Interaction of Fullerene (C60) Nanoparticles with Humic Acid and

Sep 10, 2008 - The deposition kinetics of fullerene (C60) nanoparticles onto bare silica surfaces and surfaces precoated with humic acid and alginate ...
0 downloads 0 Views 259KB Size
Download PDF
Environ. Sci. Technol. 2008, 42, 7607–7614

Interaction of Fullerene (C60) Nanoparticles with Humic Acid and Alginate Coated Silica Surfaces: Measurements, Mechanisms, and Environmental Implications K A I L O O N C H E N * ,† A N D MENACHEM ELIMELECH Department of Chemical Engineering, Environmental Engineering Program, Yale University, New Haven, Connecticut 06520-8286

Received May 4, 2008. Revised manuscript received June 21, 2008. Accepted July 14, 2008.

The deposition kinetics of fullerene (C60) nanoparticles onto bare silica surfaces and surfaces precoated with humic acid and alginate are investigated over a range of monovalent (NaCl) and divalent (CaCl2) salt concentrations using a quartz crystal microbalance. Because simultaneous aggregation of the fullerene nanoparticles occurs, especially at higher electrolyte concentrations, we normalize the observed deposition rates by the corresponding favorable (transport-limited) deposition rates to obtain the attachment efficiencies, R. The deposition kinetics of fullerene nanoparticles onto bare silica surfaces are shown to be controlled by electrostatic interactions and van der Waals attraction, consistent with the classical particle deposition behavior where both favorable and unfavorable deposition regimes are observed. The presence of dissolved humic acid and alginate in solution leads to significantly slower deposition kinetics due to steric repulsion. Precoating the silica surfaces with humic acid and alginate exerts similar steric stabilization in the presence of NaCl. In the presence of CaCl2, the deposition kinetics of fullerene nanoparticles onto both humic acid- and alginate-coated surfaces are relatively high, even at relatively low (0.3 mM) calcium concentration. This behavior is attributed to the macromolecules undergoing complex formation with calcium ions, which reduces the charge and steric influences of the adsorbed macromolecular layers.

Introduction The unique physicochemical, electrical, and mechanical properties of the Buckminsterfullerene C60 molecule have made it one of the most sought after materials for a wide range of existing and potential applications in commercial and scientific industries (1-3). As the use of fullerene C60 in products continues to grow rapidly, so does the concern for the fate of these molecules in the environment (4). It is expected that some of the fullerenes will be released into natural aquatic systems. Since fullerene C60 has extremely * Corresponding author e-mail: [email protected]; phone: (410) 516-7095. † Current address: Department of Geography and Environmental Engineering, Johns Hopkins University, Baltimore, Maryland 212182686. 10.1021/es8012062 CCC: $40.75

Published on Web 09/10/2008

 2008 American Chemical Society

low solubility in water, it is likely to exist as fullerene C60 nanoparticles once released into the aqueous phase (5, 6). Fullerene nanoparticles have been shown to be negatively charged in aqueous solutions (5, 7, 8), which has important implications for their fate and transport in natural systems. Evidence from several studies has revealed that fullerene C60 nanoparticles exhibit toxic effects on bacteria, fish, and human cells (5, 6, 9). However, the mechanisms behind their toxicity are still being determined (10-12). Recently, Tong et al. (13) showed that soil microbes exposed to fullerene nanoparticles are not significantly impacted. These results were attributed to the attachment of the nanoparticles to soil surfaces, thereby reducing their exposure to the microbes. Hence, it is apparent that the attachment/deposition behavior of fullerene nanoparticles plays a critical role in determining their bioavailability and toxicity. The mobility and transport of fullerene nanoparticles in aquatic environments are strongly dependent on their aggregation and deposition behavior. In such environments, naturally occurring particles and surfaces are ubiquitous. Hence, interactions of fullerene nanoparticles with these natural particles and surfaces are more likely to occur than fullerene-fullerene interactions. The aggregation behavior of fullerene nanoparticles was previously demonstrated to be in remarkable agreement with the classic DerjaguinLandau-Verwey-Overbeek (DLVO) theory of colloidal stability (14). However, to date, no systematic studies have been conducted on heteroaggregation of fullerene nanoparticles, that is, the aggregation and interaction of fullerene nanoparticles with other particles/surfaces. Deposition of nanoparticles can be aptly considered a form of heteroaggregation since the immobile collector is usually of a different material and can be perceived as an infinitely large stationary particle (15). Two recent column filtration studies (16, 17) showed that the deposition behavior of fullerene nanoparticles onto glass collector beads and Ottawa sand was dependent on the salt concentration. In our previous study with the quartz crystal microbalance (QCM), we found the deposition kinetics of fullerene nanoparticles onto a silica surface to be in qualitative agreement with the DLVO theory (14). Results in that study also revealed that concurrent aggregation of the fullerene nanoparticles, which occurs under high ionic strength conditions, reduces the rate of deposition due to the decrease in the convectivediffusive transport of the aggregates (14). Natural organic matter (NOM) exists in natural aquatic environments, with two of its main components being humic substances and polysaccharides (18-20). NOM readily adsorbs to solid surfaces and nanoparticles in aquatic systems, thus affecting the aggregation and deposition behavior of nanoparticles. Our recent study demonstrated that the presence of humic acid in the background solution reduces the aggregation kinetics of fullerene nanoparticles considerably due to steric repulsion induced by the humic acid macromolecules adsorbed onto the nanoparticle surfaces (21). This behavior is also evident from other studies (16, 22, 23) which reported steric stabilization of fullerene nanoparticles through NOM adsorption. The objective of this study is to further our understanding of the influence of NOMsdissolved in solution and adsorbed on solid surfacesson the deposition kinetics of fullerene nanoparticles. By employing the QCM, the interaction of fullerene nanoparticles with bare silica surfaces and silica surfaces coated with either humic acid or alginate was investigated over a range of monovalent (NaCl) and divalent (CaCl2) salt concentrations. The deposition kinetics of the VOL. 42, NO. 20, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7607

