The Formation of Silicate-Stabilized Passivating Layers on Pyrite for

Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES

Article

The formation of silicate-stabilised passivating layers on pyrite for reduced acid rock drainage Rong Fan, Michael Short, Sheng-Jia Zeng, Gujie Qian, Jun Li, Russell Schumann, Nobuyuki Kawashima, Roger St.C. Smart, and Andrea R Gerson Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.7b03232 • Publication Date (Web): 23 Aug 2017 Downloaded from http://pubs.acs.org on August 27, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 27

Environmental Science & Technology

The presence of silicate in the surface iron oxy(hydroxide) layers results in a pronounced change in surface morphology, increase in pyrite surface stability and decrease in dissolution. 86x46mm (150 x 150 DPI)

ACS Paragon Plus Environment


1

The formation of silicate-stabilised passivating

2

layers on pyrite for reduced acid rock drainage

3

Rong Fana, Michael D. Shorta, Sheng-Jia Zenga, Gujie Qiana, Jun Lia, Russell C. Schumanna,b,

4

Nobuyuki Kawashimac, Roger St.C. Smarta,d and Andrea R. Gersond* a

5

School of Natural and Built Environments, University of South Australia, Mawson Lakes,

6

South Australia 5095, Australia; b

7 8

c

Levay & Co. Environmental Services, Edinburgh, South Australia 5111, Australia;

Future Industries Institute; University of South Australia, Mawson Lakes, SA 5095, Australia d

9 10

Page 2 of 27

Blue Minerals Consultancy, Middleton, SA 5213, Australia

*Corresponding author: Tel.: 61 422 112 516; Email andrea@blueminerals consultancy.com.au

11

12

ABSTRACT

13

Acid and metalliferous release occurring when sulfide (principally pyrite)-containing rock from

14

mining activities and from some natural environments is exposed to the elements is

15

acknowledged as a major environmental problem. Acid rock drainage (ARD) management is

16

both challenging and costly for operating and legacy mine sites. Current technological solutions


1

Page 3 of 27


17

are expensive and focused on treating ARD on release rather than preventing it at source. We

18

describe here a viable, practical mechanism for reduced ARD through the formation of silicate-

19

stabilised iron hydroxide surface layers. Without silicate, oxidised pyrite particles form an over-

20

layer of crystalline goethite or lepidocrocite with porous structure. With silicate addition, a

21

smooth, continuous, coherent and apparently amorphous iron hydroxide surface layer is

22

observed, with consequent pyrite dissolution rates reduced by more than 90% at neutral pH.

23

Silicate is structurally incorporated within this layer and inhibits the phase transformation from

24

amorphous iron (oxy)hydroxide to goethite, resulting in pyrite surface passivation. This is

25

confirmed by computational simulation, suggesting that silicate-doping of a pseudo-amorphous

26

iron oxyhydroxide (ferrihydrite structure) is thermodynamically more stable than the equivalent

27

undoped structure. This mechanism and its controlling factors are described. As a consequence

28

of the greatly reduced acid generation rate, neutralisation from onsite available reactive silicate

29

minerals may be used to maintain neutral pH, after initial limestone addition to achieve neutral

30

pH, thus maintaining the integrity of these layers for effective ARD management.

31


2


Page 4 of 27

32

Introduction

33

The discharge of acid and associated dissolved metals from mine wastes and some natural

34

environments has caused extensive waterway acidification and toxic metal contamination,

35

leading to adverse water quality outcomes for natural, agricultural and domestic use1, 2. Pyrite

36

(FeS2), the primary contributor to the acid release, is the most abundant sulfide mineral in the

37

earth’s crust3; consequently its oxidation and leaching characteristics have been widely studied,

38

frequently with a view to rate reduction4-6. The oxidant (ferric iron or dissolved oxygen),

39

aqueous environment and the presence of catalysts (for example iron-oxidising bacteria), all

40

significantly impact pyrite oxidation rates and have been examined in detail7. At neutral pH, as is

41

relevant here, the dominant reactions are:

