Effect of Structural Transformation of Nanoparticulate Zero-Valent Iron

Subscriber access provided by UNIV OF TASMANIA

Article

Effect of Structural Transformation of Nanoparticulate ZeroValent Iron on Generation of Reactive Oxygen Species Di He, Jinxing Ma, Richard N. Collins, and T. David Waite Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.5b04988 • Publication Date (Web): 09 Mar 2016 Downloaded from http://pubs.acs.org on March 14, 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 33

Environmental Science & Technology

1

Effect of Structural Transformation of Nanoparticulate

2

Zero-Valent Iron on Generation of Reactive Oxygen Species

3 4 Di He†, Jinxing Ma ‡, Richard N. Collins† and T. David Waite†*

5 6

†

7

NSW 2052, Australia

8

‡

9

Science and Engineering, Tongji University, Shanghai, 200092, PR China

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

State Key Laboratory of Pollution Control and Resources Reuse, College of Environmental

10

Email addresses: [email protected] (Di He); [email protected] (Jinxing Ma);

11

[email protected] (Richard N. Collins); [email protected] (T. David Waite)

12 13 14 15 16


17

Re-submitted February, 2016

18 19 20 21 22 23

__________________________________ *Corresponding +61-2-9385-5060

author:

Professor

T.

David

Waite;

1 ACS Paragon Plus Environment

[email protected],


24

ABSTRACT

25

While it has been recognized for some time that addition of nanoparticlate zero-valent iron

26

(nZVI) to oxygen-containing water results in both corrosion of Fe0 and oxidation of

27

contaminants, there is limited understanding of either the relationship between transformation

28

of nZVI and oxidant formation or the factors controlling the lifetime and extent of oxidant

29

production. Using Fe K-edge extended X-ray absorption fine structure (EXAFS)

30

spectroscopy, we show that while nZVI particles are transformed to ferrihydrite then

31

lepidocrocite in less than two hours, oxidant generation continues for up to ten hours. The

32

major products (Fe(II) and H2O2) of the reaction of nZVI with oxygenated water are

33

associated, for the most part, with the surface of particles present with these

34

surface-associated Fenton reagents inducing oxidation of a target compound (in this study,

35

14

36

oxides on the nZVI surface with the initial formation of high surface area ferrihydrite

37

facilitating rapid and extensive adsorption of formate with co-location of this target

38

compound and surface-associated Fe(II) and H2O2 apparently critical to formate oxidation.

39

Ongoing formate oxidation long after nZVI is consumed combined with the relatively slow

40

consumption of Fe(II) and H2O2 suggest that these reactants are regenerated during the

41

nZVI-initiated heterogeneous Fenton process.

C-labelled formate). Effective oxidation of formate only occurred after formation of iron

42


Page 2 of 33

Page 3 of 33

43


INTRODUCTION

44

Zero-valent iron (ZVI) has been used to degrade a variety of contaminants. In

45

particular, reductive reactions induced by ZVI have been employed in permeable reactive

46

barriers to remediate groundwaters contaminated with chlorinated and nitro-substituted

47

organics.1, 2 Use of nano-scale zero-valent iron (nZVI) has been suggested as an alternative

48

means of exploiting the reactivity of ZVI without the need to construct passive barriers.3 The

49

main advantages of nZVI are its high reactivity and the potential for introducing the particles

50

directly into contaminated soil and groundwater.4

51

Generally, nZVI-mediated degradation of contaminants is implemented under anoxic

52

conditions since the presence of oxygen would be expected to lower the reductive efficiency

53

of the process as a result of competition with the contaminants being degraded. However, it

54

has been reported that the reaction of nZVI with O2 can produce reactive oxygen species

55

(ROS) either on the particle surface or in solution with these ROS capable of oxidizing both

56

inorganic and organic compounds.5-10 The generation of hydrogen peroxide (H2O2) during the

57

oxygenation of Fe(0) has been confirmed, with H2O2 subsequently reacting with Fe(II) to

58

produce hydroxyl radicals (HO•) and/or Fe(IV)O2+ species via the Fenton reaction.5, 6, 8, 9

