Reduced transport potential of a palladium-doped zero valent iron nanoparticle in a water saturated loamy sand.

Accepted Manuscript Reduced Transport Potential of a Palladium-doped Zero Valent Iron Nanoparticle in a Water Saturated Loamy Sand Mohan Basnet, Caroline Di Tommaso, Subhasis Ghoshal, Nathalie Tufenkji PII:

S0043-1354(14)00670-8

DOI:

10.1016/j.watres.2014.09.039

Reference:

WR 10902

To appear in:

Water Research

Received Date: 19 June 2014 Revised Date:

19 September 2014

Accepted Date: 26 September 2014

Please cite this article as: Basnet, M., Tommaso, C.D., Ghoshal, S., Tufenkji, N., Reduced Transport Potential of a Palladium-doped Zero Valent Iron Nanoparticle in a Water Saturated Loamy Sand, Water Research (2014), doi: 10.1016/j.watres.2014.09.039. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

RI PT

Pd-NZVI injection for in situ site remediation

groundwater table

loamy sand

SC

quartz sand

DNAPL

AC C

EP

TE D

M AN U

impermeable strata

ACCEPTED MANUSCRIPT

Reduced Transport Potential of a Palladium-doped Zero

M AN U

Water Research

SC

Revised and Re-submitted to:

RI PT

Valent Iron Nanoparticle in a Water Saturated Loamy Sand

September 19, 2014

2

MOHAN BASNET1, CAROLINE DI TOMMASO1, SUBHASIS GHOSHAL , and

1

TE D

NATHALIE TUFENKJI*, 1

Department of Chemical Engineering, McGill University, Montreal, Quebec H3A 0C5, Canada

2

AC C

EP

Department of Civil Engineering, McGill University, Montreal, Quebec H3A 0C3, Canada

Corresponding Author: Nathalie Tufenkji Department of Chemical Engineering McGill University 3610 University St. Montreal, Quebec, H3A 0C5 Phone: (514) 398-2999; Fax: (514) 398-6678; E-mail: [email protected]

ACCEPTED MANUSCRIPT

Abstract

2

Direct in situ injection of palladium-doped nanosized zero valent iron (Pd-NZVI) particles can

3

contribute to remediation of various environmental contaminants. A major challenge

4

encountered is rapid aggregation of Pd-NZVI and hence very limited mobility. To reduce

5

aggregation and concurrently improve particle mobility, the surface of bare Pd-NZVI can be

6

modified with stabilizing surface modifiers. Selected surface-modified Pd-NZVI has shown

7

dramatically improved stability and transport. However, little is known regarding the effects of

8

aquifer grain geochemical heterogeneity on the transport and deposition behavior of surface-

9

modified Pd-NZVI. Herein, the mobility of surface stabilized Pd-NZVI in two granular matrices

10

representative of model ground water environments (quartz sand and loamy sand) was assessed

11

over a wide range of environmentally relevant ionic strengths (IS). Carboxymethyl cellulose

12

(CMC), soybean flour and rhamnolipid biosurfactant were used as Pd-NZVI surface modifiers.

13

Our results show that, both in quartz sand and loamy sand, an increase in solution IS results in

14

reduced Pd-NZVI transport. Moreover, at a given water chemistry, Pd-NZVI transport is notably

15

attenuated in loamy sand implying that geochemical heterogeneity associated with loamy sand is

16

a key factor influencing Pd-NZVI transport potential. Experiments conducted at a higher Pd-

17

NZVI particle concentration, to be more representative of field conditions, show that

18

rhamnolipid and CMC are effective stabilizing agents even when 1 g/L Pd-NZVI is injected into

19

quartz sand.

20

chemistry, coupled with changes in aquifer geochemistry, could dramatically alter the transport

21

potential of Pd-NZVI in the subsurface environment.

22

Keywords: nanoparticle transport, nanoparticle deposition, surface modifiers, zerovalent iron,

23

loamy sand

AC C

EP

TE D

M AN U

SC

RI PT

1

Overall, this study emphasizes the extent to which variation in groundwater

1

ACCEPTED MANUSCRIPT

24

1. Introduction Palladium-doped nanosized zero valent iron (Pd-NZVI) particles are highly efficient for

26

the transformation of various environmental contaminants into innocuous products via redox

27

reactions (Zhang, 2003). The direct injection of Pd-NZVI into contaminated aquifers has been

28

suggested as a promising technique for achieving rapid remediation (Tratnyek and Johnson,

29

2006). Success of this in situ remediation technique depends on effective subsurface delivery of

30

the reactive Pd-NZVI particles to the contaminant of interest. However, rapid particle

31

aggregation and resultant limited mobility have been reported to be a major challenge in the

32

implementation of this technology (Phenrat et al., 2006). Therefore, to reduce aggregation and

33

concurrently improve transport, bare Pd-NZVI is coated with stabilizing surface modifiers (Saleh

34

et al., 2007). Furthermore, once injected for the targeted subsurface application, the particle may

35

experience various physical transformations such as aggregation, heteroaggregation with other

36

colloids, deposition on collector (grain) surfaces, as well as changes to the surface chemistry

37

resulting from reactions in the aqueous environment (Kim et al., 2012; Lowry et al., 2012;

38

Phenrat et al., 2010; Phenrat et al., 2009). Particle-particle interactions leading to aggregation can

39

have several important consequences. First, an increase in particle size due to aggregation may

40

influence particle deposition by altering the magnitude of colloidal forces between the particles

41

and interacting grain surfaces (Petosa et al., 2010). Second, particle aggregates may be more

42

prone to particle retention by physical straining or wedging (Johnson et al., 2007). Physical

43

straining has been shown to be an influencing factor limiting NZVI transport, especially for bare

44

(Pd-)NZVI (Basnet et al., 2013; Saleh et al., 2007). (Raychoudhury et al., 2014) report that

45

straining can be an important retention mechanism even for polymer-stabilized NZVI. The

46

presence of stabilizing polymers on the surface of (Pd-)NZVI can reduce both the extent of

AC C

EP

TE D

M AN U

SC

RI PT

25

2

ACCEPTED MANUSCRIPT

aggregation and deposition, the effectiveness of which depends on the type of modifier selected

48

(Fatisson et al., 2010; Saleh et al., 2008). While different surface modifiers demonstrate varied

49

stabilizing efficiency, the effectiveness of the modifier could depend on the geochemistry of the

50

site to be remediated. We have previously shown good colloidal stability and enhanced transport

51

in quartz sand of Pd-NZVI coated with CMC, rhamnolipid biosurfactant and soy protein (Basnet

52

et al., 2013). We showed that when particle aggregation is unfavorable (i.e., in the presence of

53

repulsive particle-particle interactions), all of the selected surface modifiers performed well to

54

improve Pd-NZVI mobility. However, at high ionic strength when aggregation is favorable (i.e.,

55

in the absence of repulsive interactions), rhamnolipid was more efficient at enhancing Pd-NZVI

56

transport in quartz sand (Basnet et al., 2013).

M AN U

SC

RI PT

47

A number of previous studies examined the transport and deposition behavior of

58

nanoparticles in quartz sand (Kanel et al., 2007; Kocur et al., 2013; Saleh et al., 2008; Schrick et

59

al., 2004; Tiraferri and Sethi, 2009); however, only few studies have investigated nanoparticle

60

transport in granular matrices other than quartz sand (Cornelis et al., 2013; Jaisi and Elimelech,

61

2009; Laumann et al., 2014; Petosa et al., 2013; Quevedo and Tufenkji, 2012; Zhao et al., 2012).

62

Aquifer materials can vary considerably with respect to mineral surface chemistry and grain size

63

distribution and nanoparticle transport is expected to change significantly in different granular

64

environments. For example, comparable transport behavior was observed for two types of

65

quantum dots in quartz sand (Quevedo and Tufenkji, 2012). However, these nanoparticles

66

exhibited different transport behavior in loamy sand, and this was attributed to the variable

67

affinity of the nanoparticle surface coatings to the clay surfaces in the loamy sand. Such an

68

influence of clay was also reported for ZnO nanoparticles whereby increased particle retention

69

was observed in sandy loam soil (Zhao et al., 2012). Another study comparing the transport

AC C

EP

TE D

57

3

ACCEPTED MANUSCRIPT

70

potential of cerium dioxide nanoparticles in quartz sand to that in loamy sand demonstrated

71

enhanced retention in the loamy sand, which was attributed to the presence of clays such as

72

allophane and metallic impurities in the loamy sand (Petosa et al., 2013). Although the surface modification of (Pd-)NZVI has been extensively studied in recent

74

years (Basnet et al., 2013; Phenrat et al., 2008; Tiraferri et al., 2008), their transport potential in

75

the subsurface environment remains unclear, especially through complex matrixes that contain

76

physical and geochemical heterogeneities. Kim et al. (2012) studied coated NZVI transport in a

77

heterogeneous porous matrix created by adding either kaolinite clay or silica fines to quartz sand

78

and observed greater deposition in the presence of clay minerals than with silica fines (Kim et

79

al., 2012). They hypothesized that NZVI-kaolinite heteroaggregation was the possible

80

mechanism for the observed increase in deposition. To design efficient NZVI remediation plans,

81

there is a clear need for studies on the transport potential of (Pd-)NZVI in heterogeneous

82

granular media which more closely mimic the natural environment.

TE D

M AN U

SC

RI PT

73

The objective of this study is to compare the transport potential of bare and surface-

84

modified Pd-NZVI in a geochemically complex loamy sand to that in a more homogeneous

85

quartz sand. Three types of surface modifiers were used in this study to coat the Pd-NZVI: (i)

86

CMC, (ii) rhamnolipid biosurfactant and (iii) soybean flour. These modifiers were selected based

87

on the effectiveness demonstrated in our previous work (Basnet et al., 2013) as well as their low

88

cost, biodegradability, biocompatibility and commercial availability. The transport potential of

89

Pd-NZVI was then assessed by conducting column experiments in two granular porous matrices:

90

(i) quartz sand (widely used in previous studies), and (ii) loamy sand collected from a farm near

91

Quebec City, QC. Moreover, to assess the impact of clay on the transport potential of Pd-NZVI,

92

additional column experiments were conducted with clay-supplemented quartz sand, where the

AC C

EP

83

4

ACCEPTED MANUSCRIPT

mass fraction of added clay was matched to that of the loamy sand. To our knowledge, this is one

94

of the first studies which systematically examines the combined influence of grain geochemical

95

heterogeneity and water chemistries on Pd-NZVI transport behavior. Thus, this study builds on

96

our previous work where the transport behavior of Pd-NZVI was examined in quartz sand at only

97

two solution IS.

RI PT

93

98

2. Materials and Methods

SC

99

2.1 Nanoparticle Suspension Preparation

101

Surface modification and nanoparticle suspension preparation methods were adopted from our

102

previous work (Basnet et al., 2013). Details are provided in the supplementary data. Three types

103

of anionic surface modifiers were selected as Pd-NZVI stabilizing agents: (i) the sodium salt of

104

CMC (Sigma-Aldrich), (ii) rhamnolipid JBR215 (Jeneil Biosurfactant Co., Saukville, Wisconsin,

105

USA), and (iii) soybean flour (SF) (Sigma-Aldrich). These surface modifiers were selected based

106

on their effectiveness demonstrated in our earlier studies (Basnet et al., 2013; Fatisson et al.,

107

2010). Details on the surface modifiers are presented in Table S1. The final working suspension

108

contained 0.15 g/L Pd-NZVI and 0.1 g TOC/L surface modifier. The solution IS was adjusted by

109

adding NaCl and the pH adjusted to an environmentally relevant value of 7.7 using NaOH.

