- Email: [email protected]
Transport of silver nanoparticles (AgNPs) in soil
Chemosphere 88 (2012) 670–675
Contents lists available at SciVerse ScienceDirect
Chemosphere journal homepage: www.elsevier.com/locate/chemosphere
Transport of silver nanoparticles (AgNPs) in soil Omer Sagee, Ishai Dror ⇑, Brian Berkowitz Dept. of Environmental Sciences and Energy Research, Weizmann Institute of Science, Rehovot 76100, Israel
a r t i c l e
i n f o
Article history: Received 19 December 2011 Received in revised form 12 March 2012 Accepted 17 March 2012 Available online 18 April 2012 Keywords: Silver nanoparticles Transport Soil Chlorides Humic acid
a b s t r a c t The effect of soil properties on the transport of silver nanoparticles (AgNPs) was studied in a set of laboratory column experiments, using different combinations of size fractions of a Mediterranean sandy clay soil. The AgNPs with average size of 30 nm yielded a stable suspension in water with zeta potential of 39 mV. Early breakthrough of AgNPs in soil was observed in column transport experiments. AgNPs were found to have high mobility in soil with outlet relative concentrations ranging from 30% to 70%, depending on experimental conditions. AgNP mobility through the column decreased when the fraction of smaller soil aggregates was larger. The early breakthrough pattern was not observed for AgNPs in pure quartz columns nor for bromide tracer in soil columns, suggesting that early breakthrough is related to the nature of AgNP transport in natural soils. Micro-CT and image analysis used to investigate structural features of the soil, suggest that soil aggregate size strongly affects AgNP transport in natural soil. The retention of AgNPs in the soil column was reduced when humic acid was added to the leaching solution, while a lower flow rate (Darcy velocity of 0.17 cm/min versus 0.66 cm/min) resulted in higher retention of AgNPs in the soil. When soil residual chloride was exchanged by nitrate prior to column experiments, significantly improved mobility of AgNPs was observed in the soil column. These findings point to the importance of AgNP–soil chemical interactions as a retention mechanism, and demonstrate the need to employ natural soils rather than glass beads or quartz in representative experimental investigations. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction The manufacturing and use of engineered nanoparticles (ENPs) in many commercial products has increased dramatically in recent years, together with contamination potential (Colvin, 2003; Dreher, 2004; Wiesner et al., 2006; Christian et al., 2008). After use and during production, ENPs can find their way into the soil environment; once in the soil, ENPs might be transported toward groundwater (Tian et al., 2010). Therefore, understanding ENP fate and transport in soil is of considerable importance. To date, most studies related to transport of ENPs have employed well-defined, artificial porous materials such as glass beads or acidwashed, homogeneous quartz sand in column experiments (e.g. Dunphy Guzman et al., 2006; Kanel et al., 2007; Loux and Savage, 2008; Shani et al., 2008; Baalousha, 2009; Jeong and Kim, 2009; Liu et al., 2009; Ben-Moshe et al., 2010; Shang et al., 2010; Uyusur et al., 2010). However, many soil properties such as grain size distributions, heterogeneity, and the presence of soil organic matter (SOM) are not represented in glass beads or quartz. Therefore, the relevance of these studies for transport of ENPs in the natural soil environment is in question. For example, the presence of absorbed metals that are widespread in soil minerals, such as copper and lead, ⇑ Corresponding author. Tel.: +972 8 9344230; fax: +972 8 9344124. E-mail address: [email protected] (I. Dror). 0045-6535/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.chemosphere.2012.03.055
can promote ENP aggregation (Kretzschmar and Sticher, 1997), while the presence of humic substances and dissolved organic matter can contribute to ENP suspension stability (Fang et al., 2009). It was further shown that the heterogeneous structure of the pore space in natural soils can lead to development of preferential flow paths for colloidal material, which may have a dramatic impact on transport characteristics (Sirivithayapakorn and Keller, 2003). In general, two main mechanisms can limit the transport of well-dispersed ENPs in soil: (1) straining, wherein nanoparticles are removed from solution in smaller pore spaces, and (2) physical–chemical processes which remove nanoparticles from solution by interactions between ENPs and the medium (McDowell-Boyer et al., 1986). Both mechanisms are expected to be enhanced when flow rates are lower and particles are smaller, so that the influence of diffusion is increased (Lecoanet and Wiesner, 2004). To the best of our knowledge, only a few previous studies have examined ENP transport in natural soils. Fang et al. (2009) demonstrated the retarding effect of clay content and the enhanced mobility effect of SOM on TiO2 nanoparticle transport in soil. Two other studies demonstrated that while some nanoparticles such as single-walled carbon nanotubes are practically immobilized in natural sandy soil, other nanoparticles such as surface-modified palladized iron nanoparticles migrate freely in soil (He et al., 2006; Jaisi and Elimelech, 2009). However, a systematic examination of the effect of soil properties on transport was not conducted.
O. Sagee et al. / Chemosphere 88 (2012) 670–675
The purpose of this study was to investigate the transport of silver nanoparticles (AgNPs) in natural soil. Silver nanoparticles (AgNPs) are among the most commonly produced ENPs, with many applications in clothing, cosmetics, and medical care products. AgNPs have known antibacterial properties, which makes AgNPs toxic to microorganisms in aquatic environments (Navarro et al., 2008; Tian et al., 2010; He et al., 2011). In particular, the investigation aimed to determine the effect of soil properties such as mechanical structure and presence of dissolved organic matter on the transport of AgNPs, with emphasis on understanding the relative importance of mechanical straining and physical-chemical processes of retention in transport of AgNPs in natural soil. The AgNPs used in these experiments (citrate coated AgNPs) have high negative f potential value, resulting in strong electrostatic repulsion from each other and a low tendency to aggregate. Because most natural soil minerals have negative f potential, resulting in a high-energy barrier for the attachment of AgNPs (Christian et al., 2008; Theng and Yuan, 2008; Lin et al., 2011), it was expected that the AgNPs would demonstrate relatively high mobility. 2. Experimental section 2.1. Characterization of silver nanoparticles (AgNPs) AgNPs were synthesized using the method of reduction by citrate (Dong et al., 2009), which allows control of the shape and size of AgNPs by the choice of synthesis solution pH. AgNPs were synthesized at pH 11.5 to achieve a diameter of approximately 30 nm. Suspensions from several parallel synthesis reactions were transferred to a 4 L container and kept sealed under constant stirring. This AgNP stock suspension was used in all column experiments to ensure uniform addition of AgNP suspension. Stability of the stock suspension was verified weekly using dynamic light scattering (DLS) on a Viscotek 802. Prior to each column experiment (on the same day), AgNPs were transferred to a ‘‘modified artificial rainwater’’ solution (MARW). The MARW solution contained 0.048, 0.057 and 0.035 g L 1 of sodium nitrate, potassium bicarbonate and magnesium sulfate, respectively. This composition is a modification of the artificial rainwater solution which represents the typical rainwater composition in the Negev desert in Israel (Zvikelsky et al., 2008). Two modifications to the artificial rainwater solution were made. First, calcium cations, observed in preliminary laboratory experiments to interfere with AgNP stability, were replaced with magnesium and potassium. Second, chloride anions, which cause production of silver chloride and in turn lead to its precipitation, were replaced with nitrate and sulfate. Both ion effects were also described in the literature to reduce suspension stability (e.g. El Badawy et al., 2010; Huynh and Chen, 2011) and thus interfere with elemental silver transport and ICP-MS analysis. Before each column experiment, AgNPs were transferred from the synthesis solution to the MARW solution done by centrifuging 50 mL AgNP solution and replacing 45 mL of the supernatant with MARW twice. This was done also to minimize the presence of aqueous silver ions in the solution. The solution stability of the AgNPs in the MARW was monitored by transmission electron microscopy (TEM) and DLS to confirm no change in size distribution and shape of the particles over time. Zeta potential was measured with a zetasizer nano instrument (Malvern). 2.2. Soil treatments Soil was sampled from the top 50 cm of an experimental vineyard in the Bet-Dagan area, central Israel. Two different size fractions of the soil were separated using a set of sieves; the
671
250–1000 lm diameter fraction and the <250 lm diameter fraction. Soil grains larger than 1000 lm were not used in the column experiments. From the two size fractions, three different soil mixtures were made; A – containing only the 250–1000 lm fraction, B – containing 70% (weight) of the 250–1000 lm fraction and 30% of the <250 lm fraction, and C – containing equal weights of the two fractions. 2.3. Soil properties analysis Key soil properties of the A, B, and C mixtures were determined with various soil analysis methods and are presented in Table S1 in the supplementary data (SD). Clay, silt, and sand content were measured using the hydrometer method based on Stoke’s law (Bouyoucos, 1962). Soil organic matter (SOM) content was determined by weight lost measurement after burning at 400 °C overnight (Schollenberger, 1945). Cation exchange capacity (CEC) was determined by means of ammonium acetate extraction (Bache, 1976), and specific surface area (SSA) was established by the ethylene glycol monoethyl ether (EGME) mono-layer adsorption method (Carter et al., 1965). Porosity was measured prior to each column experiment by slow saturation (from below) of the soil column with MARW solution. 2.4. X-ray microtomography (micro-CT) and image analysis X-ray microtomography (micro-CT) was used to determine structural differences among soil mixtures A, B, and C. About 2 g from each mixture was inserted into a 10 mm (i.d.) plastic holder for X-ray micro-CT imaging in a Micro XCT-400 (Xradia, California, USA) instrument. The voltage applied on the X-ray source was 40 kV and the current was 200 lA. No source filters were used. For each sample, 1500 projection images were recorded with an exposure time of 5 s and a magnification objective of 4. The final pixel size in the projection images was 4.6 lm. The volume was reconstructed with the instrument software and was then exported to Avizo Fire (VSG international) for further 3D image analysis. After de-noising and binarizing the images, individual soil aggregates could be distinguished. Example images and details of the image analysis process appear in the Supplementary data (Fig. S1). The volume of each aggregate was determined by the software and a size distribution histogram was constructed for each sample. This process was repeated three times for each of the mixtures A, B, and C. 2.5. AgNP quantification To analytically separate AgNPs from the soil colloids eluting from the column, a method for quantification of the AgNPs in their dissolved state was developed. AgNPs were dissolved in nitric acid (40.5% by weight) followed by dilution to 3.88% nitric acid and filtration by 0.22 lm cellulose acetate filter (Millipore); a correlation with R2 > 0.999 was found for calibration of the AgNP suspension by inductively coupled plasma mass spectrometry (ICP-MS) (ELAN 9000, Perkin Elmer Inc.). This method was later used to determine AgNP concentrations in column experiments. 2.6. AgNPs in soil column transport experiments Column experiments were conducted in closed, cylindrical, vertical Plexiglas columns, with length of 10 cm and inner diameter of 3.1 cm. A peristaltic pump was used to apply MARW or AgNP suspensions, at Darcy velocity of 0.66 cm min 1. Slight variations in flow rate were registered between column experiments, and taken into account in pore volume calculations. After packing the columns with the soil mixtures, the columns were slowly saturated
