Colloid Mobilization in a Fractured Soil: Effect of Pore-Water Exchange

Subscriber access provided by Weizmann Institute of Science

Article

Colloid Mobilization in a Fractured Soil: Effect of Pore Water Exchange between Preferential Flow Paths and Soil Matrix Sanjay K. Mohanty, James E Saiers, and Joseph N. Ryan Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.5b04767 • Publication Date (Web): 01 Feb 2016 Downloaded from http://pubs.acs.org on February 4, 2016

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

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

Page 1 of 30

Environmental Science & Technology

1

Colloid Mobilization in a Fractured Soil: Effect of Pore Water Exchange between Preferential Flow Paths and Soil Matrix Sanjay K. Mohanty1∗, James E. Saiers2, and Joseph N. Ryan1

1 2 3

1

Civil, Environmental and Architectural Engineering, University of Colorado Boulder School of Forestry and Environmental Studies, Yale University

2Yale


Abstract graphic

∗

Corresponding author. Phone: (509) 768-9485; fax: 215-898-0964; email: [email protected]; Present address: Earth and Environmental Science, University of Pennsylvania.

ACS Paragon Plus Environment


Page 2 of 30

2 4

Abstract

5

Exchange of water and solutes between contaminated soil matrix and bulk

6

solution in preferential flow paths has been shown to contribute to long-term release of

7

dissolved contaminants in the subsurface, but whether and how this exchange can

8

affect the release of colloids in soil are unclear. To examine this, we applied rainfall

9

solutions of different ionic strength on an intact soil core and compared resulting

10

changes in effluent colloid concentration through multiple sampling ports. Exchange of

11

water between soil matrix and the preferential flow paths leading to each port was

12

characterized on the basis of bromide (conservative tracer) breakthrough time at the

13

port. At individual ports, two rainfalls of a certain ionic strength mobilized different

14

amounts of colloids when the soil was pre-exposed to a solution of lower or high ionic

15

strength. This result indicates that colloid mobilization depended on rainfall solution

16

history, which is referred as colloid mobilization hysteresis. Extent of hysteresis was

17

increased with increases in exchange of pore water and solutes between preferential

18

flow paths and matrix. The results indicate that the soil matrix exchanged the old water

19

from the previous infiltration with new infiltrating water during successive infiltration

20

and changed the pore water chemistry in the preferential flow paths, which in turn

21

affected the release of soil colloids. Therefore, rainfall solution history and soil

22

heterogeneity must be considered to assess colloid mobilization in the subsurface. These

23

findings have implications for the release of colloids, colloid-associated contaminants

24

and pathogens from soils.


Page 3 of 30


3 25 26

Introduction Soil typically removes contaminants from infiltrating water and prevents

27

possible groundwater contamination. However, the contaminated soil may become a

28

long-term source of contaminants due to potential release of biocolloids and colloid-

29

associated chemicals.1-3 Therefore, understanding the processes that contribute to the

30

long-term release of colloids and colloid-associated contaminants in soils can help

31

assess and mitigate the risk of groundwater contamination.

32

During rainfall, colloids can be released from soil by several processes: scouring

33

by air-water interfaces, shear from flow rate increases, expansion of water films,

34

increase in pH or dissolved organic carbon concentration, and decrease in ionic strength

35

of infiltrating water.4 Among these factors, pH, ionic strength, and dissolved organic

36

carbon affect the interaction energy between colloids and collector surfaces (e.g., soil

37

grains).2 For instance, as ionic strength decreases, the interaction energy between

38

colloids and collector surface decreases below the kinetic energy of colloids, and this

39

results in a spontaneous release of colloids.5

40

Colloid mobilization also depends on soil heterogeneity.6 In general, soils are

41

physically heterogeneous and contain preferential flow paths including bedding planes,

42

cracks, and plant roots.7 During dry-wet cycles, these flow paths play a critical role in

43

generation8, distribution9 and transport10 of colloids in subsurface. Due to a large

44

difference between water infiltration rates in preferential flow paths and soil matrix,

45

water in preferential flow paths can be considered “new” and water in the soil matrix



Page 4 of 30

4 46

can be considered “old.” Old water may also reside in water packets trapped in dead-

47

end pores and intra-aggregate waters. The locations where old water resides have been

48

referred to as the immobile region.11 The old water in immobile regions can interact

49

with the new water in preferential flow paths by exchanging water and solute due to a

50

difference in their solution chemistry and water potentials (the sum of potential energy

51

contributed by gravity, surface tension, and osmosis).12 Furthermore, water may flow

52

between preferential flow paths and matrix during wetting, gravity drainage, and

53

drying.13 During gravitation drainage, high flow rate in macropore causes the

54

development of slight negative pressure, which in turns causes emptying of

55

macropores.13 Whereas during drying, capillary-induced water flow could drive water

56

flow from macro- and mesopores to hydraulically connected micropores.14 The water

57

exchange has been shown to contribute to the long-term release of dissolved

58

contaminants from contaminated soil matrix15, 16, but whether and how the exchange

59

can affect the long-term release of colloids or particulate contaminants in soil are poorly

60

understood.17-19 Experimental design in most of the studies on colloid mobilization is

61

inadequate to help probe the interaction between preferential flow paths and soil

62

matrix. For instance, most studies that examined colloid retention or release in

63

unsaturated soil used packed homogeneous soil 20, but this method destroys the

64

preferential flow paths in soil and thus results in an underestimation of colloid and

65

contaminant transport in soil.21 Only a few studies have examined the mobilization of

66

colloids in intact soil cores 6, 8, 22, 23 and they all collected a composite water sample from


Page 5 of 30


5 67

the entire soil core. Thus, they did not evaluate how colloid mobilization is affected by

68

water exchange between preferential flow paths and matrix.19

69

We examined the effect of water exchange between preferential flow paths and

70

soil matrix on the mobilization of colloids in a heterogeneous soil. We hypothesized

71

that pore water and solute exchange between preferential flow paths and soil matrix

72

would affect solution chemistry in the preferential flow paths and would consequently

73

affect colloid mobilization. To test the hypothesis, we intermittently applied rainfall

74

solution at multiple ionic strengths on an intact core of a fractured soil. Then we

75

compared resulting changes in colloid concentration and specific conductivity of

76

effluent from different preferential flow paths. On the basis of bromide transport

77

through ports during intermittent wetting cycles, exchange of pore water between

78

preferential flow paths and soil matrix was characterized and then linked to colloid

79

mobilization at the ports during applications of solutions at different ionic strength.