FIGURE 1. Deposition rates of fullerene nanoparticles onto bare silica surfaces as functions of (a) NaCl and (b) CaCl2 concentrations at pH 5.5. Each data point represents the mean of duplicate measurements at the same experimental conditions, with error bars representing standard deviations. The lines are meant to guide the eye. The inset presents the representative frequency shifts as deposition took place at 10 and 30 mM NaCl starting at 10 min. The deposition rates presented here are obtained from the slopes of the frequency shifts. fullerene nanoparticles onto the bare and modified silica surfaces were compared. Subsequently, the role of these macromolecules in controlling the interaction of nanoparticles with aquatic surfaces was elucidated and discussed.

Materials and Methods Fullerene Nanoparticle Preparation and Characterization. The fullerene nanoparticles were synthesized by the method described in our previous publication (14) and in the Supporting Information. Through high-temperature oxidation at 680 °C (TOC-V CSH, Shimadzu, Kyoto, Japan), the synthesized fullerene nanoparticle stock suspension was determined to have a total carbon content of 11.6 mg/L. Examination by transmission electron microscopy (TEM) revealed that the nanoparticles are polydisperse and mostly spherical, similar to those in the TEM image presented in our previous publication (Figure 1 in ref 14). Through dynamic light scattering (DLS) (ALV-5000, Langen, Germany), the intensity-weighted hydrodynamic radius of the fullerene nanoparticles was determined to be 50.5 nm ((1.0 nm) based on 40 measurements. The electrophoretic mobilities of the 7608

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 20, 2008

fullerene nanoparticles were also measured across a pH range of 2-12 at a background electrolyte of 1 mM KCl and temperature of 25 °C (ZetaPALS, Brookhaven Instruments Corp., Holtsville, NY). These measurements verified that the nanoparticles were negatively charged under such solution conditions (Figure S1 in the Supporting Information). Humic Acid and Alginate Solutions. We used Suwannee River humic acid (Standard II, International Humic Substances Society) and alginate (A2158, Sigma-Aldrich, St. Louis, MO) as models for naturally occurring humic substances and polysaccharides, respectively. The MWs of the humic acid and alginate employed are 1-5 and 12-80 kDa, respectively. Their chemical properties have been described in our earlier publications (21, 24). The preparation of the humic acid and alginate stock solutions is described in the Supporting Information. Solution Chemistry. The NaCl and CaCl2 electrolyte stock solutions were prepared and filtered through 0.1 µm filters (Anotop 25, Whatman, Middlesex, UK) before use. All experiments and measurements were conducted at unadjusted, ambient pH of the diluted fullerene nanoparticle suspension (pH 5.5 ( 0.1), at which the fullerene nanoparticles are stable to aggregation at low ionic strengths. All electrolyte solutions were prepared with the stock solutions, degassed through sonication for 10 min, and stored in a water bath at 27 °C before use. Quartz Crystal Microbalance. The deposition kinetics of fullerene nanoparticles onto silica and organic matter-coated surfaces were measured using a QCM. The QCM unit (D300, Q-Sense AB, Va¨stra Fro¨lunda, Sweden) comprises a measurement chamber which holds a 5 MHz AT-cut quartz sensor crystal with a silica-coated surface (QSX 303, Q-Sense). Before use, the sensor crystal was thoroughly cleaned as described previously (14). The measurement chamber takes a radial stagnation point flow configuration (15), which allows the test solution to be directed toward the center of the crystal surface before flowing out from the edges with the aid of a syringe pump (suction mode). For all experiments, the flow rate of the test solutions and nanoparticle suspensions through the measurement chamber was maintained constant at 0.1 mL/min. According to the manufacturer, such flow rate results in laminar flow in the chamber. There is also a Peltier element which maintains the solution in the chamber at a constant temperature of 25.0 °C. For each deposition experiment, the electrolyte solution, having the concentration at which deposition was to be conducted, was first flowed across the crystal surface until the normalized third overtone frequency monitored by the QCM stabilized. The normalized frequency was considered to have achieved stabilization when it ceased to drift more than 0.3 Hz in a time period of 10 min. Once a stable baseline was attained, a premeasured volume of electrolyte stock solution was introduced into the fullerene nanoparticle suspension immediately prior to the deposition experiment to obtain the desired electrolyte concentration. The suspension was then flowed through the measurement chamber. A fullerene nanoparticle concentration of 5.8 mg/L total carbon content was used for all experiments in order to achieve sufficient deposition across a range of electrolyte concentrations. As deposition of the fullerene nanoparticles on the bare or coated silica surface took place, the continuous increase in the mass of the crystal due to the deposited nanoparticles induced a continuous shift in the resonance and overtone frequencies, as described by the Sauerbrey relationship (25): ∆m ) -