59

Meanwhile, the oxidation of Fe(0) by O2 also results in the formation of Fe(II) and Fe(III)

60

species with subsequent precipitation of particulate amorphous and crystalline Fe

61

oxyhydroxides on, or in the vicinity of, the nZVI surface.11 While the formation of a surface

62

layer of Fe oxides might be expected to result in a decrease in the rate of generation of ROS

63

and, thus, the ability of nZVI to induce the oxidation of target contaminants, there has been



64

no detailed investigation of the correlation between the structural transformation of nZVI and

65

the oxidizing capacity of nZVI and any associated iron oxide products.

66

While the reaction of nZVI with target contaminants (and O2) has been emphasized

67

above, consideration must also be given to reaction of Fe(0) with water, with this reaction

68

expected to be of particular importance for nZVI where the high surface area and thus high

69

reactivity could result in the rapid corrosion of the surface and, potentially, influence the

70

ability of these reactive particles to degrade contaminants. With respect to the anaerobic

71

reaction of nZVI with water, the identification of primary products (particularly Fe(OH)2 and

72

magnetite) by 57Fe Mossbauer Spectroscopy and elucidation of the reaction mechanism have

73

been well addressed.12 In terms of the aerobic reaction of nZVI with water, however, there is

74

surprisingly limited information on both the Fe(0) oxidation kinetics and the nature of the

75

final reaction products.13 While we have implied above that formation of oxide coatings will

76

limit the extent of nZVI-mediated production of ROS, this need not be the case as it is

77

recognized that some iron oxide surfaces (including magnetite, ferrihydrite and lepidocrocite)

78

have been identified as strong sorbents for contaminants14-16 and effective substrates for

79

mediation of oxidation of these sorbed contaminants with highly reactive HO• and/or

80

Fe(IV)O2+ species generated via heterogeneous iron oxide-mediated Fenton reactions.15-19 Of

81

particular importance is the recognition in these studies that the presence of Fe(II), either at

82

the surface or within the structure of the iron oxides, is critical to the oxidative transformation

83

of the sorbed contaminants.15, 16 As such, quantitative understanding of both the structural


Page 4 of 33

Page 5 of 33


84

transformation of nZVI and surface corrosion-mediated release of Fe(II) is of great

85

importance for evaluation of the oxidizing capacity of nZVI.

86

We elucidate these effects in this study by examination of the nature of the Fe oxides

87

formed on oxygenation of nZVI particles and, in parallel, monitor both the formation/release

88

of Fe(II) and H2O2 and the oxidizing ability of the transforming nZVI assemblage over time.

89

The key questions addressed in this work include: (i) What iron oxides are formed (and how

90

fast they are formed) on exposure of nZVI to an oxygenated solution? (ii) How does the

91

concentration of Fe(II) and H2O2 change with time of exposure of nZVI to an oxygenated

92

solution? (iii) How does the oxidizing capacity of the transforming nZVI assemblage change

93

with time (and how is this capacity related to the measured concentrations of Fe(II) and

94

H2O2)? and iv) What is the relationship between the changing nature of the nZVI assemblage

95

with time and the oxidizing capacity of this assemblage?

96



97

MATERIALS AND METHODS

98

Reagents. All chemicals used in this work were analytical reagent grade and were

99

purchased from Sigma-Aldrich unless otherwise stated. All solutions were prepared in

100

ultrapure 18.2 MΩ cm Milli-Q water (Millipore). All glassware was soaked in 5% v/v

101

hydrochloric acid (HCl) for at least three days before use. 5.0 mM Fe(II) and Fe(III) stock

102

solutions were prepared by dissolving appropriate amounts of ferrous sulfate heptahydrate

103

(FeSO4·7H2O) and ferric sulphate (Fe2(SO4)3) in 5.0 mM HCl solutions, respectively. A 3.0

104

mM 1,10-phenanthroline stock solution was prepared in a buffer solution containing 0.1 M

105

acetic acid (pH = ~4) while 20 mM M H2O2 stock solutions were prepared weekly by dilution

106

of a 30% w/v H2O2 solution and standardized by UV-Vis spectrometry with a molar

107

extinction coefficient of 22.7 M-1cm-1 at 250 nm.20 Stock solutions of 200 µM Amplex Red

108

(AR; Invitrogen) mixed with 100 kU L-1 horseradish peroxidase (HRP) for H2O2

109

determination were prepared and stored at -80 oC when not in use. Sodium hydroxide (NaOH)

110

and hydrochloric acid (HCl) were used for the adjustment of pH. Experiments were

111

performed at a controlled room temperature of 22 oC.