80

Materials and Methods

81

Intact Soil Core Sampling. Intact soil cores were collected from the bottom of a

82

hill slope in the Melton Branch watershed in Oak Ridge Reservation in Tennessee, USA.

83

This site had been used in many studies to examine the effect of macropores on solute

84

transport in subsurface.9, 10, 24-29 The soil consists of fractured shale saprolite with

85

distinct bedding planes and fractures where pore sizes are classified as micropores (< 10

86

µm), mesopores (10 to 1000 µm), and macropores (>1000 µm).30 Macropore and

87

mesopores occupied nearly 0.17% of total soil volume28, but they conduct over 96% of



Page 6 of 30

6 88

infiltrating rainwater during a typical rainfall. 24, 31, 32 Micropores conduct water during

89

drying periods after gravitation drainage.30 The saturated hydraulic conductivity of the

90

soil at the field site was estimated in a previous study33, and it ranges between 4×10-6 to

91

1×10-4 m s-1. This site receives an average annual rainfall of 130 cm y-1, which accelerates

92

the weathering clay minerals and iron and manganese oxides. The infiltrating

93

rainwater carries the minerals to fracture and micro-fracture surfaces.24 These locations

94

are found to be major sources for colloids in this soil.9 Illite is the primary clay mineral

95

in the soil. Cation exchange capacity of the soil varies between 7 to 16 cmolc kg-1.24 Due

96

to extensive weathering, the soil core is expected to have limited nutrients (N, P, and C).

97

24

98 99

Using the intact soil cores collected from the field site, we have previously examined how subsurface colloid mobilization is affected by colloid size10, dry-wet

100

cycles9, and freeze-thaw cycles29. These studies provided a detailed description of intact

101

core sampling at the field site. Briefly, the core sampling involved removal of vegetation

102

and the O-layer of the soil followed by isolation of a soil cylinder within the A-horizon.

103

The cylinder was trimmed to fit inside polyvinyl chloride (PVC) pipes of 30.5 cm height

104

and 25.4 cm diameter with gaps of about 1.2 cm between the pipe wall and the soil core.

105

The gap was filled with polyurethane expandable foam (U.S. Composites, Inc., FL) to

106

secure the core to the pipe, and then the core was carefully dislodged from the

107

underlying soil and stored at 4 °C prior to its use in the experiment.


Page 7 of 30


7 108

Experimental Setup. A detailed description of the experimental setup is

109

provided elsewhere.10 The experimental setup consisted of a rainfall simulator, the soil

110

core, and a 19-port sampling grid. Rainfall simulator equipped with 85 needles (25

111

gauges) was used to distribute solutions uniformly on top of the intact core. The

112

sampling grid with 19 ports at the core base channeled pore water effluent from

113

different sections of the soil core into sample collection tubes. Each port captured water

114

that drained through the 11.4-cm2 base area under zero tension; the total collected area

115

at core base was 340 cm2 (Figure 1). To separate water that flowed near wall from the

116

water that flowed through the soil core, water drained near the perimeter of the core

117

was routed to a channel (1.3 cm width) around the sample ports and discarded. In

118

preliminary experiments with the full-height (30.0 cm) soil core, we observed that the

119

slanted bedding plane fractures carried most of the infiltrating water to the core walls.

120

To minimize this problem, we cut the soil core to decrease the height to 15.0 cm so that

121

most of the bedding planes did not extend to the foam wall. The final soil core

122

dimensions were 15.0 cm height and 25.4 cm diameter.

123

To monitor soil moisture content at the center of soil core, a soil moisture probe

124

(Delta-T Devices, Theta Probe MLX2) was inserted laterally 6 cm into the core and 7.5

125

cm below top surface of the core. The moisture data from the probe were collected at 1

126

min intervals using a data logger (DATAQ Instruments, DI 710). A variation in soil

127

moisture content at 7.5 cm depth indicates any temporal change in soil moisture at the

128

core center at different stages of rainfall. The moisture content of other parts of the soil

129

core could differ from the value measured at the center.



Page 8 of 30

8 130

Mobilization of Soil Colloids. We applied rainfall at a rate at 2.5 cm h-1, which

131

represents a typical heavy rainfall rate at Oak Ridge.34 To examine the effect of water

132

and solute exchange on colloid mobilization in preferential flow paths, the experiments

133

were conducted in three phases: conditioning, colloid mobilization, and bromide

134

application (Table 1). In conditioning phase, a solution of 0.01 mM NaCl was applied

135

for 48 h to remove any suspended colloids and equilibrate moisture content in the

136

pores. Following this rainfall application, the core was subjected to gravitational

137

drainage for 18.7 h. The colloid mobilization experiment was initiated by application of

138

a rainfall solution of 10 mM NaCl for 5.3 h followed by 18.7 h of draining by gravity.

