Enhanced Mobility of Fullerene (C60) Nanoparticles in the Presence ...

Article pubs.acs.org/est

Enhanced Mobility of Fullerene (C60) Nanoparticles in the Presence of Stabilizing Agents Yonggang Wang,† Yusong Li,‡ Jed Costanza,§ Linda M. Abriola,† and Kurt D. Pennell*,† †

Department of Civil and Environmental Engineering, Tufts University, 200 College Avenue, Medford, Massachusetts 02155, United States ‡ Department of Civil Engineering, University of NebraskaLincoln, 362R Whittier Building, Lincoln, Nebraska 68583, United States § Office of Pesticide Programs, U.S. Environmental Protection Agency, 2777 S. Crystal Dr. Arlington, Virginia 22202, United States S Supporting Information *

ABSTRACT: Experimental and mathematical modeling studies were performed to examine the effects of stabilizing agents on the transport and retention of fullerene nanoparticles (nC60) in water-saturated quartz sand. Three stabilizing systems were considered: naturally occurring compounds known to stabilize nanoparticles (Suwannee river humic acid (SRHA) and fulvic acid (SRFA)), synthetic additives used to enhance nanoparticle stability (Tween 80, a nonionic surfactant), and residual contaminants resulting from the manufacturing process (tetrahydrofuran (THF)). The results of column experiments demonstrated that the presence of THF, at concentrations up to 44.5 mg/L, did not alter nC60 transport and retention behavior, whereas addition of SRHA (20 mg C/L), SRFA (20 mg C/ L), or Tween 80 (1000 mg/L) to the influent nC60 suspensions dramatically increased the mobility of nC60, as demonstrated by coincidental nanoparticle and nonreactive tracer effluent breakthrough curves (BTCs) and minimal nC60 retention. When columns were preflushed with surfactant, nC60 transport was significantly enhanced compared to that in the absence of a stabilizing agent. The presence of adsorbed Tween 80 resulted in nC60 BTCs characterized by a declining plateau and retention profiles that exhibited hyperexponential decay. The observed nC60 transport and retention behavior was accurately captured by a mathematical model that accounted for coupled surfactant adsorption−desorption dynamics, surfactant−nanoparticle interactions, and particle attachment kinetics.

■

INTRODUCTION Fullerene (C60) nanomaterials have attracted considerable attention from scientific and industrial communities due to their unique physical and chemical properties and promising commercial applications.1,2 The rapid approach to mass production of nanomaterials and their inevitable release into the environment3 have led to concerns about their potential environmental and health impacts, necessitating a comprehensive understanding of their fate and transport in the environment.4 In this context, the transport and retention of nanoparticles in porous media are of particular interest when developing appropriate regulatory strategies, handling and disposal procedures, and effective water treatment technologies. The fate and transport of C60 nanoparticles (nC60) in watersaturated porous media are largely determined by their interactions with substances present in the aqueous phase, and are strongly influenced by physical and chemical conditions in porous media.2,5,6 For example, the effects of soil type,7−9 electrolyte species and concentration,6,10 suspension preparation method,11 flow velocity,2,12 and suspension aging13 on nC60 transport and retention have been documented. Far less attention, however, has been directed toward understanding the © 2012 American Chemical Society

potential effects of stabilizing agents on nC60 mobility in porous media. Here, the term stabilizing agent refers to compounds that act to inhibit nanoparticle agglomeration or aggregation, resulting in suspensions that are stable for prolonged periods of time (e.g., weeks to months), and may alter particle deposition on surfaces. Natural organic matter (NOM), which is ubiquitous in aquatic systems, is known to function as a stabilizing agent for colloidal suspensions. Carboxylic and phenolic groups present in NOM associate with both colloid and soil surfaces,14−16 thereby promoting colloid transport due to increased electrosteric repulsion.5,17−20 Experimental evidence of enhanced nC60 transport and reduced deposition in the presence of NOM is limited. Chen and Elimelech21 utilized a quartz crystal microbalance (QCM) to measure nC60 deposition onto flat silica surfaces in the presence of NOM, and reported a reduction in attachment efficiency in the presence of 1 mg/L Received: Revised: Accepted: Published: 11761