O. Sagee et al. / Chemosphere 88 (2012) 670–675
with MARW and porosity was determined. The columns were then washed with ten pore volumes (PV) of MARW at flow rate of 5 mL min 1. At that point, the experiment began with application of a four PV pulse of AgNP suspension diluted 1:5 by MARW. This pulse was followed by flowing four PV of MARW. The outflow solution was collected by fraction collector. After the experiments, nitric acid (69%) was added to the samples to achieve a final acid concentration of 40.5%, and the samples were stirred gently. After 4 h, samples were diluted in double distilled water (DDW) and the solution was filtered by 0.22 lm cellulose acetate filter, prior to ICP-MS analysis. Samples of the inflow solution of each column experiment were also collected and treated as described above to determine initial concentration. A calibration curve was constructed for each individual experiment to determine the C/ C0 value of each sample, for the construction of a breakthrough curves. Each soil treatment (A, B, and C) was run in triplicates. Before each experiment, the columns were opened and repacked. Fraction B of the soil was used to study the effect of three parameters on AgNP transport: flow velocity, presence of dissolved humic acid, and presence of residual chlorides in soil. In general, the experimental procedure described above was followed in all experiments; adjustments to the specific experiments are described in the following paragraph. To examine the effect of flow velocity, a set of column experiments was conducted with Darcy velocity of 0.17 cm min 1 instead of 0.66 cm min 1. The effect of dissolved humic acid on AgNP transport was studied by adding 100 ppm humic acid (Sigma-Aldrich) to the MARW and AgNP solution. The removal of residual chlorides from soil, and its replacement by nitrate, was achieved by pretreating the packed columns with potassium nitrate, slowly flowing 20 PV of 0.5 M potassium nitrate solution through them. This was done to test the hypothesis that a chemical reaction between soil residual chloride and AgNPs results in AgNP retention is soil. For this set of experiments, a Darcy velocity of 0.17 cm min 1 was applied. 2.7. AgNP transport in pure quartz sand The same procedure as for AgNP transport in soil columns was performed for a column filled with clean quartz sand. A breakthrough experiment was conducted in a column filled with homogeneous quartz sand (Accusand), with grain diameter of 210– 300 lm. Prior to packing the column, the quartz was soaked overnight in concentrated hydrochloric acid, thoroughly rinsed with DDW and heated overnight in a 400 °C oven. 2.8. Bromide tracer soil column transport experiments Bromide was used as a soluble conservative tracer in a column experiment for each of the soil mixtures A, B, and C. After saturation and washing of the columns with MARW, a four PV pulse of 500 ppm bromide (as KBr) was injected, followed by a four PV MARW pulse. Bromide concentration in the outflow samples was determined using high pressure liquid chromatography (HPLC) equipped with Alltech conductivity detector (model 650). Anions were separated with an Alltech anion 7 lm column, 100 4.6 mm. The mobile phase solution for the bromide detection was 4 mM p-Hydrobenzoic acid (pH adjusted to 7.5 with LiOH) and a flow rate of 1 mL min 1 was applied. 3. Results and discussion 3.1. Properties of silver nanoparticles (AgNPs) TEM images of the synthesized AgNPs are presented in Fig. S1 in the supplementary data. The images illustrate that most of the
synthesized AgNPs were relatively round with a diameter of approximately 40 nm; in addition, a few rod-shaped AgNPs were detected. These results were confirmed by size distribution analysis measured using DLS. The f potential for AgNPs in the MARW solution was determined to be 39 mV, suggesting that they form a stable suspension. 3.2. Soil characteristics Some basic soil properties of the three mixtures A, B, C and the <250 lm size fraction are presented in the Table S1 in the supplementary data. It is apparent that many of the soil properties do not vary significantly among the different mixtures. Specific surface area (SSA), for example, which is a measure of the surface area per unit weight of the soil solid phase, varies only in the range of 94.7–96.4 m2 g 1. This is counterintuitive given that SSA is dependent on grain size and is expected to change when size-based separation is used. However, this finding is explained by the fact that SSA determination involves breakdown of soil aggregates. Other properties show a similar trend of small variations throughout the different mixtures. When the mechanical composition of the different mixtures was determined, the sand content ranged between 62.5% and 70%. The highest sand content was found in the smallest size fraction (<250 lm). Clay content was 17.5% for soil mixture A and decreased to a value of 12.5% for the < 250 lm fraction. Therefore, while soil mixture A contains larger aggregates, the aggregates themselves tend to contain more clay. To summarize, differences among the measured properties of soil mixtures A, B, and C are assumed to be negligible when considering their possible effect on AgNP transport in soil. 3.3. X-ray microtomography (micro-CT) and image analysis Aggregate size distribution histograms for each of three soil mixture are shown in Fig. 1. Assuming spherical aggregates, the volume of each aggregate was converted into diameter. The histograms represent the total volume occupied by soil aggregates of the specified diameter. The volume units in all three histograms are normalized to the same scale. The histograms show a trend in aggregate size distribution. In all three mixtures the dominant aggregate population is in the 270–300 lm diameter range. However, mixture A contains more aggregates with a larger diameter, relative to mixtures B and C. The relative volume of aggregates with diameter smaller than 270 lm for A, B, and C are 26.1%, 45.4%, and 49.9%, respectively. Therefore, a significant structural difference between the soil mixtures is evident, in spite of the relative similarity in chemical properties. The differences in aggregate size distribution can affect AgNP transport in the soil in two possible ways: (1) soil fractions
1
Normalized Volume
672
A B C
0.8 0.6 0.4 0.2 0 0
75
165
255
345
435
525
615
705
Grain Diameter (µm) Fig. 1. Aggregate size distribution histogram for soil mixtures A (soil fraction of 1000–250 lm), B (70% (wt.) of the 250–1000 lm and 30% of the <250 lm diameter fraction), and C (containing equal weights of the two fractions).
673
O. Sagee et al. / Chemosphere 88 (2012) 670–675
containing more aggregates with smaller diameter have also smaller pores, which in turn increase mechanical straining of AgNP, and (2) larger surface area of the smaller aggregates increases the available surface area for interactions between AgNPs and aggregates. 3.4. Column experiments In most column experiments conducted in this study, a fast rise in AgNP outlet concentration was first observed; a steady plateau value of elution concentration was then reached and maintained throughout application period of the AgNP pulse. When the MARW solution was re-applied, a fast drop to zero in the AgNP outlet concentration was observed. Outlet concentrations of AgNP in the plateau stage for all of the experiments, expressed as C/C0, are given in Table S2 in the supplementary data. Note that no plateau was observed in the reduced flow speed experiments, where soil was not pretreated with potassium nitrate. The results of one representative repetition (in which plateau values were closest to that of the treatment average) for column experiments at high flow velocities (0.66 cm/min), for soil mixtures A, B, and C, are presented in Fig. 2. AgNP retention is observed to increase with decreasing aggregate size. The C/C0 average and standard deviation values for AgNPs at the column outlet for soil mixtures A, B, and C were 0.697 ± 0.041, 0.464 ± 0.035 and 0.325 ± 0.034, respectively. Mechanical straining is the simplest explanation for the increased AgNP retention in the order of A > B > C. However, in spite of the chemical similarities found among the three mixtures, chemical interaction can also account for retention in this order. This explanation is possible when considering that a reaction between AgNPs and soil occurs at the aggregate surface, and the aggregate surface area increases for the smaller aggregates. One such possible reaction is between soil residual chloride and AgNPs, as silver chloride is a readily formed precipitate, or coats the particles which in turn can change the nanoparticle suspension stability (Li et al., 2010; Huynh and Chen, 2011; Prathna et al., 2011). Theoretically, the outlet value of a conservative tracer in a soil column should reach C/C0 0.5 after one PV, assuming the tracer travels at the same velocity as the background solution. As seen in Fig. 2, AgNPs in all three soil mixtures show an earlier breakthrough behavior. The concentration of AgNPs at the column outlet almost reaches its plateau value as early as one PV after introduction of the AgNPs, suggesting that AgNPs travel faster in the soil column than the background solution. This behavior was repeated in all AgNP soil transport experiments conducted in this study.