TE D

EP

AC C

110

M AN U

100

111

2.2 Nanoparticle Characterization

112

Dynamic light scattering (DLS) and nanoparticle tracking analysis (NTA) were used to

113

characterize the hydrodynamic diameter of the particles (details are provided in the Supporting

114

Data). All characterization and transport experiments were conducted over a wide range of IS (3

115

- 100 mM NaCl).

5

ACCEPTED MANUSCRIPT

116

2.3 Column Transport Experiments

118

Bench-scale Pd-NZVI transport experiments were conducted using a glass column (GE Life

119

Sciences) with 1.6 cm internal diameter and 8.1 cm saturated packed bed length. Three sets of

120

column experiments were conducted: (i) in quartz sand to assess transport potential in a

121

relatively ideal (homogeneous) medium, (ii) in loamy sand to assess the transport potential in a

122

complex heterogeneous medium, (iii) in clay supplemented quartz sand to better understand the

123

impact of clay. Details on the sand characteristics are included in Table S2. Prior to injecting

124

nanoparticles, the breakthrough behavior of an inert tracer (10 mM KNO3) was monitored and

125

used to evaluate the bed porosity (Figure S1). All column experiments were done at least in

126

duplicate.

M AN U

SC

RI PT

117

Quartz sand column experiments were conducted as described previously (Pelley and

128

Tufenkji, 2008). Loamy sand column experiments were conducted as described previously with

129

slight modification (Jaisi and Elimelech, 2009; Petosa et al., 2013). Details on procedures for the

130

quartz sand and loamy sand column experiments are included in the Supporting Data. Particle

131

transport was also examined using columns of clay-supplemented quartz sand. Kaolinite (Sigma)

132

was selected as the model clay in this study. To ensure homogeneous clay distribution, a wet

133

mixing procedure was employed (Kim et al., 2012) using 2.05 g of clay slurry that was first

134

prepared in 10 mL DI water. The quartz sand (24.95 g) was then added into the clay slurry and

135

mixed with a bench-top vortex at full speed for 30 sec. The resultant mixture was then dried at

136

100ºC for 8 hours. The concentration of clay in the clay-supplemented quartz sand was 7.6% and

137

corresponds to the quantity of clay in the loamy sand used herein (Quevedo and Tufenkji, 2012).

AC C

EP

TE D

127

6

ACCEPTED MANUSCRIPT

138

The subsequent experimental procedures were identical to those used in the loamy sand column

139

experiments.

140

2.4 Nanoparticle Transport Potential.

142

Colloid filtration theory (CFT) was used to assess nanoparticle transport potential in the different

143

granular matrices. For each experimental condition, the Pd-NZVI particle-collector attachment

144

efficiency (αpc) was calculated using the following equation (Yao et al., 1971):

(1)

M AN U

 C  dc 2  ln  3 (1 − θ ) L η 0  C 0 

SC

α pc = −

145

RI PT

141

where dc is the mean grain diameter, θ is the bed porosity, L is the packed-bed length, and η0 is

147

the single-collector contact efficiency calculated using the Tufenkji-Elimelech equation

148

(Tufenkji and Elimelech, 2004). An advantage of calculating attachment efficiency is that it may

149

serve as a single parameter for the quantitative comparison of nanoparticle transport potential

150

across several treatments used in one study or across several studies (conducted in different labs)

151

that may have employed different granular materials, hydrodynamic conditions and nanoparticle

152

sizes. However, the shortcoming of this model is that when the breakthrough curves are

153

asymmetric and do not follow the classical CFT-type behavior, the calculated attachment

154

efficiency may only be useful for a semi-quantitative comparison.

EP

AC C

155

TE D

146

156

3. Results and Discussion

157

3.1 Electrophoretic Mobility of Pd-NZVI

158

The electrokinetic properties of bare and stabilized Pd-NZVI were evaluated by measuring Pd-

159

NZVI electrophoretic mobility (EPM) in different water chemistries. The surface coatings render

160

the particles more negatively charged over a broad range of pH (at 10 mM IS) (Figure S2). 7

ACCEPTED MANUSCRIPT

Overall, at neutral pH, all particles are negatively charged and decreases in pH up to the

162

isoelectric point (IEP) generally result in a reduced absolute EPM. The IEP varies as follows

163

(from high to low): bare > SF-coated > JBR215-coated > CMC-coated Pd-NZVI. We previously

164

confirmed adsorption of each surface modifier on the Pd-NZVI surface (Basnet et al., 2013).

RI PT

161

The IS-dependent Pd-NZVI electrokinetic behavior at pH 7.7±0.2 is shown in Table 1.

166

All the particles used in column experiments are negatively charged. At 3 mM IS, the order of

167

influent EPM (µm.cm/V.s) from less negative to more negative is: -1.9 (bare) < -2.2 (SF) < -3.8

168

(JBR215) < -4.5 (CMC). At 100 mM, the absolute EPM of influent nanoparticles decreases due

169

to more effective charge screening; the EPM (µm·cm/V·s) of SF-coated Pd-NZVI decreases to -

170

1.2 which is close to bare (-1.4) whereas JBR215-coated (-2.1) and CMC-coated (-2.5) Pd-NZVI

171

remained more negative. Finally, an inspection of Table 1 for the EPM values of particles in the

172

influent or effluent suspension does not reveal noticeable differences. These data are in

173

agreement with previously reported measurements (Basnet et al., 2013). It is worth noting that

174

the EPM of bare Pd-NZVI did not vary significantly from the lowest to the highest IS.

175

3.2 Sizing Analysis of Pd-NZVI

176

Two light scattering techniques (DLS and NTA) were employed to characterize nanoparticle size

177

(Table 1). To assess whether there is any size-dependent preferential particle retention in the

178

packed matrix, DLS size (z-average hydrodynamic diameter) measurements were done for both

179

influent and effluent suspensions. Irrespective of IS, the DLS size (dDLS) of (influent) bare Pd-

180

NZVI is > 1 µm although the nominal size of the Pd-NZVI particles was less than 100 nm, as

181

determined by TEM imaging (Figure S3). This large aggregate size can be attributed to rapid

182

aggregation of the bare Pd-NZVI due to strong particle-particle magnetic interactions (Phenrat et

183

al., 2006). In contrast, the effluent bare Pd-NZVI is < 1 µm throughout the range of IS

AC C

EP

TE D

M AN U

SC

165

8

ACCEPTED MANUSCRIPT

investigated (3-100 mM) which may be explained by the preferential deposition of larger

185

particles and/or the contribution of physical straining and/or wedging (Johnson et al., 2007) to

186

the retention of larger particles (deffluent/dinfluent 1 µm at 1 g/L for the different surface

217

modified Pd-NZVI. Moreover, a comparison of DLS obtained PSD at 0.15 g/L versus 1 g/L

218

clearly shows a marked shift in PSD towards larger particle size (Figure S5). It is also worth

219

noting the monomodal PSD for bare-, JBR215- and SF-coated Pd-NZVI. In contrast, CMC-

220

coated Pd-NZVI exhibits a bimodal PSD at 0.15 g/L Pd-NZVI which evolves to a trimodal PSD

221

at 1 g/L (Figure S5). Based on the PSDs, at 0.15 g/L, the efficiency of the stabilizers in reducing

222

aggregation ranks as follows: JBR215≈SF > CMC. At 1 g/L, marked aggregation occurs;

223

however, based on the PSDs we can infer the stabilizing effects in the following order:

224

JBR215>SF>CMC.

M AN U

TE D

EP

AC C

225

SC

213

226

3.3 Pd-NZVI Transport in Quartz Sand and Loamy Sand

227

The measured breakthrough curves (BTCs) generated from column transport experiments are

228

presented in Figure S6. Two sets of column experiments were conducted for each of the

229

experimental conditions investigated: (i) in quartz sand and (ii) in loamy sand. Bare Pd-NZVI

10

ACCEPTED MANUSCRIPT

exhibits substantial retention in both quartz and loamy sands, with elution levels (average C/C0)

231

< 15% at all IS investigated. In contrast, the surface-modified Pd-NZVI exhibit dramatically

232

higher mobility. For example, at 3 mM IS, the elution in quartz sand are 85% for JBR215-, 85%

233

for SF- and 93% for CMC-coated Pd-NZVI. However, the extent of particle elution is lower in

234

loamy sand with 60% elution for JBR215-, 63% for SF- and 72% for CMC-coated Pd-NZVI. As

235

well, at all other IS (10, 30, 100 mM), more Pd-NZVI retention occurs in loamy sand compared

236

to quartz sand (Figure S6). This result is consistent with previous studies (Kim et al., 2012;

237

Petosa et al., 2013; Quevedo and Tufenkji, 2012) where the enhanced nanoparticle retention was

238

attributed to geochemical heterogeneity of the loamy sand (e.g., presence of clay). Overall, as

239

expected, the retention of coated Pd-NZVI increases (elution decreases) with IS due to more

240

effective compression of the electrical double layers at the particle and sand surfaces (Figure S6).

M AN U

SC

RI PT

230

241

3.4 Assessing Pd-NZVI Transport Potential

243

To semi-quantitatively compare the transport behavior of the bare and coated Pd-NZVI, the

244

particle-collector attachment efficiency (αpc) was calculated (Equation 1) for all the experimental

245

conditions examined. Two approaches were used to calculate C/C0: (i) by integration of the

246

particle BTCs and (ii) the clean-bed approach which considers the average value of C/C0

247

between PVs 1.8-2. Both methods were employed because, in some cases, a stable effluent

248

particle concentration was not observed due to the dynamic nature of the BTCs. When the BTC

249

is highly asymmetric (i.e., does not follow classical CFT-type behavior), the calculated αpc

250

values should be considered valid for the purpose of semi-quantitative comparison. Moreover,

251

two approaches (DLS and NTA) were used to calculate the hydrodynamic diameter of the Pd-

252

NZVI particles. The combination of these two sizing techniques with two approaches to calculate

AC C

EP

TE D

242

11

ACCEPTED MANUSCRIPT

C/C0 results in four values of α for each experimental condition. All the calculated αpc values are

254

presented in Table S4 and for clarity only one set (using DLS sizing + BTC-integration) of

255

values is presented in graphical format in Figure 1. It is worth mentioning that, in this study, the

256

αpc values are nearly identical when calculated using the two different approaches to evaluate

257

C/C0 (for a given particle size). However, there is some variation in the magnitude of αpc

258

depending on which particle sizing technique (DLS vs NTA) is used, especially for the bare and

259

CMC-coated particles, but generally the trends between the different surface modifiers,

260

electrolyte solutions, and porous media are similar.

SC

RI PT

253

In quartz sand at 100 mM, the value of αpc of bare Pd-NZVI approaches the mass-transfer

262

limited value (≈1) and the particle deposition behaviour is termed favorable. The same

263

experimental condition in loamy sand results in αpc > 1 when NTA-based sizes are considered

264

which may be attributed to the technical limitations associated with NTA measurements and the

265

contribution of physical straining. Indeed, smaller effluent sizes (compared to influent) for bare

266

Pd-NZVI indicate preferential retention of larger aggregates (Table 1, Figure S4).