June 23, 2012 September 10, 2012 September 13, 2012 September 13, 2012 dx.doi.org/10.1021/es302541g | Environ. Sci. Technol. 2012, 46, 11761−11769

Environmental Science & Technology

Article

Effluent samples were monitored for both nC60 and stabilizing agent over time, and at the conclusion of each experiment the columns were destructively sampled to obtain solid-phase concentrations of nC60. The resulting nC60 effluent breakthrough curves (BTCs) and retention profiles were initially simulated using a mathematical model that accounted for firstorder attachment and a maximum retention capacity. Subsequent modifications to the model were undertaken to account for Tween 80 sorption−desorption dynamics and interactions between the Tween 80 and nC60.

(as total organic carbon) Suwannee River humic acid (SRHA). Espinasse et al.11 reported that the addition of 1 mg/L tannic acid reduced the attachment efficiency of nC60 by a factor of 2− 3 in columns packed with water-saturated glass beads. In a more recent column study, Wang et al.22 found that the nC60 structurally modified with ca. 8 mg/L of SRHA or Suwannee River fulvic acid (SRFA) exhibited similar transport behavior (e.g., effluent breakthrough) to unmodified nC60 at an ionic strength of 0.5 mM NaCl, pH of 6.8, and a pore-water velocity of ca. 4.5 m/d. However, the retention behavior of nC60 during transport through porous media has not been investigated in the presence of representative NOM, such as SRHA or SRFA. Surface-active-agents (surfactants) represent another important class of stabilizing agent that is often present in nanoparticle suspensions. Surfactants are often used as formulation additives to promote nanoparticle dispersions for personal care, household cleaning, and pharmaceutical products.23−25 Thus, during the use and disposal of these products, surfactants are likely to be coreleased with nanoparticles, which could facilitate or reduce the environmental transport of nanoparticles. Substantial effort has been made to assess effects of surfactants on the transport of metallic nanoparticles (e.g., zerovalent iron,26,27 copper oxide,28 silver29,30) and carbon nanotubes;30,31 however, investigations of their influence on nC60 transport and deposition in porous media are very limited. A batch experiment demonstrated that an ethoxylated nonionic surfactant was able to improve the dispersion of solid C60 and the stability of the resulting nC60 suspension.32 Wang et al.22 performed nC60 transport experiments in the presence of surfactants, but found that the influence of anionic and nonionic surfactants (0.6−8 mg/L) was negligible, while a cationic surfactant (0.6 mg/L) drastically decreased nC60 mobility in a soil. The limited data available are insufficient to provide a thorough understanding of the mechanisms governing nC60 transport and retention in the presence of stabilizing agents. In addition to commercial surfactants, residual constituents from the manufacture or preparation of engineered nanomaterials may be present in stock suspensions. In the case of nC60, tetrahydrofuran (THF) is widely employed as a transfer solvent during the preparation of stable aqueous suspensions. Although there are alternative means to prepare stable nC60 suspensions, the THF-as-vehicle method is commonly used due to the reproducibility of particle size and zeta potential relative to other methods.8 However, the use of THF inevitably leads to the presence of residual solvent and decomposition products (e.g., γ-butyrolactone or GBL) in nC60 suspensions.8,33,34 While the toxicity of one THF daughter product has been recognized,33,34 the potential influence of these residual contaminants on nC60 transport and retention in porous media has not been explored. The goal of this study was to evaluate the effects of relevant stabilizing agents on nC60 transport and retention in watersaturated quartz sand. Three types of stabilizing agents were considered: the humic and fulvic acid fractions of a naturally occurring organic matter (SRHA and SRFA), a widely used nonionic surfactant (Tween 80), and a solvent (THF) used in the suspension preparation process. A series of nC60 transport studies was conducted using nC60 influent suspensions that were prepared with SRHA, SRFA, Tween 80, or THF. An additional set of column experiments, preflushed with Tween 80 solution, was undertaken to evaluate the effects of sorbedphase surfactant on nC60 transport and retention behavior.