Two possible, complementary, explanations can be offered for this phenomenon: (1) development of preferential flow pathways in the soil column, related to the effect of soil heterogeneity and the negative charge of the particles (f potential = 39 mV), and (2) the existence of ‘‘dead pore volume’’ inside soil aggregates. In the first case, if AgNPs travel only through the larger pores where hydraulic conductivity is higher, the result is partial sampling of the pore space by the AgNPs and early breakthrough. In the second case, while soil aggregates contain solution in the inner aggregate pores, they are impenetrable to AgNPs. In this case, the volume in which AgNPs travel is smaller than the total pore volume. To examine the early breakthrough behavior, transport experiments with bromide as a conservative tracer were conducted. Furthermore, to determine if the early breakthrough phenomenon is unique to transport of AgNPs in natural soils, a transport experiment of AgNPs in a pure quartz sand column was conducted (Fig. 3). Bromide advanced in all soil columns as a conservative tracer. No clear differences in the transport properties of bromide through soil mixtures A, B and C were found, possibly because the relatively high flow rates range in the experiments masked the effect of diffusion. Moreover, it is seen that AgNPs behaved as a conservative tracer when transported through the pure quartz sand column; again, early breakthrough was not observed. These results confirm that early breakthrough is related to the properties of the porous medium and the solute, and suggest that this phenomenon is related to transport of AgNPs in natural soils. Fig. 4 shows the breakthrough curves of one representative repetition (from triplicate runs) from flow velocities 0.66 and 0.17 cm min 1 in soil fraction B. It is evident that AgNP retention in the soil column increases when flow velocity is reduced. The plateau C/C0 value for AgNPs in fraction B in the higher flow velocity was 0.464 ± 0.025, while the highest C/C0 value observed for the reduced flow velocity was smaller than 0.3. The breakthrough curve changed significantly for the reduced flow velocity. Instead of reaching a steady plateau value, as observed in all other column experiments, a declining trend is seen. In other words, the highest AgNP outlet concentrations were obtained approximately one PV after introducing the AgNPs to the column, followed by a linear decrease from C/C0 0.3 to C/C0 0.2 over four PV. Such behavior can be attributed to accumulation and consequent blocking of pore throats by AgNPs, which gradually increase AgNP filtration in the soil. This behavior was not observed for the higher flow velocities, possibly because the flow was fast enough to constantly mobilize and remove AgNPs in pore throats. The increased retention in the slower flow rate can be explained by increased mechanical straining as a result of more significant diffusion, which leads to trapping of nanoparticles not only in
1.0 1.0
A - coarse 0.8
B- medium
Bromide in A
0.8
Bromide in B
0.6
End of pulse
C/C 0
C/C 0
C- fine
0.4 0.2
Bromide in C
0.6
AgNPs in quartz
0.4
End of pulse
0.2 0.0
0.0 0
1
2
3
4
5
6
7
Pore Volumes Fig. 2. Breakthrough curves for AgNPs in different soil mixtures A (soil fraction of 250–1000 lm), B (70% (wt.) of the 250–1000 lm and 30% of the <250 lm diameter fraction), C (containing equal weights of the two fractions); Darcy velocity 0.66 cm/ min. Average aggregate size is in the order A > B > C.
0
1
2
3
4
5
6
7
Pore volumes Fig. 3. Breakthrough curves for bromide in soil mixtures A (soil fraction of 2501000 lm), B (70% (wt.) of the 250–1000 lm and 30% of the <250 lm diameter fraction), C (containing equal weights of the two fractions), and for AgNPs in pure quartz (Darcy velocity 0.66 cm min 1).
O. Sagee et al. / Chemosphere 88 (2012) 670–675
C/C 0
674
1.0
0.66 cm/mim
0.8
0.66 cm/min + humic acid End of pulse
0.6 0.4 0.2 0.0 0
1
2
3
4
5
6
7
Pore volumes
the vicinity of the flow path but also in smaller pores inside the soil aggregates. Another proposed mechanism for retention of AgNPs in soil column is chemical interaction between soil residual chloride and AgNPs, to form particles coated by a silver chloride layer; or if dissolved ions of silver are released, then the production and precipitation of AgCl is expected. In related studies, El Badawy et al. (2010) and Prathna et al. (2011) showed that the environmental conditions and capping agents have an impact on the aggregate stability in solution. This suspension stability has a strong effect on particle transport in soils. It is reasonable that this usually very fast reaction between chlorides and silver is delayed, as the presence of citrate coating on the surface of AgNPs acts as a barrier between the chlorides on soil minerals and the silver core of the AgNPs. At this stage it is not clear if the reaction between the silver and the chloride occurs with dissolved ions that were released from the nanoparticles, or if the chloride reacts with silver molecules on the nanoparticle surface, replacing the citrate coating and in turn causing removal of the nanoparticle from suspension and retardation. An additional option is attachment of silver nanoparticles to clay surfaces, as described by Cabal et al. (2010), who report attachment of AgNPs to the surface of kaolin plates with the hydroxyl groups being the main centers of adsorption. It should also be noted that while the MARW was modified by replacing Ca2+ and Cl ions, to increase stability of the AgNP suspension (Section 2.1), these ions are to a greater or lesser extent present in soils and meteoric waters. As such, the enhanced AgNP transport observed here may also be influenced, in real soils, by chemical interactions that have been excluded deliberately from this study. These results emphasize the complexity of chemical controls on AgNP transport that may be of significance in the environment. Another insight into AgNP transport mechanisms in soil can be obtained from these results. Earlier, two mechanisms were proposed for the early breakthrough of AgNPs in soil; development of preferential flow pathways, and AgNPs exclusion from inneraggregate pores. It is expected that the development of preferential flow pathways for nanoparticles in porous media is reduced with decreasing flow velocities. However, the early breakthrough of AgNPs in the soil column was similarly observed in the low flow velocity experiments. This supports the hypothesis that the early breakthrough is caused by the reduced penetration of AgNPs into the inner-aggregate space. The presence of humic acid in the background solution resulted in a dramatic increase in AgNP mobility (Fig. 5). AgNP outlet plateau C/C0 values were determined to be 0.464 ± 0.035 for the AgNPs without humic acid, and 0.749 ± 0.032 with humic acid. Humic acid, found in many natural soils, helps to stabilize AgNPs in suspension (e.g. Akaighe et al., 2011; Delay et al., 2011; Huynh
Fig. 5. Breakthrough curves for AgNPs in soil mixture B (Darcy velocity 0.66 cm min 1) with and without the presence of humic acid in solution.
1.0
C/C 0
Fig. 4. Breakthrough curves for AgNPs in soil mixture B and Darcy flow velocities of 0.66 cm min 1 and 0.17 cm min 1.
KNO3 pre-treatment rep. 1
0.8
KNO3 pre-treatment rep. 2
0.6
KNO3 pre-treatment rep. 3 No pre-treatment
0.4
End of pulse
0.2 0.0 0
1
2
3
4
5
6
7
Pore volumes Fig. 6. Breakthrough curves for AgNPs in soil mixture B (Darcy velocity 0.17 cm min 1) in columns pretreated with potassium nitrate to remove soil residual chlorine, and a control column.
and Chen, 2011). Humic acid coating formed around AgNPs can interfere with chemical interaction with the soil. On the other hand, it is assumed that the humic acid coating will not have a significant effect on mechanical straining, because AgNPs were shown to form a stable suspension even without the presence of humic acid. Therefore, the significantly improved transport of AgNPs in soil in the presence of humic acid indicates a strong influence of soil–AgNP chemical interactions on AgNP retention. It should be noted that these results do not rule out the probable combined effect of chemical interactions and mechanical straining, but rather emphasize the importance of the chemical interaction component for ENP transport. Results for the potassium nitrate pretreated columns experiments, along with a control experiment using untreated soil, are presented in Fig. 6. The lower flow velocity of 0.17 cm min 1 was used. Unlike the other experiments, variation among repetitions was too large to define one representative repetition. Therefore, Fig. 6 shows all three replicates for the potassium nitrate pretreated columns, and one representative repetition for the columns filled with untreated soil fraction B. Although the differences in AgNP outlet C/C0 values among the three pretreated columns are large, even the experiment with the smallest AgNP outlet C/C0 value (0.498) was almost double the highest C/C0 value measured in the untreated column. In one repetition, AgNPs reached a plateau C/ C0 value of one, indicating complete soil mobility. The average of the three repetitions for the pretreated columns was 0.731 ± 0.228, which is more than double the average plateau value for AgNPs in the faster flow rate experiments for the same soil fraction.