TE D

M AN U

261

A key observation that can be made by inspection of Figure 1 is that surface modification

268

of Pd-NZVI can significantly reduce αpc, by at least one order of magnitude. At a given IS, for

269

coated Pd-NZVI, the value of αpc is generally larger in loamy sand compared to quartz sand. The

270

geochemical heterogeneity and potential for physical straining of the loamy sand could

271

contribute to this observed higher retention as described in earlier studies where larger αpc values

272

for different quantum dots and cerium dioxide nanoparticles were reported in the same loamy

273

sand (Petosa et al., 2013; Quevedo and Tufenkji, 2012) than in quartz sand.

AC C

EP

267

274

As described above, for the bare and CMC-coated Pd-NZVI, we note a difference in αpc

275

depending on the particle sizing technique used (Table S4). At a given solution IS, a biased

12

ACCEPTED MANUSCRIPT

larger DLS size will yield an underestimated αpc value (and consequently over predict transport

277

potential), whereas a biased smaller NTA size could overestimate the same. Hence, a suitable

278

approach might be to use the range of αpc (min – max), for the different methods of estimation, at

279

any given experimental condition of interest when predicting the transport potential of these

280

polydisperse nanoparticles in the environment. An illustration of the predicted Pd-NZVI

281

transport distance (based on αpc) is presented in Figure S7 that clearly shows significant

282

attenuation in travel distance in loamy sand.

283

SC

RI PT

276

3.5 Effect of Particle Concentration on Pd-NZVI transport

285

The comparative breakthrough behavior of bare and coated Pd-NZVI at two different influent

286

mass concentrations (0.15 and 1 g/L) is presented in Figure 2. For bare Pd-NZVI, elution is

287

negligible at both 1 g/L and 0.15 g/L and thus, bare Pd-NZVI is not suitable for real field

288

application where a greater mobility to reach the contaminant zone is required. Similar

289

conclusions have been reported from field and laboratory studies for bare and polymer-coated

290

NZVI (He et al., 2010; Johnson et al., 2013; Kocur et al., 2014; Su et al., 2013). For CMC- and

291

JBR215-coated Pd-NZVI, the elution is comparable at both particle concentrations, while the

292

elution is slightly attenuated for SF-coated Pd-NZVI at higher particle concentration. Moreover,

293

the deposition behavior is dynamic for SF-coated Pd-NZVI whereby the effluent concentration

294

after ~1.5 PVs decreases with time. Concentration induced Pd-NZVI aggregation may be

295

responsible for this dynamic deposition behaviour (discussed in the next section). These results

296

thereby suggest that the selected surface modifiers perform well even at higher Pd-NZVI

297

concentration (i.e., at g/L) that are in the range of NZVI doses that are considered essential for

298

site remediation (Jiemvarangkul et al., 2011; Phenrat et al., 2010). It is important to note that an

AC C

EP

TE D

M AN U

284

13

ACCEPTED MANUSCRIPT

increase in particle concentration induces substantial NZVI aggregation as observed in our

300

previous study (Raychoudhury et al., 2012). The extent of concentration-dependent aggregation

301

is depicted by the change in the intensity-weighted PSD (Figure S5) which shows considerable

302

shift of the PSD towards larger sizes for both bare and coated Pd-NZVI.

RI PT

299

The calculated αpc at 0.15 and 1 g/L Pd-NZVI mass concentration are compared in Figure

304

2c. Because Pd-NZVI aggregate sizes could not be estimated by NTA for the 1 g/L suspensions,

305

the αpc values referred here are those estimated on the basis of DLS particle size measurements.

306

The difference in αpc is negligible for SF- and JBR215-coated Pd-NZVI. In contrast, the αpc for

307

the CMC-stabilized Pd-NZVI is approximately an order of magnitude lower at 1 g/L than at 0.15

308

g/L; this marked reduction in αpc can be attributed to the variation in the sizes of injected

309

particles. The η0 value does not change much within a large range of particle sizes (147-1023

310

nm) used in this study (Figure S8). For example, for JBR215-coated Pd-NZVI, the η0 value does

311

not change when the particle concentration is increased from 0.15 g/L (dp=161 nm, η0=0.03) to 1

312

g/L (dp=1023 nm, η0=0.03) (Figure S8). Hence, similar αpc values are obtained for these

313

treatments. However, for CMC-coated Pd-NZVI, the considerably larger aggregate size (> 5 µm)

314

at 1 g/L results in a considerably higher η0 value which significantly influences αpc. It is worth

315

noting that the calculated attachment efficiencies should be interpreted with caution due to

316

challenges in adequately measuring particle size.

M AN U

TE D

EP

AC C

317

SC

303

Although the steric stabilization of the selected surface modifiers against nanoparticle

318

deposition is well established, the use of extended DLVO theory (which considers the

319

contribution of steric interactions) is not straightforward. Moreover, measurement of the surface

320

modifier layer thickness on the Pd-NZVI from TEM imaging is prone to artifacts due to sample

321

preparation. Thus, in this study, we deemed it inappropriate to present extended DLVO

14

ACCEPTED MANUSCRIPT

calculations which would be based on a number of assumptions. Based on simple classical

323

DLVO calculations, an increase in particle (aggregate) size results in an increase in the height of

324

the repulsive energy barrier upon approach of a particle to a collector surface. An increase in

325

particle concentration from 0.15 to 1 g/L (and associated increase in aggregate size) thus results

326

in increases in the height of the repulsive energy barrier in the interaction energy profile (by

327

factors of 1.1, 5.3 and 7.2, respectively for bare, CMC- and JBR215-coated Pd-NZVI) (Figure

328

S9a, b, c). In contrast, SF-coated Pd-NZVI shows the opposite behavior, whereby the energy

329

barrier becomes smaller with increasing particle concentration (and size) (Figure S9d). This may

330

explain the observed higher deposition of SF-coated Pd-NZVI at 1 g/L (versus 0.15 g/L).

331

Furthermore, at 1 g/L, the higher elution observed for CMC-coated Pd-NZVI is consistent with a

332

previous study (Saleh et al., 2007) that reports higher elution of triblock copolymer modified

333

NZVI at higher particle concentration (> 1 g/L) in comparison to the elution obtained at lower

334

particle concentration (180 mg/L). It is important to note that, for bare Pd-NZVI at 1 g/L,

335

straining is a major retention mechanism as supported by the significant difference in DLS size

336

in the influent (2731 nm) versus effluent suspension (866 nm) (deffluent/dinfluent ≤ 1) (Figure S10)

337

and is in agreement with the straining argument made at 0.15 g/L (Figure S4a). We previously

338

reported physical straining of bare Pd-NZVI during transport through 651 µm quartz sand

339

(Basnet et al., 2013). A recent study (Raychoudhury et al., 2014) also demonstrates grain size

340

dependent straining behavior with greater retention of CMC-coated NZVI in finer sand having a

341

mean size similar to that used in this study, i.e., 256 µm.

343

SC

M AN U

TE D

EP

AC C

342

RI PT

322

3.6 Effect of Clay on Pd-NZVI Transport and Deposition

15

ACCEPTED MANUSCRIPT

To further investigate the role of clay present in the loamy sand on the observed deposition

345

behavior, a separate series of transport experiments (at 10 mM IS) was conducted whereby the

346

quartz sand was supplemented with a model clay (kaolinite) at a concentration equivalent to that

347

measured in the loamy sand (Quevedo and Tufenkji, 2012). Although kaolinite is not the exact

348

same clay material as that present in the loamy sand (allophane), it was selected herein as a

349

reasonable surrogate. The results of these transport experiments are compared with those

350

obtained in clean quartz sand and loamy sand (Figure 3). It can be observed that Pd-NZVI elutes

351

much earlier from the kaolinite supplemented quartz sand, likely due to an increase in pore water

352

velocity as the addition of kaolinite may have altered the pore network connectivity that resulted

353

in reduced bed porosity as seen by much lower effective porosity (0.12) and much higher

354

dispersion coefficient (0.75 cm2/min) (Table S2). Secondly, Pd-NZVI elution from the kaolinite

355

supplemented sand is markedly lower compared to the identical conditions in quartz sand and

356

loamy sand (Figure S11), and may be attributed to the heterogeneities created with the addition

357

of clay and in agreement with previous observations reported by (Kim et al., 2012). Finally,

358

dynamic deposition behavior is observed in the heterogeneous porous matrices compared to the

359

quartz sand. A summary of transport studies demonstrating dynamic deposition behavior is

360

presented in Table S5.

SC

M AN U

TE D

EP

JBR215-coated Pd-NZVI exhibits “ripening” type deposition behavior whereby

AC C

361

RI PT

344

362

deposited particles act as additional collectors for the deposition of approaching particles

363

resulting in a temporal decrease in C/C0 (Figure 3d, g). Overall, the “ripening” behavior seems to

364

be quite significant in loamy sand and kaolinite amended quartz sand. This result suggests that

365

the addition of kaolinite in quartz sand may contribute to capturing the dynamic deposition

366

behavior observed in loamy sand. We have shown previously, that for the same treatment

16

ACCEPTED MANUSCRIPT

(JBR215-coated Pd-NZVI), an increase in solution IS results in marked variation from a pseudo-

368

ripening type behavior at lower IS (10 mM Na+) to considerable ripening at higher IS (300 mM

369

Na+) (Basnet et al., 2013) (Table S5). Likewise, for guar gum-coated NZVI, it was shown that

370

the increase in approach velocity may also influence the dynamic behavior, whereby higher fluid

371

velocity led to a clear ripening type breakthrough curve (Tiraferri and Sethi, 2009) (Table S5).

RI PT

367

SF-coated Pd-NZVI exhibits “blocking” type deposition behavior whereby deposition of

373

approaching particles is blocked by previously deposited particles, resulting in a temporal

374

increase in C/C0 (Figure 3b, e, h). The dynamic behavior is again amplified in the kaolinite

375

amended column in the same manner as was observed for JBR215-coated Pd-NZVI, also

376

mirroring the behavior in loamy sand. Such a “blocking” type behavior was previously observed

377

in several studies (Table S5) including one that reports aggregation as a contributing factor for

378

the observed dynamic behaviour for polyacrylic acid-coated NZVI (Raychoudhury et al., 2010).

379

However, for SF-coated NZVI, the influence of aggregation is less important, as suggested by a

380

smaller dDLS at 10 mM (Table 1). It is very likely that the complex physical and geochemical

381

heterogeneity of the collector surface and variable affinities of the SF molecules for the collector

382

surface might have contributed to the observed dynamic deposition process.

EP

TE D

M AN U

SC

372

The deposition rate of CMC-coated Pd-NZVI appears constant in each granular matrix

384

tested herein (Figure 3c, f, i). This behavior is consistent with previous observations for CMC-

385

stabilized NZVI through sand and sandy soil granular matrices (Basnet et al., 2013; He et al.,

386

2009). With CMC coating, the steady-state breakthrough behavior is observed even at higher

387

(g/L) NZVI concentration in a sand packed column (Kocur et al., 2013), pseudo-steady state to

388

“blocking type” behavior in a kaolinite-supplemented column (Jung et al., 2014), and in column

389

packed with different sand sizes (Raychoudhury et al., 2014).

AC C

383

17

ACCEPTED MANUSCRIPT

390

4. Conclusions

392

The direct injection of Pd-NZVI into contaminated aquifers has been suggested as a promising

393

technique for achieving rapid in situ remediation (Tratnyek and Johnson, 2006; Zhang, 2003).

