Influence of Intrinsic Colloid Formation on ... - ACS Publications

Subscriber access provided by UNIV OF NEBRASKA - LINCOLN

Article

Influence of intrinsic colloid formation on migration of cerium through fractured carbonate rock Emily L. Tran, Ofra Klein-BenDavid, Nadya Teutsch, and Noam Weisbrod Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.5b03383 • Publication Date (Web): 13 Oct 2015 Downloaded from http://pubs.acs.org on October 18, 2015

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 23

Environmental Science & Technology

1 2

Influence of intrinsic colloid formation on migration of cerium through fractured carbonate rock

3

Emily L. Tran1, Ofra Klein-BenDavid2, Nadya Teutsch3, Noam Weisbrod1*

4 5 6 7

1

Zuckerberg Institute for Water Research, Jacob Blaustein Institutes for Desert Research, Ben Gurion University of the Negev, Midreshet Ben Gurion, 08990 Israel 2 Nuclear Research Center of the Negev, P.O. Box 9100, 84190 Beer Sheva, Israel

8

3

9

*Corresponding author. Email: [email protected], phone +972-8-65 96903

Geological Survey of Israel, Jerusalem, 95501 Israel

10 11

Abstract

12

Migration of colloids may facilitate the transport of radionuclides leaked from near

13

surface waste sites and geological repositories. Intrinsic colloids are favorably formed by

14

precipitation with carbonates in bicarbonate-rich environments, and their migration may be

15

enhanced through fractured bedrock. The mobility of Ce(III) as an intrinsic colloid was

16

studied in an artificial rain water solution through a natural discrete chalk fracture. The results

17

indicate that at variable injection concentrations (between 1 and 30 mg/L), nearly all of the

18

recovered Ce takes the form of an intrinsic colloid of >0.45 µm diameter, including in those

19

experiments in which the inlet solution was first filtered via 0.45 µm. In all experiments, these

20

intrinsic colloids reached their maximum relative concentrations prior to that of the Br

21

conservative tracer. Total Ce recovery from experiments using 0.45 µm filtered inlet solutions

22

was only about 0.1%, and colloids of >0.45 µm constituted the majority of recovered Ce.

23

About 1% of Ce was recovered when colloids of >0.45 µm were injected, indicating the

24

enhanced mobility and recovery of Ce in the presence of bicarbonate.

25

1 ACS Paragon Plus Environment


26

Introduction

27

Geological disposal of radioactive waste is the internationally agreed-upon, long term

28

solution for the disposal of high level waste, long lived radionuclides and spent fuel.

29

Corrosion of waste canisters during storage periods of hundreds of thousands of years can

30

lead to leakage of radionuclides, which may make their way into the subsurface and

31

groundwater.1–3 Previous investigations have revealed that radionuclide migration away from

32

the repository is faster than predicted by traditional advection-dispersion based models.4 At

33

the Nevada test site, actinides migrated 1.3 km from their source in only 30 years.5 This has

34

since been attributed to colloid-facilitated transport in which radionuclides sorb to mobile

35

pseudocolloids composed of clay-minerals, iron-oxide colloids, or natural organic matter and

36

migrate unretarded along a flow path as a result of size exclusion and charge exclusion.6–8

37

However, the extent to which intrinsic colloids contributed to this migration remains unclear.9

38

Colloids have been defined in the literature as either inorganic or organic solids of

39

~1nm to a few micrometers in size that remain suspended in water.10,11 Intrinsic colloids are

40

those that are formed by the polymerization of compounds of a specific element.10 Intrinsic

41

colloids may also be formed when readily precipitable complexes become supersaturated as a

42

result of changes of geochemical parameters such as pH, redox potential, or CO2 partial

43

pressure.11 These intrinsic colloids may be responsible for faster migration and increased

44

recovery of leaked actinides in the environment. Möri et al.12 found that actinides migrate

45

through a fractured flow path faster than the iodine conservative tracer, even in the absence of

46

carrier colloids such as bentonite. This faster migration was attributed to the formation of

47

homogeneous or heterogeneous intrinsic radiocolloids.12

48

As the ions OH- and CO32- are present in almost all groundwater types, they are often

49

the most important ligands in the complexation of radionuclides.13 Carbonate speciation in

50

particular becomes enhanced in deep underground repositories, where CO2 partial pressure

51

reaches up to 10-2 atm in contrast to only 3x10-4 atm in the ambient atmosphere.13 Ikeda et


Page 2 of 23

Page 3 of 23


52

al.14 postulate that carbonate complexes are of great importance in the overall migration of

53

radionuclides in the environment, as carbonates are abundant in the groundwater and form

54

strong complexes with actinides. Even in cases of competition between humic acid and