O. Sagee et al. / Chemosphere 88 (2012) 670–675
These results emphasize the important role of soil residual chloride in retention of AgNPs in soil. More generally, this study emphasizes the important role of chemistry in ENP transport in soil. The results reported here also demonstrate the need to further investigate transport of other ENPs in real soil, rather than glass beads and pure quartz. These results also emphasize the need for a better understanding of the complex mechanisms influencing chemical interactions of ENPs in soils, which in turn will enable development of realistic models for ENP transport in natural soil. 4. Conclusions AgNPs were generally found to be mobile in natural soil, indicating a possible migration of particles to groundwater, which may lead to its contamination. The transport of AgNPs in soil was observed to increase in the presence of larger soil aggregates. Humic acid in solution increased mobility of AgNPs, and removal of residual chlorides from the soil also resulted in enhanced transport. It can be concluded that chloride-AgNP chemistry has a significant role in AgNP retention in soil, probably with a combined effect of mechanical straining. The early breakthrough of AgNPs observed through the soil column can be related to the inability of AgNPs to enter the inner aggregate pore space. Tracer-like AgNP transport through quartz sand columns indicates the important role of natural soil for the understanding of AgNP transport in the subsurface. In general, it is suggested that transport of ENPs in natural soil is not represented appropriately by use of pure quartz or glass beads. Therefore, further systematic study of ENPs in natural soils is required. Acknowledgements Two anonymous referees are thanked for helpful comments. The financial support of The Dr. Scholl Center for Water and Climate Research is gratefully acknowledged. Brian Berkowitz holds the Sam Zuckerberg Professorial Chair in Hydrology. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.chemosphere. 2012.03.055. References Akaighe, N., MacCuspie, R.I., Navarro, D.A., Aga, D.S., Banerjee, S., Sohn, M., Sharma, V.K., 2011. Humic acid-induced silver nanoparticle formation under environmentally relevant conditions. Environ. Sci. Technol. 45, 3895–3901. Baalousha, M., 2009. Aggregation and disaggregation of iron oxide nanoparticles: influence of particle concentration, pH and natural organic matter. Sci. Total Environ. 407, 2093–2101. Bache, B.W., 1976. The measurement of cation exchange capacity of soils. J. Sci. Food Agric. 27, 273–280. Ben-Moshe, T., Dror, I., Berkowitz, B., 2010. Transport of metal oxide nanoparticles in saturated porous media. Chemosphere 81, 387–393. Bouyoucos, G.J., 1962. Hydrometer method improved for making particle size analyses of soils. Agron. J. 54, 464–465. Cabal, B., Torrecillas, R., Malpartida, F., Moya, J.S., 2010. Heterogeneous precipitation of silver nanoparticles on kaolinite plates. Nanotechnology 21, 475705. Carter, D.L., Heliman, M.D., Gonzales, C.L., 1965. Ethylene glycol monoethyl ether for determining surface area of silicate minerals. Soil Sci. 100, 356–360. Christian, P., Von der Kammer, F., Baalousha, M., Hofmann, T., 2008. Nanoparticles: structure, properties, preparation and behaviour in environmental media. Ecotoxicology 17, 326–343. Colvin, V.L., 2003. The potential environmental impact of engineered nanomaterials. Nat. Biotechnol. 21, 1166–1170.
675
Delay, M., Dolt, T., Woellhaf, A., Sembritzki, R., Frimmel, F.H.J., 2011. Interactions and stability of silver nanoparticles in the aqueous phase: influence of natural organic matter (NOM) and ionic strength. Chromatogr. A 1218, 4206–4212. Dong, X., Ji, X., Wu, H., Zhao, L., Li, J., Yang, W., 2009. Shape control of silver nanoparticles by stepwise citrate reduction. J. Phys. Chem. C 113, 6573–6576. Dreher, K.L., 2004. Health and environmental impact of nanotechnology: toxicological assessment of manufactured nanoparticles. Toxicol. Sci. 77, 3–5. Dunphy Guzman, K.A., Finnegan, M.P., Banfield, J.F., 2006. Influence of surface potential on aggregation and transport of titania nanoparticles. Environ. Sci. Technol., 7688–7693. El Badawy, A.M., Luxton, T.P., Silva, R.G., Scheckel, K.G., Suidan, M., Tolaymat, T.M., 2010. Impact of environmental conditions (pH, ionic strength, and electrolyte type) on the surface charge and aggregation of silver nanoparticles suspensions. Environ. Sci. Technol. 44, 1260–1266. Fang, J., Shan, X.-Q., Wen, B., Lin, J.-M., Owens, G., 2009. Stability of titania nanoparticles in soil suspensions and transport in saturated homogeneous soil columns. Environ. Pollut. 157, 1101–1109. He, F., Zhao, D., Liu, J., Roberts, C.B., 2006. Stabilization of Fe–Pd nanoparticles with sodium carboxymethyl cellulose for enhanced transport and dechlorination of trichloroethylene in soil and groundwater. Ind. Eng. Chem. Res. 46, 29–34. He, D., Jones, A.M., Garg, S., Pham, A.N., Waite, T.D., 2011. Silver nanoparticle– reactive oxygen species interactions: application of a charging–discharging model. J. Phys. Chem. C 115, 5461–5468. Huynh, K.A., Chen, K.L., 2011. Aggregation kinetics of citrate and polyvinylpyrrolidone coated silver nanoparticles in monovalent and divalent electrolyte solutions. Environ. Sci. Technol. 45, 5564–5571. Jaisi, D.P., Elimelech, M., 2009. Single-walled carbon nanotubes exhibit limited transport in soil columns. Environ. Sci. Technol. 43, 9161–9166. Jeong, S.W., Kim, S.D., 2009. Aggregation and transport of copper oxide nanoparticles in porous media. J. Environ. Monit. 11, 1595–1600. Kanel, S., Nepal, D., Manning, B., Choi, H., 2007. Transport of surface-modified iron nanoparticle in porous media and application to arsenic(III) remediation. J. Nanopart. Res. 9, 725–735. Kretzschmar, R., Sticher, H., 1997. Transport of humic-coated iron oxide colloids in a sandy soil: influence of Ca2+ and trace metals. Environ. Sci. Technol. 31, 3497– 3504. Lecoanet, H.F., Wiesner, M.R., 2004. Velocity effects on fullerene and oxide nanoparticle deposition in porous media. Environ. Sci. Technol. 38, 4377–4382. Li, X., Lenhart, J., Walker, H.W., 2010. Dissolution-accompanied aggregation kinetics of silver nanoparticles. Langmuir 26, 16690–16698. Lin, S., Cheng, Y., Bobcombe, Y., Jones, K.L., Liu, J., Wiesner, M.R., 2011. Deposition of silver nanoparticles in geochemically heterogeneous porous media: predicting affinity from surface composition analysis. Environ. Sci. Technol. 45, 5209– 5215. Liu, X., O’Carroll, D.M., Petersen, E.J., Huang, Q., Anderson, C.L., 2009. Mobility of multiwalled carbon nanotubes in porous media. Environ. Sci. Technol 43, 8153– 8158. Loux, N., Savage, N., 2008. An assessment of the fate of metal oxide nanomaterials in porous media. Water Air Soil Pollut. 194, 227–241. McDowell-Boyer, L.M., Hunt, J.R., Sitar, N., 1986. Particle transport through porous media. Water Resour. Res. 22, 1901–1921. Navarro, E., Piccapietra, F., Wagner, B., Marconi, F., Kaegi, R., Odzak, N., Sigg, L., Behra, R., 2008. Toxicity of silver nanoparticles to chlamydomonas reinhardtii. Environ. Sci. Technol. 42, 8959–8964. Prathna, T.C., Chandrasekaran, N., Mukherjee, A., 2011. Studies on aggregation behaviour of silver nanoparticles in aqueous matrices: effect of surface functionalization and matrix composition. Colloids and Surfaces A 390, 216– 224. Schollenberger, C.J., 1945. Determination of soil organic matter. Soil Sci. 59, 53–56. Shang, J., Liu, C., Wang, Z., Wu, H., Zhu, K., Li, J., Liu, J., 2010. In-situ measurements of engineered nanoporous particle transport in saturated porous media. Environ. Sci. Technol. 44, 8190–8195. Shani, C., Weisbrod, N., Yakirevich, A., 2008. Colloid transport through saturated sand columns: Influence of physical and chemical surface properties on deposition. Colloids Surf. A 316, 142–150. Sirivithayapakorn, S., Keller, A., 2003. Transport of colloids in saturated porous media: a pore-scale observation of the size exclusion effect and colloid acceleration. Water Resour. Res. 39, 1109. Theng, B.K.G., Yuan, G., 2008. Nanoparticles in the soil environment. Elements 4, 395–399. Tian, Y., Gao, B., Silvera-Batista, C., Ziegler, K., 2010. Transport of engineered nanoparticles in saturated porous media. J. Nanopart. Res. 12, 2371–2380. Uyusur, B., Darnault, C.J.G., Snee, P.T., Kokën, E., Jacobson, A.R., Wells, R.R., 2010. Coupled effects of solution chemistry and hydrodynamics on the mobility and transport of quantum dot nanomaterials in the vadose zone. J. Contam. Hydrol. 118, 184–198. Wiesner, M.R., Lowry, G.V., Alvarez, P., Dionysiou, D., Biswas, P., 2006. Assessing the risks of manufactured nanomaterials. Environ. Sci. Technol. 40, 4336–4345. Zvikelsky, O., Weisbrod, N., Dody, A., 2008. A comparison of clay colloid and artificial microsphere transport in natural discrete fractures. J. Colloid Interface Sci. 323, 286–292.
