Enhanced transport of CeO2 nanoparticles in porous media by macropores

Science of the Total Environment 543 (2016) 223–229

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Enhanced transport of CeO2 nanoparticles in porous media by macropores Jing Fang a,⁎, Min-hao Wang a, Dao-hui Lin b, Bing Shen a a b

School of Environmental Science and Engineering, Zhejiang Gongshang University, Hangzhou 310012, PR China Department of Environmental Science, Zhejiang University, Hangzhou 310058, PR China

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• It is the first time to explore the effect of macropores on nCeO2 transport. • nCeO2 transport was enhanced by vertical macropores in quartz sand and soil. • The effect of preferential flow was greater in soil than it was in quartz sand.

a r t i c l e

i n f o

Article history: Received 11 August 2015 Received in revised form 9 October 2015 Accepted 7 November 2015 Available online 14 November 2015 Editor: Thomas Kevin V Keywords: nCeO2 Transport Quartz sand Soil Macropores

a b s t r a c t This is the first study to investigate the effect of macropores on the transport of CeO2 nanoparticles (nCeO2) in quartz sand and soil. The artificial macropore types are the vertical continuous macropore (O–O), and the vertical discontinuous macropore (O–C). The results indicated that the mobility of nCeO2 was significantly enhanced by the macropore in both quartz sand and soil, and the enhancement was greater in the continuous macropore than in the discontinuous macropore. Compared with the homogeneous column, both the O–O and O–C macropores in quartz sand favored an earlier breakthrough and a larger initial effluent recovery rate of nCeO2. However, there was little influence on the plateau concentration and the total effluent recovery rate. In soil, both types of macropores significantly shortened nCeO2 breakthrough time, and favored a higher plateau concentration, and a larger initial and total effluent recovery rate. The O–O macropore which accounted for only 1% of the total pore volume had doubly increased the total mobility of nCeO2 in soil; even the mobility was increased by 30% with the O–C macropore. It was found that the effect of preferential flow on nCeO2 transport was greater in soil than it was in quartz sand. © 2015 Elsevier B.V. All rights reserved.

1. Introduction ⁎ Corresponding author. E-mail address: [email protected] (J. Fang).

http://dx.doi.org/10.1016/j.scitotenv.2015.11.039 0048-9697/© 2015 Elsevier B.V. All rights reserved.

Cerium oxide nanoparticles (nCeO2) are widely used in consumer products that take advantage of the high oxygen storage and the UV

224

J. Fang et al. / Science of the Total Environment 543 (2016) 223–229

absorbing capacity of nCeO2, and the low redox potential of the CeIV/CeIII redox couple (Karen et al., 2009). Consequently, these nanoparticles are also used as a colloidal reactive barrier material in groundwater remediation (Kanel et al., 2006). However, comparable with the other nanomaterials, the widespread use of nCeO2 would inevitably bring about their release into the environment. Based on a 2001 report on the human health risks of cerium from diesel fuel, the average worldwide estimated level of cerium in soil was 20–60 mg kg−1 (Health Effects Institute, 2001). Another study reported that the modelpredicted nCeO2 concentration for soils near highways was 102– 103 μg kg−1 considering the nCeO2 accumulated during a 40-year period (Gottschalk et al. 2013). Furthermore, some typical characteristics of severe biological toxicity are associated with nCeO2 (Cassee et al., 2011; Leung et al., 2015) and these include substantial bioaccumulation in the liver of zebra fish exposed via ingestion (Johnston et al., 2010), disturbance in the plant defense mechanisms (Majumdar et al., 2014), and a significant lipid peroxidation and cell membrane damage of the lung cells (Lin et al., 2006). Because of the toxicity and potential acute health risk of nCeO2 to humans and the ecosystem, our understanding of their transport, particularly in natural porous media, should be improved to enhance the assessment of the life cycle of these engineered nanoparticles. Over the last decade, a number of researchers have explored the transport behavior of nCeO2. Li et al. (2011) found that nCeO2 in porous media is transported preferably in neutral and alkaline conditions, while it is deposited in acidic and high ionic strength conditions. The presence of low levels (1–6 mg/L) of Suwannee River humic acid, fulvic acid, alginate, citric acid, and carboxymethyl cellulose significantly enhanced the stability and mobility of nCeO2 (Liu et al., 2012; Lv et al., 2014). It was proposed that the retention of nCeO2 in the soil was associated with naturally occurring colloids, such as Al, Si, and Fe oxides (Cornelis et al., 2011). Petosa et al. (2013) studied the mobility of polyacrylic acid-coated nCeO2 in saturated quartz sand and loamy sand columns and observed the enhanced particle retention in the loamy sand column in comparison with the quartz sand column. Hassan et al. (2013) tried to quantitatively investigate the transport parameters of nCeO2 flowing through porous media and intended to develop a simulation applicable technique for designing and operating studies. Nonetheless, these studies were typically limited to the investigation of the transport and retention of nCeO2 in homogeneous porous media (lacking any macropores). Heterogeneity is an inherent feature of natural porous media that plays a crucial role in the transport processes of colloids and solutes (Sheng et al., 2014). Natural soil is a complex porous medium that consists of various types of macropores, such as plant root channels, earthworm holes that formed by biological forces, and fractures, cracks, and fissures that formed by physical processes. Macropores could act as preferential flow paths for water and solute when rainfall or irrigation occurs. It is well known that preferential flow is a non-equilibrium movement of water and solute through soils that bypass a portion of the soil matrix. Preferential flow limits the sorption capacity of the soil, reduces solute residence time and changes the distribution of elements in soil, posing an increased risk of ground water contamination (Garrido et al., 2014; Jassogne et al., 2009, Patil and Das, 2013). The introduction of preferential pathways could also reduce the transport time of colloids or microbe. It revealed that a network-like system of macropores was essential for the rapid transport of polystyrene latex microspheres (Ullum, 2001). The preferential pathway almost halved the front-arrival time and increased the recovery of colloids in heterogeneous and partially saturated sand columns in saturation conditions (Mishurov et al., 2008). Wang et al. (2013, 2014) conducted a systematic study of the effect of preferential flow on the transport of Escherichia coli in sandy columns. They found that a higher ionic strength enhanced the importance of preferential flow on the microbe, while the length and continuity of a macropore significantly affected the preferential transport.