394

However, upon injection, the mobility of highly reactive Pd-NZVI may be influenced by several

395

environmental factors pertaining to (i) nanoparticle surface chemistry and its aggregation state,

396

(ii) aquifer grain physical and geochemical heterogeneities and (iii) ground water chemistry. This

397

study has addressed such factors, the combination of which results in either improved or limited

398

Pd-NZVI transport. Well-controlled column experiments were carried out to investigate the

399

transport potential of bare and coated (rhamnolipid, soy flour and CMC) Pd-NZVI in granular

400

porous matrices of varied complexity. Our results show that the selected surface modifiers are

401

efficient in stabilizing and concurrently improving Pd-NZVI mobility in quartz sand and loamy

402

sand. The direct application of surface modifiers to Pd-NZVI reduced the value of αpc by at least

403

1 order of magnitude (Basnet et al., 2013). At a given salt concentration, Pd-NZVI exhibits more

404

affinity to the loamy sand compared to the quartz sand, emphasizing the critical role that grain

405

surface geochemical heterogeneity may play in controlling Pd-NZVI mobility. The role of clay

406

was further examined by supplementing the quartz sand with a model clay. Interestingly,

407

surface-modified Pd-NZVI exhibits a distinctive behavior for each type of surface coating

408

examined; with JBR215 coating, the particle deposition rate tends to decrease with time

409

(ripening type behavior); with SF coating, the particle deposition rate tends to increase with time

410

(blocking type behavior); and with CMC coating, the particle deposition rate is relatively

411

constant with time. These trends are particularly clear in the heterogeneous matrices.

AC C

EP

TE D

M AN U

SC

RI PT

391

18

ACCEPTED MANUSCRIPT

The observed variation in the affinity of coated Pd-NZVI for heterogeneous collector

413

surfaces (loamy sand) in contrast to the relatively homogeneous quartz sand (on which several

414

previous studies are based) contributes to the uncertainty associated to predicting and/or

415

engineering Pd-NZVI mobility for a real field application. Therefore, site specific transport

416

potential assessment considering both hydrogeology and geochemistry of the site is necessary.

417

This study provides a starting point for understanding to what extent the variation in aquifer

418

heterogeneity coupled with a fluctuation in ground water chemistry can influence Pd-NZVI

419

transport behavior.

SC

RI PT

412

M AN U

420

Acknowledgements

422

This research was supported by NSERC Strategic Grant no. 365253 and NSERC Discovery,

423

Golder Associates Ltd., FQRNT, a MEDA to M.B., and the CRC program. C.D.T. was supported

424

by a McGill SURE award.

TE D

421

425

Appendix. Supplementary Data

427

EP

426

References

429 430 431 432

Basnet, M., Ghoshal, S., Tufenkji, N., 2013. Rhamnolipid Biosurfactant and Soy Protein Act as Effective Stabilizers in the Aggregation and Transport of Palladium-Doped Zerovalent Iron Nanoparticles in Saturated Porous Media. Environmental Science & Technology 47 (23), 1335513364.

433 434 435

Cornelis, G., Pang, L., Doolette, C., Kirby, J.K., McLaughlin, M.J., 2013. Transport of silver nanoparticles in saturated columns of natural soils. Science of the Total Environment 463–464 (0), 120-130.

436 437

Domingos, R.F., Tufenkji, N., Wilkinson, K.J., 2009. Aggregation of Titanium Dioxide Nanoparticles: Role of a Fulvic Acid. Environmental Science & Technology 43 (5), 1282-1286.

AC C

428

19

ACCEPTED MANUSCRIPT

Fatisson, J., Ghoshal, S., Tufenkji, N., 2010. Deposition of Carboxymethylcellulose-Coated Zero-Valent Iron Nanoparticles onto Silica: Roles of Solution Chemistry and Organic Molecules. Langmuir 26 (15), 12832-12840.

441 442 443

He, F., Zhang, M., Qian, T., Zhao, D., 2009. Transport of carboxymethyl cellulose stabilized iron nanoparticles in porous media: Column experiments and modeling. Journal of Colloid and Interface Science 334 (1), 96-102.

444 445 446

He, F., Zhao, D., Paul, C., 2010. Field assessment of carboxymethyl cellulose stabilized iron nanoparticles for in situ destruction of chlorinated solvents in source zones. Water Research 44 (7), 2360-2370.

447 448

Jaisi, D.P., Elimelech, M., 2009. Single-Walled Carbon Nanotubes Exhibit Limited Transport in Soil Columns. Environmental Science & Technology 43 (24), 9161-9166.

449 450 451

Jiemvarangkul, P., Zhang, W.-x., Lien, H.-L., 2011. Enhanced transport of polyelectrolyte stabilized nanoscale zero-valent iron (nZVI) in porous media. Chemical Engineering Journal 170 (2–3), 482-491.

452 453 454 455

Johnson, R.L., Nurmi, J.T., O’Brien Johnson, G.S., Fan, D., O’Brien Johnson, R.L., Shi, Z., Salter-Blanc, A.J., Tratnyek, P.G., Lowry, G.V., 2013. Field-Scale Transport and Transformation of Carboxymethylcellulose-Stabilized Nano Zero-Valent Iron. Environmental Science & Technology 47 (3), 1573-1580.

456 457 458

Johnson, W.P., Li, X., Yal, G., 2007. Colloid Retention in Porous Media: Mechanistic Confirmation of Wedging and Retention in Zones of Flow Stagnation. Environmental Science & Technology 41 (4), 1279-1287.

459 460 461

Jung, B., O'Carroll, D., Sleep, B., 2014. The influence of humic acid and clay content on the transport of polymer-coated iron nanoparticles through sand. Science of the Total Environment 496, 155-164.

462 463 464

Kanel, S., Nepal, D., Manning, B., Choi, H., 2007. Transport of surface-modified iron nanoparticle in porous media and application to arsenic(III) remediation. Journal of Nanoparticle Research 9 (5), 725-735.

465 466 467

Kim, H.J., Phenrat, T., Tilton, R.D., Lowry, G.V., 2012. Effect of kaolinite, silica fines and pH on transport of polymer-modified zero valent iron nano-particles in heterogeneous porous media. Journal of Colloid and Interface Science 370 (1), 1-10.

AC C

EP

TE D

M AN U

SC

RI PT

438 439 440

20

ACCEPTED MANUSCRIPT

Kocur, C.M., Chowdhury, A.I., Sakulchaicharoen, N., Boparai, H.K., Weber, K.P., Sharma, P., Krol, M.M., Austrins, L., Peace, C., Sleep, B.E., O’Carroll, D.M., 2014. Characterization of nZVI Mobility in a Field Scale Test. Environmental Science & Technology 48 (5), 2862-2869.

471 472

Kocur, C.M., O'Carroll, D.M., Sleep, B.E., 2013. Impact of nZVI stability on mobility in porous media. Journal of Contaminant Hydrology 145 (0), 17-25.

473 474

Laumann, S., Micić, V., Hofmann, T., 2014. Mobility enhancement of nanoscale zero-valent iron in carbonate porous media through co-injection of polyelectrolytes. Water Research 50, 70-79.

475 476

Lowry, G.V., Gregory, K.B., Apte, S.C., Lead, J.R., 2012. Transformations of Nanomaterials in the Environment. Environmental Science & Technology 46 (13), 6893-6899.

477 478 479 480

Nobbmann, U., Connah, M., Fish, B., Varley, P., Gee, C., Mulot, S., Chen, J., Zhou, L., Lu, Y., Sheng, F., Yi, J., Harding, S.E., 2007. Dynamic light scattering as a relative tool for assessing the molecular integrity and stability of monoclonal antibodies. Biotechnology and Genetic Engineering Reviews 24, 117-128.

481 482 483

Pelley, A.J., Tufenkji, N., 2008. Effect of particle size and natural organic matter on the migration of nano- and microscale latex particles in saturated porous media. Journal of Colloid and Interface Science 321 (1), 74-83.

484 485 486

Petosa, A.R., Brennan, S.J., Rajput, F., Tufenkji, N., 2012. Transport of two metal oxide nanoparticles in saturated granular porous media: Role of water chemistry and particle coating. Water Research 46 (4), 1273-1285.

487 488 489

Petosa, A.R., Jaisi, D.P., Quevedo, I.R., Elimelech, M., Tufenkji, N., 2010. Aggregation and Deposition of Engineered Nanomaterials in Aquatic Environments: Role of Physicochemical Interactions. Environmental Science & Technology 44 (17), 6532-6549.

490 491 492

Petosa, A.R., Öhl, C., Rajput, F., Tufenkji, N., 2013. Mobility of nanosized cerium dioxide and polymeric capsules in quartz and loamy sands saturated with model and natural groundwaters. Water Research 47 (15), 5889-5900.

493 494 495 496

Phenrat, T., Kim, H.-J., Fagerlund, F., Illangasekare, T., Lowry, G.V., 2010. Empirical correlations to estimate agglomerate size and deposition during injection of a polyelectrolytemodified Fe0 nanoparticle at high particle concentration in saturated sand. Journal of Contaminant Hydrology 118 (3–4), 152-164.

AC C

EP

TE D

M AN U

SC

RI PT

468 469 470

21

ACCEPTED MANUSCRIPT

Phenrat, T., Kim, H.-J., Fagerlund, F., Illangasekare, T., Tilton, R.D., Lowry, G.V., 2009. Particle Size Distribution, Concentration, and Magnetic Attraction Affect Transport of PolymerModified Fe0 Nanoparticles in Sand Columns. Environmental Science & Technology 43 (13), 5079-5085.

501 502 503 504

Phenrat, T., Saleh, N., Sirk, K., Kim, H.-J., Tilton, R., Lowry, G., 2008. Stabilization of aqueous nanoscale zerovalent iron dispersions by anionic polyelectrolytes: adsorbed anionic polyelectrolyte layer properties and their effect on aggregation and sedimentation. Journal of Nanoparticle Research 10 (5), 795-814.

505 506 507

Phenrat, T., Saleh, N., Sirk, K., Tilton, R.D., Lowry, G.V., 2006. Aggregation and Sedimentation of Aqueous Nanoscale Zerovalent Iron Dispersions. Environmental Science & Technology 41 (1), 284-290.

508 509 510

Quevedo, I.R., Tufenkji, N., 2012. Mobility of Functionalized Quantum Dots and a Model Polystyrene Nanoparticle in Saturated Quartz Sand and Loamy Sand. Environmental Science & Technology 46 (8), 4449-4457.

511 512 513

Raychoudhury, T., Naja, G., Ghoshal, S., 2010. Assessment of transport of two polyelectrolytestabilized zero-valent iron nanoparticles in porous media. Journal of Contaminant Hydrology 118 (3-4), 143-151.

514 515 516

Raychoudhury, T., Tufenkji, N., Ghoshal, S., 2012. Aggregation and deposition kinetics of carboxymethyl cellulose-modified zero-valent iron nanoparticles in porous media. Water Research 46 (6), 1735-1744.

517 518 519

Raychoudhury, T., Tufenkji, N., Ghoshal, S., 2014. Straining of polyelectrolyte-stabilized nanoscale zero valent iron particles during transport through granular porous media. Water Research 50 (0), 80-89.

520 521 522

Saleh, N., Kim, H.-J., Phenrat, T., Matyjaszewski, K., Tilton, R.D., Lowry, G.V., 2008. Ionic Strength and Composition Affect the Mobility of Surface-Modified Fe0 Nanoparticles in WaterSaturated Sand Columns. Environmental Science & Technology 42 (9), 3349-3355.

523 524 525

Saleh, N., Sirk, K., Liu, Y., Phenrat, T., Dufour, B., Matyjaszewski, K., Tilton, R.D., Lowry, G.V., 2007. Surface modifications enhance nanoiron transport and NAPL targeting in saturated porous media. Environmental Engineering Science 24 (1), 45-57.