Contents lists available at SciVerse ScienceDirect
Chemosphere journal homepage: www.elsevier.com/locate/chemosphere
Transport of silver nanoparticles (AgNPs) in soil Omer Sagee, Ishai Dror ⇑, Brian Berkowitz Dept. of Environmental Sciences and Energy Research, Weizmann Institute of Science, Rehovot 76100, Israel
a r t i c l e
i n f o
Article history: Received 19 December 2011 Received in revised form 12 March 2012 Accepted 17 March 2012 Available online 18 April 2012 Keywords: Silver nanoparticles Transport Soil Chlorides Humic acid
a b s t r a c t The effect of soil properties on the transport of silver nanoparticles (AgNPs) was studied in a set of laboratory column experiments, using different combinations of size fractions of a Mediterranean sandy clay soil. The AgNPs with average size of 30 nm yielded a stable suspension in water with zeta potential of 39 mV. Early breakthrough of AgNPs in soil was observed in column transport experiments. AgNPs were found to have high mobility in soil with outlet relative concentrations ranging from 30% to 70%, depending on experimental conditions. AgNP mobility through the column decreased when the fraction of smaller soil aggregates was larger. The early breakthrough pattern was not observed for AgNPs in pure quartz columns nor for bromide tracer in soil columns, suggesting that early breakthrough is related to the nature of AgNP transport in natural soils. Micro-CT and image analysis used to investigate structural features of the soil, suggest that soil aggregate size strongly affects AgNP transport in natural soil. The retention of AgNPs in the soil column was reduced when humic acid was added to the leaching solution, while a lower flow rate (Darcy velocity of 0.17 cm/min versus 0.66 cm/min) resulted in higher retention of AgNPs in the soil. When soil residual chloride was exchanged by nitrate prior to column experiments, significantly improved mobility of AgNPs was observed in the soil column. These findings point to the importance of AgNP–soil chemical interactions as a retention mechanism, and demonstrate the need to employ natural soils rather than glass beads or quartz in representative experimental investigations. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction The manufacturing and use of engineered nanoparticles (ENPs) in many commercial products has increased dramatically in recent years, together with contamination potential (Colvin, 2003; Dreher, 2004; Wiesner et al., 2006; Christian et al., 2008). After use and during production, ENPs can find their way into the soil environment; once in the soil, ENPs might be transported toward groundwater (Tian et al., 2010). Therefore, understanding ENP fate and transport in soil is of considerable importance. To date, most studies related to transport of ENPs have employed well-defined, artificial porous materials such as glass beads or acidwashed, homogeneous quartz sand in column experiments (e.g. Dunphy Guzman et al., 2006; Kanel et al., 2007; Loux and Savage, 2008; Shani et al., 2008; Baalousha, 2009; Jeong and Kim, 2009; Liu et al., 2009; Ben-Moshe et al., 2010; Shang et al., 2010; Uyusur et al., 2010). However, many soil properties such as grain size distributions, heterogeneity, and the presence of soil organic matter (SOM) are not represented in glass beads or quartz. Therefore, the relevance of these studies for transport of ENPs in the natural soil environment is in question. For example, the presence of absorbed metals that are widespread in soil minerals, such as copper and lead, ⇑ Corresponding author. Tel.: +972 8 9344230; fax: +972 8 9344124. E-mail address: [email protected] (I. Dror). 0045-6535/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.chemosphere.2012.03.055
can promote ENP aggregation (Kretzschmar and Sticher, 1997), while the presence of humic substances and dissolved organic matter can contribute to ENP suspension stability (Fang et al., 2009). It was further shown that the heterogeneous structure of the pore space in natural soils can lead to development of preferential flow paths for colloidal material, which may have a dramatic impact on transport characteristics (Sirivithayapakorn and Keller, 2003). In general, two main mechanisms can limit the transport of well-dispersed ENPs in soil: (1) straining, wherein nanoparticles are removed from solution in smaller pore spaces, and (2) physical–chemical processes which remove nanoparticles from solution by interactions between ENPs and the medium (McDowell-Boyer et al., 1986). Both mechanisms are expected to be enhanced when flow rates are lower and particles are smaller, so that the influence of diffusion is increased (Lecoanet and Wiesner, 2004). To the best of our knowledge, only a few previous studies have examined ENP transport in natural soils. Fang et al. (2009) demonstrated the retarding effect of clay content and the enhanced mobility effect of SOM on TiO2 nanoparticle transport in soil. Two other studies demonstrated that while some nanoparticles such as single-walled carbon nanotubes are practically immobilized in natural sandy soil, other nanoparticles such as surface-modified palladized iron nanoparticles migrate freely in soil (He et al., 2006; Jaisi and Elimelech, 2009). However, a systematic examination of the effect of soil properties on transport was not conducted.
O. Sagee et al. / Chemosphere 88 (2012) 670–675
The purpose of this study was to investigate the transport of silver nanoparticles (AgNPs) in natural soil. Silver nanoparticles (AgNPs) are among the most commonly produced ENPs, with many applications in clothing, cosmetics, and medical care products. AgNPs have known antibacterial properties, which makes AgNPs toxic to microorganisms in aquatic environments (Navarro et al., 2008; Tian et al., 2010; He et al., 2011). In particular, the investigation aimed to determine the effect of soil properties such as mechanical structure and presence of dissolved organic matter on the transport of AgNPs, with emphasis on understanding the relative importance of mechanical straining and physical-chemical processes of retention in transport of AgNPs in natural soil. The AgNPs used in these experiments (citrate coated AgNPs) have high negative f potential value, resulting in strong electrostatic repulsion from each other and a low tendency to aggregate. Because most natural soil minerals have negative f potential, resulting in a high-energy barrier for the attachment of AgNPs (Christian et al., 2008; Theng and Yuan, 2008; Lin et al., 2011), it was expected that the AgNPs would demonstrate relatively high mobility. 2. Experimental section 2.1. Characterization of silver nanoparticles (AgNPs) AgNPs were synthesized using the method of reduction by citrate (Dong et al., 2009), which allows control of the shape and size of AgNPs by the choice of synthesis solution pH. AgNPs were synthesized at pH 11.5 to achieve a diameter of approximately 30 nm. Suspensions from several parallel synthesis reactions were transferred to a 4 L container and kept sealed under constant stirring. This AgNP stock suspension was used in all column experiments to ensure uniform addition of AgNP suspension. Stability of the stock suspension was verified weekly using dynamic light scattering (DLS) on a Viscotek 802. Prior to each column experiment (on the same day), AgNPs were transferred to a ‘‘modified artificial rainwater’’ solution (MARW). The MARW solution contained 0.048, 0.057 and 0.035 g L 1 of sodium nitrate, potassium bicarbonate and magnesium sulfate, respectively. This composition is a modification of the artificial rainwater solution which represents the typical rainwater composition in the Negev desert in Israel (Zvikelsky et al., 2008). Two modifications to the artificial rainwater solution were made. First, calcium cations, observed in preliminary laboratory experiments to interfere with AgNP stability, were replaced with magnesium and potassium. Second, chloride anions, which cause production of silver chloride and in turn lead to its precipitation, were replaced with nitrate and sulfate. Both ion effects were also described in the literature to reduce suspension stability (e.g. El Badawy et al., 2010; Huynh and Chen, 2011) and thus interfere with elemental silver transport and ICP-MS analysis. Before each column experiment, AgNPs were transferred from the synthesis solution to the MARW solution done by centrifuging 50 mL AgNP solution and replacing 45 mL of the supernatant with MARW twice. This was done also to minimize the presence of aqueous silver ions in the solution. The solution stability of the AgNPs in the MARW was monitored by transmission electron microscopy (TEM) and DLS to confirm no change in size distribution and shape of the particles over time. Zeta potential was measured with a zetasizer nano instrument (Malvern). 2.2. Soil treatments Soil was sampled from the top 50 cm of an experimental vineyard in the Bet-Dagan area, central Israel. Two different size fractions of the soil were separated using a set of sieves; the
671
250–1000 lm diameter fraction and the <250 lm diameter fraction. Soil grains larger than 1000 lm were not used in the column experiments. From the two size fractions, three different soil mixtures were made; A – containing only the 250–1000 lm fraction, B – containing 70% (weight) of the 250–1000 lm fraction and 30% of the <250 lm fraction, and C – containing equal weights of the two fractions. 2.3. Soil properties analysis Key soil properties of the A, B, and C mixtures were determined with various soil analysis methods and are presented in Table S1 in the supplementary data (SD). Clay, silt, and sand content were measured using the hydrometer method based on Stoke’s law (Bouyoucos, 1962). Soil organic matter (SOM) content was determined by weight lost measurement after burning at 400 °C overnight (Schollenberger, 1945). Cation exchange capacity (CEC) was determined by means of ammonium acetate extraction (Bache, 1976), and specific surface area (SSA) was established by the ethylene glycol monoethyl ether (EGME) mono-layer adsorption method (Carter et al., 1965). Porosity was measured prior to each column experiment by slow saturation (from below) of the soil column with MARW solution. 2.4. X-ray microtomography (micro-CT) and image analysis X-ray microtomography (micro-CT) was used to determine structural differences among soil mixtures A, B, and C. About 2 g from each mixture was inserted into a 10 mm (i.d.) plastic holder for X-ray micro-CT imaging in a Micro XCT-400 (Xradia, California, USA) instrument. The voltage applied on the X-ray source was 40 kV and the current was 200 lA. No source filters were used. For each sample, 1500 projection images were recorded with an exposure time of 5 s and a magnification objective of 4. The final pixel size in the projection images was 4.6 lm. The volume was reconstructed with the instrument software and was then exported to Avizo Fire (VSG international) for further 3D image analysis. After de-noising and binarizing the images, individual soil aggregates could be distinguished. Example images and details of the image analysis process appear in the Supplementary data (Fig. S1). The volume of each aggregate was determined by the software and a size distribution histogram was constructed for each sample. This process was repeated three times for each of the mixtures A, B, and C. 2.5. AgNP quantification To analytically separate AgNPs from the soil colloids eluting from the column, a method for quantification of the AgNPs in their dissolved state was developed. AgNPs were dissolved in nitric acid (40.5% by weight) followed by dilution to 3.88% nitric acid and filtration by 0.22 lm cellulose acetate filter (Millipore); a correlation with R2 > 0.999 was found for calibration of the AgNP suspension by inductively coupled plasma mass spectrometry (ICP-MS) (ELAN 9000, Perkin Elmer Inc.). This method was later used to determine AgNP concentrations in column experiments. 2.6. AgNPs in soil column transport experiments Column experiments were conducted in closed, cylindrical, vertical Plexiglas columns, with length of 10 cm and inner diameter of 3.1 cm. A peristaltic pump was used to apply MARW or AgNP suspensions, at Darcy velocity of 0.66 cm min 1. Slight variations in flow rate were registered between column experiments, and taken into account in pore volume calculations. After packing the columns with the soil mixtures, the columns were slowly saturated