■

MATERIALS AND METHODS Materials. Fullerene (C60, 99.9%; purified by sublimation) was obtained from the Materials Electronics Research Corp. (Tuscon, AZ). Tetrahydrofuran (THF), methanol, toluene, calcium chloride, sodium chloride, and sodium hydroxide, at certified ACS grade or higher, were purchased from Fisher Scientific (Fair Lawn, NJ). Magnesium perchlorate (ACS reagent grade) was purchased from MP Biomedicals, Inc. (Solon, OH). Tween 80 (Lot 2398A) was purchased from Uniqema (New Castle, DE). The molecular weight and critical micelle concentration (CMC) of Tween 80 are 1310 g/mol and ca. 13 mg/L, respectively.35 SRHA and SRFA (standard II), two well-characterized NOM fractions, were obtained from the International Humic Substances Society (St. Paul, MN). Ottawa sand (40−50 mesh), with a mean diameter (d50) of 0.36 mm, was obtained from U.S. Silica (Berkeley Springs, WV). Prior to use, the Ottawa sand was thoroughly cleaned using a sequential acid wash, water rinse, ultrasonication, and oven-drying procedure.8 Preparation of nC60 Stock Suspension. Stable, monodisperse aqueous suspensions of nC60 were prepared following an established solvent dissolution procedure.8,36 Briefly, the C60-saturated THF solution was manually poured at a rate of 1 L/min into an equal volume (250 mL) of deionized (DI) water (>18.0 MΩ-cm, Nanopure Model D4741, Barnstead International, Dubuque, IA) and mixed at 1000 rpm on a stirrer plate (Fisher Scientific) at room temperature (23 ± 1 °C). The resulting solution underwent three cycles of vacuum evaporation process at 75 °C to minimize THF residual. The nC60 suspension was then vacuum-filtered through a 0.22 μm cellulose acetate membrane (Corning Inc., Corning, NY), and the resulting filtrate was collected and stored in the dark for subsequent use as the nC60 stock suspension with a shelf life of approximately 5 months. Preparation of nC60 Influent Suspensions. Influent nC60 suspensions were prepared by adding 100 mL nC60 stock suspension to an equal volume of background electrolyte solution at a mixing rate of 120 mL/min to achieve an nC60 concentration of ca. 4.5 mg/L in 1 × 10−3 M CaCl2 and 5 × 10−5 M NaHCO3. To investigate the potential effects of stabilizing agents on nC60 transport behavior, nC60 influent suspensions were prepared with concentrated stock solutions containing either SRHA, SRFA, THF, or Tween 80. Stock solutions of SRHA and SRFA were prepared by dissolving 41.5 mg of SRHA powder and 44.3 mg of SRFA powder, respectively, in 100 mL of DI water. After mixing for 24 h, both solutions were adjusted to pH 7.0 with 0.1 M NaOH and filtered through a 0.22 μm cellulose acetate membrane. The total organic carbon content of the SRHA and SRFA stock solutions were determined to be 181.1 and 195.7 mg C/L, respectively, using a Shimadzu TOC-Vw (wet oxidation) analyzer (Columbia, MD). To prepare influent suspensions, 11762

dx.doi.org/10.1021/es302541g | Environ. Sci. Technol. 2012, 46, 11761−11769


Article

Table 1. Experimental Conditions and Fitted Simulation Parameters for nC60 and Stabilizing Agent Column Studies Conducted in 40−50 Mesh Ottawa Sand with 1 × 10−3 M CaCl2 Buffered to pH 7 Using 5 × 10−5 M NaHCO3, at a Pore-Water Velocity of 7.8 m/d C0a (mg/L)

ξb (mV)

dac (nm)

retd (%)