139

During rainfall application, water samples were collected at multiple ports in glass test

140

tubes at 20 min intervals and immediately analyzed. During drainage by gravity,

141

effluent flow was insufficient for analysis. The mobilization experiment was repeated

142

by applying a total of ten rainfall solutions of different concentrations of NaCl to the

143

same core in the following order: 10, 5, 1, 0.1, 0.01, 0.01, 0.1, 1, 5, and 10 mM. Between

144

each rainfall application for 5.7 h, the flow was interrupted for 18.7 h for drainage by

145

gravity. We used solutions of ionic strength between 0.01 and 10 mM to simulate

146

infiltrating rainwater. A typical ionic strength of infiltrating rainwater is 0.1 mM 35,

147

although the ionic strength of infiltrating water can increase or decrease based on the

148

runoff solution chemistry. 36

149

We applied bromide (a conservative tracer) to the soil core after completion of

150

colloid mobilization experiments to measure the residence time of water that drained at

151

different ports. We employed flow interruption technique to compare the extent of


Page 9 of 30


9 152

water exchange between preferential flow paths and soil matrix that contributed to flow

153

at all ports.37 The application of a 1 mM NaBr solution for 10 h was preceded by

154

application of a 0.01 mM NaCl solution for 24 h to flush pore water from the core and

155

followed by application of a 0.01 mM of NaCl for another 10 h to fluh bromide from soil

156

pores. Water samples were collected at multiple ports in glass test tubes at 20 min

157

intervals and immediately analyzed for bromide. To confirm whether the exchange of

158

water between fractures and matrix occurred during gravity drainage in the colloid

159

mobilization experiments, we interrupted the flow for the same duration (18.7 h, as

160

used in the colloid mobilization experiment) once during the plateau of the bromide

161

breakthrough curve and twice during the breakthrough tail and let the pore water

162

drained by gravity. As bromide does not adsorb on or release from the soil, any change

163

in concentration of bromide during flow interruption at the port would be indicative of

164

water and solute exchange between pore waters in fractures and soil matrix.37 To

165

measure a change in bromide concentration during flow interruption, we compared

166

bromide concentration of effluent before the flow interruption with bromide

167

concentration of effluent after flow was reinitiated. Residence time for the flow paths

168

reaching each of the 19 ports through which pore water flowed was estimated by

169

measuring the time required for the effluent bromide concentration to reach 50% of the

170

influent concentration.

171

Analysis of Water Samples. We calculated the volume of water drained at each

172

port by subtracting the weight of the empty tube from the tube with water sample. Flow

173

rate at ports was calculated by dividing sample volume with time interval for the



Page 10 of 30

10 174

sample collection and converted to water flux by dividing the flow rate with the area of

175

a port. During rainfall application, flux was measured after water started to drip from

176

the bottom of the core. Samples collected from active ports were analyzed for pH (Orion

177

8102BNU combination electrode and Orion 720A meter), specific conductivity (Orion

178

011050 probe and Orion 105 meter), turbidity (Hach 2100N), and bromide concentration

179

(bromide-selective probe, Accumet 13-620-524, and Orion 720A meter). Specific

180

conductivity was used as a proxy for the ionic strength of pore water. Turbidity was

181

used as a proxy for colloid concentration. Turbidity was converted to colloid

182

concentration (mg L-1) using a calibration curve of total suspended sediment

183

concentration versus turbidity as described elsewhere.29

184

Results

185

Water Flow through Fractures. Water drained through 6 of the 19 ports below

186

the core (Figure 2). Outflow flux through the active ports varied spatially by a factor of

187

1.9 and exceeded the rainfall application rate. During a rainfall event, water flux

188

through active ports was steady (flux varied by no more than 10% from one sampling

189

interval to another), but during the 5th rainfall, the flux through port R270 decreased,

190

and water began to flow from an adjacent port R240, which was previously inactive.

191

The change in flow rate at two adjacent ports indicates that some of the preferential

192

flow paths leading to port R270 become connected to flow paths leading to port R240.

193

During all rainfall events, moisture content at the core center was 65 ± 1% (average ±

194

standard deviation), which was reduced to 59 ± 1% after 18.7 h of gravity drainage.


Page 11 of 30


11 195

With increases in the ionic strength of the feed solutions, the specific

196

conductivity of the effluent increased, and with decreases in the ionic strength of the

197

feed solutions, effluent conductivity decreased (Figure 2). At the beginning of a rainfall

198

event (within first 1 h), effluent specific conductivity depended on the ionic strength of

199

the previous rainfall event, but as rainfall progressed, the specific conductivity of

200

effluent became similar to influent conductivity. During first 1 h of a rainfall, effluent

201

conductivity was greater than the rainfall solution conductivity if a solution at high

202

ionic strength was injected in the previous rainfall. On the other hand, effluent

203

conductivity was lower than the rainfall solution conductivity if a solution at low ionic

204

strength was injected in the previous rainfall. Effluent pH varied spatially during

205

rainfall event and temporally between rainfall events (Figure 2). In particular, effluent

206

pH at all ports increased with a decrease in ionic strength of rainfall solution.

207

Bromide breakthrough time and plateau concentration varied between ports

208

(Figure 3). Flow interruption during the plateau of bromide breakthrough curve caused

209

a decrease in bromide concentration, whereas flow interruptions during the tail of a

210

bromide breakthrough curve resulted in an increase in bromide concentration. The

211

change in concentration of bromide due to flow interruption was greatest at the port

212

with longest breakthrough time, and the change in concentration of bromide was least

213

for the port with the shortest breakthrough time. Bromide breakthrough time was

214

uncorrelated (R2 = 0.011) to the flux at the ports (insert in Figure 3).



Page 12 of 30

12 215

Mobilization of Colloids. During rainfall applications, colloid concentration

216

exhibited a rapid initial peak and a gradual decrease (Figure 2). Typically, colloid

217

mobilization was high when the ionic strength of the rainfall was low, and low when

218

the ionic strength of the rainfall was high. Colloid mobilization varied spatially between

219

ports.

220

The peak in colloid concentration or the total amount of colloids mobilized

221

during two rainfalls at same ionic strength was not matched (Figure 2). Comparing

222

colloid mobilization peaks in rainfall at same ionic strength (e.g., 2 vs. 9, 3 vs. 8, and 4

223

vs. 7), we found that pre-exposure to lower ionic strength water increased colloid

224

mobilization from soil, whereas pre-exposure to higher ionic strength water lowered the

225

colloid mobilization. The dependence of the colloid mobilization on the history of

226

rainfall solution is referred as colloid mobilization hysteresis and plotted in Figure 4.

227

Colloid mobilization during infiltration of 10 mM NaCl solution (rainfall 1 and 10) was

228

negligible due to increase in attractive interaction between soil colloids and stationary

229

soil grains.38 Thus, pre-exposure to different ionic strength water had negligible effect

230

on the amount of colloids mobilized during rainfall 1 and 10. Extent of colloid

231

mobilization hysteresis varied between ports and maximum degree of hysteresis was

232

observed at port R30.