O. Sagee et al. / Chemosphere 88 (2012) 670–675
with MARW and porosity was determined. The columns were then washed with ten pore volumes (PV) of MARW at flow rate of 5 mL min 1. At that point, the experiment began with application of a four PV pulse of AgNP suspension diluted 1:5 by MARW. This pulse was followed by flowing four PV of MARW. The outflow solution was collected by fraction collector. After the experiments, nitric acid (69%) was added to the samples to achieve a final acid concentration of 40.5%, and the samples were stirred gently. After 4 h, samples were diluted in double distilled water (DDW) and the solution was filtered by 0.22 lm cellulose acetate filter, prior to ICP-MS analysis. Samples of the inflow solution of each column experiment were also collected and treated as described above to determine initial concentration. A calibration curve was constructed for each individual experiment to determine the C/ C0 value of each sample, for the construction of a breakthrough curves. Each soil treatment (A, B, and C) was run in triplicates. Before each experiment, the columns were opened and repacked. Fraction B of the soil was used to study the effect of three parameters on AgNP transport: flow velocity, presence of dissolved humic acid, and presence of residual chlorides in soil. In general, the experimental procedure described above was followed in all experiments; adjustments to the specific experiments are described in the following paragraph. To examine the effect of flow velocity, a set of column experiments was conducted with Darcy velocity of 0.17 cm min 1 instead of 0.66 cm min 1. The effect of dissolved humic acid on AgNP transport was studied by adding 100 ppm humic acid (Sigma-Aldrich) to the MARW and AgNP solution. The removal of residual chlorides from soil, and its replacement by nitrate, was achieved by pretreating the packed columns with potassium nitrate, slowly flowing 20 PV of 0.5 M potassium nitrate solution through them. This was done to test the hypothesis that a chemical reaction between soil residual chloride and AgNPs results in AgNP retention is soil. For this set of experiments, a Darcy velocity of 0.17 cm min 1 was applied. 2.7. AgNP transport in pure quartz sand The same procedure as for AgNP transport in soil columns was performed for a column filled with clean quartz sand. A breakthrough experiment was conducted in a column filled with homogeneous quartz sand (Accusand), with grain diameter of 210– 300 lm. Prior to packing the column, the quartz was soaked overnight in concentrated hydrochloric acid, thoroughly rinsed with DDW and heated overnight in a 400 °C oven. 2.8. Bromide tracer soil column transport experiments Bromide was used as a soluble conservative tracer in a column experiment for each of the soil mixtures A, B, and C. After saturation and washing of the columns with MARW, a four PV pulse of 500 ppm bromide (as KBr) was injected, followed by a four PV MARW pulse. Bromide concentration in the outflow samples was determined using high pressure liquid chromatography (HPLC) equipped with Alltech conductivity detector (model 650). Anions were separated with an Alltech anion 7 lm column, 100 4.6 mm. The mobile phase solution for the bromide detection was 4 mM p-Hydrobenzoic acid (pH adjusted to 7.5 with LiOH) and a flow rate of 1 mL min 1 was applied. 3. Results and discussion 3.1. Properties of silver nanoparticles (AgNPs) TEM images of the synthesized AgNPs are presented in Fig. S1 in the supplementary data. The images illustrate that most of the
synthesized AgNPs were relatively round with a diameter of approximately 40 nm; in addition, a few rod-shaped AgNPs were detected. These results were confirmed by size distribution analysis measured using DLS. The f potential for AgNPs in the MARW solution was determined to be 39 mV, suggesting that they form a stable suspension. 3.2. Soil characteristics Some basic soil properties of the three mixtures A, B, C and the <250 lm size fraction are presented in the Table S1 in the supplementary data. It is apparent that many of the soil properties do not vary significantly among the different mixtures. Specific surface area (SSA), for example, which is a measure of the surface area per unit weight of the soil solid phase, varies only in the range of 94.7–96.4 m2 g 1. This is counterintuitive given that SSA is dependent on grain size and is expected to change when size-based separation is used. However, this finding is explained by the fact that SSA determination involves breakdown of soil aggregates. Other properties show a similar trend of small variations throughout the different mixtures. When the mechanical composition of the different mixtures was determined, the sand content ranged between 62.5% and 70%. The highest sand content was found in the smallest size fraction (<250 lm). Clay content was 17.5% for soil mixture A and decreased to a value of 12.5% for the < 250 lm fraction. Therefore, while soil mixture A contains larger aggregates, the aggregates themselves tend to contain more clay. To summarize, differences among the measured properties of soil mixtures A, B, and C are assumed to be negligible when considering their possible effect on AgNP transport in soil. 3.3. X-ray microtomography (micro-CT) and image analysis Aggregate size distribution histograms for each of three soil mixture are shown in Fig. 1. Assuming spherical aggregates, the volume of each aggregate was converted into diameter. The histograms represent the total volume occupied by soil aggregates of the specified diameter. The volume units in all three histograms are normalized to the same scale. The histograms show a trend in aggregate size distribution. In all three mixtures the dominant aggregate population is in the 270–300 lm diameter range. However, mixture A contains more aggregates with a larger diameter, relative to mixtures B and C. The relative volume of aggregates with diameter smaller than 270 lm for A, B, and C are 26.1%, 45.4%, and 49.9%, respectively. Therefore, a significant structural difference between the soil mixtures is evident, in spite of the relative similarity in chemical properties. The differences in aggregate size distribution can affect AgNP transport in the soil in two possible ways: (1) soil fractions
1
Normalized Volume
672
A B C
0.8 0.6 0.4 0.2 0 0
75
165
255
345
435
525
615
705
Grain Diameter (µm) Fig. 1. Aggregate size distribution histogram for soil mixtures A (soil fraction of 1000–250 lm), B (70% (wt.) of the 250–1000 lm and 30% of the <250 lm diameter fraction), and C (containing equal weights of the two fractions).
673
O. Sagee et al. / Chemosphere 88 (2012) 670–675
containing more aggregates with smaller diameter have also smaller pores, which in turn increase mechanical straining of AgNP, and (2) larger surface area of the smaller aggregates increases the available surface area for interactions between AgNPs and aggregates. 3.4. Column experiments In most column experiments conducted in this study, a fast rise in AgNP outlet concentration was first observed; a steady plateau value of elution concentration was then reached and maintained throughout application period of the AgNP pulse. When the MARW solution was re-applied, a fast drop to zero in the AgNP outlet concentration was observed. Outlet concentrations of AgNP in the plateau stage for all of the experiments, expressed as C/C0, are given in Table S2 in the supplementary data. Note that no plateau was observed in the reduced flow speed experiments, where soil was not pretreated with potassium nitrate. The results of one representative repetition (in which plateau values were closest to that of the treatment average) for column experiments at high flow velocities (0.66 cm/min), for soil mixtures A, B, and C, are presented in Fig. 2. AgNP retention is observed to increase with decreasing aggregate size. The C/C0 average and standard deviation values for AgNPs at the column outlet for soil mixtures A, B, and C were 0.697 ± 0.041, 0.464 ± 0.035 and 0.325 ± 0.034, respectively. Mechanical straining is the simplest explanation for the increased AgNP retention in the order of A > B > C. However, in spite of the chemical similarities found among the three mixtures, chemical interaction can also account for retention in this order. This explanation is possible when considering that a reaction between AgNPs and soil occurs at the aggregate surface, and the aggregate surface area increases for the smaller aggregates. One such possible reaction is between soil residual chloride and AgNPs, as silver chloride is a readily formed precipitate, or coats the particles which in turn can change the nanoparticle suspension stability (Li et al., 2010; Huynh and Chen, 2011; Prathna et al., 2011). Theoretically, the outlet value of a conservative tracer in a soil column should reach C/C0 0.5 after one PV, assuming the tracer travels at the same velocity as the background solution. As seen in Fig. 2, AgNPs in all three soil mixtures show an earlier breakthrough behavior. The concentration of AgNPs at the column outlet almost reaches its plateau value as early as one PV after introduction of the AgNPs, suggesting that AgNPs travel faster in the soil column than the background solution. This behavior was repeated in all AgNP soil transport experiments conducted in this study.