55

carbonates for lanthanide complexation, carbonates successfully compete as ligands under

56

alkaline conditions and at high CO32- concentrations.15,16 The formation of carbonate

57

complexes alone, however, does not necessarily mobilize radionuclides, as dissolved

58

complexes may sorb to negatively-charged surfaces. Zavarin et al.17 observed retention of

59

dissolved Sm(CO3)2- and SmCO3+ due to sorption onto fracture walls when Sm(III) was

60

injected into carbonate rock fractures. However, some low-solubility actinide-carbonate

61

complexes can become large molecular clusters and ultimately precipitate out as intrinsic

62

colloids in response to changes in the geochemical gradient and ultimately be subjected to

63

colloid-facilitated transport.11,13 It has been shown that under conditions of high ionic strength

64

and carbonate concentration, carbonate competes effectively with hydrolysis reactions to

65

cause trivalent actinides to precipitate and form An2(CO3)3·nH20 where An is a trivalent

66

actinide.13 These reactions therefore should be included in solubility and mobility predictions

67

for leaked actinides in the vicinity of nuclear repositories under these conditions.

68

Numerous studies have proved that colloids can be particularly mobile through

69

fractures that serve as preferential pathways for groundwater flow,10,12,18,19 even in cases

70

where fractures account for a small portion of the total porosity of a matrix. 20 This is in part

71

due to the higher shear forces associated with water movement through a fracture, which

72

cause mobilization and increased migration capacity of the colloids.20 Groundwater flow

73

through fractures has been found to be a significant factor in the migration of radionuclides

74

originating from nuclear waste repositories in particular.21–23 As nuclear waste repositories are

75

planned to be established deep within fractured rock formations,21,23,24 investigation of these

76

intrinsic colloids' transport should be focused on movement within fractures and fracture

77

networks. Previous studies that have utilized rock column experiments to investigate colloid-



78

facilitated radionuclide transport used mainly granitic fractures.12,19,24–26 However, very little

79

is known about the migration of radionuclides in areas dominated by carbonate lithology.27

80

In areas where granitic rocks are absent or rare, carbonate rocks may be the dominant

81

bedrock used for geological disposal of nuclear wastes. This is the case in Israel, where a

82

future geological repository for spent fuel is currently being sited. Previous studies have

83

postulated that formation of carbonate solids should be considered in prediction of nuclear

84

contamination mobility through carbonate rocks without producing evidence to support this

85

conjecture.27,28 It is therefore important to investigate how leaked nuclear waste would behave

86

in the local carbonate environment.

87

Due to their similar chemical behaviors and redox states, lanthanides are commonly

88

used as analogues of these actinides in studies of radionuclide migration.29–33 Ce specifically

89

has been used as an analogue of trivalent and tetravalent actinides,34–36 as its +III and +IV

90

oxidation states make it a good analogue for Pu.34 This study therefore aimed to elucidate the

91

impacts of intrinsic colloid formation through precipitation of radionuclides with carbonates

92

on the migration and recovery of radionuclides through a chalk fracture, using Ce as an

93

analogue for redox active actinides such as Pu.

94

Materials and Methods

95

Solution preparation

96

All experiments were run in a background solution of artificial rainwater (ARW)

97

containing concentrations of major ions representative of the chemical composition of

98

rainwater in the Negev Desert as described by Zvikelsky and Weisbrod37 and can be found in

99

the SI (Table S1). ARW has previously been used in other chalk fracture transport

100

studies.23,37–39 Br was chosen as a conservative (non-sorbing) tracer. For all injected tracer

101

solutions and batch experiments, stock solutions of Br and Ce were prepared according to

102

procedures detailed in the SI.


Page 4 of 23

Page 5 of 23

103 104


Colloid Characterization Several methods were used to determine the exact speciation of intrinsic colloids

105

which precipitate out of solution in ARW. Batch experiments, performed according to

106

procedures detailed in the SI, were conducted by adding Ce in stepwise function to an ARW

107

solution in order to determine the rate at which Ce precipitates out of solution. Additionally,

108

XRD analysis was performed on precipitate colloids that had formed in the mixed Ce and

109

ARW solution. Colloids were filtered out of solution using 0.45 µm filters (Wattmann) and

110

analyzed with X-ray powder diffraction (XRPD) (Figure S1). Finally, the chemical speciation

111

modeling program Visual MINTEQ40 (ver 3.1) was used to determine the likely speciation of

112

all elements in solution (Figures S2-S3; details of XRPD included in SI).

113

Particle size distribution and zeta potential

114

Size fractions of precipitate colloids were determined using series filtrations as

115

outlined in the SI. Polycarbonate membrane filters (Sterlitech, Kent, WA) were used in a

116

syringe filtration assembly. Two solutions of 30 mg/L Ce in ARW containing precipitate

117

colloids were prepared. One was left sitting overnight, and one was filtered immediately.