C∆fn n

(1)

where ∆m is the mass of fullerene nanoparticles deposited, ∆fn is the shift in resonance (or overtone) frequency, with n

being the overtone number (1, 3, 5, 7), and C is the crystal constant (17.7 ng/(Hz-cm2)). Time periods of between 20 and 60 min were provided to allow sufficient deposition for determining the nanoparticle deposition rate. All deposition experiments were duplicated for each experimental condition. Determination of Initial Deposition Kinetics. Since the changes in resonance and overtone frequencies are proportional to the mass deposited on the crystal, the rate of deposition is indicated by the rate of frequency shift. In our study, we monitored the shifts in the normalized third overtone frequency (∆f(3) ) ∆f3/3) to determine the deposition rates at different electrolyte concentrations. The attachment efficiency, R, is used to quantify the deposition kinetics at each electrolyte concentration: R)

d∆f(3) ⁄ dt

(d∆f(3) ⁄ dt)fav

(2)

Here, the numerator represents the rate of shift in the normalized third resonance frequency at the electrolyte concentration of interest. The denominator represents the corresponding rate under favorable (nonrepulsive) deposition conditions obtained at the same electrolyte and nanoparticle concentrations. Since we are interested in the initial (early stage) deposition kinetics, the initial slopes of the shift in the normalized third overtone frequency were used in cases where the rate of frequency shift was not constant with time. One possible reason for the changing rate of frequency shift with time is the simultaneous occurrence of significant aggregation of the fullerene nanoparticles during deposition (14). Also, simultaneous aggregation of the nanoparticles will lead to a drop in the initial rate of deposition due to a lower rate of convective-diffusive transport of the aggregates toward the silica surface (14). Thus, it is necessary to normalize the rate of frequency shift at the electrolyte concentration of interest by the favorable rate determined at the same electrolyte concentration in order to determine the attachment efficiency independent of the influence of aggregation, as presented in eq 2. This favorable rate has been determined experimentally as described below. Because the fullerene nanoparticles and silica surface are both negatively charged at the studied solution conditions (pH 5.5), favorable deposition conditions were induced by modifying the silica surface through precoating with a layer of cationic poly-L-lysine (PLL) polyelectrolyte (P-1274, SigmaAldrich, St. Louis, MO) as described elsewhere (ref 26 and Supporting Information). The PLL-coated surface was exposed to 2 mL of the electrolyte solution of interest before the fullerene nanoparticles were deposited onto the PLLcoated surface in the presence of the same electrolyte solution. Under such conditions, it is reasonable to assume that all fullerene nanoparticles or aggregates contacting the oppositely charged PLL-coated surface would result in a permanent attachment. Precoating Silica Surfaces with Humic Acid and Alginate. For deposition experiments conducted on surfaces precoated with humic acid and alginate, the bare silica surfaces were first coated with a layer of PLL, as discussed in the previous section. This PLL layer was then rinsed with 2 mL of HEPES solution followed by 2 mL of 1 mM NaCl. A second layer of negatively charged macromolecules was adsorbed onto the PLL layer by flowing 2 mL of humic acid or alginate (30 mg/L TOC) prepared in 1 mM NaCl solution across the PLL-coated surface (26). The fast adsorption of humic acid and alginate under such conditions usually leads to sharp frequency shifts ∆f(3) of about 5 Hz. A representative profile of the frequency shift is presented in Figure S2. The baseline stabilized within 10 min of exposing the PLL layer to the humic acid or alginate solutions, indicating complete coverage of the PLL layer by the macromolecules. The formed layer of humic acid or alginate macromolecules was rinsed with 2 mL of 1 mM NaCl

solution. Finally, the electrolyte of interest was flowed across the coated surface until a constant baseline was obtained, before the nanoparticles prepared in the same electrolyte solution were introduced for the deposition experiment.