Two possible, complementary, explanations can be offered for this phenomenon: (1) development of preferential flow pathways in the soil column, related to the effect of soil heterogeneity and the negative charge of the particles (f potential = 39 mV), and (2) the existence of ‘‘dead pore volume’’ inside soil aggregates. In the first case, if AgNPs travel only through the larger pores where hydraulic conductivity is higher, the result is partial sampling of the pore space by the AgNPs and early breakthrough. In the second case, while soil aggregates contain solution in the inner aggregate pores, they are impenetrable to AgNPs. In this case, the volume in which AgNPs travel is smaller than the total pore volume. To examine the early breakthrough behavior, transport experiments with bromide as a conservative tracer were conducted. Furthermore, to determine if the early breakthrough phenomenon is unique to transport of AgNPs in natural soils, a transport experiment of AgNPs in a pure quartz sand column was conducted (Fig. 3). Bromide advanced in all soil columns as a conservative tracer. No clear differences in the transport properties of bromide through soil mixtures A, B and C were found, possibly because the relatively high flow rates range in the experiments masked the effect of diffusion. Moreover, it is seen that AgNPs behaved as a conservative tracer when transported through the pure quartz sand column; again, early breakthrough was not observed. These results confirm that early breakthrough is related to the properties of the porous medium and the solute, and suggest that this phenomenon is related to transport of AgNPs in natural soils. Fig. 4 shows the breakthrough curves of one representative repetition (from triplicate runs) from flow velocities 0.66 and 0.17 cm min 1 in soil fraction B. It is evident that AgNP retention in the soil column increases when flow velocity is reduced. The plateau C/C0 value for AgNPs in fraction B in the higher flow velocity was 0.464 ± 0.025, while the highest C/C0 value observed for the reduced flow velocity was smaller than 0.3. The breakthrough curve changed significantly for the reduced flow velocity. Instead of reaching a steady plateau value, as observed in all other column experiments, a declining trend is seen. In other words, the highest AgNP outlet concentrations were obtained approximately one PV after introducing the AgNPs to the column, followed by a linear decrease from C/C0 0.3 to C/C0 0.2 over four PV. Such behavior can be attributed to accumulation and consequent blocking of pore throats by AgNPs, which gradually increase AgNP filtration in the soil. This behavior was not observed for the higher flow velocities, possibly because the flow was fast enough to constantly mobilize and remove AgNPs in pore throats. The increased retention in the slower flow rate can be explained by increased mechanical straining as a result of more significant diffusion, which leads to trapping of nanoparticles not only in
1.0 1.0
A - coarse 0.8
B- medium
Bromide in A
0.8
Bromide in B
0.6
End of pulse
C/C 0
C/C 0
C- fine
0.4 0.2
Bromide in C
0.6
AgNPs in quartz
0.4
End of pulse
0.2 0.0
0.0 0
1
2
3
4
5
6
7
Pore Volumes Fig. 2. Breakthrough curves for AgNPs in different soil mixtures A (soil fraction of 250–1000 lm), B (70% (wt.) of the 250–1000 lm and 30% of the <250 lm diameter fraction), C (containing equal weights of the two fractions); Darcy velocity 0.66 cm/ min. Average aggregate size is in the order A > B > C.
0
1
2
3
4
5
6
7
Pore volumes Fig. 3. Breakthrough curves for bromide in soil mixtures A (soil fraction of 2501000 lm), B (70% (wt.) of the 250–1000 lm and 30% of the <250 lm diameter fraction), C (containing equal weights of the two fractions), and for AgNPs in pure quartz (Darcy velocity 0.66 cm min 1).
O. Sagee et al. / Chemosphere 88 (2012) 670–675
C/C 0
674
1.0
0.66 cm/mim
0.8
0.66 cm/min + humic acid End of pulse
0.6 0.4 0.2 0.0 0
1
2
3
4
5
6
7
Pore volumes
the vicinity of the flow path but also in smaller pores inside the soil aggregates. Another proposed mechanism for retention of AgNPs in soil column is chemical interaction between soil residual chloride and AgNPs, to form particles coated by a silver chloride layer; or if dissolved ions of silver are released, then the production and precipitation of AgCl is expected. In related studies, El Badawy et al. (2010) and Prathna et al. (2011) showed that the environmental conditions and capping agents have an impact on the aggregate stability in solution. This suspension stability has a strong effect on particle transport in soils. It is reasonable that this usually very fast reaction between chlorides and silver is delayed, as the presence of citrate coating on the surface of AgNPs acts as a barrier between the chlorides on soil minerals and the silver core of the AgNPs. At this stage it is not clear if the reaction between the silver and the chloride occurs with dissolved ions that were released from the nanoparticles, or if the chloride reacts with silver molecules on the nanoparticle surface, replacing the citrate coating and in turn causing removal of the nanoparticle from suspension and retardation. An additional option is attachment of silver nanoparticles to clay surfaces, as described by Cabal et al. (2010), who report attachment of AgNPs to the surface of kaolin plates with the hydroxyl groups being the main centers of adsorption. It should also be noted that while the MARW was modified by replacing Ca2+ and Cl ions, to increase stability of the AgNP suspension (Section 2.1), these ions are to a greater or lesser extent present in soils and meteoric waters. As such, the enhanced AgNP transport observed here may also be influenced, in real soils, by chemical interactions that have been excluded deliberately from this study. These results emphasize the complexity of chemical controls on AgNP transport that may be of significance in the environment. Another insight into AgNP transport mechanisms in soil can be obtained from these results. Earlier, two mechanisms were proposed for the early breakthrough of AgNPs in soil; development of preferential flow pathways, and AgNPs exclusion from inneraggregate pores. It is expected that the development of preferential flow pathways for nanoparticles in porous media is reduced with decreasing flow velocities. However, the early breakthrough of AgNPs in the soil column was similarly observed in the low flow velocity experiments. This supports the hypothesis that the early breakthrough is caused by the reduced penetration of AgNPs into the inner-aggregate space. The presence of humic acid in the background solution resulted in a dramatic increase in AgNP mobility (Fig. 5). AgNP outlet plateau C/C0 values were determined to be 0.464 ± 0.035 for the AgNPs without humic acid, and 0.749 ± 0.032 with humic acid. Humic acid, found in many natural soils, helps to stabilize AgNPs in suspension (e.g. Akaighe et al., 2011; Delay et al., 2011; Huynh
Fig. 5. Breakthrough curves for AgNPs in soil mixture B (Darcy velocity 0.66 cm min 1) with and without the presence of humic acid in solution.
1.0
C/C 0
Fig. 4. Breakthrough curves for AgNPs in soil mixture B and Darcy flow velocities of 0.66 cm min 1 and 0.17 cm min 1.
KNO3 pre-treatment rep. 1
0.8
KNO3 pre-treatment rep. 2
0.6
KNO3 pre-treatment rep. 3 No pre-treatment
0.4
End of pulse
0.2 0.0 0
1
2
3
4
5
6
7
Pore volumes Fig. 6. Breakthrough curves for AgNPs in soil mixture B (Darcy velocity 0.17 cm min 1) in columns pretreated with potassium nitrate to remove soil residual chlorine, and a control column.
and Chen, 2011). Humic acid coating formed around AgNPs can interfere with chemical interaction with the soil. On the other hand, it is assumed that the humic acid coating will not have a significant effect on mechanical straining, because AgNPs were shown to form a stable suspension even without the presence of humic acid. Therefore, the significantly improved transport of AgNPs in soil in the presence of humic acid indicates a strong influence of soil–AgNP chemical interactions on AgNP retention. It should be noted that these results do not rule out the probable combined effect of chemical interactions and mechanical straining, but rather emphasize the importance of the chemical interaction component for ENP transport. Results for the potassium nitrate pretreated columns experiments, along with a control experiment using untreated soil, are presented in Fig. 6. The lower flow velocity of 0.17 cm min 1 was used. Unlike the other experiments, variation among repetitions was too large to define one representative repetition. Therefore, Fig. 6 shows all three replicates for the potassium nitrate pretreated columns, and one representative repetition for the columns filled with untreated soil fraction B. Although the differences in AgNP outlet C/C0 values among the three pretreated columns are large, even the experiment with the smallest AgNP outlet C/C0 value (0.498) was almost double the highest C/C0 value measured in the untreated column. In one repetition, AgNPs reached a plateau C/ C0 value of one, indicating complete soil mobility. The average of the three repetitions for the pretreated columns was 0.731 ± 0.228, which is more than double the average plateau value for AgNPs in the faster flow rate experiments for the same soil fraction.