233

Discussion

234 235

Water flow through a heterogeneous soil core. The spatial variation of water flow observed in this study is consistent with results observed in previous field studies


Page 13 of 30


13 236

on the Oak Ridge soil in laboratory9, 10 and at the field site.28, 32 In our study, only six

237

ports drained water during rainfall events and the outflow flux at these ports exceeded

238

rainfall application rate (or corresponding inflow flux on the core top). Water applied at

239

the top of soil core flowed through a smaller area at the core base, thereby causing the

240

outflow flux at the active ports to exceed the rainfall application rate. These results

241

indicate that most of the applied water infiltrated through preferential flow paths such

242

as bedding planes, fractures, and micro-fractures in the soil core. Using dye injection in

243

another soil core from the Oak Ridge site, we previously showed that fractures in the

244

soil varied by length and pore size and were distributed heterogeneously along with

245

soil matrix.10 Based on our previous study9, drying during 18.7 h of gravitation drainage

246

is negligible. During gravity drainage, pore water is expected to drain from macropores

247

due to development of negative water pressure during high flow through macropore.13

248

On the other hand, during drying periods after gravitation drainage, water flows in

249

mesopores and micropores and from mesopores to micropore due to evaporation

250

induced capillary flow.14 This process is relevant for exchange of water, solute and

251

colloids between preferential flow paths and soil matrix.

252

Lack of correlation between flow rate and bromide breakthrough time at ports

253

indicates that each port received water from soil flow paths of different macroporosity.

254

For instance, at constant rainfall application rate, the flux at two ports that received

255

water from either one large fracture or several small fractures could be similar, although

256

water in smaller fractures infiltrates at smaller velocity than water in large fractures.

257

Therefore, hydraulic residence time—average time taken to infiltrate a pore volume of



Page 14 of 30

14 258

water through fractures—would be short for the large fracture due to limited

259

interaction of infiltrating water with the surrounding soil matrix.39 Therefore, we use

260

hydraulic residence time, not water flux, to characterize the interaction of pore water

261

between fractures and matrix. Stopping the flow during the bromide breakthrough

262

plateau decreased the concentration of bromide in the water, and stopping the flow

263

during the breakthrough tail increased the concentration of bromide in the pore water.

264

Stopping the rainfall application interrupted advective transport of infiltrating water

265

through fractures; therefore, any change in the concentration of bromide in the pore

266

water can be attributed to exchange of pore water or solute between the fractures (new

267

water) and the soil matrix (old water).37 The port where bromide concentration changed

268

the most during flow interruption also exhibited the longest bromide breakthrough

269

time and the shortest plateau concentration. This indicates that this port had received a

270

greater amount of old water from immobile region. Therefore, we surmised that

271

preferential flow paths that exhibited longer residence time (or bromide breakthrough

272

time) had a greater fraction of the pore water exchanged with water in the surrounding

273

soil matrix.

274

Mobilization of Colloids. Colloid mobilization varied spatially between ports,

275

which is expected due to difference in soil physical heterogeneity.10 As explained

276

earlier, each port received water from flow paths with different permeability, which

277

could affect colloid transport and mobilization during a rainfall.10 Similar to the results

278

of the previous studies9, 10, colloid mobilization (peaks and total amount of colloids)

279

generally increased with increases in flux at ports, possibly due to increase in shear on


Page 15 of 30


15 280

attached colloids.40 In the previous studies 9, 10, we characterized the colloids mobilized

281

from the soil cores for their size distribution, surface charge, elemental composition,

282

and mineral properties. We found that colloids mobilized due to dry-wet cycles had

283

73.7% clay minerals, 3.5% quartz, and 4.6% iron-bearing minerals, and more than 90%

284

of the mobilized colloids had size less than 10 µm.9 The colloids exhibits net negative

285

surface charge in the pH ranges of pore water.10 We assumed that similar types of

286

colloids were mobilized in this study.

287

We observed that colloid mobilization increased with a decrease in the ionic

288

strength of rainfall solution. This is in accordance with Derjaguin-Landau-Verwey-

289

Overbeek (DLVO) theory41, which predicts that a decrease in ionic strength increases

290

the repulsion between the surfaces of colloid and grains of the same charge. The

291

increase in repulsion results in colloid mobilization in soil 42. A decrease in ionic

292

strength also increased pore water pH, which could also contribute to colloid

293

mobilization. However, different amounts of colloids were mobilized during two

294

rainfalls of same solution ionic strength following an exposure to a lower or higher ionic

295

strength water. This result indicates that colloid mobilization during a rainfall

296

depended on the rainfall solution history. We attributed this colloid mobilization

297

discrepancy to a difference in pore water ionic strength as a result of mixing of the

298

infiltrating pore water (new water) with pore water in the matrix (old water). A

299

difference in the initial conductivity of effluents during first hour of two rainfalls with

300

the same infiltrating solution confirmed that the ionic strength of pore water was

301

different during the beginning of both rainfalls. This initial period was critical for the



Page 16 of 30

16 302

mobilization of colloids by advancing wetting cycle.43 Furthermore, colloids could be

303

released from or near soil matrix and accumulate in pore water during gravity

304

drainage9 or flow pause.23 Consequently, the exchange of water between matrix and

305

preferential flow path could also carry some of these colloids and thus affect the

306

amount of colloids transported through preferential flow paths during the ensuing

307

rainfall. For instance, colloid mobilization peak was high in port R30, which also had

308

the longest bromide residence time. A long residence time indicates that water

309

exchange between matrix and preferential flow paths at this port was greater than the

310

exchange at the other ports. Thus, colloid mobilization due to exchange of water is

311

expected to be high at port R30.