112

Synthesis of nZVI, Lepidocrocite and Ferrihydrite. nZVI was synthesized daily in

113

an anaerobic chamber (855 Series; Plas-Lab Inc.) according to a protocol modified from

114

literature.5 In brief, a 0.25 M ferrous chloride tetrahydrate (FeCl2·4H2O) solution was

115

prepared in a 30% v/v methanol solution, and a 0.5 M sodium borohydride (NaBH4)

116

solution was added dropwise to the Fe(II) solution with magnetic stirring. The freshly

117

synthesized nZVI particles were then rinsed three times with deaerated methanol, dried 6 ACS Paragon Plus Environment

Page 6 of 33

Page 7 of 33


118

overnight at 50 °C and finally ground into a fine powder in the anaerobic chamber prior to

119

use.

120

Lepidocrocite (Lpd) was synthesized by oxygenation of nZVI in aqueous phase for 24

121

hours with the Fe oxide product harvested by centrifugation three times at 3000 rpm for 3

122

minutes (with the product washed with Milli-Q water between centrifugations) and

123

subsequently freeze-dried at -45 °C. Ferrihydrite (Fhy) was synthesized by the addition of 1

124

M NaOH to 5 mM Fe(III) solution to a final pH of 7.5.21 The freshly-made slurry was

125

rapidly mixed at 150 rpm for 30 seconds followed by slow mixing at 30 rpm for 30 minutes,

126

then centrifuged three times at 3000 rpm for 3 minutes (washed with Milli-Q water between

127

centrifugations) and freeze-dried at -45 °C.

128

Experimental Setup. Simultaneous measurements of adsorption capacity and oxidant

129

production of nZVI were conducted in a gas tight suite, as illustrated in SI Figure S1, using

130

H14COO− as a probe compound. The system consisted of a reaction vessel fitted with a

131

sparger input port and an outflow line to a trapping vessel containing 1.0 M CO2-free NaOH

132

solution. Each assay was initiated by adding a pre-weighed amount of nZVI (or, in some

133

studies, lepidocrocite or ferrihydrite) to 100 mL air-saturated Milli-Q water containing 1

134

µM H14COO−. Continuous air-sparging drove oxidized H14COO− (i.e.

135

downstream trapping vessel followed by trapping in the form of

136

solution with an increase in the concentration of 14CO32− in the trapping vessel representing

137

“oxidized” H14COO−. Following 24 h reaction, 20 mL of an acidic quench solution (~12 M


14

14

CO2) to the

CO32− in 1.0 M NaOH


Page 8 of 33

138

HCl and 1 mM formic acid) was added to the reaction vessel with continuous air-sparging

139

of the system. After the solids completely dissolved, the release of H14COOH into the

140

solution within the reaction vessel represents “adsorbed” H14COO− while an increase in the

141

concentration of

142

form of

143

experiments have confirmed that, under acidic conditions (pH = ~1), ~100% H14COOH

144

remained in the reaction vessel with negligible

145

vessel during 5 h air sparging (SI Figure S2). Samples were collected from both reaction

146

and trapping vessels at predetermined time intervals for subsequent analysis. Experiments

147

were undertaken in unbuffered systems with the pH changes following nZVI addition to

148

these oxygenated solutions (i.e. 6.2 to 8.3) typical of those expected on addition of nZVI to

149

natural waters. .