AC C

EP

TE D

M AN U

SC

RI PT

497 498 499 500

22

ACCEPTED MANUSCRIPT

Schrick, B., Hydutsky, B.W., Blough, J.L., Mallouk, T.E., 2004. Delivery Vehicles for Zerovalent Metal Nanoparticles in Soil and Groundwater. Chemistry of Materials 16 (11), 21872193.

529 530 531

Su, C., Puls, R.W., Krug, T.A., Watling, M.T., O'Hara, S.K., Quinn, J.W., Ruiz, N.E., 2013. Travel distance and transformation of injected emulsified zerovalent iron nanoparticles in the subsurface during two and half years. Water Research 47 (12), 4095-4106.

532 533 534

Tiraferri, A., Chen, K.L., Sethi, R., Elimelech, M., 2008. Reduced aggregation and sedimentation of zero-valent iron nanoparticles in the presence of guar gum. Journal of Colloid and Interface Science 324 (1-2), 71-79.

535 536

Tiraferri, A., Sethi, R., 2009. Enhanced transport of zerovalent iron nanoparticles in saturated porous media by guar gum. Journal of Nanoparticle Research 11 (3), 635-645.

537 538

Tratnyek, P.G., Johnson, R.L., 2006. Nanotechnologies for environmental cleanup. Nano Today 1 (2), 44-48.

539 540 541

Tufenkji, N., Elimelech, M., 2004. Correlation Equation for Predicting Single-Collector Efficiency in Physicochemical Filtration in Saturated Porous Media. Environmental Science & Technology 38 (2), 529-536.

542 543

Yao, K.-M., Habibian, M.T., O'Melia, C.R., 1971. Water and waste water filtration. Concepts and applications. Environmental Science & Technology 5 (11), 1105-1112.

544 545

Zhang, W.-x., 2003. Nanoscale Iron Particles for Environmental Remediation: An Overview. Journal of Nanoparticle Research 5 (3), 323-332.

546 547 548

Zhao, L., Peralta-Videa, J.R., Hernandez-Viezcas, J.A., Hong, J., Gardea-Torresdey, J.L., 2012. Transport and retention behavior of ZnO nanoparticles in two natural soils: Effect of surface coating and soil composition. Journal of Nano Research 17, 229-242.

550

SC

M AN U

TE D

EP

AC C

549

RI PT

526 527 528

23

ACCEPTED MANUSCRIPT

Table 1. Physico-chemical characterization of Pd-NZVI used in the transport experiments. influent suspension EPM

d DLS

(mM NaCl)

(µm.cm/V.s)

(nm)

3

10

30

1043 ± 126

-1.9±0.3

-2.4±0.2

1389 ±235

-1.6±0.1

1167 ±100

effluent suspension PDI

column type

EPM

d DLS

(µm.cm/V.s)

(nm)

0.6 quartz sand -2.4 ± 0.1

515 ± 53

0.7

loamy sand -1.6 ± 0.2

596 ± 76

0.6

0.4 quartz sand -2.9 ± 0.1

973 ± 82

0.3


605 ± 51

0.5

0.5 quartz sand -1.0 ± 0.2 loamy sand -1.0 ± 0.1

100

-1.4±0.1

1332 ±223

0.5 quartz sand -1.5 ± 0.1 loamy sand -1.0 ± 0.3

JBR215-coated

3

10

-3.8±0.1

179 ± 5

-3.7±0.1

161 ± 54

-3.6±0.1

181 ± 3

30

100

CMC-coated

3

10

30

147 ± 6

-2.0±0.1

159 ± 3

-1.2±0.1

-4.5±0.0

-4.3±0.1

-3.9±0.1

253 ± 34

-2.5±0.1

AC C

100

-2.5±0.1

647 ± 91

1415 ± 189

1298 ± 94

1499 ± 175

516 ± 56

175 ± 5

0.3 quartz sand -3.7 ± 0.1

0.5 quartz sand -3.6 ± 0.2

0.6 quartz sand -2.1 ± 0.1

0.6

0.7

0.3

141 ± 41

0.3

171 ± 26

0.3

139 ± 3

0.4

171 ± 4

0.4

0.4

252 ± 15

0.5

0.3 quartz sand -1.8 ± 0.1

167 ± 38

0.3


188 ± 3

0.3

0.4 quartz sand -2.3 ± 0.1

155 ± 3

0.3


130 ± 6

0.3

0.3 quartz sand -2.1 ± 0.1

146 ± 3

0.3


187 ± 8

0.3

0.4 quartz sand -1.1 ± 0.1

180 ± 11

0.3


286 ± 36

0.4

0.7 quartz sand -4.4 ± 0.1

579 ± 63

0.7


707 ± 36

0.7

1492 ± 141

0.5


1401 ± 82

0.5

0.6 quartz sand -3.9 ± 0.2

1278 ± 63

0.6


799 ± 65

0.7


495 ± 46

0.6

1115 ± 148

0.5

0.4

10

-2.1±0.1

2731 ± 303

0.5 quartz sand -1.0 ± 0.2

866 ± 306

JBR215-coated*

10

-4.0±0.1

1023 ± 48

0.3 quartz sand -4.0 ± 0.1

853 ± 35

0.4

SF-coated*

10

-1.2 ±0.2

1841 ±104

0.4 quartz sand -1.1 ±0.1

1154 ±60

0.3

CMC-coated*

10

-4.8±0.1

5916 ±444

0.3 quartz sand -5.5 ± 0.1

8295 ±163

0.6

Bare*

*influent Pd-NZVI mass concentration = 1 g/L (for all other conditions, the particle concentration was 0.15 g/L ) n.d.: not determined

the presented sizing and EPM values represent mean ± 95% confidence interval

1

d 90

(nm)

(nm)

144 ± 21

239 ± 33

373 ± 49

200 ± 22

306 ± 23

411 ± 63

234 ± 24

358 ± 27

559 ± 95

251 ± 21

399 ± 71

537 ± 100

81 ± 18

210 ± 17

385 ± 53

90 ± 18

191 ± 23

348 ± 64

97 ± 18

211 ± 39

397 ± 29

103 ± 13

185 ± 9

323 ± 8

34 ± 1

107 ± 8

273 ± 32

44 ± 15

152 ± 63

345 ± 78

48 ± 15

120 ± 21

264 ± 9

80 ± 27

194 ± 64

398 ± 48

103 ± 26

229 ± 41

363 ± 26

118 ± 42

278 ± 32

518 ± 51

172 ± 22

300 ± 26

497 ± 129

179 ± 33

280 ± 9

397 ± 60

0.3

178 ± 3

0.6 quartz sand -2.5 ± 0.0

d 50 (or d NTA)

0.4


0.6 quartz sand -4.1 ± 0.1

d 10 (nm)

0.5

M AN U

10

196 ± 10

-2.2±0.1

920 ± 196

172 ± 3

TE D

3

263 ± 19

EP

SF-coated

-2.1±0.1

618 ± 52


loamy sand -3.3 ± 0.2 100

829 ± 204

0.4 quartz sand -3.5 ± 0.1

loamy sand -3.5 ± 0.1 30

NTA size analysis (influent suspension) PDI

RI PT

Bare

ionic strength

SC

particle type

n.d.

ACCEPTED MANUSCRIPT

loamy sand

(a) 10

0

10

-1

10

-2

1

10

100

Ionic Strength (mM)

(b) 10

0

10

-1

10

-2

1

RI PT

Bare JBR215 SF CMC

Attachment Efficiency (αpc)


quartz sand

10

100

SC

Ionic Strength (mM)

AC C

EP

TE D

M AN U

Fig. 1. Calculated particle-collector attachment efficiency (αpc) as a function of solution IS for bare, JBR215-, SF- and CMC-coated Pd-NZVI suspensions at an influent concentration of 150 mg/L in (a) quartz sand and (b) loamy sand. The calculation is based on DLS measured particle hydrodynamic diameter with C/C0 calculated from the numerical integration of the BTC. The dashed lines are included to guide the eye.

ACCEPTED MANUSCRIPT

CPd-NZVI = 0.15 g/L

0.6

C/C0

C/C0

0.8

0.4

0.8 0.6 0.4 0.2

0.2 0.0 0

(b)

1.0

RI PT

Bare JBR215 SF CMC

(a)

1

2

3

4

0.0 0

5

10

-2

10

-3

2

3

4

5

0.15 g/L 1.00 g/L

M AN U

-1

(c)

Bare JBR215 SF

TE D


0

10

1

Pore Volumes

Pore Volumes 10

SC

1.0

CPd-NZVI = 1.00 g/L

CMC

AC C

EP

Fig. 2. Effect of influent Pd-NZVI mass concentration on the elution of bare, CMC-, JBR215and SF-coated Pd-NZVI in a quartz sand-packed column at 10 mM IS: (a) at 0.15 g/L and (b) at 1.0 g/L. (c) Calculated particle-collector attachment efficiency (αpc) based on DLS-measured particle (aggregate) sizes with C/C0 calculated from the numerical integration of the BTC.

ACCEPTED MANUSCRIPT

JBR215-coated

SF-coated 1.0

(a)

0.8

0.6

0.4 0.2

0

10

20

0.0

30

0.2

0

10

Time (min)

20

Time (min)

C/C0

0

10

20

30

(h)

0.8

0.6 0.4 0.2

C/C0

0.6

C/C0

C/C0

0.2

1.0

(g)

0.8

0.4 0.2

0

10

20

0.4

0.0

0

10

Time (min)

1.0

clay-amended quartz sand

0.4

0.0

30

30

0

10

20

Time (min)

TE D

Time (min)

0.0

30

0.6

M AN U

10

20

(f)

0.8

0.2 0

10

SC

C/C0

C/C0

loamy sand

0.2

0

1.0

0.6

0.4

0.0

Time (min)

(e)

0.8

0.6

0.0

30

1.0

(d)

0.8

0.0

20

Time (min)

1.0

0.4

RI PT

C/C0

C/C0

0.2

(c)

0.8

0.6

0.4

0.0

1.0

(b)

0.8

0.6

C/C0

quartz sand

1.0

CMC-coated

30

20

30

Time (min) 1.0

(i)

0.8 0.6 0.4 0.2 0.0

0

10

20

30

Time (min)

AC C

EP

Fig. 3. Measured breakthrough curves for JBR215-coated Pd-NZVI (a, d, g), SF-coated PdNZVI (b, e, h) and CMC-coated Pd-NZVI (c, f, i) in quartz sand (a, b, c), loamy sand (d, e, f) and clay (kaolinite)-supplemented quartz sand (g, h, i) at 10 mM IS. Other experimental conditions are: packed bed length 8.1 cm, approach (Darcy) velocity 7.46 × 10-5 m/s and Pd-NZVI concentration (0.15 g/L).