In particular, understanding the effect of the macropore on the transport behavior of engineered nanoparticles is crucial to assessing their mobility in field porous media. However, relevant information is scarce. To our knowledge, this is the first study to investigate the effect of macropores on the transport of engineered nanoparticles in porous media. The main objectives of this research were therefore to (1) compare the effect of macropores on nCeO2 transport in two types of porous media (quartz sand versus disturbed soil), and (2) examine the influence of the physical features (being continuous or discontinuous) of macropores on the preferential flow and the subsequent transport of nCeO2. The results of this study are expected to provide helpful information on the better understanding of the transport of nanoparticles in natural soil with macropores and the effects of preferential flow. 2. Materials and methods 2.1. Porous media Two different materials (quartz sand and natural soil) were investigated in this study. Quartz sand (Shanghai Ling Feng Chemical Reagent Co. Ltd., China) consists of 99.8% SiO2 (quartz) and trace amounts of metal oxides, with a mean diameter of 325 μm. The quartz sand was cleaned sequentially by concentrated hydrochloric acid (HCl) and 10% sodium hydroxide (NaOH) to remove the metal impurities. The sand was subsequently rinsed repeatedly with deionized distilled water (DDW) to a neutral condition and was afterwards oven-dried at 105 °C. The zeta potential of the quartz sand, sonicated (100 W, 25 °C, 1 h) in 1 mM NaNO3 solution, was measured with a Zetasizer 3000 HSa (Malvern Instruments Ltd., UK). The zeta potential was −35.4 mV, indicating that the quartz was negatively charged. The soil of the surface layer (0–15 cm) was sampled from agricultural land in Hangzhou, China, and was air-dried and sieved (b 1 mm) prior to use. The soil pH was 7.71, measured at a soil to DDW ratio of 1:10 (w/ v) by an Orion (Model 250A+) pH meter. The soil organic matter was 1.94%, measured by using the Walkley–Black procedure (Nelson and Sommers, 1982). The soil texture was sandy loam, with 77.6% sand, 16.0% silt, and 6.4% clay, measured by the pipette method (Konert and Vandenberghe, 1997). The zeta potential of the soil, sonicated (100 W, 25 °C, 1 h) in 1 mM NaNO3 solution, was measured with a Zetasizer 3000 HSa (Malvern Instruments Ltd., UK) and was found to be −17.7 mV, indicating that the soil was also negatively charged. 2.2. Preparation of nCeO2 suspensions A concentrated nCeO2 suspension (≥40% [w/v], without dispersant), with the nCeO2 being 99% pure and the primary average particle size being 30 nm (provided by the company), was bought from Hangzhou Wanjing New Material Ltd., China, and was used without further surface modification. The hydrodynamic size and zeta potential of the nCeO2 suspension was 160 nm and −43.2 mV, respectively, measured using a Zetasizer (Nano-ZS90, Malvern, England). For each experiment, the concentrated nCeO2 suspension was diluted with DDW to achieve the desired concentration (approximately 160 mg L−1). This was followed by stirring and immediate bath sonication (100 W, 25 °C) for 15 min at room temperature. 2.3. Construction of macropores In the literature, various techniques for creating artificial macropores, depending on the objectives of the research, soil type, and column dimensions, have been described (Hu and Brusseau, 1995; Li and Masoud, 1997; Czapar et al., 1992). In this study, the macropores were created by inserting a glass rod (3 mm o.d.) before packing the columns and pulling out the glass rod after saturating the columns. A 40 mesh steel wire was used to stabilize the structure of the artificial macropores. Two types of columns with different

J. Fang et al. / Science of the Total Environment 543 (2016) 223–229

macropores were constructed (shown in Fig. 1). We defined the columns as the continuous vertical macropore (macropore open both at the packed column surface and the bottom, open-open, O–O), and the discontinuous vertical macropore (macropore open at the top surface and closed at the down surface of the packed column, open-closed, O– C). The column without an artificial macropore (homogenous column) was used as a control column.