312

Mechanism of Colloid Mobilization Hysteresis. We quantified the discrepancy

313

in colloid mobilization between rainfalls at a certain ionic strength (I) with a hysteresis

314

index (HII, Equation 1). We defined hysteresis index (HII) as the ratio of concentration

315

difference in colloids concenration of the colloid mobilization peaks in first (C1) and

316

second rainfall (C2) at a given ionic strength (I) to the concentration of the colloid peak

317

(C1) during the first rainfall. ∆C C2 − C1 = C1 C1

318

HI I =

319

The hysteresis index is a measure of colloid mobilization discrepancy between

(Equation 1)

320

two rainfalls when the same ionic strength solution was used in two rainfalls. When HII

321

is zero, there would be no discrepancy; when HII > 0, colloid mobilization would


Page 17 of 30


17 322

increase from first to second rainfall; when HII < 0, colloid mobilization would decrease

323

from first to second rainfall. We used colloid mobilization peaks, instead of total

324

amount of colloids mobilized during the entire rainfall, because the peak accounts for

325

the colloids accumulated from soil matrix or mobilized due to exchange of water during

326

flow interruption.9, 23, 44 As rainfall continues, the contribution of the exchanged water to

327

outflow is expected to decrease or become negligible due to a large difference in flow

328

rate between matrix and preferential flow path.13

329

The hysteresis index varied between ports and was correlated to bromide

330

breakthrough time at the port (Figure 5). Because bromide breakthrough time or

331

hydraulic residence time is an indicator of exchange between fractures and the soil

332

matrix 45, we attributed hysteresis in colloid mobilization to exchange of infiltrating

333

pore water between fractures and the soil matrix (or new and old water, respectively).

334

The colloid mobilization in preferential flow paths would be independent of old water

335

or the previous rainfall solution chemistry if less water is exchanged between

336

preferential flow paths and the surrounding soil matrix. On the other hand, the colloid

337

mobilization would be dependent on the old water if more water is exchanged. This

338

theory is confirmed by high degree of hysteresis at the ports (e.g., port R30) that

339

exhibited a slow change in the specific conductivity of the pore water during the

340

application of rainfall of a different ionic strength. This slow change in the specific

341

conductivity indicates a high degree of exchange of pore water between fractures and

342

the soil matrix.



Page 18 of 30

18 343

To investigate the cause of colloid release hysteresis, Torkzaban et al. 46 subjected

344

saturated homogeneous media (sand and glass bead) with varying surface roughness to

345

intermittent infiltration of solutions at different ionic strengths and observed that

346

colloid retention and release hysteresis was increased with increases in collector surface

347

roughness and colloid size. Thus, they attributed colloid retention and release

348

hysteresis to nanoscale heterogeneity on collector surface. In our study, we observed

349

that colloid mobilization hysteresis varied between ports even though nanoscale

350

heterogeneity is expected to be similar in different regions of the soil core. Furthermore,

351

the mechanism of colloid retention or mobilization hysteresis is expected to be different

352

in unsaturated soil during dry-wet cycles due to the presence of pore scale

353

heterogeneity, immobile water, and unsaturated condition. Therefore, we surmise that

354

exchange of pore waters between matrix and fractures contributed to colloid

355

mobilization hysteresis in our study.

356

Environmental Implications. Comparing extents of colloid mobilization through

357

the preferential flow paths, we resolved the effect of pore water interaction between

358

preferential flow paths and soil matrix on colloid transport. We showed that the soil

359

matrix could serve as a memory of previous solution chemistry in a soil by trapping

360

and releasing solutions from previous rainfall events (old water). An exchange of the

361

old water with infiltrating water (new water) changes the solution chemistry of bulk

362

pore water, which in turns influences colloid mobilization in the soil. This advances the

363

understanding of how macropore-matrix interaction affects colloid mobilization in

364

heterogeneous soil.19 The mobilized colloids could reintroduce colloid-associated


Page 19 of 30


19 365

chemical contaminants to infiltrating water 29. Our result has direct implication on

366

colloid-facilitated mobilization of low-level radioactive wastes,47 which reside in

367

shallow storage trenches and pits in soil horizons overlying the fractured bedrock at

368

Oak Ridge National Laboratory (ORNL)—the location from where the soil core was

369

collected. Furthermore, interaction between new and old water has an implication for

370

the long-term release or retention of microbial contaminants in subsurface soil.48

371

Because the old water can be a significant component of vadose zone outflow into

372

streams or aquifers49, the exchange of old water with new or infiltrating water and

373

resulting changes in colloid mobilization in the subsurface could have implications on

374

water quality of streams and aquifers. A change in pore water solution chemistry can

375

also affect sorption and desorption of contaminants and dissolution of minerals 50, 51.

376

Thus, the study that aims to examine the effect of ionic strength on colloid mobilization

377

and contaminant sorption and desorption in heterogeneous soils must account for the

378

solution chemistry of the previous rainfall.

379

Acknowledgements

380

We thank the U.S. Department of Energy Environmental Science and

381

Management Program for supporting this research under grant DOE-FG02-08ER64639.

382

We thank Timothy Dittrich and Mark Serravezza (University of Colorado at Boulder)

383

and Tonia Melhorn and Philip Jardine (Oak Ridge National Laboratory) for assistance

384

during sampling of soil cores from Oak Ridge, Tennessee.



Page 20 of 30

20 385

Reference

386 387 388

1. McCarthy, J. F.; Zachara, J. M., Subsurface transport of contaminants: Mobile colloids in the subsurface environment may alter the transport of contaminants. Environ. Sci. Technol. 1989, 23, (5), 496-502.

389 390

2. Kretzschmar, R.; Borkovec, M.; Grolimund, D.; Elimelech, M., Mobile subsurface colloids and their role in contaminant transport. Advances in Agronomy 1999, 66, 121-193.

391 392 393

3. Mohanty, S. K.; Torkelson, A. A.; Dodd, H.; Nelson, K. L.; Boehm, A. B., Engineering solutions to improve the removal of fecal indicator bacteria by bioinfiltration systems during intermittent flow of stormwater. Environ. Sci. Technol. 2013, 47, (19), 10791-10798.

394 395

4. DeNovio, N. M.; Saiers, J. E.; Ryan, J. N., Colloid movement in unsaturated porous media: Recent advances and future directions. Vadose Zone J. 2004, 3, (2), 338-351.