ACCEPTED MANUSCRIPT

Basnet et al., 2014 Research Highlights

• Comparison of Pd-NZVI transport in quartz, loamy, and clay-amended sands

RI PT

• Grain geochemical heterogeneity influences Pd-NZVI transport and dynamic deposition behavior

• Selected surface modifiers dramatically enhance Pd-NZVI transport at lower IS • At a given IS, reduced Pd-NZVI transport in loamy sand versus quartz sand

AC C

EP

TE D

M AN U

SC

• Selected surface modifiers are effective at high concentration (1 g/L) of Pd-NZVI

ACCEPTED MANUSCRIPT

RI PT

Supplementary Data

M AN U

SC

Reduced Transport Potential of a Palladium-doped Zero Valent Iron Nanoparticle in a Water Saturated Loamy Sand

MOHAN BASNET1, CAROLINE DI TOMMASO1, SUBHASIS GHOSHAL2, and NATHALIE TUFENKJI*, 1

Department of Chemical Engineering, McGill University, Montreal, Quebec H3A 0C5, Canada Department of Civil Engineering, McGill University, Montreal, Quebec H3A 0C3, Canada

AC C

EP

2

TE D

1

Corresponding Author: Nathalie Tufenkji Department of Chemical Engineering McGill University 3610 University St. Montreal, Quebec, H3A 0C5 Phone: (514) 398‐2999; Fax: (514) 398‐6678; E‐mail: [email protected]

ACCEPTED MANUSCRIPT

SI1. Surface Modifier Preparation

M AN U

SC

RI PT

The stock CMC solution was prepared by dissolving (overnight) measured CMC mass in a filtered (0.22 µm) DI water yielding 1.65 g/L TOC (total organic carbon) stock concentration. The TOC content was determined using a TOC analyzer (Shimadzu Corporation). Rhamnolipid stock solution (1 g/L TOC) was prepared by simply diluting the original stock (received in liquid form) in a filtered (0.22 µm) DI water. The stock SF solution was prepared as follows. First, the SF powder was dissolved overnight in DI water (mass concentration 10 g/L). Next, to facilitate SF dissolution, the solution pH was raised to 12 by adding NaOH and kept the solution in mild stirring for 1 hour. After this alkaline treatment, the solution pH was adjusted to 7.7 by adding HCl. To remove any undissolved larger particles, the solution was then sequentially filtered through Whatman filter papers, 0.45 µm and 0.22 µm acetate filter. The final SF stock yielded 1.88 g/L TOC and the same stock was used in all the transport/characterizations experiments. All the stock surface modifiers solutions were deoxygenated with N2 purging, sealed and stored in dark at 4°C until further use. The stocks were only opened in N2 purged glove box for further manipulation i.e. to coat the bare Pd-NZVI. Table S1. Details of the surface modifiers used in this study.

Carboxymethyl cellulose

composition

CMC

polyelectrolyte with sodium as a counter ion to the carboxyl group

JBR215

molecular weight

aqueous solution containing mixture of mono- and dirhamnolipid

AC C

Soy flour

SF

defatted soybean flour with ≈52% protein

structure

700 kDa

mono-rhamnolipid: 504 Da di-rhamnolipid: 650 Da

EP

Jeneil rhamnolipid biosurfactant

abbreviation

TE D

surface modifiers

mono-rhamnolipid

di-rhamnolipid n.d.

The chemical structure for JBR215 is taken from (Mulligan, 2005) and for SF, as a representative of soy protein, from (Sun, 2005).

1

ACCEPTED MANUSCRIPT

SI2. Pd-NZVI Suspension Preparation

SC

RI PT

An aqueous slurry of bare NZVI particles was obtained from Golder Associates Ltd. (Montreal, Canada). Upon receipt, the stock slurry was vacuum-dried and stored in a N2 purged hypoxic chamber. For each experiment, a stock suspension (10 g/L) was prepared by suspending a known mass of NZVI powder in deionized (DI) water (ultrapure N2 purged) and sonicated (Misonix Sonicator, S-4000) for 20 min. Palladium doping was done according to a method described previously (Zhang, 2003) by mixing freshly dispersed bare NZVI particles with an ethanol solution of palladium acetate. The dose of palladium was 0.47 wt% of NZVI mass. For each experiment, the stock (10 g/L) Pd-NZVI suspension was diluted in deoxygenated DI water and sonicated (10 min), suspended in the presence of surface modifiers, and equilibrated for ~20 hrs at room temperature using an end-over-end rotator. SI3. Pd-NZVI Characterization

M AN U

Two complementary techniques were employed to measure particle hydrodynamic diameter: (i) dynamic light scattering (DLS) and (ii) nanoparticle tracking analysis (NTA). DLS measurements (Z-average reported) were performed using ZetaSizer Nano ZS (Malvern, UK) operating at 633 nm (red laser) with the refractive index set at 2.87 (Phenrat et al., 2009); and NTA measurements were performed using NanoSight LM10 (NanoSight, UK) operating at 405 nm (blue laser). To determine nominal Pd-NZVI size, high-resolution transmission electron microscopy (TEM) images of dried Pd-NZVI suspensions were also captured.

EP

TE D

The electrokinetic properties of bare and coated Pd-NZVI were assessed by measuring electrophoretic mobility (EPM) using laser Doppler velocimetry (ZetaSizer Nano ZS, Malvern, UK). Measured EPMs were converted to zeta potential using the Henry’s equation (Ohshima, 1994). All the sizing and EPM measurements were done at least in triplicate of 2-4 independent samples for each experimental condition at room temperature (≈22ºC). The EPM and DLS sizing measurements were performed using samples of the influent suspensions (collected prior to injecting into the column) and effluent suspensions collected after two pore volumes (PVs).

AC C

SI4. Column Transport Experiments Quartz sand column experiments were conducted as described previously (Pelley and Tufenkji, 2008). Briefly, the steps involved are wet packing the column, equilibration with background electrolyte (~10 PVs), Pd-NZVI injection (~3 PVs at concentration C0) and post-injection with ~3 PVs particle free background electrolyte. The effluent Pd-NZVI concentration was monitored in real-time using a flow cell mounted in a UV-visible spectrophotometer (Agilent 8453) at a wavelength of 508 nm (Basnet et al., 2013; Saleh et al., 2008). During injection, the reservoir containing Pd-NZVI suspension was continuously mixed using a shaker (INFORS HT), and a constant C0 confirmed with UV-vis measurements over the time span of column experiment. The pH of the column influent and effluent suspensions was monitored and found to range between 7.5 and 7.9.

2

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Loamy sand column experiments were conducted as described previously with slight modification (Jaisi and Elimelech, 2009; Petosa et al., 2013). The column was purged with CO2 in an upward flow direction for 20 min to improve water saturation as suggested previously (Kretzschmar and Sticher, 1997). Next, 20 mM CaCl2 was injected in an upward flow direction at a flow rate of 1.3 mL/min for 40 PVs, and was employed to achieve soil colloid stabilization (Jaisi and Elimelech, 2009). The flow was then switched to the electrolyte of interest for equilibration (> 500 PVs) of the packed column. Before the start of column experiment, using the same electrolyte, the column was further equilibrated in downward flow direction for at least 10 PVs. Finally, the flow rate was adjusted to the experimental flow rate of 0.9 mL/min and equilibrated further with at least 6 PVs (particle free) background electrolyte (with or without the supplemented surface modifiers). After this step, the subsequent experimental procedure was identical to that used for the quartz sand column.

3

ACCEPTED MANUSCRIPT

Table S2. Summary of the physicochemical characteristics and the experimental conditions used in column transport experiments. quartz sand

loamy sand

10th percentile (d 10)

200 μm

114 μm

mean grain size (d 50)

256 μm

225 μm

90 percentile (d 90)

396 μm

653 μm

span ((d 90-d 10)/d 50)

0.77

2.40

th

chemical composition

porosity (θ)

0.76

sand (92 %), clay (8 %)

8.1 cm

8.1 cm

1.66 g/cm3

1.47 g/cm3

1.66 g/cm3

0.37

0.35

0.12

6.03 mL

5.70 mL

1.92 mL

7.46 × 10‐5 m/s

7.46 × 10‐5 m/s

EP

pore volume (PV)

364 μm

8.1 cm

TE D

packing density

236 μm

sodium calcium aluminum silicon oxide (Na0.98Ca0.02Al1.02Si2.98O8), quartz (SiO2), aluminum quartz (SiO2), potassium silicate (Al2Si2O5(OH)4) aluminum silicate (KAlSi3O8)

quartz (SiO2)

packed bed packed bed length

sand ( 88 %), silt ( 4%), clay (8 %)

sand (99.9 %)

184 μm

M AN U

grain size classification

SC

granular media


RI PT

parameters

experimental conditions

AC C

approach velocity

dispersion coefficient (D)

7.46 × 10‐5 m/s 2

2

2

0.16 cm /min

0.18 cm /min

0.75 cm /min

Pd-NZVI concentration

0.15 and 1 g/L

0.15 g/L

0.15 g/L

ionic strength

3-100 mM NaCl

3-100 mM NaCl

10 mM NaCl

Note: particle size distribution with size classification in quartz sand and loamy sand was adopted from (Quevedo and Tufenkji, 2012). The effective porosity of the granular materials was determined from the tracer breakthrough curves.

4

ACCEPTED MANUSCRIPT

C/C0

C/C0

1.0

0.5

experiment model

0.0 0.0

0.5

1.0

1.5

loamy sand

0.5

experiment model

0.0

2.0

0.0

0.5

Pore Volumes

0.0 0.0

1.5

2.0

SC


0.5

1.0

Pore Volumes

M AN U

C/C0

1.0

RI PT

quartz sand

1.0

experiment model

0.5

1.0

1.5

2.0

Pore Volumes

EP

Bare JBR215-coated

2 0 -2

4

pH

(b)

CMC-coated SF-coated

2 0 -2 -4

-4

2

4

EPM(m.cm/V.s)

(a)

AC C

EPM (µm.cm/V.s)

4

TE D

Fig.S1. Breakthrough curves for NO3- tracer and the model fit based on advection-dispersion equation.

6

2

8

4

6

8

pH

Fig. S2. Variation in electrophoretic mobility (EPM) as a function of pH for (a) Bare and JBR215-coated Pd-NZVI, and (b) CMC and SF-coated Pd-NZVI at 10 mM IS. The dashed lines are included to guide the eye.

5

ACCEPTED MANUSCRIPT

SI5. Imaging of Pd-NZVI

TE D

M AN U

SC

RI PT

Representative TEM images (Figure S3) of Pd-NZVI were obtained using Philips CM200 TEM equipped with a 2k×2k CCD camera (Advanced Microscopy Techniques Corp., MA, USA) operating at 200 kV. These high resolution images are helpful in estimating nominal sizes and shapes of dried Pd-NZVI suspension. A careful inspection of TEM images reveals the nominal Pd-NZVI sizes to be < 100 nm. The chain-like structure of dried Pd-NZVI (on the surface of TEM grid) observed in this study is consistent with previous studies (Cirtiu et al., 2011; Fatisson et al., 2010; Raychoudhury et al., 2012), and is likely an artifact associated with TEM sample preparation, i.e. drying the sample prior to TEM imaging (Domingos et al., 2009).

AC C

EP

Fig. S3. Representative TEM images for bare, CMC- and JBR215-coated Pd-NZVI at 10 mM IS.

6

ACCEPTED MANUSCRIPT

1.2

Bare

deffluent/dinfluent

0.4

0.0 1

0.8

0.4

quartz sand loamy sand

(a) 10

0.0 1

100

Ionic Strength(mM) SF-coated

1.2

0.4

0.0 1


(c) 10

100

Ionic Strength(mM)

10

CMC-coated

0.8

M AN U

0.8


(b)

100

Ionic Strength(mM)

deffluent/dinfluent

deffluent/dinfluent

1.2

RI PT

0.8

JBR215-coated

SC

deffluent/dinfluent

1.2

0.4

0.0 1

(d)

10


100

Ionic Strength(mM)

AC C

EP

TE D

Fig. S4. Effluent to influent particle size ratio (deffluent/dinfluent) as a function of IS for the bare (a), JBR215-coated (b), SF-coated (c) and CMC-coated (d) Pd-NZVI used in the transport experiments. The sizing data were DLS measured Pd-NZVI sizes during the column experiments (Table 1). The conditions likely to result in straining (deffluent/dinfluent ≤ 1) are shaded in grey.