2.4. Column experiments The transport experiments were performed in open vertical glass columns, of which the inner diameter was 5 cm and the height 25 cm. The columns were dry packed with quartz sand or soil. The packed densities of the quartz sand and soil columns were 1.45 and 1.04 g cm−3, respectively, and the length of the packed columns was about 15 cm. The columns were initially saturated with DDW, injected in an upward direction to remove any air entrapment. Once saturated, the columns were equilibrated by sequentially pumping DDW in a downward direction for approximately 4 pore volumes (PVs), using a peristaltic pump (Longer BT100-2J), to stabilize the columns. After equilibration, the input solution was switched to an nCeO2 suspension or a KBr solution. The effluent samples of nCeO2 or KBr were collected at discrete time intervals from the bottom of the column. When the feed dispersion was depleted, nCeO2 or KBr was flushed out by adding several additional PVs of DDW. At the same time, another quartz sand or soil column was leached with DDW through the whole time, which was used as blank. During the whole leaching experiment, constant ponding head of 1 cm was maintained at the column surface by controlling revolution of the peristaltic pump. The absorbance of the nCeO2 suspensions and effluent was measured by the UV–vis spectrophotometer (Agilent 8453) at 460 nm. A calibration curve was created by diluting the 160 mg L−1 nCeO2 suspension, which was linear within the range of 0–160 mg L−1. The determination limit of this method for nCeO2 was 0.6 mg L−1. The concentration of nCeO2 in the outflow was calculated by obtaining the difference between the concentration in the nCeO2 transport experiment and that in the blank. The dissolution of nCeO2 during the column experiment was disregarded in this study because it was reported to be much lower than or close to the detection limits of the neutral or alkaline conditions (Cornelis et al., 2011). A concentration of 30 mg L−1 KBr was used as a tracer. Both the influent and effluent concentrations of the bromide were detected by a precision ion meter (PXS-450), with a bromide selective electrode (Shanghai KangYi instrument Co., China). The concentrations of the influent and effluent nCeO2 suspensions (or bromide) were defined as C0 and C, respectively. C/C0 was used as

225

a parameter to generate the breakthrough curves (BTCs) as a function of PVs passing through the columns. All the experiments were conducted in duplicate and the results were found to be statistically similar (p N 0.05). All the experimental data were analyzed by the statistical analysis software SPSS (SPSS Inc., USA). 2.5. DLVO calculations The total interaction energy of the nCeO2 upon approach to the quartz sand and soil was subsequently calculated, using the Derjaguin–Landau–Verwey–Overbeek theory (DLVO) and a sphere– plate assumption (Derjaguin and Landau, 1941; Verwey and Overbeek, 1948). For the sphere–plate interactions (nCeO2 collector), the corresponding interaction energies (Eq. 1), equivalent to the sum of the electrostatic double layer (ΦEDL) (Eq. 2) (Hogg et al., 1966) and van der Waals (ΦVDW) (Eq. 3) interactions (Gregory, 1981; Petosa et al., 2010), were calculated using the following equations: ΦDLVO ðhÞ ¼ ΦEDL ðhÞ þ ΦVDW ðhÞ

ð1Þ


     1 þ expð−khÞ þ ψ2p þ ψ2c ln ½1− expð−2khÞ ΦEDL ðhÞ ¼ πε0 εr p 2ψp ψc ln 1 þ expð−khÞ

ð2Þ ΦVDW ðhÞ ¼ −

A r  123 p 14h 6h 1 þ λ

ð3Þ

where ΦDLVO, ΦEDL, and ΦVDW are the total, electrostatic, and van der Waals interaction energies, respectively, and h is the separation distance between the nCeO2 and the interface of interest [m]. ΦDLVO, ΦEDL, and ΦVDW are commonly made dimensionless by dividing by the product of the Boltzmann constant (k = 1.38 × 10−23 J K−1) and the absolute temperature [T]. ε is the dielectric constant of the medium (78.5 for water at 20 °C), ε0 is the vacuum permittivity (8.854 × 10−12 C2 N−1 m−2), rp is the average radius of nCeO2 [m]. ψp and ψc are the surface potentials of the nCeO2 and the collector (e.g., quartz sand and soil). The electrostatic double-layer interactions were quantified by employing the expression derived by Hogg et al. (1966), using zeta potentials in place of surface potentials. k is the Debye–Huckel parameter [m−1] and λ is the characteristic wavelength (100 nm) (Elimelech et al., 1995). The Hamaker constant (A) for the nCeO2–water–sand (A123 = 0.46 × 10−20 J) and the nCeO2–water–soil (A123 = 0.27 × 10−20 J) (Eq. 4) was calculated according to the Hamaker constant equation of the individual materials under saturated conditions (Israelachvili, 1992): A123 ¼

pffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffipffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffi A33 − A22 A11 − A22

ð4Þ

where A11, A22, and A33 are the Hamaker constants for nCeO2 (5.57 × 10− 20 J) (Israelachvili, 1992), water (3.70 × 10− 20 J) (Bergstrom, 1997), and collector (quartz sand 8.86 × 10−20 J and soil 6.50 × 10−20 J) (Israelachvili, 1992), respectively. 3. Results and discussion 3.1. Transport behavior of bromide in quartz sand

Fig. 1. Position of macropores in the columns of different treatments (a. homogeneous column; b. column with O–O vertical continuous macropore, c. column with O–C vertical discontinuous macropore).