42

FeS2 + 15/4O2 + 7/2H2O → Fe(OH)3 + 2H2SO4 → FeOOH + H2O + 2H2SO4

43

More correctly, the initially-formed amorphous iron hydroxide Fe(OH)3 transforms, possibly via

44

crypto-crystalline ferrihydrite8 (formally 5Fe2O3.9H2O, by loss of 6H2O from 10 units of

45

Fe(OH)3), to the crystalline oxyhydroxide (goethite or lepidocrocite) FeOOH (by loss of a further

46

H2O)9, 10. For simplicity, we will refer to the initial Equation 1 oxidation product as amorphous

47

iron hydroxide and the intermediate and final products as semi-amorphous and crystalline iron

48

oxyhydroxide respectively. Where the nature of the product is uncertain, we use the terminology

49

(oxy)hydroxide.

50

Reducing the concentration of oxidant at the pyrite surface, which can be achieved by the

51

formation of surface layers, can yield significantly decreased oxidation rates4, 5, 11-15. The

52

formation of an iron (oxy)hydroxide-silicate layer, via precipitation onto framboidal pyrite

53

surfaces, was observed to reduce the pyrite dissolution rate by one order of magnitude12, but this


(1)

3

Page 5 of 27


54

effect had limited duration. On the basis of long-term column leaching research and field studies

55

of acid rock drainage (ARD) from waste rock dumps with limestone additions at the Grasberg

56

mine (Indonesia), it was proposed that silicate, from silicate mineral dissolution, may stabilise

57

the iron (oxy)hydroxide over-layer at neutral pH, reducing pyrite oxidation rates by 50–95%

58

through restricted O2 and water access to pyrite surfaces16, 17. It has been found that these

59

passivating layers can be preserved in a continuous, coherent and stable form by providing

60

sufficient alkalinity to maintain the pH above 618. The form of this surface layer, as observed to

61

date, has been described as an amorphous iron hydroxide-containing silica11, 18. The inclusion of

62

silicate is known to significantly affect the structural characteristics of iron oxyhydroxide

63

precipitates from solution (without pyrite) and the transformation of semi-amorphous iron

64

hydroxide (ferrihydrite) to more stable iron oxyhydroxide and oxide phases (goethite and

65

hematite (Fe2O3))19-21.

66

These observations suggest that soluble silicate may have a significant influence on the nature of

67

iron (oxy)hydroxide surface layers formed on pyrite during oxidation at neutral pH, and

68

consequently, alter the efficacy of these surface layers in inhibiting further oxidation. However,

69

the formation mechanism, structure and, importantly, stability of these passivating layers on

70

pyrite has not been defined. This paper provides detailed characterisation of these surface layers

71

and their formation mechanisms with the goal of understanding their potential role in control of

72

acid and metalliferous drainage.

73

Materials and Methods

74

Sample and dissolution tests. Pyrite (FeS2) was supplied by Geo Discoveries (NSW, Australia).

75

It contained 42.8 wt.% Fe and 51.5 wt.% S with minor trace impurities of 0.01 wt.% Al, 0.01


4


Page 6 of 27

76

wt.% Ca, 0.01 wt.% Cu, 0.01 wt.% Zn and 0.16 wt.% C. Powder X-ray diffraction indicated that

77

the sample contained no mineral phases other than pyrite. The pyrite was crushed, ground, and

78

screened to a particle size range of 38−75 µm. Sonication was applied to remove fine particles

79

adhering to the larger ones. The sample was washed with 3 M HCl solution for a few minutes

80

and subsequently washed with ethanol and dried in a vacuum oven overnight. The surface area

81

was 0.35 m2 g−1 according to BET analysis.