MBe (%)

kattf (1/h)

Smaxg (μg/g)

αh

η0i

C60

4.7

−23.5

94.2

29.0

92.9

0.064

4.7

−25.8

95.4

27.4

89.5

0.105

0.065

C60 + SRHA

3.9

−19.1

96.9

1.2

96.7

0.017

0.063

C60 + SRFA

4.4

−23.6

95.6

0.8

100.1

0.020

0.065

C60 + TW80-1

4.9

−15.1

92.5

0.0

103.1

0.019

0.068

C60 + TW80-2

5.0

−17.0

92.6

0.0

103.1

0.016

0.068

C60, TW80-1

4.2

−24.2

90.6

6.3

96.1

0.004

0.069

C60, TW80-2

4.3

−22.9

93.6

7.1

96.4

1.80 (1.69, 1.95) 1.8 (1.7, 1.92) 0.10 (0.04, 0.16) 0.089 (0.06, 0.11) 0.102 (0.06, 0.15) 0.094 (−0.01, 0.2) 1.5 (−19, 22) 1.4 (−8.2, 11)

0.116

C60 + THF

6.45 (5.53, 7.81) 5.90 (4.92, 6.90) 0.92 (0.05, 1.80) 1.14 (0.66, 1.62) 1.11 (0.33, 1.90) 0.96 (−0.8, 2.7) 0.26 (0.12, 0.39) 0.35 (0.21, 0.49)

0.006

0.067

column identifier

a

Input concentration of nC60, SRHA, SRFA, or Tween 80. bAverage zeta potential of nC60. cnC60 mean hydrodynamic diameter. dMeasured mass percentage of retention. eOverall mass balance. fAttachment rate coefficient. gMaximum retention capacity. hCollision efficiency factor. iSingle collector efficiency. The designations -1 and -2 indicate replicate experiments. The 95% confidence limits of fitted katt and Smax are presented in parentheses. The relatively large 95% confidence intervals for Smax in the last two experiments indicate that the model predictions are not sensitive to the first order attachment rate and maximum retention capacity under the experimental conditions.

the nC60 stock suspension was first mixed with a SRHA, SRFA, THF, or Tween 80 stock solution in a 200 mL volumetric flask. A predetermined volume of background electrolyte stock solution (0.5 M CaCl2 and 0.01 M NaHCO3) was introduced to the mixture, and degassed DI water was then added to achieve the desired final concentration of SRHA (20 mg C/L), SRFA (20 mg C/L), THF (44.5 mg/L), or Tween 80 (1000 mg/L), in 1 × 10−3 M CaCl2 and 5 × 10−5 M NaHCO3. The ionic strength (IS) of these solutions was 3.05 mM, consistent with the 1−10 mM range typically observed in groundwater. Column Experiments. A total of 11 column experiments were conducted following procedures described by Wang et al.8 To summarize, a borosilicate glass column (15 cm length × 2.5 cm i.d., Kontes, Vineland, NJ) was dry packed with 136 g of Ottawa sand in 1-cm increments, yielding a soil bulk density of 1.68 g/cm3 and a porosity of 0.37. The dry column was flushed with CO2 gas to facilitate dissolution of any gas entrapped during the water imbibition process. At least 10 pore volumes (PVs) of degassed background electrolyte solution (1 × 10−3 M CaCl2 and 5 × 10−5 M NaHCO3) were introduced in an upflow mode at a rate of ca. 1 mL/min using a Dynamax SD-200 pump (Varian Inc., Palo Alto, CA) equipped with a 25 mL pump head and a pulse dampener. Following complete saturation of the porous medium with background electrolyte, nonreactive tracer tests were conducted to assess water flow and hydrodynamic dispersion. Three pore volumes (ca. 90 mL) of a 1 × 10−3 M CaBr2 and 5 × 10−5 M NaHCO3 solution were introduced at a flow rate of ca. 1 mL/ min, followed by three pore volumes of 1 × 10−3 M CaCl2 and 5 × 10−5 M NaHCO3 at the same flow rate. Effluent samples were collected continuously and bromide concentrations were measured using an ion selective bromide electrode (ColeParmer Instrument Co., Vernon Hills, IL) connected to an Accumet Model 50 pH meter (Fisher Scientific, Fair Lawn, NJ). The resulting nonreactive tracer breakthrough curves (BTCs) were expressed as relative concentration (C/Co, where Co is the