396 397

5. Tufenkji, N.; Elimelech, M., Breakdown of colloid filtration theory: Role of the secondary energy minimum and surface charge heterogeneities. Langmuir 2005, 21, (3), 841-852.

398 399

6. Jacobsen, O. H.; Moldrup, P.; Larsen, C.; Konnerup, L.; Petersen, L. W., Particle transport in macropores of undisturbed soil columns. J. Hydrol. 1997, 196, (1-4), 185-203.

400 401

7. Beven, K.; Germann, P., Macropores and water-flow in soils. Water Resour. Res. 1982, 18, (5), 1311-1325.

402 403

8. Majdalani, S.; Michel, E.; Di-Pietro, L.; Angulo-Jaramillo, R., Effects of wetting and drying cycles on in situ soil particle mobilization. Eur. J. Soil. Sci. 2008, 59, (2), 147-155.

404 405 406

9. Mohanty, S. K.; Saiers, J. E.; Ryan, J. N., Colloid mobilization in a fractured soil during dry-wet cycles: Role of drying duration and flow path permeability. Environ. Sci. Technol. 2015, 49, (15), 9100-9106.

407 408 409

10. Mohanty, S. K.; Bulicek, M. C. D.; Metge, D. W.; Harvey, R. W.; Ryan, J. N.; Boehm, A. B., Mobilization of microspheres from a fractured soil during intermittent infiltration events. Vadose Zone J. 2015, 14, (1), doi:10.2136/vzj2014.05.0058.

410 411 412

11. Simunek, J.; Jarvis, N. J.; van Genuchten, M. T.; Gardenas, A., Review and comparison of models for describing non-equilibrium and preferential flow and transport in the vadose zone. J. Hydrol. 2003, 272, (1-4), 14-35.

413 414 415

12. Kapetas, L.; Dror, I.; Berkowitz, B., Evidence of preferential path formation and path memory effect during successive infiltration and drainage cycles in uniform sand columns. J. Contam. Hydrol. 2014, 165, (0), 1-10.

416 417 418

13. Gwo, J. P.; Jardine, P. M.; Wilson, G. V.; Yeh, G. T., Using a multiregion model to study the effects of advective and diffusive mass transfer on local physical nonequilibrium and solute mobility in a structured soil. Water Resour. Res. 1996, 32, (3), 561-570.


Page 21 of 30


21 419 420

14. Lehmann, P.; Assouline, S.; Or, D., Characteristic lengths affecting evaporative drying of porous media. Physical Review E 2008, 77, (5), doi:10.1103/PhysRevE.77.056309.

421 422 423

15. Seyedabbasi, M. A.; Newell, C. J.; Adamson, D. T.; Sale, T. C., Relative contribution of DNAPL dissolution and matrix diffusion to the long-term persistence of chlorinated solvent source zones. J. Contam. Hydrol. 2012, 134, 69-81.

424 425

16. Feenstra, S.; Cherry, J.; Sudicky, E.; Haq, Z., Matrix diffusion effects on contaminant migration from an injection well in fractured sandstone. Ground Water 1984, 22, (3), 307-316.

426 427

17. Leij, F. J.; Bradford, S. A., Colloid transport in dual-permeability media. J. Contam. Hydrol. 2013, 150, 65-76.

428 429

18. Oswald, J. G.; Ibaraki, M., Migration of colloids in discretely fractured porous media: effect of colloidal matrix diffusion. J. Contam. Hydrol. 2001, 52, (1-4), 213-244.

430 431

19. McCarthy, J. F.; McKay, L. D., Colloid transport in the subsurface: Past, present, and future challenges. Vadose Zone J. 2004, 3, (2), 326-337.

432 433 434

20. Bradford, S. A.; Torkzaban, S., Colloid transport and retention in unsaturated porous media: A review of interface-, collector-, and pore-scale processes and models. Vadose Zone J. 2008, 7, (2), 667-681.

435 436 437

21. D'Alessio, M.; Vasudevan, D.; Lichwa, J.; Mohanty, S. K.; Ray, C., Fate and transport of selected estrogen compounds in Hawaii soils: Effect of soil type and macropores. J. Contam. Hydrol. 2014, 166, 1-10.

438 439 440

22. Kjaergaard, C.; Moldrup, P.; de Jonge, L. W.; Jacobsen, O. H., Colloid mobilization and transport in undisturbed soil columns. II. The role of colloid dispersibility and preferential flow. Vadose Zone J. 2004, 3, (2), 424-433.

441 442 443

23. Schelde, K.; Moldrup, P.; Jacobsen, O. H.; de Jonge, H.; de Jonge, L. W.; Komatsu, T., Diffusion-limited mobilization and transport of natural colloids in macroporous soil. Vadose Zone J. 2002, 1, (1), 125-136.

444 445 446

24. Jardine, P. M.; Jacobs, G. K.; Wilson, G. V., Unsaturated transport processes in undisturbed heterogeneous porous media: I. Inorganic contaminants. Soil Sci. Soc. Am. J. 1993, 57, (4), 945-953.

447 448 449

25. Jardine, P. M.; Sanford, W. E.; Gwo, J. P.; Reedy, O. C.; Hicks, D. S.; Riggs, J. S.; Bailey, W. B., Quantifying diffusive mass transfer in fractured shale bedrock. Water Resour. Res. 1999, 35, (7), 2015-2030.

450 451 452

26. Jardine, P. M.; Wilson, G. V.; Luxmoore, R. J.; Mccarthy, J. F., Transport of inorganic and natural organic tracers through an isolated pedon in a forest watershed. Soil Sci. Soc. Am. J. 1989, 53, (2), 317-323.



Page 22 of 30

22 453 454 455

27. Mayes, M. A.; Jardine, P. M.; Larsen, I. L.; Brooks, S. C.; Fendorf, S. E., Multispecies transport of metal-EDTA complexes and chromate through undisturbed columns of weathered fractured saprolite. J. Contam. Hydrol. 2000, 45, (3-4), 243-265.

456 457

28. Wilson, G. V.; Luxmoore, R. J., Infiltration, macroporosity, and mesoporosity distributions on two forested watersheds. Soil Sci. Soc. Am. J. 1988, 52, (2), 329-335.