82

A concentrated sodium metasilicate (Na2SiO3) solution (D serial sodium silicate, pH 11-12,

83

Na2O 14.7 wt.%; SiO2 29.4 wt.%; solid 44.1 wt.%; dissolved Si 22,000 mg L−1 confirmed by

84

ICP-MS) supplied by PQ Corporation (Malvern, PA, USA) was used as the source of dissolved

85

silicate. The concentrated sodium metasilicate solution was diluted in Milli-Q water to 0.8 mM,

86

i.e. 22 mg L−1 Si, which is much less than the concentrations used in previous studies using

87

silicate additions 12. With these initial silicate concentrations, the portion of polymeric silica that

88

forms in the pH range from 3.5 to 7.5 is less than 20% of the total Si and most is present as

89

silicate monomers 22. For simplicity and relation to Si assay, the silicate concentrations will be

90

expressed as mM Si. The concentration was selected because the passivation effect becomes

91

pronounced at about 50 mg L−1 sodium metasilicate (0.4 mM Si) according to our preliminary

92

study. 0.01 M KCl was used as the background electrolyte to ensure sure the solution ionic

93

strengths were approximately constant throughout the experiment 23.

94

The natural concentration of silicate in drainage waters is highly dependent on local geological

95

and hydrological condition with the Si concentration chosen for application in this study being of

96

similar magnitude to Si concentrations reported in neutral mine drainage or natural waters at pH

97

6–8 of 3.5–15 mg L−1 24, 25. A further study carried out by some of the authors of this paper has


5

Page 7 of 27


98

reported Si concentration of 55 mg L−1 at a Tasmanian mine site in waste rock dump seepage

99

(pH 3.3) and 3.7 mg L−1 at neutral pH in the downstream catchment. This suggests that the

100

concentration of silicate used may be achievable some ARD systems but in others silicate

101

amendments may be required. The minimum concentration of silicate required to achieve the

102

same behaviours as reported here has not been examined as yet.

103

Experiments were conducted at three different solution pH values (3.0 ± 0.1, 5.0 ± 0.2 and 7.4 ±

104

0.4) without and with addition of dissolved silicate (0.8 mM Si). The pH values chosen are

105

relevant to natural and mine waste ARD environments and to remediated systems. Our objective

106

was to minimise the amount of alkaline amendment required for effective remediation to provide

107

an economically feasible strategy that is practical for large-scale implementation. For this reason

108

alkaline systems were not investigated. The solution pH was manually adjusted daily using 1 M

109

HCl or 1 M NaOH solution. For experiments at neutral pH, 1.66 g calcite was used to buffer the

110

solution pH. The phase purity of calcite of 98.7% is estimated from inorganic C content in the

111

bulk assay and also confirmed by powder X-ray diffraction. The calcite was kept separate from

112

the pyrite in a Nylon mesh bag (mesh size 31 µm) so there was no physical contact. No gypsum

113

(CaSO4.2H2O) was observed using scanning electron microscopy (SEM) for these experimental

114

conditions. Speciation calculations (PHREEQC) also indicated that the system was

115

undersaturated with respect to gypsum. One litre of solution with 2 g of pyrite added was used

116

for each experiment; 10 mL was sampled periodically and membrane filtered (0.45 µm) for

117

inductively coupled plasma optical emission spectrometry (ICP-OES) analysis. The aliquots

118

removed were not replaced but totalled no more than 110 mL over the course of the experiments.


6


Page 8 of 27

119

A 1 L wide-mouth high-density polyethylene (HDPE) bottle was used for each dissolution

120

experiment to exclude the possibility of silicate leaching from glass containers11. Each bottle was

121

sealed using a HDPE lid with a 5 mm hole ensuring full atmospheric contact and minimal

122

evaporation. Milli-Q water was added based on daily weighing to compensate for evaporation

123

losses. All experiments were conducted at room temperature without agitation. Periodic solution

124

samples were taken for solution redox (Eh, SHE) and pH measurements, as well as Fe, S and Si

125

concentration analyses by ICP-OES. Experimental errors, less than 10% for one standard

126

deviation, were estimated from uncertainty of solution analysis as relative standard deviation.

127

Ultrafiltration of dissolution supernatant. Total concentrations of Fe, S, and Si were measured

128

on a portion of each sample digested with nitric acid prior to analysis by ICP-OES. The

129

remaining sample (not digested) was filtered through a 0.45 µm nylon membrane. A portion of

130

the filtrate was analysed by ICP-OES and the remainder subjected to ultrafiltration.