influent or applied concentration and C is measured effluent concentration) versus the number of dimensional pore volumes of influent solution introduced into the column. These data were fit to a one-dimensional (1-D) form of the advective− dispersive−reactive (ADR) transport equation using CXTFIT ver 2.1,37 to yield hydrodynamic dispersivity values (αD = DH/ vp, where DH is the hydrodynamic dispersion coefficient and vp is the pore-water velocity) that ranged from 0.057 to 0.064 cm, with a retardation factor (RF) of 1.0 ± 0.02 (Figure S1, Supporting Information). Immediately after the nonreactive tracer test, a pulse (ca. 5 PVs) of nC60 suspension (ca. 4.5 mg/L) was introduced into each column at a flow rate of ca. 1 mL/min using a syringe pump (Model 22, Harvard Apparatus, Inc., Holliston, MA), followed by the introduction of ca. 3 PVs of nC60-free solution at the same flow rate. The applied flow rate corresponds to a pore-water velocity of 7.8 m/day, yielding a column residence time of ca. 28 min and a Reynolds number of 0.013 (Re = ρlqd50/μ, where ρl is the liquid density, q is the Darcy velocity, d50 is the mean grain diameter, and μ is the dynamic viscosity of the fluid), which is several orders-of-magnitude below the limit of laminar flow in packed beds (Re < 10). This protocol was followed except in the case of Tween 80 preflood experiments (Exp. TW80, C60-1,2), for which 10 PVs of nC60-free Tween 80 solution (1000 mg/L + 1 × 10−3 M CaCl2 + 5 × 10−5 M NaHCO3) followed by three PVs of background electrolyte solution were introduced into the column, prior to Tween 80free nC60 injection. During each column experiment effluent samples were collected continuously in 15-mL sterile plastic centrifuge tubes (Fisher Scientific) using a Retriever II fraction collector (Teledyne Isco Inc., Lincoln, NE). At the conclusion of each transport experiment, the columns were destructively sectioned into 1.5-cm increments. Retained nC60 was extracted from each solid sample by placing it into ca. 10 mL of DI water, shaking for 3 h (Labquake, Barnstead International), and ultasonicating for 1 min. Experimental conditions and mass 11763