Results and Discussion Deposition on Bare Silica Surfaces. The rates of fullerene nanoparticle deposition onto bare silica surfaces over a range of NaCl and CaCl2 concentrations are presented in Figure 1 as rates of frequency shifts, d∆f(3)/dt. The inset in Figure 1a shows representative frequency shifts when deposition of fullerene nanoparticles occurred at 10 and 30 mM NaCl. At 30 mM NaCl, the rate of frequency shift was much higher compared to that at 10 mM NaCl, indicating that the rate of deposition was much higher at 30 mM NaCl. The trend for the deposition rates as functions of NaCl and CaCl2 concentrations shown in this figure is similar to what we have observed previously (14). Within the lower range of electrolyte concentrations, as the concentration increases for both electrolytes, it results in a greater degree of charge screening of the fullerene nanoparticles and silica surfaces, leading to an increase in the rate of deposition. However, as the electrolyte concentrations are further increased, significant aggregation of the suspended nanoparticles occurs during the deposition run. The increase in the size of the aggregates results in reduced convective-diffusive transport of the aggregates toward the silica surface. To derive the deposition kinetics in the form of attachment efficiencies as described in eq 2, the fullerene nanoparticles were deposited onto PLL-coated silica surfaces (favorable deposition conditions) over a range of NaCl and CaCl2 concentrations. Figure 2 presents the rates of fullerene nanoparticle deposition under such conditions. We observe maximum deposition rates at low electrolyte concentrations (1-10 mM NaCl and 0.3-1.0 mM CaCl2) where the fullerene nanoparticles are reasonably stable to aggregation within the time period of the deposition run. The average maximum deposition rates observed in the presence of low NaCl and CaCl2 concentrations are relatively close (3.88 and 3.60 Hz/ min, respectively), indicating fast transport of mostly unaggregated fullerene nanoparticles toward the PLL layer. These values correspond to mass deposition rates of 68.7 and 63.7 ng/(min-cm2), respectively, calculated using eq 1. At concentrations above 10 mM NaCl and 1 mM CaCl2, the rate of deposition decreases with increasing electrolyte concentrations. At these electrolyte concentrations, significant aggregation of the fullerene nanoparticles occurs, as illustrated by the schematics in Figure 2. By conducting timeresolved dynamic light scattering (DLS) measurements on the fullerene nanoparticles at the same nanoparticle concentration employed for the deposition experiments, considerable aggregation was shown to occur at higher electrolyte concentrations. For instance, the hydrodynamic radius of the nanoparticles increased from 50.5 to over 220 nm in the presence of 60 mM NaCl within 20 min of introducing the premeasured amount of NaCl stock solution into the stable nanoparticle suspension. In the presence of 10 mM NaCl, the hydrodynamic radius increased only slightly to about 52 nm over the same time period. The stark difference in the deposition kinetics under favorable conditions at these two NaCl concentrations as a result of the occurrence of aggregation is clearly reflected in Figure 2a. The aggregation profiles obtained from time-resolved DLS and deposition profiles obtained from the QCM measurements at these two NaCl concentrations are also presented in Figure S3. By normalizing the nanoparticle deposition rates (Figure 1) by the favorable deposition rates at the corresponding electrolyte concentrations (Figure 2), the attachment efficiencies over the range of NaCl and CaCl2 concentrations are derived (Figure 3). The deposition inverse stability profiles VOL. 42, NO. 20, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7609

FIGURE 2. Favorable deposition rates of fullerene nanoparticles onto PLL-coated silica surfaces as functions of (a) NaCl and (b) CaCl2 concentrations at pH 5.5. Duplicate measurements were conducted at 1 and 10 mM NaCl, with error bars representing standard deviations, while single measurements were conducted at other electrolyte concentrations. The lines are meant to guide the eye. The illustration shows that the fullerene nanoparticles remain stable to aggregation at low electrolyte concentrations while aggregating at higher electrolyte concentrations. The images of discrete fullerene nanoparticles and aggregates are obtained under the transmission electron microscope (TEM). (i.e., the log of attachment efficiencies against the log of electrolyte concentrations) are typical of classical deposition behavior in which the deposition kinetics are controlled by the electrostatic interactions and van der Waals attraction between the particles and solid surfaces (15). At low electrolyte concentrations, electrostatic repulsion occurs between the negatively charged fullerene nanoparticles and silica surface, resulting in unfavorable deposition (i.e., R < 1). As the electrolyte concentration is increased, the degree of charge screening and double layer compression of the nanoparticles and silica surfaces increases, resulting in higher deposition kinetics. When the electrolyte concentration equals and exceeds the critical deposition concentration (CDC), sufficient charge screening occurs to allow for favorable deposition (i.e., R ≈ 1). The intersections between the extrapolations from the unfavorable and favorable deposition regimes yield the CDC values of 32.1 mM for NaCl and 0.7 mM for CaCl2. It is useful to point out that this normalization allows for the isolation of the influence of solution chemistry and surface properties of the nanoparticles and solid surface on the deposition kinetics by eliminating the effects of nanoparticle 7610