7

ACCEPTED MANUSCRIPT

Table S3. Calculated zeta potential for the Pd-NZVI particles and the collector surfaces used in transport experiments. The measured electrophoretic mobility (EPM) was converted into zeta potential (ZP) using Henry’s equation (Ohshima, 1994). ionic strength (mM NaCl)

EPM (μm.cm/V.s)

3

-1.9 ±

0.3

10

-2.4 ±

0.2

30

-1.6 ±

0.1

100

-1.4 ±

0.1

-19.3 ± 1.1

3

-3.8 ±

0.1

-53.3 ± 1.2

10

-3.7 ±

0.1

-51.0 ± 0.9

30

-3.6 ±

0.1

-49.3 ± 1.4

100

-2.1 ±

0.1

-28.1 ± 1.4

3

-2.2 ±

0.1

-30.4 ± 1.1

10

-2.5 ±

0.1

-35.1 ± 1.1

30

-2.0 ±

0.1

-27.7 ± 1.6

TE D

particle

100

-1.2 ±

0.1

-15.9 ± 0.8

3

-4.5 ±

0.0

-61.0 ± 0.5

10

-4.3 ±

0.1

-57.5 ± 0.9

30

-3.9 ±

0.1

-52.6 ± 1.5

100

-2.5 ±

0.1

-33.7 ± 0.8

SF-coated Pd-NZVI

collector

EP

CMC-coated Pd-NZVI

AC C

quartz sand*

kaolinite

3

-32.9 ± 2.8 -21.3 ± 1.1

-58.7

10

-56.4

n/a

30

-50.6

100

-37.0

3

-3.5 ±

0.3

-48.0 ± 4.2

10

-4.0 ±

0.4

-53.9 ± 5.9

30

-3.4 ±

0.3

-46.2 ± 4.5

100

-3.0 ±

0.2

-40.3 ± 2.5

8

-26.4 ± 3.4

SC

JBR215-coated Pd-NZVI

M AN U

Bare Pd-NZVI

ZP (mV)

RI PT

particle/collector

ACCEPTED MANUSCRIPT

40

10

1000

Size (nm)

30 20 10

at 0.15 g/L pk 1: 110 nm (1%) pk 2: 1105 nm (99%) at 1 g/L pk 1: 243 nm (1%) pk 2: 1122 nm (4%) pk 3: 4801 nm (95%)

0 10

10000

100

(c) JBR215-coated

20

10

1000

Size (nm)

0.15 g/L 1 g/L

20

10

0 100

10000

TE D

0 100

30

% Intensity

0.15 g/L 1 g/L

10000

(d) SF-coated

M AN U

30

1000

Size (nm)

40

40

% Intensity

0.15 g/L 1 g/L

RI PT

20

0 100

(b) CMC-coated

40

0.15 g/L 1 g/L

% Intensity

% Intensity

30

50

SC

(a) bare

1000

Size (nm)

10000

AC C

EP

Figure S5. DLS measured intensity weighted particle size distribution (PSD) for (a) bare (b) CMC-, (c) JBR215-, and (d) SF-coated Pd-NZVI (0.15 and 1.0 g/L mass concentration) at 10 mM IS.

9

ACCEPTED MANUSCRIPT

quartz sand Tracer 3 mM 30 mM 100 mM

(a)

0.6 0.4 0.2 2

3

4

Pore Volumes

0.0 0

5

Tracer 3 mM 30 mM 100 mM

(c)

1

2

3

4

5

Pore Volumes

(d)

0.8 0.6 0.4

M AN U

0.4 0.2

0.2

1

2

3

4

Pore Volumes


(e)

0.8

TE D

0.6 0.4 0.2 0.0 0

1

2

3

4

Pore Volumes

0.8

1.0

4

5

1

2

3

4

5

1

2

3

4

5

0.6 0.4

1.0

0.2

Pore Volumes

(h)

0.8

C/C0

AC C 0.4

1

3

(f)

0.0 0

0.6

0.0 0

2

Pore Volumes

0.2


(g)

1

0.8

5

EP

1.0

0.0 0

5

C/C0

0.0 0

C/C0

1.0

C/C0

0.6

C/C0

0.4

SC

1

0.8

C/C0

JBR215-coated

1.0

1.0

SF-coated

0.6

0.2

0.0 0

CMC-coated

(b)

0.8

C/C0

C/C0

Bare

0.8

1.0

RI PT

1.0

loamy sand

0.6 0.4 0.2

2

3

4

Pore Volumes

0.0 0

5

Pore Volumes

Fig. S6. Representative measured breakthrough curves (BTCs) for bare Pd-NZVI in (a) quartz sand and (b) loamy sand; JBR215-coated Pd-NZVI in (c) quartz sand and (d) loamy sand; SFcoated Pd-NZVI in (e) quartz sand and (f) loamy sand; CMC-coated Pd-NZVI in (g) quartz sand and (h) loamy sand at selected salt concentrations (3, 30 and 100 mM IS). The breakthrough behavior of a conservative tracer (NO3−) is also presented (solid line). 10

ACCEPTED MANUSCRIPT

Table S4. Calculated particle-collector attachment efficiencies for the experimental conditions used in the transport experiments.

3

quartz sand

0.22 ± 0.01

0.26 ± 0.01

0.19 ± 0.04

0.40 ± 0.08

30

0.34 ± 0.02

0.59 ± 0.04

100

0.52 ± 0.33

1.16 ± 0.74

0.25 ± 0.01

0.30 ± 0.02

0.18 ± 0.03

0.39 ± 0.06

0.39 ± 0.00

0.69 ± 0.00

0.43 ± 0.07

0.95 ± 0.16

0.26 ± 0.02

0.28 ± 0.02

0.31 ± 0.09

0.60 ± 0.17

0.31 ± 0.02

0.50 ± 0.04

0.27 ± 0.04

0.29 ± 0.09

0.56 ± 0.18

30

0.28 ± 0.02

0.45 ± 0.03

100

0.61 ± 0.03

1.24 ± 0.07

0.63 ± 0.00

1.29 ± 0.00

0.02 ± 0.00

0.02 ± 0.00

0.02 ± 0.00

0.03 ± 0.00

10

0.02 ± 0.01

0.02 ± 0.01

0.02 ± 0.00

0.02 ± 0.00

30

0.03 ± 0.01

0.04 ± 0.01

0.04 ± 0.01

0.04 ± 0.01

100

0.06 ± 0.01

0.05 ± 0.01

0.08 ± 0.02

0.06 ± 0.02

0.05 ± 0.01 0.04 ± 0.00

0.05 ± 0.01

0.06 ± 0.01

0.06 ± 0.01

0.04 ± 0.01

0.04 ± 0.01

0.05 ± 0.01

30

0.06 ± 0.01

0.07 ± 0.01

0.07 ± 0.01

0.08 ± 0.01

100

0.15 ± 0.01

0.12 ± 0.00

0.16 ± 0.01

0.13 ± 0.01

0.02 ± 0.01

0.01 ± 0.00

0.03 ± 0.00

0.02 ± 0.00

3

quartz sand

loamy sand

3

quartz sand

10

TE D

30 100 3 10

loamy sand

EP

30 100 3

quartz sand

0.02 ± 0.01

0.02 ± 0.01

0.04 ± 0.01

0.04 ± 0.01

0.02 ± 0.01

0.01 ± 0.00

0.02 ± 0.00

0.01 ± 0.00

0.07 ± 0.01

0.06 ± 0.01

0.08 ± 0.00

0.07 ± 0.00

0.05 ± 0.01

0.03 ± 0.00

0.06 ± 0.01

0.04 ± 0.01

0.04 ± 0.00

0.04 ± 0.00

0.05 ± 0.00

0.06 ± 0.00

0.13 ± 0.00

0.11 ± 0.00

0.16 ± 0.01

0.13 ± 0.01

0.60 ± 0.23

0.50 ± 0.20

0.57 ± 0.06

0.47 ± 0.05

0.01 ± 0.01

0.01 ± 0.01

0.02 ± 0.01

0.02 ± 0.01

10

0.02 ± 0.00

0.05 ± 0.00

0.03 ± 0.00

0.06 ± 0.00

30

0.04 ± 0.01

0.07 ± 0.01

0.04 ± 0.01

0.07 ± 0.01 0.16 ± 0.05

AC C

CMC-coated

αNTA

0.25 ± 0.03

3

loamy sand

αDLS

10

10

SF-coated

αNTA

10

3

JBR215-coated

αDLS

RI PT

column type

SC

Bare

+

(mM Na )

M AN U

particle type

average C /C 0 from 1.8-2 PVs approach**

BTC-integration approach*

ionic strength

100

0.06 ± 0.02

0.15 ± 0.05

0.07 ± 0.02

0.05 ± 0.03 0.04 ± 0.02

0.03 ± 0.02

0.06 ± 0.02

0.05 ± 0.02

0.08 ± 0.04

0.04 ± 0.02

0.08 ± 0.04

30

0.05 ± 0.00

0.08 ± 0.01

0.05 ± 0.00

0.09 ± 0.00

100

0.13 ± 0.02

0.27 ± 0.04

0.16 ± 0.01

0.32 ± 0.01

3

10

loamy sand

Two methods were used to determine the normalized effluent nanoparticle concentration (C/C0): (i) numerical integration of entire BTC (*) and (ii) average C/C0 between 1.8–2.0 PV (**) (Tufenkji and Elimelech, 2005). The calculation is based on both DLS and NTA measured particle sizes in influent suspensions of 150 mg/L. 11

ACCEPTED MANUSCRIPT

1.0

(a) quartz sand Bare JBR215 SF CMC

0.6 0.4 0.2 0.0 1

2

3

4

5

SC

0

RI PT

C/C0

0.8

M AN U

Travel Distance (m) 1.0

(b) loamy sand

Bare JBR215 SF CMC

0.6 0.4

TE D

C/C0

0.8

0.2 0.0

EP

0

1

2

3

4

5

Travel Distance (m)

AC C

Fig. S7. Predicted transport distance after the injection of Pd-NZVI in a model groundwater environment: (a) quartz sand, and (b) loamy sand. The predicted concentration profiles do not account for dynamic deposition behaviour and are calculated based on αpc values calculated at 3 mM IS.

12

ACCEPTED MANUSCRIPT

 D  I

0.3

 G  0

0.2

RI PT

Single-Collector Contact Efficiency ()

0.4

0.1 0.0 100

1000

SC

Particle Size (nm)

M AN U

Fig. S8. Variation in the single-collector contact efficiency (ƞ0) as a function of Pd-NZVI sizes calculated using the Tufenkji-Elimelech equation (Tufenkji and Elimelech, 2004). Size dependent variation in each transport mechanism is also highlighted where ƞI is the transport by interception; ƞG is the transport by gravity and ƞD is the transport by diffusion (ƞ0 = ƞI + ƞG + ƞD)

1000

(a) 500

-500 -1000

5

10

15

VT(KBT)

0

-5000 -10000

20

5

10

20

1000

(d)

SF (0.15 g/L) SF (1.00 g/L)

VT(KBT)

500

-2000

15

Separation Distance (nm)

JBR215 (0.15 g/L) JBR215 (1.00 g/L)

EP

0

(c)

AC C

VT(KBT)

2000

CMC (0.15 g/L) CMC (1.00 g/L)

0

Separation Distance (nm) 4000

(b)

5000

TE D

VT(KBT)

10000

bare (0.15 g/L) bare (1.00 g/L)

0 -500

-4000

5

10

15

-1000

20


5

10

15

20


Fig. S9. Calculated particle-collector total interaction energy profiles as a function of separation distance for bare, CMC-, JBR- and SF-coated Pd-NZVI at 10 mM IS. The calculation is based on the DLS sizing, and the EPM data (converted into zeta potential using Henry’s equation) taken from Table S3.