The representative plots of the observed BTCs for bromide in both the homogeneous and the quartz sand columns with macropores are shown in Fig. 2a. Bromide acted as a good conservative tracer, with good mass recovery and breakthrough at 1 PV in the homogeneous (control) quartz sand column. In the column with the O–O macropore, bromide appeared in the effluent at 0.15 PV, which was much earlier than was that of the control at 0.71 PV. The recovery rate of the effluent bromide at an early stage (b 1 PV) in the O–O column was

226

J. Fang et al. / Science of the Total Environment 543 (2016) 223–229

Fig. 2. Representative breakthrough curves of (a) Br and (b) nCeO2 in quartz sand columns with and without macropores.

approximately 16.1%, whereas it was only 5.2% in the control (Table 1). These results suggested that a large portion of bromide rapidly moved through the O–O macropore. An early tracer breakthrough is a reliable indicator for preferential flow and transport (Koestel and Jorda, 2014; Knudby and Carrera, 2005). In the column with the O–C macropore, the effluent concentration of bromide was also significantly higher (p b 0.05) than was the control at the early stage (b1 PV), although the initial breakthrough of bromide just appeared a little earlier than the control. It was expected that bromide would have to pass through an 8 cm matrix below the macropore, which limited its breakthrough time. Knudby and Carrera (2005) also found the better the continuity of the vertical macropore, the shorter would be the migration time of a solute. On the contrary to the control column, the slope of the BTCs of bromide in columns with O–O and O–C tended to decrease after initial rapid rise in concentration, and such decrease of O–O column was more than that of O–C column (see pore volumes 1–2, Fig. 2a). The plateau concentration of bromide in the O–O and O–C macropore columns was 0.94 and 0.98, respectively, which was less than that of the control (0.99). Previous studies indicated that the breakthrough in columns with macropores was additionally controlled by horizontal transport of the tracer via diffusion, which leads to the delayed reaching of a maximal concentration in column effluent (Jorgensen et al., 1998; Allaire-Leung et al., 2000; Wang et al., 2013). In this study, the flattening of BTCs suggested the diffusion of bromide from the macropores into the quartz sand matrix. This effect is also visible from stronger tailing of the breakthrough curve in columns with macropores compared to the homogenous column after start of flushing the column with DDW (see pore volumes between 4 and 6, Fig. 2a). After the start of flushing there was a rapid drop in concentration followed by slowly decreasing concentrations of bromide due to continuing diffusion of bromide out

of the matrix and back into the macropores. Results indicated that the better the connectivity of macropore was, the stronger the diffusion effect was. 3.2. Transport behavior of nCeO2 in quartz sand Fig. 2b presents the representative plots of BTCs observed for nCeO2 in both the homogeneous column and in the quartz sand columns containing macropores. The relevant transport information is presented in Table 1. The breakthrough of nCeO2 in the homogeneous column was rapid and reached a plateau effluent concentration as high as 0.97, indicating the high mobility and slight deposition of nCeO2 in the quartz sand columns. The high interaction energy barrier of 160 kT between the nCeO2 and the quartz sand, without secondary minima, demonstrated an unfavorable attachment condition for nCeO2 on the quartz sand surface (Fig. 3). Under such unfavorable conditions for attachment, the nCeO2–grain interaction is the rate-limiting step for attachment (Ryan and Elimelech, 1996). Likewise, Li et al. (2011) found rapid and complete BTCs of nCeO2 at a low ionic strength of 1 and 2 mM, where the normalized effluent concentration rapidly reached 0.96. For the columns with macropores, the initial breakthrough of nCeO2 was at 0.15 PV and 0.46 PV for the O–O and the O–C column, respectively, which was much earlier than was that of the control at 0.73 PV. This result demonstrated the rapid movement of nCeO2 through the O–O

Table 1 Transport information of Br and nCeO2 in quartz sand column experiments.

Br

nCeO2

a

Structure

Average Darcy velocity (cm/h)

Plateau value (C/C0)

Recovery in initial effluent (%)a

Recovery in total effluent (%)

Control O–O O–C Control O–O O–C

6.10 6.10 6.10 6.10 6.10 6.10

0.99 0.94 0.98 0.97 0.94 0.96

5.2 16.1 7.2 5.2 17.4 8.4

99.0 94.1 98.1 95.2 94.1 93.5

Recovery in the first 1 PV effluent.

Fig. 3. DLVO derived interaction energy profiles for nCeO2-quartz sand and nCeO2–soil as a function of the separation distance in the collector system.