Article

film thickness, J&W Scientific Inc., Folsom, CA) connected to an Agilent 5975 mass selective detector (MSD). For THF analysis, approximately 2 mL of influent or effluent sample were introduced into 22 mL headspace vials containing 2 g of NaCl and then crimp sealed with Teflon-lined septa followed by loading into an HT3 headspace autosampler (TeledyneTekmar, Mason, OH) that was connected to the GC-MSD. The aqueous concentration of GBL was determined by extraction with an equal volume of methylene chloride for 30 min on a Labquake oscillating shaker which was analyzed by direct liquid injection into the GC-MSD. A 4-point calibration curve for GBL quantification was prepared over a concentration range 20−250 mg/L, with a lower detection limit of 10 mg/L. The THF concentration in the nC60 stock was determined to be 1.6 mg/L and GBL was not detected. The mean diameter, size distribution, and electrophoretic mobility of nC60 aggregates in aqueous suspension were determined by dynamic light scattering (DLS) using a Zetasizer Nano ZS analyzer (Malvern Instruments Ltd., Southborough, MA) operated in noninvasive back scattering (NIBS) mode at an angle of 173°. Approximately 1 mL of nC60 suspension was loaded into a disposable cuvette (DTS0012, Malvern Instruments Ltd., Southborough, MA) and analyzed using a green laser at a wavelength of 532 nm and a refractive index of 1.33. The presence of Tween 80, SRHA or SRFA did not alter the refractive index of the aqueous phase. All particle size and electrophoretic mobility measurements were performed in triplicate, and the operation of the system was verified using a monodisperse suspension of polystyrene spheres (3100A, Nanosphere Size Standards, Duke Scientific Corp., Palo Alto, CA) with a mean diameter of 97 nm and a zeta potential standard (−50 mV, DTS1050, Malvern Instruments Ltd., Southborough, MA). The zeta potential of nC60 aggregates was calculated from electrophoretic mobility using the Henry equation.39 The specific surface area of the sand was determined using the multipoint Brunauer−Emmett−Teller (BET) equation based on krypton (Kr) gas adsorption obtained using an ASAP 2405 analyzer (Micromeritics, Norcross, GA) operated at 77 K. The surface composition of clean sand was characterized using a SSX-100 X-ray photoelectron spectrometer (XPS) that was equipped with an Al Kα monochromatized X-ray source and a hemispherical electron energy detector, and operated under a pass energy of 100 V and spot size of 600 μm. The XPS results indicate that the surface of the clean sand used in this study consists of only oxygen and silicon (Figure S2, Supporting Information). Mathematical Modeling. Given uniform packing and laminar flow conditions and assuming the absence of particle− particle interactions, the transport and deposition of particles in the 1-D sand columns can be expressed as30

recoveries for each nC60 column transport experiment are summarized in Table 1. Analytical Methods. The concentration of nC60 in aqueous suspensions in the absence of SRHA or SRFA was determined using a Cary 3E UV−vis spectrophotometer (Varian Inc., Palo Alto, CA) scanned over wavelength range 190−500 nm. The absorbance response at a wavelength of 344 ± 1 nm was used for quantification based on a 5-point calibration curve prepared by serial dilution of a known nC60 stock solution (9.5 mg/L). Concentrations of nC60 in the stock solution and aqueous samples in the presence of SRHA or SRFA were determined using a liquid−liquid extraction process.7 An aliquot (2 mL) of aqueous sample was destabilized with 0.8 mL of 0.1 M Mg(ClO4)2, to which 2.0 mL of neat toluene was added to extract C60 from the aqueous phase. The extraction vials were mixed at 400 rpm for 10 min, followed by mixing at 250 rpm for 12 h (Innova 2100, New Brunswick Scientific Co., Inc., Edison, NJ). After mixing, the neat toluene phase was transferred to a 1.5-mL centrifuge tube (VWR International, West Chester, PA) and separated at 5000 rpm for 10 min (Eppendorf Model 5415D, Brinkmann Instruments Inc., Westbury, NY) to remove large particles (e.g., >0.45 μm) that could clog the high-performance liquid chromatography (HPLC) column. An aliquot (0.3 mL) of toluene was transferred into 1.2 mL methanol, and the combined solution was mixed for 1 min on vortex (Touch mixer model 232, Fisher Scientific). Quantitative analysis of C60 in the toluene− methanol solution was performed using an Agilent 1100 HPLC equipped with an Agilent G1313A autosampler, an Alltima C18 column (150 mm length × 4.6 mm i.d., 5 μm particle size), and a diode array detector (DAD) operated at a wavelength of 334 nm. The HPLC was operated at a constant flow rate of 1 mL/min with an isocratic mobile phase consisting of 55% toluene and 45% methanol. Reference C60 standards were prepared by placing 2 mg of fullerene powder into an amber vial containing 10 mL of toluene, which was placed a sonication bath (Fisher Scientific) for 3 h. A five-point calibration curve over a concentration range of 0.1 to 5 mg/L was prepared by dilution of the C60 reference standard. The detection limits of UV, solvent-extraction, and solid-phase extraction methods for nC60 were 0.03 mg/L, 0.01 mg/L, and 0.1 μg/g, respectively, based on the method of Hubaux and Vos.38 To quantify the concentration of SRHA or SRFA, the aqueous phase obtained from the extraction vial was analyzed using the Cary 3E UV−vis spectrophotometer at a wavelength of 290 nm, where minimum interference of magnesium perchlorate was observed. To determine the concentration of Tween 80 in the presence of nC60, effluent samples were filtered through a 0.02 μm membrane filter (Whatman International Ltd., Maidstone, Kent, U.K.) to remove nC60. The filtrate was then analyzed using the Cary 3E UV−vis spectrophotometer operated at a wavelength of 234 nm. Quantification of Tween 80 was obtained using a 5-point calibration curve prepared over a concentration range 20−600 mg/L; when necessary, samples were diluted 2-fold to fall within this concentration range. Uncorrected linear regression coefficient (r2) values obtained for all analyte calibration curves ranged from 0.96 to 0.99. Analysis of THF and GBL was performed using an Agilent 6890 gas chromatograph (GC) equipped with a DB-1 ms column (30 m length × 0.25 mm outside diameter × 0.25 μm