150

14

14

CO32− in the trapping vessel represents “precipitated” carbonate (in the

C-labelled siderite (FeCO3(s)) arising from the oxidation of H14COO−. Blank

14

C-labelled material found in the trapping

Analyses of Fe(II), H2O2 and H14COO−. Fe(II) concentrations were determined

151

spectrophotometrically at 510 nm using the phenanthroline method,22 with

152

concentration of Fe(II) in the aqueous phase quantified by addition of phenanthroline

153

following filtration through 0.22 µm PVDF filters (Millipore) while the concentration of

154

total Fe(II), including free and surface-associated Fe(II), was measured by addition of

155

phenanthroline prior to filtration. It should be noted that the concentration of total Fe(II) is

156

only available when nZVI particles are completely consumed (after 2 h reaction) as the

157

relatively low pH (~4) of phenanthroline stock solutions can cause rapid dissolution of


the

Page 9 of 33


158

nZVI particles, leading to dramatic discrepancies between the actual and measured

159

concentrations of total Fe(II).

160

H2O2 concentrations were measured using the Amplex Red method.23 Specifically,

161

the samples were diluted with 10.0 mM MOPS buffer (pH 7.0) and mixed in a 1 cm quartz

162

cuvette with AR/HRP stock solution at final AR and HRP concentrations of 2.0 µM and

163

1.0 kU L−1 respectively with the resorufin produced due to the oxidation of AR by H2O2 in

164

the presence of HRP exhibiting a fluorescence emission maxima at 587 nm upon excitation

165

at 563 nm. The concentration of aqueous H2O2 was determined by addition of AR/HRP into

166

diluted samples after filtration while the concentration of total H2O2, which was attributed

167

mainly to H2O2 on the surface of particles, was quantified by addition of AR/HRP into

168

diluted samples before filtration. Calibration was performed daily by the addition of H2O2

169

into MOPS buffer at pH 7.0 over the concentration range of 0~800.0 nM. H2O2 calibration

170

was unaffected by the addition of lepidocrocite (2g L-1 as Fe) indicating that the presence of

171

Fe oxyhydroxide has negligible effect on the AR/HRP fluorescence signal (SI Figure S3).

172

H14COO− in the reaction vessel and

14

CO32− in the trapping vessel were measured

173

using a Packard Tri-Carb 2100TR scintillation counter following the addition of 1 mL of

174

sample into 5 mL of liquid scintillation fluid (Ultima GOLD, PACKARD).24

175

Characterization of nZVI and oxidation products. Solid phase Fe speciation of

176

nZVI was investigated using Fe K-edge extended X-ray absorption fine structure (EXAFS)

177

spectroscopy. EXAFS measurements were conducted at the XAS beamline (ID 12) at the



178

Australian Synchrotron (Clayton, Australia). The energy was selected with a Si(111) liquid

179

nitrogen-cooled double crystal fixed-exit monochromator and the beam was focused (fully

180

tuned) both vertically and horizontally with a rhodium-coated toroidal mirror. The beam

181

energy was monitored with a Fe reference foil and found to vary by < 0.2 eV over the three

182

day period of measurements. The freeze-dried powdered references (ferrihydrite and

183

lepidocrocite) and samples (fresh and oxygenated nZVI) were diluted with boron nitride and

184

packed into aluminum slides for EXAFS transmission analyses at ~7 K. Spectra were energy

185

calibrated with the software Average and normalized and background corrected with the

186

standard features of ATHENA.25 Linear combination fits (LCF) of oxygenated nZVI samples

187

were also carried out in ATHENA over k = 0-12 Å-1 (k3-weighted), using ferrihydrite,

188

lepidocrocite and fresh nZVI sample as standards, in order to quantitatively determine the

189

composition of oxygenated nZVI samples. Consideration was given to the effects of particle

190

size in these analyses. ARTEMIS was used to perform non-linear least-squares fitting of

191

k3-weighted EXAFS nZVI spectra in R space.25 Phase and amplitude functions used to fit

192

the spectra of the references and samples were generated with FEFF6 (within ARTEMIS)

193

from the crystallographic data of Fe(0) bulk metal and nZVI.25, 26

194

A JEOL ARM200F aberration-corrected Scanning Transmission Electron Microscope

195

(STEM) was employed to identify the morphology and structure of nZVI following exposure