J. Fang et al. / Science of the Total Environment 543 (2016) 223–229

and O–C macropore via the preferential flow path. At the early stage (b 1 PV), the recovery rate of nCeO2 was approximately 17.4% and 8.4% for the O–O and the O–C columns, respectively, which was almost 3.3 and 1.6 times higher than was that of the control column (5.2%) (Table 1). This result indicated that the mobility of the nCeO2 was markedly enhanced by the macropores, especially the O–O continuous macropore. The continuous macropore invariably favored the faster movement of solute through the soil in contrast with the discontinuous macropore which created more disruptions to the flow and transport pathways than did the continuous macropore (Allaire-Leung et al., 2000; Zhou et al., 2013; Wang et al., 2013). Very similar to the BTCs of bromide, the slope of the BTCs of nCeO2 in columns with O–O and O–C also tended to decrease after initial rapid rise in concentration, and the strong tailing of the BTCs was observed after start of flushing the column with DDW (see pore volumes 1–2 and 4–6, Fig. 2b). This demonstrated that there was significant diffusion effect of nCeO2 between the macropores and the quartz sand matrix. In addition, the total recovery rate of nCeO2 was quite similar for both the homogeneous and the macropore columns (O–O and O–C), which indicated that final contribution of the effluent nCeO2 from the macropores was negligible compared with those from the matrix. This is easily attributable to the quite low retention of nCeO2 in the quartz sand matrix in this study. 3.3. Transport behavior of bromide in soil columns As shown in Fig. 4a, bromide acted as a good conservative tracer in soil, with a quick breakthrough and the plateau bromide concentration reaching almost one in the homogeneous (control) soil column. At the early stage (b 1 PV), the order of effluent bromide was O–O N O–C N control, with a recovery rate in the initial effluent of 7.5%, 6.3%, and 5.3%, respectively (Table 2). These results suggest that a portion of bromide rapidly moved through the O–O and O–C macropores. The slope of the BTCs of bromide in soil columns with O–O and O–C decreased after initial rapid rise in concentration, and the very strong tailing of the BTCs was observed, which was due to the significant diffusion of Br between the macropores and the soil matrix (see pore volumes 1–2 and 4–6, Fig. 4a). The diffusion effect in O–O soil column was also found to be stronger than that in O–C column. In addition, the recovery rate in the total effluent of bromide was much lower for the O–O and O–C macropore soil columns than was that of the control. This was attributed to the significant tailing of bromide in the macropore soil columns and the incomplete washing during the elution phase. 3.4. Transport behavior of nCeO2 in soil columns The breakthrough of soil colloids in the absence of nCeO2 was monitored as a blank control in this study. The absorbance (at 460 nm) of the

227

Table 2 Transport information of Br and nCeO2 in soil column experiments.

Br

nCeO2

a b

Structure

Average Darcy velocity (cm/h)

Plateau value (C/C0)

Recovery in initial effluent (%)

Recovery in total effluent (%)

Control O–O O–C Control O–O O–C

3.68 6.74 5.64 5.02 7.74 5.78

0.99 0.97 0.97 0.31 0.61 0.39

5.3a 7.5a 6.3a 7.6b 19.5b 10.8b

96.9 88.4 88.7 27.5 57.6 35.8

Recovery in the first 1 PV effluent. Recovery in the first 2 PV effluent.

outflow was 0.019–0.043. These results indicate that the amount of soil colloids migrating was limited and that their effect on the transport of nCeO2 was negligible. Fig. 4b shows the representative BTCs for nCeO2 in both the homogenous and the macropore soil columns. For the homogenous soil column, nCeO2 gradually flowed out after 0.5 PV and reached the plateau concentration of 0.31 at 1.5 PV, with a total recovery in the effluent of 27.5%. Breakthrough time of nCeO2 demonstrated that the mobility of nCeO2 in the soil was significantly weaker than was the mobility in the quartz sand columns. As calculated by DLVO, there was an energy barrier of 50.2 kT between nCeO2 and the soil surface (Fig. 3), likewise suggesting an unfavorable deposition condition for nCeO2. However, compared with quartz sand, soil is a significantly more complicated porous medium. Soil has a greater heterogeneous surface charge and smaller particle size, larger surface area, more irregular and rougher shape, and different types of minerals and organic matters. These factors could have a significant effect on the transport of nanoparticles. It was found that the retention/deposition of nanoparticles was invariably higher in the soil than it was in the quartz sand under the same solution chemistry (Quevedo and Tufenkji, 2012; Zhang et al., 2012; Sun et al., 2015). Straining takes place through hetero-aggregation with soil impurities, such as clay minerals and metal oxides (Cornelis et al., 2013). In addition, surface charge heterogeneities could have contributed to the observed lower nCeO2 transport in the soil. As has been recognized, various metal oxide impurities (e.g. Al2O3, Fe2O3) on a soil surface acted as a source of positive surface charge heterogeneity, providing favorable deposition sites for negatively charged particles (Tufenkji and Elimelech, 2005). For the macropore soil columns, significant two-pulse effluent of nCeO2 was found, especially at the O–O system. The first pulse (peak at 1.5 PVs) was associated with the transport through the macropore, whereas the second pulse (N2 PVs) was mainly from the matrix. Two-

Fig. 4. Representative breakthrough curves of (a) Br and (b) nCeO2 in soil columns with and without macropores.