ρ ∂S ∂C ∂ 2C ∂C + b = DH 2 − vp ∂t θw ∂t ∂x ∂x

ρb ∂S θw ∂t

= kC

(1)

(2)

where C and S are particle concentrations in the aqueous and solid phases, respectively, ρb is the solid bulk density, θw is the porosity, x is the travel distance, t is time, and k is the effective rate of particle attachment. Previous studies2,6 suggest that the nC60 particle attachment rate depends linearly on the surface sites available for deposition. When the solid phase 11764



Article

Figure 1. Measured and simulated effluent breakthrough curves and retention profiles for nC60 transport in water-saturated 40−50 mesh Ottawa sand at a pore-water velocity of 7.8 m/d for input suspensions containing: nC60 alone or THF (44.5 mg/L) (A, B); premixed with SRHA (20 mg C/ L) or SRFA (20 mg C/L) (C, D); and Tween 80 (1000 mg/L) (E, F). The mathematical model used to simulate the experimental results accounted for first-order nanoparticle attachment kinetics and a maximum retention capacity. The detection limit of solid-phase extraction method for C60 was 0.1 μg/g.

(4)

described above. Values for katt and Smax were obtained by minimizing the sum of squares residuals between measured and modeled nC60 effluent concentration and retention profile data using a nonlinear least-squares optimization function provided by MATLAB R2009b (The MathWorks, Inc., Natick, MA). Breakthrough and retention profile data were weighted equally in the optimization and 95% confidence intervals for the fitted parameters were calculated using a nonlinear regression parameter confidence interval function provided by MATLAB. The collision efficiency factor (α) was then calculated on the basis of the fitted katt using eq 4.

Here, dc is the mean diameter of sand and the collision efficiency factor (α) represents the fraction of nC60 that remains attached after collision. The single collector efficiency (η0) represents the frequency of nC60 collisions with the porous medium grain surfaces, which can be approximated by the correlation equation developed by Tufenkji and Elimelech.41 Equations 1−4 were solved numerically to simulate the column experiment results. Here, DH was obtained independently by fitting to the measured nonreactive tracer BTCs, as

RESULTS AND DISCUSSION Baseline nC60 Transport. An initial column experiment was performed in the absence of stabilizing agent to provide baseline nC60 transport and retention data for comparison to subsequent column studies. At a pore-water velocity of 7.8 m/ day and ionic strength of 3.05 mM, nC60 appeared in the effluent at ca. 1.2 PV and the relative concentration gradually increased over the course of nanoparticle pulse injection,

concentration approaches the maximum retention capacity Smax, no additional attachment occurs. The resulting effective attachment rate k can be expressed as2 k = katt