196

to oxygenated water. The nZVI suspensions were diluted with ethanol before mounting on a

197

copper grid coated with carbon film in an anaerobic chamber. Images were obtained in high

198

angle annular dark field (HAADF) and bright field (BF) modes. X-ray microanalysis was


Page 10 of 33

Page 11 of 33


199

carried out using a Noran System, coupled to a JEOL large area (1sr) silicon drift X-ray

200

detector. X-ray elemental maps (256×256) were acquired in spectrum image mode, using a

201

0.1 nm probe under 0.15 nA condition. Data was analyzed using the routines in the NSS

202

software, which included principal component analysis (PCA) and phase analysis.



203

Page 12 of 33

RESULTS AND DISCUSSION

204

Production of Fe(II) and H2O2. As can be seen in Figures 1 and 2, both Fe(II) and

205

H2O2 were generated following exposure of nZVI to oxygenated water at circumneutral pH.

206

The release of Fe(II) could be attributed to the ubiquitous reaction of nZVI with water (Eq. 1)

207

and the oxygenation of nZVI (Eq. 2) while the oxygenation of nZVI also results in the

208

production of H2O2 through the reduction of O2 at the nZVI surface (Eq. 2). It can be clearly

209

observed from Figures 1a and b that surface-associated Fe(II) (Fe0n-1-FeII) accounts for the

210

majority of total Fe(II) generated following the addition of nZVI to oxygenated water,

211

implying that the equilibrium between Fe0n-1-FeII and Fe(II) in solution (FeII(aq)) (Eq. 3)

212

remains predominantly to the left, at least under the conditions used in the studies described

213

here. This is consistent with our observation (Figures 2a and b) that most H2O2 also exists on

214

the surface of nZVI.

215

Fe 0n + 2H 2O → Fe 0n−1 -Fe II + H 2 + 2OH −

216

Fe0n + O 2 2H  → Fe0n−1 -Fe II + H 2O 2

+

(1) (2)

217

(3)

218

As shown in Eq. 2, the reaction between Fe0 and O2 also results in the formation of

219

H2O2, with most of this reactive oxygen species, like Fe(II), present in surface-associated

220

form (Figure 2). Interestingly, while the time dependence of the increase in concentration of

221

Fe(II) and H2O2 immediately following exposure of nZVI to oxygenated water are similar,

222

the measured concentrations of both surface-associated and dissolved H2O2 are somewhat


Page 13 of 33


223

less than that of surface-associated and dissolved Fe(II) suggesting the presence of a sink

224

(possibly Fe0) for H2O2.

225

With further exposure (after 1 h) of nZVI to oxygenated water, a dramatic decrease in

226

the concentration of H2O2 and Fe(II) was observed (see Figures 1 and 2), suggesting that the

227

consumption of H2O2 and Fe(II) outcompetes the generation of H2O2 and Fe(II). This

228

decrease can be attributed to (i) the loss of Fe(0), (ii) oxygenation of Fe(II) and/or (iii) the

229

interaction between Fe(II) and H2O2 over time. The consumption of H2O2 by Fe(II)

230

presumably occurred via heterogeneous and homogeneous Fenton reactions (Eqs. 4 and 5)

231

whereas the interaction of H2O2 with Fe(0) resulted in production of Fe(II) and water with

232

this reaction exhibiting no capacity for generation of highly reactive intermediates (Eq. 6).

233

The oxygenation of Fe(II) represents an alternative sink for both dissolved and

234

surface-located Fe(II) (Eqs. 7 and 8) under the aerobic and circumneutral pH conditions

235

employed in these studies.

236

Fe0n−1 -FeII + H2O2 → Fe0n−1 -FeIII + HO• +OH−

(4)

237

FeII + H2 O2 → FeIII + HO• +OH−

(5)

238

2H Fe0n + H2O2  → Fen0−1 -FeII + 2H2O

+

(6)

239

(7)

240

(8)

241

+

•− 2H O•− → H 2 O2 + O 2 2 + O2 

(9)

242

Structural Transformation of nZVI. The solid Fe species resulting from exposure

243

of nZVI to oxygen were examined, as a function of exposure time, by EXAFS spectroscopy.