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 20, 2008

FIGURE 3. Attachment efficiencies of fullerene nanoparticles onto bare silica surfaces as functions of (a) NaCl and (b) CaCl2 concentrations at pH 5.5. Each data point represents the mean of duplicate measurements at the same experimental conditions, with error bars representing standard deviations. The dashed lines are extrapolated from the favorable and unfavorable deposition regimes, and their intersections yield the respective CDC of 32.1 mM for NaCl and 0.7 mM for CaCl2. aggregation. In other words, this normalization would yield the same results as deposition experiments employing extremely low nanoparticle concentrations such that aggregation would be insignificant even at high electrolyte concentrations. However, it would not be realistic to conduct such experiments as deposition would have to take place over a long period of time in order to achieve sufficient deposition for the determination of deposition kinetics. Deposition on Silica Surfaces in the Presence of Background Humic Acid and Alginate. Deposition experiments were conducted in the presence of background humic acid and alginate (1 mg/L TOC) at electrolyte concentrations of 30 mM NaCl and 1 mM CaCl2. These electrolyte concentrations are close to the CDC values determined in the absence of background macromolecules as discussed in the previous section. Prior to these deposition experiments, the bare silica surfaces were rinsed with the electrolytes of interest, followed by humic acid and alginate solutions (1 mg/L TOC) prepared in the same electrolytes. No significant frequency shifts were observed when the humic acid and alginate solutions were introduced at 30 mM NaCl, indicating that the macromolecules do not adsorb onto the silica surfaces under such conditions. However, in the presence of 1 mM CaCl2, significant adsorption of both macromolecules occurred due to complex formation with Ca2+ resulting in charge neutralization (24, 27). In this case, the adsorption of macromolecules caused frequency shifts, ∆f(3), of about 15 and 40 Hz over 70-80 min for humic acid and alginate, respectively. The rate of adsorption of macromolecules was observed to decrease with time. When the baseline stabilized after 70-80

FIGURE 4. Attachment efficiencies of fullerene nanoparticles prepared in humic acid and alginate solutions (both at 1 mg/L TOC) onto bare silica surfaces in the presence of 30 mM NaCl and 1 mM CaCl2 at pH 5.5. The attachment efficiencies for deposition in the absence of macromolecules in the solutions are reproduced from Figure 3. Each bar represents the mean of duplicate measurements at the same experimental conditions, with error bars representing standard deviations. Humic acid and alginate are not observed to adsorb on bare silica surfaces at 30 mM NaCl, as presented in the illustration (not drawn to scale). At 1 mM CaCl2, humic acid and alginate macromolecules take a compact conformation and adsorb onto bare silica surfaces significantly. The fullerene nanoparticle suspensions prepared in the presence of the respective macromolecules were only introduced after adsorption of the macromolecules on the silica surfaces ceased. min, indicating surface saturation and no more adsorption of macromolecules on the silica surface, the nanoparticles prepared in the humic acid or alginate solutions were introduced into the QCM for the deposition run. Insignificant aggregation of the fullerene nanoparticles occurs in the presence of background humic acid and alginate at 30 mM NaCl and 1 mM CaCl2 as verified through timeresolved DLS (Figure S4). Hence, the attachment efficiencies of the fullerene nanoparticles in the presence of these macromolecules were calculated using the maximum favorable deposition rate determined at 1 mM NaCl (4.03 Hz/ min, Figure 2) as the denominator in eq 2. These attachment efficiencies are presented in Figure 4. The schematic in the figure illustrates that the deposition behavior is dominated by the interactions between the coated fullerene nanoparticles and bare silica surface in the presence of 30 mM NaCl. Conversely, the interactions between the coated nanoparticles and coated silica surface control the deposition kinetics in the presence of 1 mM CaCl2. At 30 mM NaCl, the attachment efficiency in the presence of humic acid is about 3 orders of magnitude lower than that in the absence of humic acid. This drastic decrease in the deposition kinetics is due to the fast adsorption of humic acid onto the fullerene nanoparticles, resulting in steric repulsion between the coated nanoparticles and bare silica

surfaces. These results correspond to our previous findings that the aggregation kinetics of fullerene nanoparticles are greatly decreased in the presence of humic acid due to steric repulsion between the humic acid-coated nanoparticles (21). In comparison, the presence of alginate in the solution led to a decrease in the attachment efficiency by almost 2 orders of magnitude. The greater impact of humic acid is possibly due to humic acid adsorbing more readily onto fullerene nanoparticles through π-π interactions between their aromatic rings (22, 28) compared to alginate, which adsorbs onto fullerene nanoparticles likely through hydrophobic interactions. Other evidence that supports the higher propensity for adsorption of humic acid over alginate is the greater degree of stabilization of the fullerene nanoparticles with humic acid at 30 mM NaCl as well as 1 mM CaCl2, compared to the corresponding results with alginate (Figure S4). In the presence of 1 mM CaCl2, the adsorption of humic acid and alginate onto both fullerene nanoparticles and silica surfaces results in some degree of steric repulsion. Both humic acid and alginate are able to form complexes with Ca2+, which may result in both macromolecules taking a more compact conformation (24, 27). Also, some bridging may occur between the macromolecules on the fullerene nanoparticles and silica surface in the presence of Ca2+ (21, 24). The combination of both these factors may explain why the degree of stabilization in the presence of both macromolecules at 1 mM CaCl2 is not as pronounced as that observed in the presence of humic acid at 30 mM NaCl. Deposition on Silica Surfaces Coated with Humic Acid. Figure 5 presents the attachment efficiencies of the fullerene nanoparticles on silica surfaces precoated with humic acid. The schematic in Figure 5 illustrates that the interactions between the bare fullerene nanoparticles and humic acidcoated surfaces play an important role in controlling the deposition behavior. In the presence of 10 and 30 mM NaCl with no background humic acid in the solution, the deposition kinetics for the coated silica surfaces are about 1 order of magnitude lower than those for bare silica surfaces. The humic acid layer on the silica surface effectively retarded the deposition of fullerene nanoparticles predominantly through steric repulsion. The complexity of the system was increased when deposition of the fullerene nanoparticles onto the humic acid-coated surface was conducted in the presence of background humic acid (1 mg/L TOC) in solution at 30 mM NaCl. This was performed by rinsing the humic acidcoated surface with the electrolyte of interest (i.e., 30 mM NaCl in this case) followed by the same electrolyte with a background humic acid until a constant baseline was achieved. Following that, the fullerene nanoparticles prepared in the presence of background humic acid at 30 mM NaCl were introduced into the QCM module. The adsorption of humic acid on the fullerene nanoparticles gave rise to a greater degree of steric stabilization on top of the humic acid layer on the silica surface, as was evident from the further decrease in the attachment efficiency from 0.11 to 0.038. In the presence of 0.3 and 1.0 mM CaCl2, the attachment efficiencies of fullerene nanoparticles onto humic acidcoated surfaces are fairly high at 0.24 and 0.38, respectively. Since the humic acid layers were rinsed with the respective electrolytes until the frequency shifts ∆f(3) ceased to change, they were likely to have reached equilibrium by the end of the rinse for both 0.3 and 1.0 mM CaCl2 solutions, where Ca2+ had been bound to almost all available carboxyl functional groups of the humic acid macromolecules (27). Hence, regardless of the concentration of CaCl2 employed in the rinse solutions, the conformational and charge properties of the humic acid layers should be comparable at the end of the rinses. This rinsing effect on the layers may lead to the comparable deposition kinetics onto the coated surfaces at VOL. 42, NO. 20, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7611