458 459

29. Mohanty, S. K.; Saiers, J. E.; Ryan, J. N., Colloid-facilitated mobilization of metals by freeze-thaw cycles. Environ. Sci. Technol. 2014, 48, (2), 977-984.

460 461

30. Luxmoore, R. J., Microporosity, mesoporosity, and macroporosity of soil. Soil Sci. Soc. Am. J. 1981, 45, (3), 671-672.

462 463 464

31. Luxmoore, R. J.; Jardine, P. M.; Wilson, G. V.; Jones, J. R.; Zelazny, L. W., Physical and chemical controls of preferred path flow through a forested hillslope. Geoderma 1990, 46, (1-3), 139-154.

465 466

32. Watson, K. W.; Luxmoore, R. J., Estimating macroporosity in a forest watershed by use of a tension infiltrometer. Soil Sci. Soc. Am. J. 1986, 50, (3), 578-582.

467 468

33. Wilson, G. V.; Jardine, P. M.; O'Dell, J. D.; Collineau, M., Field-scale transport from a buried line source in variably saturated soil. J. Hydrol. 1993, 145, (1–2), 83-109.

469 470

34. Wilson, G. V.; Jardine, P. M.; Luxmoore, R. J.; Jones, J. R., Hydrology of a forested hillslope during storm events. Geoderma 1990, 46, (1-3), 119-138.

471 472 473

35. Chandra Mouli, P.; Venkata Mohan, S.; Reddy, S. J., Rainwater chemistry at a regional representative urban site: influence of terrestrial sources on ionic composition. Atmos. Environ. 2005, 39, (6), 999-1008.

474 475 476

36. Grebel, J. E.; Mohanty, S. K.; Torkelson, A. A.; Boehm, A. B.; Higgins, C. P.; Maxwell, R. M.; Nelson, K. L.; Sedlak, D. L., Engineering infiltration systems for urban stormwater reclamation. Environ. Eng. Sci. 2013, 30, (8), 437-454.

477 478

37. Brusseau, M. L.; Hu, Q.; Srivastava, R., Using flow interruption to identify factors causing nonideal contaminant transport. J. Contam. Hydrol. 1997, 24, (3-4), 205-219.

479 480

38. Grolimund, D.; Borkovec, M., Release of colloidal particles in natural porous media by monovalent and divalent cations. J. Contam. Hydrol. 2006, 87, (3-4), 155-175.

481 482 483

39. Reedy, O.; Jardine, P.; Wilson, G.; Selim, H., Quantifying the diffusive mass transfer of nonreactive solutes in columns of fractured saprolite using flow interruption. Soil Sci. Soc. Am. J. 1996, 60, (5), 1376-1384.

484 485

40. Kaplan, D. I.; Bertsch, P. M.; Adriano, D. C.; Miller, W. P., Soil-borne mobile colloids as influenced by water flow and organic carbon. Environ. Sci. Technol. 1993, 27, (6), 1193-1200.


Page 23 of 30


23 486 487 488

41. Torkzaban, S.; Bradford, S. A.; Walker, S. L., Resolving the coupled effects of hydrodynamics and DLVO forces on colloid attachment in porous media. Langmuir 2007, 23, (19), 9652-9660.

489 490

42. Zhang, H.; Selim, H. M., Colloid mobilization and arsenite transport in soil columns: Effect of ionic strength. J. Environ. Qual. 2007, 36, (5), 1273-1280.

491 492 493

43. Zhuang, J.; McCarthy, J. F.; Tyner, J. S.; Perfect, E.; Flury, M., In situ colloid mobilization in hanford sediments under unsaturated transient flow conditions: Effect of irrigation pattern. Environ. Sci. Technol. 2007, 41, (9), 3199-3204.

494 495

44. Jacobsen, O. H.; Moldrup, P.; de Jonge, H.; de Jonge, L. W., Mobilization and transport of natural colloids in a macroporous soil. Physics and Chemistry of the Earth 1998, 23, (2), 159-162.

496 497 498

45. Clothier, B.; Vogeler, I.; Green, S.; Scotter, D., Transport in unsaturated soil: Aggregates, macropores, and exchange. In Physical Nonequilibrium in Soils: Modeling and Application, Selim, H. M.; Ma, L., Eds. Ann Arbor Press: Chelsea, Michigan, 1998; pp 273-296.

499 500 501

46. Torkzaban, S.; Kim, H. N.; Simunek, J.; Bradford, S. A., Hysteresis of colloid retention and release in saturated porous media during transients in solution chemistry. Environ. Sci. Technol. 2010, 44, (5), 1662-1669.

502 503

47. US NRC, Advice on the Department of Energy's Cleanup Technology Roadmap : Gaps and Bridges. National Academies Press: Washington, D.C., 2009; p 271.

504 505 506

48. Safadoust, A.; Mahboubi, A. A.; Mosaddeghi, M. R.; Gharabaghi, B.; Unc, A.; Voroney, P.; Heydari, A., Effect of regenerated soil structure on unsaturated transport of Escherichia coli and bromide. J. Hydrol. 2012, 430–431, (0), 80-90.

507 508 509

49. Collins, R.; Jenkins, A.; Harrow, M., The contribution of old and new water to a storm hydrograph determined by tracer addition to a whole catchment. Hydrological Processes 2000, 14, (4), 701-711.

510 511 512

50. Zachara, J. M.; Serne, J.; Freshley, M.; Mann, F.; Anderson, F.; Wood, M.; Jones, T.; Myers, D., Geochemical processes controlling migration of tank wastes in Hanford's vadose zone. Vadose Zone J. 2007, 6, (4), 985-1003.

513 514 515

51. Dittrich, T. M.; Reimus, P. W., Uranium transport in a crushed granodiorite: Experiments and reactive transport modeling. J. Contam. Hydrol. 2015, 175–176, 44-59.



Page 24 of 30

24 516

Table 1. Experimental phases and the underlying purpose for each phase. Phase Conditioning

Colloid mobilization a

Treatment condition Infiltration of 0.1 mM NaCl solution for 48 h.