118

Aliquots of each solution were filtered through filters of successively smaller pore sizes, and

119

the final concentration of Ce from each filter was analyzed.

120

A particle size and zeta potential analyzer (90plus, Brookhaven Instruments Co.,

121

Holtsville NY) was used to determine the effective colloid diameter as a function of time by

122

dynamic light scattering. The same instrument was applied to determine the zeta potential of

123

these solutions. All analytical procedures are detailed in the SI.

124

The dissolved fraction of a solution is commonly defined as that which passes

125

through a 0.45 µm filter.41,42 However, we are aware that smaller intrinsic colloids are present

126

in solutions passed through 0.45 µm filters investigated in this study. Suspensions containing

127

small colloids (0.02 µm) have been shown to exhibit lower recoveries in fractures due to



Page 6 of 23

128

higher colloid collision frequency leading to deposition.37 Furthermore, small colloids are

129

prone to increased flow through stagnation zones or diffusion into matrix pore water which

130

further increases retention.37 Finally, large-size particles have been shown to exhibit greater

131

size exclusion effects leading to increased dispersion and thus earlier breakthrough.43 As

132

preliminary batch experiments demonstrated that approximately 80-90% of 1 mg/L Ce added

133

to ARW could be removed by a 0.45µm filter (Table1), 0.45 µm filters were chosen in this

134

study to differentiate between the larger and smaller colloidal fractions of Ce.

135

Fractured core and hydraulic measurements

136

A 38.5 cm-long and 20 cm-diameter chalk core containing a natural fracture spanning

137

the width of the core was used for all subsequent experiments, as depicted in Figure S4. The

138

core was drilled from the Eocene-age bedrock in the northern Negev Desert. The core was

139

oriented horizontally so that the fracture was parallel to the surface to limit gravitational

140

effects on flow and particle transport. No more than two days prior to each transport

141

experiment, a hydraulic test was conducted to determine the equivalent hydraulic aperture of

142

the fracture according to the cubic law:44

143

2𝑏 3 =

144

where 2b is the equivalent hydraulic aperture (L), Q is the fluid discharge (L3/T), µ is the

145

dynamic viscosity of the fluid (M/LT), 𝜌 is the fluid density (M/L3), g is the gravitational

146

constant (m/s2), W is the fracture width under saturated conditions, and

147

gradient. Thus, an equivalent fracture volume (EFV) was established to normalize the

148

injection time required for each tracer experiment according to the following equation:

149

𝑙 × 𝑊 × 2𝑏 = 𝐸𝐹𝑉

150

in which l is the fracture length, W is the fracture width and 2b is the equivalent hydraulic

151

aperture. As this calculation assumes two parallel plates and does not account for the internal

152

fracture structure heterogeneities, this calculation obviously only gives a rough estimation of

𝑄12µ 𝑑𝑙 ∗ 𝜌𝑔𝑊 𝑑𝑥

(1)

𝑑𝑙 𝑑𝑥

is the hydraulic

(2)


Page 7 of 23


153

the fracture volume within an order of magnitude.39 Zheng et al.45 showed that using the cubic

154

law technique consistently underestimates the true fracture aperture in comparison to that

155

calculated by a mass balance technique. Thus, it is assumed that the true fracture volume in

156

this study is actually greater than that given here. The calculation used here, however, is

157

useful for normalizing the injection time of each experiment as a result of changes in

158

hydraulic aperture due to precipitation/dissolution processes. Even though the EFV is not a

159

true fracture volume and the times at which the C/C0 reaches their maximum are reported

160

according to this definition, the deviation between true fracture volume and EFV is similar for

161

all tracers. Details of core origins, preparation and hydraulic aperture measurements are

162

included in the SI.

163

Column Transport Experiments

164

Four sets of experimental parameters were used in this study (Table 1), and each was

165

carried out in duplicate. In the first two, high concentrations of Ce (C0~30 mg/L) were

166

injected into the core. High concentrations of radionuclides have previously been used to

167

demonstrate radionuclide migration and emphasize the difference between colloidal and

168

dissolved species migration.39,46 These high concentrations of Ce were used in this study to

169

ensure high concentrations of intrinsic colloid precipitation and observe their effects,

170

especially in cases where HCO3- was a limiting factor. In the first of these experiments, the Ce

171

was allowed to precipitate with the HCO3- in the ARW, and the tracer solution was injected

172

into the core including the intrinsic colloids precipitated in the inflow prior to injection. In the

173

latter, a high concentration tracer (C0~45 mg/L Ce) was prepared, and the entire solution was