131

Ultrafiltration was carried out using a Model 8400 Amicon stirred ultrafiltration cell (Millipore

132

Corporation). Each sample was fractionated using regenerated cellulose membranes (Millipore

133

Corporation) with nominal molecular weight cut-off (1 kDa NMWCO, Filter Code PLAC07610).

134

Fe, S and Si concentrations were measured in the filtrate and retentate to determine the mass

135

fraction of each element passing 1.5 nm (1kD NMWCO).

136

Surface analyses are described in Supporting Information S1 and statistical analyses in

137

Supporting Information S2.

138 139


7

Page 9 of 27


140

Results and Discussion

141

Pyrite dissolution

142

Experiments were conducted at three solution pH values (3.0 ± 0.1, 5.0 ± 0.2 and 7.4 ± 0.4)

143

without and with addition of dissolved silicate (0.8 mM Si) in quiescent conditions to mimic

144

ARD control conditions. The pyrite dissolution rate, as moles of pyrite per unit surface area

145

(Supporting Information Table 1), for each solution condition, was calculated from the measured

146

solution total S concentration across 290 days (FIGURE 1Figure 1).

147

At pH 3.0, there was no significant difference (p = 0.695) in the pyrite dissolution rate

148

(Supporting Information Table 1) with or without the addition of silicate during 290 days (Figure

149

1a). At pH 5.0, the pyrite dissolution rate in the presence of added silicate is significantly smaller

150

than without silicate addition (FIGURE 1Figure 1b). At pH 7.4 in the absence of dissolved

151

silicate (Figure 1c), the pyrite dissolution rate (Supporting Information Table 1) was essentially

152

constant during 290 days leaching, with approximately 1.3 mmol·m−2 dissolved at 290 days. In

153

contrast, the equivalent pyrite dissolution rate in the presence of silicate (Figure 1c) was

154

effectively zero (p = 0.143), with only ≈0.03 mmol·m−2 pyrite dissolved after 290 days,

155

representing a dissolution rate reduction of >97%. It is clear that the presence of dissolved

156

silicate significantly affects pyrite dissolution at pH 7.4. To more clearly demonstrate this, the

157

pyrite dissolved in the presence of silicate minus that dissolved in the absence of silicate, at each

158

pH, is plotted as a function of time (Figure 1d).


8


Page 10 of 27

159 160

FIGURE 1. Pyrite dissolution as a function of time at pH 3.0 (a), pH 5.0 (b) and pH 7.4 (c).

161

Pyrite dissolution for the system with added silicate (initially 0.8 mM Si) minus pyrite

162

dissolution with no added silicate (d). Rate data fitted with linear regression lines ± 95%

163

confidence bands and level of statistical significance for differences in rates (Supporting

164

Information S3) given as either not significant (ns), significant at p < 0.001 (***) or p < 0.0001

165

(****).

166

The pyrite dissolution rate (5.00×10−11 mol m−2 s−1) at circum-neutral pH 7.4 without silicate is

167

significantly greater than at pH 3.0 (H = 7.20; p < 0.05) (Supporting Information, Table 1). Long

168

term acid rock leaching tests have shown that the greatest sulfate release rate is obtained in the

169

presence of limestone in rock wastes12, 26. It has been proposed that pyrite oxidation rates are


9

Page 11 of 27


170

enhanced by the addition of bicarbonate/carbonate ions27 via a cycle of ferrous-pyrite/ferric-

171

carbonate redox couples, resulting in increased rates of transferral of electrons from S in pyrite to

172

dissolved O2. In ref. 27, 0.1–1.0 M carbonate was used, whereas our carbonate concentrations,

173

equilibrated with air, would not exceed 0.001 M.

174

The log rate of pyrite dissolution during the 290 days (y; mol m−2 s−1) without silicate when

175

plotted as a function of pH (x; 3.0, 5.0; 7.4) (Supporting Information Table 1) is significantly

176

linear (log y = (0.0669)x − 10.764; R2 = 0.995). This near-perfect linearity corresponds to the

177

leach trend found for the pH region where pyrite dissolution rate is determined by dissolved O228.

178

This suggests that the effect of the presence of dissolved carbonate at pH 7.4 is negligible and

179

that the increased leach rate is due predominantly to pH effects.