Page 14 of 33

244

The EXFAS fitting results of both Fe(0) metal foil and nZVI are shown in Figure 3a and the

245

structural parameters are summarized in SI Table S1. The spectrum for Fe(0) metal foil

246

showed a pronounced first Fe-Fe shell at R + ∆= 2.469±0.016 Å from CN = 8 Fe atoms while

247

the spectrum of nZVI displayed the first Fe-Fe shell at very similar R + ∆= 2.454±0.009 Å

248

from CN = 2.7±0.4 (SI Table S1), indicating that (i) Fe(0) dominates the speciation of Fe in

249

nZVI and (ii) the coordination number of the Fe-Fe shell strongly depends on particle size.

250

Figure 3b shows the k3-weighted EXAFS spectra for nZVI, ferrihydrite, lepidocrocite and

251

oxidized nZVI following exposure to oxygenated water over a 24 h time period. The LCF

252

results (Figure 3c) indicate that nZVI transformed to lepidocrocite via an intermediate

253

ferrihydrite phase following exposure of nZVI to oxygenated water with essentially all nZVI

254

transformed to iron oxides after two hours of exposure to the oxic aqueous environment. This

255

is consistent with our STEM results (Figure 4) which show that the amorphous assemblages

256

initially grow out of the nZVI spherical core with subsequent transformation to “platelet-like”

257

crystalline forms.

258

unreacted nZVI while the matchable oxygen distribution (Figures 4b and c) was clearly

259

observed following exposure of nZVI to oxygenated water over time, further confirming the

260

conversion of nZVI to Fe oxides in oxygenated water.

261 262 263 264

STEM-XEDS elemental mapping only revealed the abundance of Fe for

Based on these general observations, the following scheme was developed in order to quantify the transformation kinetics of the various mineral phases (Eqs. 9 and 10): k

1 Fe0n−1 -Fe III   → Fhy

k2 Fhy  → Lpd Fe(II)


(9) (10)

Page 15 of 33


265

In this scheme, oxidized nZVI initially undergoes transformation to form ferrihydrite

266

followed by the Fe(II)-catalyzed transformation of this amorphous Fe(III) oxide to

267

lepidocrocite. No assumption was made as to how these processes occur; simply that they can

268

be considered as pseudo-first order processes with rate constants remaining constant

269

throughout a reaction for each particular treatment. For such a scheme, the decrease in nZVI

270

concentration, increase in lepidocrocite concentration and change in ferrihydrite

271

concentration over time can be expressed as shown in Eqs. 11-13 (further details relating to

272

the kinetic calculation can be found in SI Section S2):

273

[nZVI] = [nZVI]0 e− k1t

(11)

274

[Lpd] = [nZVI]0 (1 − e − k1k2t /( k1 + k2 ) )

(12)

275

[Fhy] = [nZVI]0 − [nZVI] − [Lpd]

(13)

276

where [nZVI]0 is the initial nZVI concentration and k1 and k2 represent the pseudo-first order

277

rate constants for transformation between nZVI and ferrihydrite and transformation from

278

ferrihydrite to lepidocrocite, respectively. The values of k1 and k1k2/(k1+k2) were calculated

279

by conducting a linear least squares regression analysis of ln([nZVI]/[nZVI]0) and

280

ln(1−[Lpd]/[nZVI]0) as a function of time (5 h), respectively. As demonstrated in SI Section

281

S2, the values of k1 and k2 are calculated as 6.1×10-3 and 5.9×10-3 s-1, respectively.

282

It is perhaps a little surprising that Fe(II)-catalyzed transformation of ferrihydrite to

283

goethite does not occur following the exposure of nZVI to oxygenated water. The absence of

284

this phase however could be related to the relatively low molar ratio of Fe(II) to Fe oxide (

Effect of Structural Transformation of Nanoparticulate Zero-Valent Iron

Recommend Documents