228

J. Fang et al. / Science of the Total Environment 543 (2016) 223–229

pulse transport was always found in the porous media with the macropores (Li and Masoud, 1997; Wang et al., 2013, 2014). The first breakthrough time of nCeO2 in the O–O system was much earlier than was that of the control. The effluent recovery rate of nCeO2 at the early stage (b2 PV) was 17.6%, which was much higher than was that of the control at 7.6%. The value of plateau concentration and the total effluent recovery rate reached 0.61 and 57.6% for the soil columns with the O–O macropore, respectively, which was almost twice as high as were those of the control (0.31 and 27.5%, respectively). These results demonstrated the rapid movement nCeO2 through preferential flow path of the vertical continuous macropore (O–O) which accounted for only 1% of the total pore volume had doubly increased the total mobility of nCeO2 in soil. Likewise for the O–C soil columns, the effluent concentration of nCeO2 was significantly higher (p b 0.05) than was the control during the entire leaching process, although the initial breakthrough of nCeO2 did not appear significantly earlier compared with the control. As mentioned before, nCeO2 had to be transported through the soil matrix below the macropore, which limited its breakthrough time. The value of plateau concentration and the total effluent recovery rate was 0.39 and 35.8% for the soil columns with the O–C macropore, respectively, indicating that it was increased by 30% with the vertical discontinuous macropore in soil which accounted for 0.5% of the total pore volume. In addition, similarly to Br, visible strong tailing of BTCs of nCeO2 in soil columns with macropores were also obtained after start of flushing the column with DDW due to the diffusion of nCeO2 from the soil matrix back into the macropores (see pore volumes between 4 and 6, Fig. 4b). Numerous studies have indicated there were three basic requirements to inducing significant preferential flow in porous media, namely (Li and Masoud, 1994; Allaire-Leung et al., 2000; Edvina et al., 2009): i) macropores, ii) different velocities of water and solute in porous media, and iii) the controlling role of the macropore domain in the process of solute migration. In this study, the vertical macropores in the soil had significantly reduced the breakthrough time (O–O macropore), increased the effluent plateau concentration, and enhanced the total effluent recovery rate of nCeO2 compared with those of the control. However, the O–O macropores in the quartz sand did not increase the effluent plateau concentration and the total effluent recovery rate of nCeO2. It demonstrated that the movement of nCeO2 through the macropores had made a greater contribution to the mobility in the soil compared with that in the quartz sand. Previous studies have reported that the influence of the macropore was generally greater on sorption solutes than it was on non-adsorption solutes (Allaire-Leung et al., 2000). Our study confirmed that in the soil the macropores played a more important role in the mobility of the nCeO2 compared with their role in the quartz sand. In other words, it was found that the effect of preferential flow on the transport of nCeO2 was enhanced in more favorable deposition porous media. 4. Conclusions In particular, compared with the homogeneous column, the mobility of nCeO2 was significantly enhanced by the vertical macropores in both the quartz sand and the soil. Additionally, this enhancement was more pronounced in the continuous macropore (O–O system) than it was in the discontinuous macropore (O–C system). In the quartz sand, the initial breakthrough of nCeO2 in both the O–O and the O–C macropores occurred much earlier than did that of the control, demonstrated by the rapid movement of nCeO2 through the macropore via the preferential flow path. However, the value of effluent plateau concentration and the total effluent recovery rate of nCeO2 were not increased by the macropores in the quartz sand. In the soil, significant two-pulse effluent of nCeO2 was found in both the O–O and the O–C columns, indicating a large portion of nCeO2 moving through the macropores. Both the O–O and the O–C macropores reduced the nCeO2 breakthrough time and favored a higher plateau concentration. The O–O macropore which