180

Williamson and Rimstidt28 have shown that at pH 3.0 and above, where ferric iron

181

concentrations are very low due to precipitation, dissolved O2 is the dominant oxidant. In our

182

quiescent conditions, the rate measured for all three pH conditions is around one order of

183

magnitude slower than their rates at the same pH in their study where solution stirring was

184

applied, suggesting that the reaction is to some extent bulk diffusion-controlled. A similar

185

conclusion has been made from hydrologic simulation of a mine tailing waste29.

186

Iron hydroxide formation and silicate adsorption

187

Adsorption of silicate on iron oxyhydroxides has been demonstrated by observation of an FTIR

188

absorption frequency in the range 950−1000 cm−1 assigned to Fe-O-Si19, 21, 30-33, with the extent

189

of adsorption determined by the aqueous silicate concentration and pH. Silicate is preferentially

190

adsorbed on ferrihydrite surfaces at pH 8−11. At silicate to iron mole ratios >1, or at pH values

191

below 8 or above 11, silicate adsorption decreases34, 35. Transformation studies using iron


10


Page 12 of 27

192

oxyhydroxide precipitates showed that the proportion of semi-amorphous ferrihydrite in the final

193

products, which also included crystalline goethite and/or hematite, after 24 h increased from 1%

194

to nearly 100% as silicate concentration was increased from 10−5 to 10−3 M19. Hence, the

195

incorporation of silicate into the structure of iron oxyhydroxide species inhibits the

196

crystallisation of ferric-oxyhydroxides and consequently, the presence of amorphous ferric-

197

(oxy)hydroxide may be maintained36.

198

Recent research on iron precipitation in water samples from a legacy radioactive waste site has

199

confirmed this mechanism of inhibition of Fe(II) oxidation derived from elevated silica

200

concentrations in the circum-neutral pH range37. Their results have shown that, as Si:Fe ratios

201

increase, the primary Fe(III) oxidation product transitions from lepidocrocite to a

202

ferrihydrite/silica-ferrihydrite composite. Competitive desorption experiments suggested that

203

Fe(II) was associated with more weakly bound, outer-sphere complexes on silica-ferrihydrite,

204

rather than lepidocrocite, conferring decreased ability for Fe(II) to undergo surface-induced

205

hydrolysis inhibiting the heterogeneous Fe(II) oxidation mechanism.

206

Figure 2 shows the dissolution, precipitation and silicate adsorption behaviour over time for each

207

pH condition. There is a strong anticorrelation between the solution pH and iron concentration

208

without (r = −0.650; p = 0.002) and with silicate addition (r = −0.813; p < 0.0001), with iron

209

concentration decreasing with increasing pH. During the first 10 days of pyrite oxidation at pH

210

3.0, the ratio of S-to-Fe in solution was nearly two both in the presence and absence of silicate.

211

Subsequently, this ratio increased to greater than two, indicative of Fe(oxy)hydroxide

212

precipitation38. The ratio of S-to-Fe in solution at pH 5.0 and 7.4 was greater than two from

213

commencement of leaching.


11

Page 13 of 27


214

At pH 3.0 there is slightly less aqueous iron in the presence of silicate than in the absence of

215

silicate (Figure 2a). In contrast, at pH 5.0 the iron concentrations are greater in the presence as

216

compared to the absence of silicate (Figure 2b), while the iron concentrations at pH 7.4 are less

217

than the detection limit in most instances due to iron-containing precipitation. It has been

218

observed previously that the iron concentration retained in solution increases with silicate to iron

219

ratio36. This is believed to be due to the presence of ultra-fine, iron hydroxide colloidal particles,

220

stabilised by the presence of silicate, resulting in increased apparent iron concentration39.

221

However, others have quantitatively demonstrated formation of the aqueous FeSiO(OH)32+

222

species in dilute acidic aqueous solutions using UV-visible absorption spectroscopy40. Our

223

ultrafiltration measurements at pH 3.0 showed 99% of the iron was in species

The Formation of Silicate-Stabilized Passivating Layers on Pyrite for

Recommend Documents