13

ACCEPTED MANUSCRIPT

0.15 g/L 1.00 g/L

1.0 0.5 0.0

Bare JBR215 SF

RI PT

deffluent/dinfluent

1.5

CMC

M AN U

SC

Fig. S10. Ratio of particle sizes (deffluent/dinfluent) for the transport experiments conducted at 10 mM IS using bare, JBR215-, SF- and CMC-coated Pd-NZVI suspension at 0.15 and 1 g/L. deffluent and dinfluent were DLS measured Pd-NZVI sizes during the column experiments (Table 1). Straining is more likely when deffluent/dinfluent ≤ 1.

1.0

quartz sand loamy sand quartz sand + kaolinite

0.6

TE D

C/C0

0.8

0.4 0.2

JBR215

EP

0.0

SF

CMC

Treatment

AC C

Fig. S11. The elution of JBR215-, SF- and CMC-coated Pd-NZVI through quartz sand, loamy sand and clay-supplemented quartz sand column at 10 mM IS. The elution through each column was calculated by the numeral integration of BTCs presented in Figure 3.

14

ACCEPTED MANUSCRIPT

Table S5. Summary of NZVI and Pd-NZVI transport studies demonstrating dynamic deposition behavior. column parameters

150 mg/L

quartz sand (651 µm)

CMC coated Pd-NZVI

R95-coated Pd-NZVI


PV3A stabilized NZVI

PAA-coated NZVI

2 g/L

0.1 - 3.0 g/L

8.1 cm

8 cm

quartz sand (600- 30 cm 850 µm)

quartz sand (375 µm)

4.5 cm

silica sand (250 µm)

7 cm

1.0 × 10-4 m/s

n/a

0.30

3.33 × 10-6 to 1.67 × 10-4 m/s

0.1-1 mM Na+ (pH 7.4)

0.36

n/a

-4

0.36

1.29 × 10 m/s

silica sand 61.3 cm (300 µm)

0.33

1.06 × 10-4 m/s

AC C

30 mg/L

1.27 × 10-4 m/s

sandy soil (< 2 mm)

EP

200 mg/L

sand (250-420 µm)

15

this study

resembles steady state

Jiemvarangkul et al., 2011

strongly resembles to blocking type and was attributed to temporal Raychoudhury et al., decrease in single collector 2010 efficiency (ƞ 0) caused by rapid PAANZVI aggregation

resembles steady state behavior at lower velocity (2.76 × 10-5 m/s); in contrast, the ripening type BTC at higher velocity (1.38 × 10-4 m/s) with uncertain reason

Tiraferri and Sethi, 2009

resembles steady state in both type of granular matrices

Feng et al., 2009

resembles steady to blocking type at 1-100 mM lower IS (1, 10 mM); appears slightly Na+ ripening type at higher IS (40, 100 (pH 7.7) mM)

Saleh et al., 2008

0.37

1.38 × 10-4 m/s

guar gum-coated NZVI

variation in collector grain geochemical heterogeneity influenced the BTC dynamic behavior (ripening type for JBR215and blocking type for SF-coated PdNZVI), whereas CMC-coated PdNZVI exhibits steady state behaviour

references

increase in the IS (from 10 to 300 10 and 300 mM) influenced the BTC dynamic + behavior (blocking type for R95- and Basnet et al., 2013 mM Na (pH 7.7) ripening type for JBR215-coated PdNZVI)

0.40

TE D

154 mg/L

SDBS modified NZVI

7.46 × 10-5 m/s

2.76 × 10-5 10 mM Na+ m/s (pH 7.5)

polymer-coated NZVI

CMC-coated NZVI

0.38

7.46 × 10-5 10 mM Na+ m/s (pH 7.7)

dynamic deposition behavior (i.e. the shape of BTCs)

RI PT

SF coated Pd-NZVI

quartz sand (0.37), loamy sand (0.47)

solution chemistry

SC

150 mg/L

quartz sand (w/ and w/o kaolinite) (256 µm), loamy sand (225 µm)


length

approach (Darcy) porosity velocity

M AN U

nanoparticle type

nanoparticle collector concentration grain

1, 10 mM Na+ (pH 7.5)

1 mM Na+ (pH 7.5)

ACCEPTED MANUSCRIPT

References

RI PT

Basnet, M., Ghoshal, S., Tufenkji, N., 2013. Rhamnolipid Biosurfactant and Soy Protein Act as Effective Stabilizers in the Aggregation and Transport of Palladium-Doped Zerovalent Iron Nanoparticles in Saturated Porous Media. Environmental Science & Technology 47 (23), 1335513364.

SC

Cirtiu, C.M., Raychoudhury, T., Ghoshal, S., Moores, A., 2011. Systematic comparison of the size, surface characteristics and colloidal stability of zero valent iron nanoparticles pre- and postgrafted with common polymers. Colloids and Surfaces A: Physicochemical and Engineering Aspects 390 (1-3), 95-104.

M AN U

Domingos, R.F., Baalousha, M.A., Ju-Nam, Y., Reid, M.M., Tufenkji, N., Lead, J.R., Leppard, G.G., Wilkinson, K.J., 2009. Characterizing manufactured nanoparticles in the environment: Multimethod determination of particle sizes. Environmental Science & Technology 43 (19), 7277-7284. Fatisson, J., Ghoshal, S., Tufenkji, N., 2010. Deposition of Carboxymethylcellulose-Coated Zero-Valent Iron Nanoparticles onto Silica: Roles of Solution Chemistry and Organic Molecules. Langmuir 26 (15), 12832-12840.

TE D

Jaisi, D.P., Elimelech, M., 2009. Single-Walled Carbon Nanotubes Exhibit Limited Transport in Soil Columns. Environmental Science & Technology 43 (24), 9161-9166.

EP

Kretzschmar, R., Sticher, H., 1997. Transport of Humic-Coated Iron Oxide Colloids in a Sandy Soil: Influence of Ca2+ and Trace Metals. Environmental Science & Technology 31 (12), 34973504.

AC C

Mulligan, C.N., 2005. Environmental applications for biosurfactants. Environmental Pollution 133 (2), 183-198. Ohshima, H., 1994. A Simple Expression for Henry's Function for the Retardation Effect in Electrophoresis of Spherical Colloidal Particles. Journal of Colloid and Interface Science 168 (1), 269-271. Pelley, A.J., Tufenkji, N., 2008. Effect of particle size and natural organic matter on the migration of nano- and microscale latex particles in saturated porous media. Journal of Colloid and Interface Science 321 (1), 74-83.

16

ACCEPTED MANUSCRIPT

Petosa, A.R., Öhl, C., Rajput, F., Tufenkji, N., 2013. Mobility of nanosized cerium dioxide and polymeric capsules in quartz and loamy sands saturated with model and natural groundwaters. Water Research 47 (15), 5889-5900.

RI PT

Phenrat, T., Kim, H.-J., Fagerlund, F., Illangasekare, T., Tilton, R.D., Lowry, G.V., 2009. Particle Size Distribution, Concentration, and Magnetic Attraction Affect Transport of PolymerModified Fe0 Nanoparticles in Sand Columns. Environmental Science & Technology 43 (13), 5079-5085.

SC

Quevedo, I.R., Tufenkji, N., 2012. Mobility of Functionalized Quantum Dots and a Model Polystyrene Nanoparticle in Saturated Quartz Sand and Loamy Sand. Environmental Science & Technology 46 (8), 4449-4457.

M AN U

Raychoudhury, T., Tufenkji, N., Ghoshal, S., 2012. Aggregation and deposition kinetics of carboxymethyl cellulose-modified zero-valent iron nanoparticles in porous media. Water Research 46 (6), 1735-1744. Saleh, N., Kim, H.-J., Phenrat, T., Matyjaszewski, K., Tilton, R.D., Lowry, G.V., 2008. Ionic Strength and Composition Affect the Mobility of Surface-Modified Fe0 Nanoparticles in WaterSaturated Sand Columns. Environmental Science & Technology 42 (9), 3349-3355.

TE D

Sun, X.S. (2005) Bio-Based Polymers and Composites. Wool, R.P. and Sun, X.S. (eds), pp. 292326, Academic Press, Burlington.

EP

Tufenkji, N., Elimelech, M., 2005. Breakdown of Colloid Filtration Theory: Role of the Secondary Energy Minimum and Surface Charge Heterogeneities. Langmuir 21 (3), 841-852.

AC C

Tufenkji, N., Elimelech, M., 2004. Correlation Equation for Predicting Single-Collector Efficiency in Physicochemical Filtration in Saturated Porous Media. Environmental Science & Technology 38 (2), 529-536. Zhang, W.-x., 2003. Nanoscale Iron Particles for Environmental Remediation: An Overview. Journal of Nanoparticle Research 5 (3), 323-332.

17

Transport and deposition of stabilized engineered silver nanoparticles in water saturated loamy sand and silty loam.

Transport of carbon colloid supported nanoscale zero-valent iron in saturated porous media.

Efficient removal of uranium from aqueous solution by zero-valent iron nanoparticle and its graphene composite.

Transport of nano zero-valent iron supported by mesoporous silica microspheres in porous media.

Deactivation of nanoscale zero-valent iron by humic acid and by retention in water.

A MD simulation and analysis for aggregation behaviors of nanoscale zero-valent iron particles in water via MS.

Nanosilver and Nano Zero-Valent Iron Exposure Affects Nutrient Exchange Across the Sediment-Water Interface.

Continuous phosphorus removal from water by physicochemical method using zero valent iron packed column.

Potential environmental implications of nanoscale zero-valent iron particles for environmental remediation.

Antimicrobial and Genotoxicity Effects of Zero-valent Iron Nanoparticles.

Capture and storage of hydrogen gas by zero-valent iron.

Can zero-valent iron nanoparticles remove waterborne estrogens?

Phosphate removal from aqueous solutions by nanoscale zero-valent iron.

Weak magnetic field accelerates chromate removal by zero-valent iron.

C micro-electrolysis assisted by zero-valent copper and zero-valent aluminium.

Hexavalent chromium reduction in contaminated soil: A comparison between ferrous sulphate and nanoscale zero-valent iron.

Interaction between Cu2+ and different types of surface-modified nanoscale zero-valent iron during their transport in porous media.

Surface coating with Ca(OH)2 for improvement of the transport of nanoscale zero-valent iron (nZVI) in porous media.

Treatment of acid rock drainage using a sulfate-reducing bioreactor with zero-valent iron.

Toxicity of zero-valent iron nanoparticles to a trichloroethylene-degrading groundwater microbial community.

Transformation and composition evolution of nanoscale zero valent iron (nZVI) synthesized by borohydride reduction in static water.

Removal of As(III) and As(V) from aqueous solutions using nanoscale zero valent iron-reduced graphite oxide modified composites.

Enhanced paramagnetic Cu²⁺ ions removal by coupling a weak magnetic field with zero valent iron.

The role of magnetite nanoparticles in the reduction of nitrate in groundwater by zero-valent iron.