174

filtered prior to injection through 0.45 µm syringe filters to remove all intrinsic colloids that

175

are larger than this diameter. Thus, it is assumed that only the 0.45

186

µm colloidal Ce. Filtered samples were immediately acidified to 0.1 M HNO3 in preparation

187

for analysis. Unfiltered samples were digested using concentrated HNO3 before being diluted

188

to 0.1 M HNO3. All samples were analyzed for Ce and Br concentrations by ICP-MS

189

(Nexion, Perkin Elmer, USA). A schematic diagram of the flow system (Figure S4) and

190

further experimental and analytical details are presented in the SI. Normalized concentrations

191

of Ce and Br (C/C0) from the filtered and unfiltered samples were plotted to obtain

192

breakthrough curves. In total, 17 experiments were run in the core over the course of this

193

research, including preliminary experiments during which experimental parameters were

194

tested, and only eight experiments are reported here (Tables 1-2). After 13 experiments, a pH

195

6.5 phosphate buffer was flushed through the core once to attempt to remove sorbed Ce from

196

the fracture walls.

197

Results and Discussion

198

Precipitation of intrinsic Ce Colloids

199

Batch experiments in which Ce (1-10 mg/L) was added to ARW showed that the

200

HCO3- precipitates out of solution at a rate that correlates linearly with the addition of Ce

201

(Figure 1). This stoichiometric relationship strongly indicates that the Ce present in solution is

202

nearly all precipitated out as carbonate phases. When presented in molar terms, the slope of

203

the linear graph is -3.0 rather than -1.5, as would be expected if only Ce2(CO3)3·6H2O was

204

precipitated. Thus, additional precipitates are likely present as well, which may include


Page 8 of 23

Page 9 of 23


205

NaCe(CO3)2 and Ce(OH)CO336 in addition to some amorphous carbonate phase. The XRD

206

analysis performed on these precipitates also indicates the presence of the carbonate

207

compound Ce2(CO3)3·6H2O plus some amorphous substances (Figure S1).

208

Data from an REE thermodynamic database47 was used for speciation and solubility

209

calculations. The chemical reaction for cerium (III) carbonate precipitation is given as:

210

Ce2 (CO3 )3 ∙ 8H2 O(s) ↔ 2Ce3+ + 3CO3 2+ + 8H2 O

211

Specifics of Ce solubility in a carbonate system are discussed by Ferri et al36 who denote the

212

reactions controlling the equilibria of the Ce(III) and carbonate:

213

Ce2 (CO3 )3 (s) + 6H + ↔ 2Ce2+ + 3CO2 (g) + 3H2 O

logK = 21.8 ± 0.05

(4)

214

Ce3+ + H2 O + CO2 (g) ↔ CeCO3 + + 2H +

logβ2,1 = −11.3 ± 0.05

(5)

215

Ce3+ + 2H2 O + 2CO2 (g) ↔ Ce(CO3 )2 + 4H +

logβ4,2 = −24.1 ± 0.1

(6)

216

The LogK values for these solids were added to a Visual MINTEQ40 database and used to

217

verify the solubility of Ce(III) in ARW as described in the supporting information. Figures S2

218

and S3 show that Ce(III) precipitates out as Ce2(CO3)3·8H2O(s) over the concentration and pH

219

ranges used in our experiments, which is consistent with the batch and XRD analysis results.

220

The saturation index of this precipitate for 1 mg/L Ce is 4.3 at pH 7.5 and 25ºC (Figure S2a),

221

indicating that the precipitate is well over saturation. Even at a lower theoretical near-field

222

repository concentration of 0.01 mg/L, this solid is expected to precipitate in ARW under

223

these conditions (saturation index of 0.32).

224

Particle size distribution and zeta potential:

225

−

logK = −35.1

(3)

The series filtrations of the precipitate colloids performed immediately after solution

226

preparation indicated that 46±1% of the Ce stayed in dissolved or colloidal form with a

227

diameter of 5µm in diameter. After 12 hours, only 19±1% of the

230

Ce remained within the smallest (5 µm (Figure S5). This latter size fractionation represents the distribution

232

of colloid sizes injected into the fracture, as the tracer solution was always prepared the day

233

prior to injection.

234

Time dependent particle size analysis indicated a strong exponential trend between

235

time elapsed from colloidal creation and the average diameter of the colloids in the

236

suspension. The colloidal particle size, measured by the particle size analyzer immediately

237

after preparation of the suspension, was 176±3 nm. Average colloidal diameter increased over

238

time (Figure 2), and after sitting for 24 hours, the average particle size was no longer within

239

the limit of the detection of the instrument (

Influence of Intrinsic Colloid Formation on ... - ACS Publications

Recommend Documents