FIGURE 5. Attachment efficiencies of fullerene nanoparticles onto humic acid-coated silica surfaces in the absence and presence of humic acid in solution as functions of (a) NaCl and (b) CaCl2 concentrations at pH 5.5. Each bar represents the mean of duplicate measurements at the same experimental conditions, with error bars representing standard deviations. The attachment efficiencies for deposition onto bare silica surfaces are reproduced from Figure 3. The humic acid-coated surfaces are prepared by first coating the silica surface with a positively charged PLL layer, before coating the PLL layer with a second layer of humic acid, as shown in the illustration (not drawn to scale). both CaCl2 concentrations. The saturation of the humic acid layers with Ca2+ in both cases could also result in substantial reduction in the charge of the layers, thereby allowing for relatively high deposition kinetics. The addition of background humic acid led to a drastic drop in attachment efficiency of fullerene nanoparticles from 0.38 to 0.014 in 1.0 mM CaCl2. This attachment efficiency of 0.014 is of the same order of magnitude as that for the deposition in the presence of background humic acid in 1.0 mM CaCl2 (R ) 0.038, Figure 4). This result is reasonable since both systems are similar with humic acid macromol7612

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 20, 2008

FIGURE 6. Attachment efficiencies of fullerene nanoparticles onto bare and alginate-coated silica surfaces in the absence of alginate in solution as functions of (a) NaCl and (b) CaCl2 concentrations at pH 5.5. Each bar represents the mean of duplicate measurements at the same experimental conditions, with error bars representing standard deviations. The attachment efficiencies for deposition onto bare silica surfaces are reproduced from Figure 3. The inset presents the representative frequency shift as deposition onto the alginate-coated surface took place at 10 mM NaCl starting at 6 min. The initial slope (represented by the dashed line) is used to derive deposition rate. ecules adsorbing on both the bare and PLL-modified silica surfaces readily in the presence of CaCl2. Deposition on Silica Surfaces Coated with Alginate. The attachment efficiencies of fullerene nanoparticles onto alginate-coated silica surfaces are presented in Figure 6. All deposition experiments were conducted in the absence of background alginate in solution. At 10 mM NaCl, we observed two regimes in the frequency shift when deposition took place onto the alginate-coated surface, as presented in the inset of Figure 6a. Since we are interested in the initial deposition kinetics, we reported the attachment efficiency at 10 mM NaCl based on the initial slope represented by the dashed line. Because the alginate macromolecules are larger than humic acid, the alginate layer is likely much rougher than the humic acid layer, especially at the lower NaCl

