from the desorption of

499

H2O to the gas phase when samples are dry.

500 501

Figure 4. Areas of O-H stretching bands of G and LL as a function of Si loading. Those of RL

502

could not be included due insufficient band intensity in the band of –OH. (a) Band areas of –OH

503

of GT (3667 cm-1 region) and µ-OH of LL (3626 cm-1). (b) Band areas of 3720 and 3740 cm-1

504

arising from O-H stretches of hydroxyl groups of bound silicates.

505 506

Figure 5. The 3630-3690 cm-1 region showing splitting of the 3661 cm-1 band in GT and red-

507

shift of the 3667 cm-1 band of RL with Si loading.

508 509

Figure 6. Top and middle: Representative hydrogen bonding environments of the the (001) face

510

of lepidocrocite (top) and (110) face of GT (middle) with ~1 Si/nm2 mononuclear monodentate

511

complexes, both extracted from a snapshot of MD simulations. Bottom: Schematic representation

512

of plausible configurations of oligomers on GT.

513 514 515 516 517

ACS Paragon Plus Environment

21

Environmental Science & Technology

Page 22 of 28

518 519 520

Table of Contents Graphic

521

ACS Paragon Plus Environment

22

Page 23 of 28

Environmental Science & Technology

Schematic representation of crystal habits of rod lepidocrocite (RL), lath lepidocrocite (LL) and goethite (GT) particles under study. See Fig. S1 for transmission electron images. Disposition of surface hydroxo groups on dominant crystallographic faces, taken from results of Song and Boily.19 LL and RL particles expose neutrally-charged µ-OH groups on the (010) face, and ~0.9/nm2 of –OH (~10% of the µ-OH population) from imperfections. The (100) face of LL and GT and the (110) and (100) faces of GT consist of rows of −OH, µ–OH, and µ3–OH groups. About 15% of –OH groups in LL and RL and about 50% of these groups in GT donate a hydrogen bond to a neighboring –OH. The greater population in GT is cause by a permanent hydrogen bond involving and underlying µ3–OH group of structural origin. Terminations of LL and RL consist of (001)-like faces consisting of µ3–OH and bare Fe (Lewis acid) sites (not shown). In GT, the (021) face exposes, on the other hand, both –OH and µ–OH groups, that are strongly hydrogen bonded. (see Fig. S1 and Song and Boily19 for further details). 112x96mm (300 x 300 DPI)

ACS Paragon Plus Environment

Environmental Science & Technology

Adsorption isotherms of Si on GT, RL and LL particle surface (298 K, no background electrolyte), determined by ICP-OES. Curves were generated using Eq. 1 with the modeling parameters of Table S3. Horizontal grey dashes lined and shaded area indicate the crystallographic density (ρ -OH) of singly-coordinated groups. 235x170mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 24 of 28

Page 25 of 28

Environmental Science & Technology

ATR-FTIR spectra of O-H stretching region of N2-dry (298 K) GT, RL and LL particles under study . (a) No silicate present. (b-d) Exposed to aqueous solutions of silicic acid at pH 7 for 3 d, then dried under N2(g) at 298 K. See Figs. S5-S7 for full data sets of pH (4 and 7) and reaction time (3 and 30 d). We note that bands of –OH groups in samples prepared at pH 4 (Figs. S5-S7) are always of lower intensities that those prepared at pH 7 because protonation of –OH to –OH2 alters the hydrogen bonding network and creates vacancies from the desorption of H2O to the gas phase when samples are dry. 347x85mm (300 x 300 DPI)

ACS Paragon Plus Environment

Environmental Science & Technology

Areas of O-H stretching bands of G and LL as a function of Si loading. Those of RL could not be included due insufficient band intensity in the band of –OH. (a) Band areas of –OH of GT (3667 cm-1 region) and µ-OH of LL (3626 cm-1). (b) Band areas of 3720 and 3740 cm-1 arising from O-H stretches of hydroxyl groups of bound silicates. 232x210mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 26 of 28

Page 27 of 28

Environmental Science & Technology

The 3630-3690 cm-1 region showing splitting of the 3661 cm-1 band in GT and red-shift of the 3667 cm-1 band of RL with Si loading. 299x99mm (300 x 300 DPI)

ACS Paragon Plus Environment

Environmental Science & Technology

Top and middle: Representative hydrogen bonding environments of the the (001) face of lepidocrocite (top) and (110) face of GT (middle) with ~1 Si/nm2 mononuclear monodentate complexes, both extracted from a snapshot of MD simulations. Bottom: Schematic representation of plausible configurations of oligomers on GT.

ACS Paragon Plus Environment

Page 28 of 28