accounted for only 1% of the total pore volume had increased the total mobility of nCeO2 by 2 times in soil, even it was increased by 30% with the O–C macropore which accounted for 0.5% of the total pore volume. Although the present study highlighted the influence of the O–O and O– C macropores on nCeO2 transport, our knowledge on assessing the transport routes of nanoparticles in heterogeneous porous media remains limited. Further studies are needed to investigate the influence of macropore tortuosity, the neighboring macropores and the solution chemistry on the nanoparticle transport, which could help to improve our understanding of natural environments. Acknowledgments This work was funded by the 973 program of China (No 2014CB441104), and the National Natural Science Foundation of China (No 21007057). References Allaire-Leung, S.E., Gupta, S.C., Moncrief, J.F., 2000. Water and solute movement in soil as influenced by macropore characterstic 1. Macropore continuity. J. Contam. Hydrol. 41 (3), 283–301. Bergstrom, L., 1997. Hamaker constants of inorganic materials. Adv. Colloid Interf. Sci. 70, 125–169. Cassee, F.R., Van Balen, E.C., Singh, C., Green, D., Muijser, H., Weinstein, J., Dreher, K., 2011. Exposure, health and ecological effects review of engineered nanoscale cerium and cerium oxide associated with its use as a fuel additive. Crit. Rev. Toxicol. 41 (3), 213–229. Cornelis, G., Ryan, B., McLaughlin, M.J., Kirby, J.K., Beak, D., Chittleborough, D., 2011. Solubility and batch retention of CeO2 nanoparticles in soils. Environ. Sci. Technol. 45 (7), 2777–2782. Cornelis, G., Pang, L., Doolette, C., Kirby, J.K., McLaughlin, M.J., 2013. Transport of silver nanoparticles in saturated columns of natural soils. Sci. Total Environ. 463-464, 120–130. Czapar, G.F., Horton, R., Fawcett, R.S., 1992. Herbicide and tracer movement in soil columns containing an artificial macropore. J. Environ. Qual. 21 (1), 110–115. Derjaguin, B., Landau, L., 1941. Theory of the stability of strongly charged lyophobic sols and of the adhesion of strongly charged particles in solutions of electrolytes. URSS. Acta Physicochim. 14, 733–762. Edvina, L., Laurent, L., Beatrice, B., Herve, A., 2009. Modeling the influence of an artificial macropore in sandy columns on flow and solute transfer. J. Hydrol. 376 (4), 392–402. Elimelech, M., Gregory, J., Jia, X., Williams, R.A., 1995. Particle Deposition and Aggregation: Measurement, Modeling and Simulation. Butterworth Heinemann, Oxford, U.K. Garrido, F., Serrano, S., Barrios, L., Uruñuela, J., Helmhart, M., 2014. Preferential flow and metal distribution in a contaminated alluvial soil from São Domingos mine (Portugal). Geoderma 213, 103–114. Gottschalk, F., Sun, T.Y., Nowack, B., 2013. Environmental concentrations of engineered nanomaterials: review of modeling and analytical studies. Environ. Pollut. 181, 287–300. Gregory, J., 1981. Approximate expressions for retarded van der Waals interaction. J. Colloid Interface Sci. 83 (1), 138–145. Hassan, A.A., Zhen, L., Sahle-Demessie, E., Sorial, G.A., 2013. Computational fluid dynamics simulation of transport and retention of nanoparticles in saturated sand filters. J. Hazard. Mater. 244-245, 251–258. Health Effects Institute (HEI), 2001. Evaluation of human health risk from cerium added to diesel fuel (March 28, 2008); http://pubs.healtheffects.org/view.php?id=172. Hogg, R., Healy, T.W., Ferstenau, D.W., 1966. Mutual coagulation of colloidal dispersions. Trans. Faraday Soc. 62, 1638–1651. Hu, Q.H., Brusseau, M.L., 1995. Effect of solute size on transport in structured porous media. Water Resour. Res. 31 (7), 1637–1646. Israelachvili, J.N., 1992. Intermolecular and Surface Forces. 2nd Ed. Academic Press, London. Jassogne, L., Hettiarachchi, G., Ann McNeill, D.C., 2009. Distribution and speciation of nutrient elements around micropores. Soil Sci. Soc. Am. J. 73 (4), 1319–1326. Johnston, B.D., Scown, T.M., Moger, J., Cumberland, S.A., Baalousha, M., Linge, K., van Aerle, R., Jarvis, K., Lead, J.R., Tyler, C.R., 2010. Bioavailability of nanoscale metal oxides TiO2, CeO2, and ZnO to fish. Environ. Sci. Technol. 44 (3), 1144–1151. Jorgensen, P.R., Mckay, L.D., Spliid, N.H., 1998. Evaluation of chloride and pesticide transport in a fractured clayey till using large undisturbed columns and numerical modeling. Water Resour. Res. 34 (4), 539–553. Kanel, S.R., Greneche, J.M., Choi, H., 2006. Arsenic(V) removal from groundwater using nano scale zero-valent iron as a colloidal reactive barrier material. Environ. Sci. Technol. 40, 2045–2050. Karen, V.H., Joris, T.K.Q., Joanna, M.B., Karel, A.C., De, S., Andreas, E., Paul, V.M., Clifford, B., George, M., Vyvyan H., C., Dik, V.D.M., Konrad, R., Kenneth, A.D., Anna, S., Anna, L., Iseult, L., Geert, S., Björn, D.S., Laszlo, V., Colin, R.J., 2009. Fate and effects of CeO2 nanoparticles in aquatic ecotoxicity tests. Environ. Sci. Technol. 43 (12), 4537–4546. Knudby, C., Carrera, J., 2005. On the relationship between indicators of geostatistical, flow and transport connectivity. Adv. Water Resour. 28, 405–421. Koestel, J., Jorda, H., 2014. What determines the strength of preferential transport in undisturbed soil under steady-state flow? Geoderma 217-218, 144–160.