O. Sagee et al. / Chemosphere 88 (2012) 670–675
These results emphasize the important role of soil residual chloride in retention of AgNPs in soil. More generally, this study emphasizes the important role of chemistry in ENP transport in soil. The results reported here also demonstrate the need to further investigate transport of other ENPs in real soil, rather than glass beads and pure quartz. These results also emphasize the need for a better understanding of the complex mechanisms influencing chemical interactions of ENPs in soils, which in turn will enable development of realistic models for ENP transport in natural soil. 4. Conclusions AgNPs were generally found to be mobile in natural soil, indicating a possible migration of particles to groundwater, which may lead to its contamination. The transport of AgNPs in soil was observed to increase in the presence of larger soil aggregates. Humic acid in solution increased mobility of AgNPs, and removal of residual chlorides from the soil also resulted in enhanced transport. It can be concluded that chloride-AgNP chemistry has a significant role in AgNP retention in soil, probably with a combined effect of mechanical straining. The early breakthrough of AgNPs observed through the soil column can be related to the inability of AgNPs to enter the inner aggregate pore space. Tracer-like AgNP transport through quartz sand columns indicates the important role of natural soil for the understanding of AgNP transport in the subsurface. In general, it is suggested that transport of ENPs in natural soil is not represented appropriately by use of pure quartz or glass beads. Therefore, further systematic study of ENPs in natural soils is required. Acknowledgements Two anonymous referees are thanked for helpful comments. The financial support of The Dr. Scholl Center for Water and Climate Research is gratefully acknowledged. Brian Berkowitz holds the Sam Zuckerberg Professorial Chair in Hydrology. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.chemosphere. 2012.03.055. References Akaighe, N., MacCuspie, R.I., Navarro, D.A., Aga, D.S., Banerjee, S., Sohn, M., Sharma, V.K., 2011. Humic acid-induced silver nanoparticle formation under environmentally relevant conditions. Environ. Sci. Technol. 45, 3895–3901. Baalousha, M., 2009. Aggregation and disaggregation of iron oxide nanoparticles: influence of particle concentration, pH and natural organic matter. Sci. Total Environ. 407, 2093–2101. Bache, B.W., 1976. The measurement of cation exchange capacity of soils. J. Sci. Food Agric. 27, 273–280. Ben-Moshe, T., Dror, I., Berkowitz, B., 2010. Transport of metal oxide nanoparticles in saturated porous media. Chemosphere 81, 387–393. Bouyoucos, G.J., 1962. Hydrometer method improved for making particle size analyses of soils. Agron. J. 54, 464–465. Cabal, B., Torrecillas, R., Malpartida, F., Moya, J.S., 2010. Heterogeneous precipitation of silver nanoparticles on kaolinite plates. Nanotechnology 21, 475705. Carter, D.L., Heliman, M.D., Gonzales, C.L., 1965. Ethylene glycol monoethyl ether for determining surface area of silicate minerals. Soil Sci. 100, 356–360. Christian, P., Von der Kammer, F., Baalousha, M., Hofmann, T., 2008. Nanoparticles: structure, properties, preparation and behaviour in environmental media. Ecotoxicology 17, 326–343. Colvin, V.L., 2003. The potential environmental impact of engineered nanomaterials. Nat. Biotechnol. 21, 1166–1170.
675
Delay, M., Dolt, T., Woellhaf, A., Sembritzki, R., Frimmel, F.H.J., 2011. Interactions and stability of silver nanoparticles in the aqueous phase: influence of natural organic matter (NOM) and ionic strength. Chromatogr. A 1218, 4206–4212. Dong, X., Ji, X., Wu, H., Zhao, L., Li, J., Yang, W., 2009. Shape control of silver nanoparticles by stepwise citrate reduction. J. Phys. Chem. C 113, 6573–6576. Dreher, K.L., 2004. Health and environmental impact of nanotechnology: toxicological assessment of manufactured nanoparticles. Toxicol. Sci. 77, 3–5. Dunphy Guzman, K.A., Finnegan, M.P., Banfield, J.F., 2006. Influence of surface potential on aggregation and transport of titania nanoparticles. Environ. Sci. Technol., 7688–7693. El Badawy, A.M., Luxton, T.P., Silva, R.G., Scheckel, K.G., Suidan, M., Tolaymat, T.M., 2010. Impact of environmental conditions (pH, ionic strength, and electrolyte type) on the surface charge and aggregation of silver nanoparticles suspensions. Environ. Sci. Technol. 44, 1260–1266. Fang, J., Shan, X.-Q., Wen, B., Lin, J.-M., Owens, G., 2009. Stability of titania nanoparticles in soil suspensions and transport in saturated homogeneous soil columns. Environ. Pollut. 157, 1101–1109. He, F., Zhao, D., Liu, J., Roberts, C.B., 2006. Stabilization of Fe–Pd nanoparticles with sodium carboxymethyl cellulose for enhanced transport and dechlorination of trichloroethylene in soil and groundwater. Ind. Eng. Chem. Res. 46, 29–34. He, D., Jones, A.M., Garg, S., Pham, A.N., Waite, T.D., 2011. Silver nanoparticle– reactive oxygen species interactions: application of a charging–discharging model. J. Phys. Chem. C 115, 5461–5468. Huynh, K.A., Chen, K.L., 2011. Aggregation kinetics of citrate and polyvinylpyrrolidone coated silver nanoparticles in monovalent and divalent electrolyte solutions. Environ. Sci. Technol. 45, 5564–5571. Jaisi, D.P., Elimelech, M., 2009. Single-walled carbon nanotubes exhibit limited transport in soil columns. Environ. Sci. Technol. 43, 9161–9166. Jeong, S.W., Kim, S.D., 2009. Aggregation and transport of copper oxide nanoparticles in porous media. J. Environ. Monit. 11, 1595–1600. Kanel, S., Nepal, D., Manning, B., Choi, H., 2007. Transport of surface-modified iron nanoparticle in porous media and application to arsenic(III) remediation. J. Nanopart. Res. 9, 725–735. Kretzschmar, R., Sticher, H., 1997. Transport of humic-coated iron oxide colloids in a sandy soil: influence of Ca2+ and trace metals. Environ. Sci. Technol. 31, 3497– 3504. Lecoanet, H.F., Wiesner, M.R., 2004. Velocity effects on fullerene and oxide nanoparticle deposition in porous media. Environ. Sci. Technol. 38, 4377–4382. Li, X., Lenhart, J., Walker, H.W., 2010. Dissolution-accompanied aggregation kinetics of silver nanoparticles. Langmuir 26, 16690–16698. Lin, S., Cheng, Y., Bobcombe, Y., Jones, K.L., Liu, J., Wiesner, M.R., 2011. Deposition of silver nanoparticles in geochemically heterogeneous porous media: predicting affinity from surface composition analysis. Environ. Sci. Technol. 45, 5209– 5215. Liu, X., O’Carroll, D.M., Petersen, E.J., Huang, Q., Anderson, C.L., 2009. Mobility of multiwalled carbon nanotubes in porous media. Environ. Sci. Technol 43, 8153– 8158. Loux, N., Savage, N., 2008. An assessment of the fate of metal oxide nanomaterials in porous media. Water Air Soil Pollut. 194, 227–241. McDowell-Boyer, L.M., Hunt, J.R., Sitar, N., 1986. Particle transport through porous media. Water Resour. Res. 22, 1901–1921. Navarro, E., Piccapietra, F., Wagner, B., Marconi, F., Kaegi, R., Odzak, N., Sigg, L., Behra, R., 2008. Toxicity of silver nanoparticles to chlamydomonas reinhardtii. Environ. Sci. Technol. 42, 8959–8964. Prathna, T.C., Chandrasekaran, N., Mukherjee, A., 2011. Studies on aggregation behaviour of silver nanoparticles in aqueous matrices: effect of surface functionalization and matrix composition. Colloids and Surfaces A 390, 216– 224. Schollenberger, C.J., 1945. Determination of soil organic matter. Soil Sci. 59, 53–56. Shang, J., Liu, C., Wang, Z., Wu, H., Zhu, K., Li, J., Liu, J., 2010. In-situ measurements of engineered nanoporous particle transport in saturated porous media. Environ. Sci. Technol. 44, 8190–8195. Shani, C., Weisbrod, N., Yakirevich, A., 2008. Colloid transport through saturated sand columns: Influence of physical and chemical surface properties on deposition. Colloids Surf. A 316, 142–150. Sirivithayapakorn, S., Keller, A., 2003. Transport of colloids in saturated porous media: a pore-scale observation of the size exclusion effect and colloid acceleration. Water Resour. Res. 39, 1109. Theng, B.K.G., Yuan, G., 2008. Nanoparticles in the soil environment. Elements 4, 395–399. Tian, Y., Gao, B., Silvera-Batista, C., Ziegler, K., 2010. Transport of engineered nanoparticles in saturated porous media. J. Nanopart. Res. 12, 2371–2380. Uyusur, B., Darnault, C.J.G., Snee, P.T., Kokën, E., Jacobson, A.R., Wells, R.R., 2010. Coupled effects of solution chemistry and hydrodynamics on the mobility and transport of quantum dot nanomaterials in the vadose zone. J. Contam. Hydrol. 118, 184–198. Wiesner, M.R., Lowry, G.V., Alvarez, P., Dionysiou, D., Biswas, P., 2006. Assessing the risks of manufactured nanomaterials. Environ. Sci. Technol. 40, 4336–4345. Zvikelsky, O., Weisbrod, N., Dody, A., 2008. A comparison of clay colloid and artificial microsphere transport in natural discrete fractures. J. Colloid Interface Sci. 323, 286–292.