Purpose To remove any suspended colloids and equilibrate moisture content in the pores.

Gravitational drainage for 18.7 h.

To simulate drying cycle before wetting events for colloid mobilization.

Day 1: Application of 10 mM NaCl solution for 5.3 h followed by 18.7 h of gravitational drainage.

To compare mobilization of colloids from different preferential flow paths as a results of transient in solution ionic strength.

Day 2-5: The experiments were repeated at lower ionic strengths; solutions of 5, 1, 0.1, 0.01 mM NaCl were applied on Day 2, 3, 4, and 5, respectively. Day 6-10: The experiments were repeated with increasing ionic strengths. Solutions of 0.01, 0.1, 1, 5, and 10 mM NaCl were applied on Day 6, 7, 8, 9, and 10, respectively. Bromide application

517 518 519

Application of 0.01 MM NaCl for 24 h.

To flush pore water (10 mM NaCl) from the previous rainfall.

To obtain bromide breakthrough Application of 1 mM NaBr for 10 h followed by application of 0.01 mM NaCl at different ports. for 10 h. a After each rainfall application (5.3 h), the flow was interrupted for 18.7 h for drainage by gravity.


Page 25 of 30


25 520

List of Figures

521 522 523 524 525

Figure 1. Design of the 19-port grid for water sampling at the bottom of the soil core. The 19 ports (3.81 cm diameter each) were arranged in two concentric hexagons around a center port. The soil core was placed above the sampling port and below the rainfall reservoir where 85 needles were also arranged hexagonally to distribute rainwater uniformly on the top of the core.

526 527 528 529 530

Figure 2. Effluent colloid concentration (mg L-1), specific conductivity (µS cm-1), pH, and flux (cm h-1) at six active ports during 10 rainfall events at different ionic strengths. The six active ports are listed in the legend in the upper right corner. Flow in port R240 commenced during the fifth rainfall application. Between rainfall events, the core was drained by gravity for 18.7 h.

531 532 533 534 535 536

Figure 3. Concentration of bromide normalized to the influent concentration (C/C0) at the six active ports. The six active ports are listed in the legend in the upper right corner. Vertical dashed lines indicate 18.7 h flow interruptions. The x-axis represents cumulative rainfall duration without including flow interruption periods. The insert presents the correlation between bromide breakthrough time or residence time of water and flux at ports.

537 538 539 540 541 542 543 544

Figure 4. Colloid mobilization hysteresis at four ports in response to transient in solutions ionic strength. Colloid peak concentration in two rainfalls at same ionic strength varied based on the ionic strength of the previous rainfalls: colloid concentration was high when the previous rainfall had low ionic strength solution and low when the previous rainfall had high ionic strength solution. Results from Port R240 and 270 are not included because of inactive or unsteady flow at both ports during some of the 10 rainfall events. Filled symbols indicate the first mobilization experiment (10 mM) and arrows indicate the succession in which the experiments were carried out.

545 546 547 548 549 550 551 552

Figure 5. Degree of hysteresis or hysteresis index (HII) as a function of the bromide breakthrough time at the six active ports. Hysteresis index (HII) is defined as the ratio of difference in concentration of colloid peak (∆C= C2-C1) during first and second rainfalls at a certain ionic strength (I) to the concentration of the peak (C1) during the first rainfall. The y-axis scales are different. A larger slope indicates a greater discrepancy in colloid mobilization between two rainfalls at the ionic strength (I). Results from Port R240 and 270 are not included because of inactive or unsteady flow at both ports during some of the 10 rainfall events.

553



Page 26 of 30

26

554 555 556 557 558 559

Figure 1. Design of the 19-port grid for water sampling at the bottom of the soil core. The 19 ports (3.81 cm diameter each) were arranged in two concentric hexagons around a center port. The soil core was placed above the sampling port and below the rainfall reservoir where 85 needles were also arranged hexagonally to distribute rainwater uniformly on the top of the core.

560


Page 27 of 30


27

561 562 563 564 565 566

Figure 2. Effluent colloid concentration (mg L-1), specific conductivity (µS cm-1), pH, and flux (cm h-1) at six active ports during 10 rainfall events at different ionic strengths. The six active ports are listed in the legend in the upper right corner. Flow in port R240 commenced during the fifth rainfall application. Between rainfall events, the core was drained by gravity for 18.7 h.

567



Page 28 of 30

28 568

569 570 571 572 573 574 575

Figure 3. Concentration of bromide normalized to the influent concentration (C/C0) at the six active ports. The six active ports are listed in the legend in the upper right corner. Vertical dashed lines indicate 18.7 h flow interruptions. The x-axis represents cumulative rainfall duration without including flow interruption periods. The insert presents the correlation between bromide breakthrough time or residence time of water and flux at ports.

576


Page 29 of 30


29 577

578 579 580 581 582 583 584 585 586

Figure 4. Colloid mobilization hysteresis at four ports in response to transient in solutions ionic strength. Colloid peak concentration in two rainfalls at same ionic strength varied based on the ionic strength of the previous rainfalls: colloid concentration was high when the previous rainfall had low ionic strength solution and low when the previous rainfall had high ionic strength solution. Results from Port R240 and 270 are not included because of inactive or unsteady flow at both ports during some of the 10 rainfall events. Filled symbols indicate the first mobilization experiment (10 mM) and arrows indicate the succession in which the experiments were carried out.



Page 30 of 30

30

587 588 589 590 591 592 593 594 595

Figure 5. Degree of hysteresis or hysteresis index (HII) as a function of the bromide breakthrough time at the six active ports. Hysteresis index (HII) is defined as the ratio of difference in concentration of colloid peak (∆C= C2-C1) during first and second rainfalls at a certain ionic strength (I) to the concentration of the peak (C1) during the first rainfall. The y-axis scales are different. A larger slope indicates a greater discrepancy in colloid mobilization between two rainfalls at the ionic strength (I). Results from Port R240 and 270 are not included because of inactive or unsteady flow at both ports during some of the 10 rainfall events.

596


Colloid Mobilization in a Fractured Soil: Effect of Pore-Water Exchange

Recommend Documents