J. Fang et al. / Science of the Total Environment 543 (2016) 223–229 Konert, M., Vandenberghe, J., 1997. Comparison of laser grain-size analysis with pipette and sieve analysis: a solution for the underestimation of the clay fraction. Sedimentology 44, 523–535. Leung, Y.H., Yung, M.N., Ng, A.M.C., Ma, A.P.Y., Wong, S.W.Y., Chan, C.M.N., Ng, Y.H., Djurisic, A.B., Guo, M., Wong, M.T., Leung, F.C.C., Chan, W.K., Leung, K.M.Y., Lee, H.K., 2015. Toxicity of CeO2 nanoparticles — the effect of nanoparticle properties. J. Photochem. Photobiol. B Biol. 145, 48–59. Li, M.Y., Masoud, G., 1994. Preferential transport of nitrate through soil columns containing root channels. Soil Sci. Soc. Am. J. 58 (3), 653–659. Li, M.Y., Masoud, G., 1997. Preferential transport of solute through soil columns containing constructed macropores. Soil Sci. Soc. Am. J. 61, 1308–1317. Li, Z., Sahle-Demessie, E., Hassan, A.A., Sorial, G.A., 2011. Transport and deposition of CeO2 nanoparticles in water saturated porous media. Water Res. 45 (15), 4409–4418. Lin, W., Huang, Y., Zhou, X., Ma, Y., 2006. Toxicity of cerium oxide nanoparticles in human lung cancer cells. Int. J. Toxicol. 25 (6), 451–457. Liu, X., Chen, G., Su, C., 2012. Influence of collector surface composition and water chemistry on the deposition of cerium dioxide nanoparticles: QCM-D and column experiment approaches. Environ. Sci. Technol. 46 (12), 6681–6688. Lv, X.Y., Gao, B., Sun, Y.Y., Shi, X.Q., Xu, H.X., Wu, J.C., 2014. Effects of humic acid and solution chemistry on the retention and transport of cerium dioxide nanoparticles in saturated porous media. Water Air And Soil Pollution 22 (10), 1–9. Majumdar, S., Peralta-Videa, J.R., Bandyopadhyay, S., Castillo-Michel, H., HernandezViezcas, J.A., Sahi, S., Gardea-Torresdey, J.L., 2014. Exposure of cerium oxide nanoparticles to kidney bean shows disturbance in the plant defense mechanisms. J. Hazard. Mater. 278, 279–287. Mishurov, M., Yakirevich, A., Weisbrod, N., 2008. Colloid transport in a heterogeneous partially saturated sand column. Environ. Sci. Technol. 42 (4), 1066–1071. Nelson, D.W., Sommers, L.E., 1982. Total carbon, organic carbon, and organic matter. Part 2. In: Page, A.L. (Ed.), Methods of Soil Analysis, second ed. ASA and SSSA, Madison, WI pp, pp. 539–579. Patil, M.D., Das, B.S., 2013. Assessing the effect of puddling on preferential flow processes through under bund area of lowland rice field. Soil Tillage Res. 134, 61–71. 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. Environ. Sci. Technol. 44 (17), 6532–6549.

229

Petosa, A.R., Ohl, 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 Res. 47 (15), 5889–5900. Quevedo, I.R., Tufenkji, N., 2012. Mobility of functionalized quantum dots and a model polystyrene nanoparticle in saturated quartz sand and loamy sand. Environ. Sci. Technol. 46 (8), 4449–4457. Ryan, J.N., Elimelech, M., 1996. Colloid mobilization and transport in groundwater. Colloids and Surface A 107, 1–56. Sheng, F., Liu, H.H., Wang, K., Zhang, R.D., Tang, Z.H., 2014. Investigation into preferential flow in natural unsaturated soils with field multiple-tracer infiltration experiments and the active region model. J. Hydrol. 508, 137–146. Sun, P.D., Shijirbaatar, A., Fang, J., Owens, G., Lin, D.H., Zhang, K.K., 2015. Distinguishable transport behavior of zinc oxide nanoparticles in silica sand and soil columns. Sci. Total Environ. 505, 189–198. 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. Ullum, M., 2001. Effects of Water Content and Soil Structure on Colloid Transport in Porous Media. Ph. D dissertation, Technical Unversity of Denmark, Copenhagen. Verwey, E.J.M., Overbeek, J.T.G., 1948. Theory of the Stability of Lyophobic Colloids. The Netherlands, Amsterdam. Wang, Y.S., Bradford, S.A., Simunek, J., 2013. Transport and fate of microorganisms in soils with preferential flow under different solution chemistry conditions. Water Resour. Res. 49, 2424–2436. Wang, Y.S., Bradford, S.A., Simunek, J., 2014. Physicochemical factors influencing the preferential transport of Escherichia coli in soils. Vadose Zone J. 13 (1), 1–10. http://dx.doi. org/10.2136/vzj2013.07.0120. Zhang, L., Hou, L., Wang, L., Kan, A.T., Chen, W., Tomson, M.B., 2012. Transport of fullerene nanoparticles (nC60) in saturated sand and sandy soil: controlling factors and modeling. Environ. Sci. Technol. 46 (13), 7230–7238. Zhou, B.B., Li, Y., Wang, Q.J., Jiang, Y.L., Li, S., 2013. Preferential water and solute transport through sandy soil containing artificial macropores. Environmental Earth Science 70, 2371–2379.