concentration of 10 mM where the adsorbed alginate could take a more extended conformation from the silica surface. This irregular surface may invariably allow for the fullerene nanoparticles to be trapped within the depressions of the alginate layer. After all the depressions are filled with nanoparticles, further incoming nanoparticles do not deposit as readily, which explains the higher initial deposition rate and the lower deposition rate in the later stages. At 30 mM NaCl, we observed a constant rate of deposition with time, most likely because the alginate layer becomes more compacted at higher salt concentration due to charge screening of the alginate carboxyl groups, resulting in a smoother surface. The compaction of the alginate layer is consistent with our earlier findings (24). Through DLS, we previously showed that the hydrodynamic radius of alginatecoated hematite nanoparticles decreased by about 5 nm when the NaCl concentration was raised from 0 to 20 mM. This result leads us to believe that the thickness of the alginate layer on the silica surface may also vary within the same order of magnitude with electrolyte concentration, which will substantially affect the surface roughness. In addition, the greater degree of charge screening of the fullerene nanoparticles and alginate-coated surfaces at 30 mM NaCl will also result in faster deposition kinetics than at 10 mM NaCl. The trends for the deposition kinetics of the fullerene nanoparticles onto alginate-coated surfaces in the presence of CaCl2 are similar to those for humic acid-coated surfaces. The deposition kinetics onto the alginate-coated surfaces are relatively high, likely due to alginate forming complexes with Ca2+ which leads to charge neutralization of the alginate layer. Alginate also undergoes gelation in the presence of Ca2+, which results in the modification of the physical properties and surface morphology of the alginate layer (26). This surface modification is likely to exert some influence over the deposition behavior of the fullerene nanoparticles. Implications for Fate and Transport in Aquatic Environments. The deposition kinetics of fullerene nanoparticles onto bare silica surfaces are consistent with classical deposition behavior. Favorable deposition of the fullerene nanoparticles on silica surfaces starts to occur at ∼30 mM NaCl and 0.7 mM CaCl2, which is higher than typical monovalent cation concentrations of freshwaters, but within the concentration range of Ca2+ (29, 30). Hence, in freshwaters where NOM concentration is very low, the affinity of fullerene nanoparticles to attach to silica surfaces may be controlled by divalent cations such as Ca2+. In most natural aquatic systems, however, NOM is present both in the aqueous phase as well as on the surface of minerals and suspended particles. In freshwater systems, NOM concentration is typically on the order of several mg/L (TOC). Even at a conservative humic acid and alginate concentration (1 mg/L TOC), the deposition kinetics of fullerene nanoparticles can be reduced by up to 3 orders of magnitude through steric repulsion. In contrast, preadsorbed NOM on silica surface can either retard or enhance the deposition kinetics, depending on the solution ionic composition and the physicochemical properties of the NOM macromolecules. The elucidation of the influence each component of NOM exerts on the deposition behavior of fullerene nanoparticles is the first step to understanding and eventually predicting their fate and transport in real systems where all components coexist in varying proportions.

Acknowledgments Funding was provided by the National Science Foundation (BES 0504258).

Supporting Information Available Electrophoretic mobilities of fullerene nanoparticles as a function of pH (Figure S1); frequency shift when silica surface is precoated with humic acid (Figure S2); aggregation and

deposition profiles of fullerene nanoparticles at 10 and 60 mM NaCl (Figure S3); aggregation profiles of fullerene nanoparticles in humic acid and alginate solutions (Figure S4). Also presented are details on procedures for synthesis of fullerene nanoparticles, preparation of humic acid and alginate stock solutions, and modification of silica surface with PLL. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Eckert, J. F.; Nicoud, J. F.; Nierengarten, J. F.; Liu, S. G.; Echegoyen, L.; Barigelletti, F.; Armaroli, N.; Ouali, L.; Krasnikov, V.; Hadziioannou, G. Fullerene-oligophenylenevinylene hybrids: Synthesis, electronic properties, and incorporation in photovoltaic devices. J. Am. Chem. Soc. 2000, 122, 7467–7479. (2) Venturini, J.; Koudoumas, E.; Couris, S.; Janot, J. M.; Seta, P.; Mathis, C.; Leach, S. Optical limiting and nonlinear optical absorption properties of C60-polystyrene star polymer films: C60 concentration dependence. J. Mater. Chem. 2002, 12, 2071– 2076. (3) Bosi, S.; Da Ros, T.; Spalluto, G.; Prato, M. Fullerene derivatives: an attractive tool for biological applications. Eur. J. Med. Chem. 2003, 38, 913–923. (4) Colvin, V. L. The potential environmental impact of engineered nanomaterials. Nat. Biotechnol. 2003, 21, 1166–1170. (5) Fortner, J. D.; Lyon, D. Y.; Sayes, C. M.; Boyd, A. M.; Falkner, J. C.; Hotze, E. M.; Alemany, L. B.; Tao, Y. J.; Guo, W.; Ausman, K. D.; Colvin, V. L.; Hughes, J. B. C60 in water: Nanocrystal formation and microbial response. Environ. Sci. Technol. 2005, 39, 4307–4316. (6) Sayes, C. M.; Fortner, J. D.; Guo, W.; Lyon, D.; Boyd, A. M.; Ausman, K. D.; Tao, Y. J.; Sitharaman, B.; Wilson, L. J.; Hughes, J. B.; West, J. L.; Colvin, V. L. The differential cytotoxicity of water-soluble fullerenes. Nano Lett. 2004, 4, 1881–1887. (7) Deguchi, S.; Alargova, R. G.; Tsujii, K. Stable dispersions of fullerenes, C60 and C70, in water. Preparation and characterization. Langmuir 2001, 17, 6013–6017. (8) Brant, J.; Lecoanet, H.; Hotze, M.; Wiesner, M. Comparison of electrokinetic properties of colloidal fullerenes (n-C60) formed using two procedures. Environ. Sci. Technol. 2005, 39, 6343– 6351. (9) Oberdo¨rster, E. Manufactured nanomaterials (Fullerenes, C60) induce oxidative stress in the brain of juvenile largemouth bass. Environ. Health Perspect. 2004, 112, 1058–1062. (10) Sayes, C. M.; Gobin, A. M.; Ausman, K. D.; Mendez, J.; West, J. L.; Colvin, V. L. Nano-C60 cytotoxicity is due to lipid peroxidation. Biomaterials 2005, 26, 7587–7595. (11) Porter, A. E.; Gass, M.; Muller, K.; Skepper, J. N.; Midgley, P.; Welland, M. Visualizing the uptake of C60 to the cytoplasm and nucleaus of human monocyte-derived macrophage cells using energy-filtered transmission electron microscopy and electron tomography. Environ. Sci. Technol. 2007, 41, 3012–3017. (12) Dhawan, A.; Taurozzi, J. S.; Pandey, A. K.; Shan, W. Q.; Miller, S. M.; Hashsham, S. A.; Tarabara, V. V. Stable colloidal dispersions of C60 fullerenes in water: Evidence for genotoxicity. Environ. Sci. Technol. 2006, 40, 7394–7401. (13) Tong, Z. H.; Bischoff, M.; Nies, L.; Applegate, B.; Turco, R. F. Impact of fullerene (C60) on a soil microbial community. Environ. Sci. Technol. 2007, 41, 2985–2991. (14) Chen, K. L.; Elimelech, M. Aggregation and deposition kinetics of fullerene (C60) nanoparticles. Langmuir 2006, 22, 10994– 11001. (15) Elimelech, M.; Gregory, J.; Jia, X.; Williams, R. A. Particle Deposition and Aggregation: Measurement, Modelling and Simulation; Butterworth-Heinemann: Oxford, England, 1995. (16) Espinasse, B.; Hotze, E. M.; Wiesner, M. R. Transport and retention of colloidal aggregates of C60 in porous media: Effects of organic macromolecules, ionic composition, and preparation method. Environ. Sci. Technol. 2007, 41, 7396–7402. (17) Wang, Y.; Li, Y.; Fortner, J. D.; Hughes, J. B.; Abriola, L. M.; Pennell, K. D. Transport and retention of nanoscale C60 aggregates in water-saturated porous media. Environ. Sci. Technol. 2008, 42, 3588–3594. (18) Wilkinson, K. J.; Balnois, E.; Leppard, G. G.; Buffle, J. Characteristic features of the major components of freshwater colloidal organic matter revealed by transmission electron and atomic force microscopy. Colloids Surf., A 1999, 155, 287–310. (19) Walker, H. W.; Bob, M. M. Stability of particle flocs upon addition of natural organic matter under quiescent conditions. Water Res. 2001, 35, 875–882. VOL. 42, NO. 20, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7613

(20) Kozarac, Z.; Cosovic, B.; Mobius, D.; Dobric, M. Interaction of polysaccharides with lipid monolayers. J. Colloid Interface Sci. 2000, 226, 210–217. (21) Chen, K. L.; Elimelech, M. Influence of humic acid on the aggregation kinetics of fullerene (C60) nanoparticles in monovalent and divalent electrolyte solutions. J. Colloid Interface Sci. 2007, 309, 126–134. (22) Terashima, M.; Nagao, S. Solubilization of [60] fullerene in water by aquatic humic substances. Chem. Lett. 2007, 36, 302–303. (23) Duncan, L. K.; Jinschek, J. R.; Vikesland, P. J. C60 colloid formation in aqueous systems: Effects of preparation method on size, structure, and surface charge. Environ. Sci. Technol. 2008, 42, 173–178. (24) Chen, K. L.; Mylon, S. E.; Elimelech, M. Aggregation kinetics of alginate-coated hematite nanoparticles in monovalent and divalent electrolytes. Environ. Sci. Technol. 2006, 40, 1516–1523. (25) Sauerbrey, G. Verwendung Von Schwingquarzen Zur Wagung Dunner Schichten Und Zur Mikrowagung. Z. Phys. 1959, 155, 206–222.

7614

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 20, 2008

(26) de Kerchove, A. J.; Elimelech, M. Formation of polysaccharide gel layers in the presence of Ca2+ and K+ ions: Measurements and mechanisms. Biomacromolecules 2007, 8, 113–121. (27) Hong, S. K.; Elimelech, M. Chemical and physical aspects of natural organic matter (NOM) fouling of nanofiltration membranes. J. Membr. Sci. 1997, 132, 159–181. (28) Hyung, H.; Fortner, J. D.; Hughes, J. B.; Kim, J. H. Natural organic matter stabilizes carbon nanotubes in the aqueous phase. Environ. Sci. Technol. 2007, 41, 179–184. (29) Sayer, M. D. J.; Reader, J. P.; Dalziel, T. R. K. Freshwater acidification: effects on the early life stages of fish. Rev. Fish Biol. Fish. 1993, 3, 95–132. (30) Gorham, E. The Ionic Composition of Some Lowland Lake Waters from Cheshire, England. Limnol. Oceanogr. 1957, 2, 22–27.

ES8012062