Smax − S Smax

(3)

where katt is the particle attachment rate calculated as katt =

3(1 − θw )vp 2dc

αη0

40

■

11765



Article

low solid-phase concentrations of nC60 and the fact that effluent BTCs and retention profiles were weighted equally in the model fitting routine. The corresponding values of the attachment rate coefficient (katt) and the maximum retention capacity (Smax) were approximately 6- and 20-fold smaller, respectively, than those obtained from the baseline experiment (Table 1). These findings demonstrate that the presence of SRHA or SRFA in the influent suspension greatly enhances nC60 mobility, in agreement with the results of previous studies conducted under similar conditions.11,21 The observed enhancement in nC60 mobility was similar for SRHA and SRFA (e.g., similar α values) even though these organic macromolecules are significantly different in terms of molecular weight, aromatic and aliphatic components.42,43 In addition, the value of the collision efficiency factor (α ≈ 0.02) was similar to those measured using QCM (ca. 0.03) at a similar ionic strength and NOM concentration (1 mg C/L),21 suggesting the potential utility of the QCM for assessing nanoparticle transport parameters. To evaluate the impact of electrostatic interactions on nC60 mobility, Derjaguin−Landau−Verwey−Overbeek (DLVO) interaction energies, including electronic double layer repulsion and van der Waals attractive interactions, were computed.44 In the absence of any stabilizing agent and under similar experimental conditions, the primary energy barrier of the Ottawa sand-nC60 system was ca. 28 kT, with a negligible secondary energy minimum. Streaming potential measurements (see Supporting Information) indicate that the sand surfaces become more negative in the presence of 20 mg C/L SRHA or SRFA, increasing from −31 mV to −36 and −40 mV, respectively. However, DLVO energy profiles obtained for the Ottawa sand−nC60 system in the presence of 20 mg C/L SRHA and SRFA exhibited primary energy barriers of ca. 27 and 28 kT, respectively (Figure S3, Supporting Information). The nearly identical values of the primary energy barrier suggest that similar nanoparticle retention should have been observed in the presence of either SRHA or SRFA compared to the baseline experiment. Results obtained using a QCM also showed reduced rates of nC60 deposition in the presence of SRHA, which was attributed to steric repulsion.21,45 Thus, the reduced C60 attachment (and increased mobility) observed in the column experiment was attributed to steric repulsion or “stabilization” of nC60 associated with SRHA or SRFA coatings on the nanoparticles and collector surfaces.5,15,19,45 Effects of Surfactant on nC60 Mobility. Addition of a nonionic surfactant (Tween 80, 1000 mg/L) to the nC60 suspension resulted in effluent BTCs that coincided with that of the nonreactive tracer, and yielded nondetectable levels of nC60 retention (Figure 1E,F). The modified filtration model accurately captured changes in nC60 effluent concentrations throughout the pulse injection, and the greatly reduced retention of nC60 by the sand (dash lines in Figure1E,F). Consistent with the observed enhancement in nC60 mobility, the fitted attachment rate coefficients (katt) decreased to a value 6 times smaller than that obtained from the nC60 baseline experiment (Table 1). Additionally, the model showed that premixing with Tween 80 reduced the nC60 maximum retention capacity (Smax) on the quartz sand by more than one order-of-magnitude (from 1.8 to 0.1 μg/g). The primary DLVO energy barrier calculated in the presence of surfactant was approximately 18 kT, far less than the value obtained in the absence of stabilizing agents (Figure S3). This sizable reduction in the primary energy barrier is not consistent

reaching a maximum value of 0.97 (Figure 1A). Immediately following the reintroduction of nC60-free background solution (ca. 6 PVs), effluent nC60 concentrations decreased sharply to the detection limit, consistent with previous studies.6 The gradual increase in nC60 effluent concentration is indicative of rate-limited nC60 attachment on the quartz sand, while the rapid decline in nC60 effluent concentrations following the pulse injection is consistent with irreversible attachment. Approximately 29% of introduced nC60 mass was retained by the quartz sand, which was distributed relatively uniformly over the length of the column (Figure 1B). Although slightly lower retention was observed near the column inlet, the overall trend of the retention profile is consistent with a maximum or limiting nC60 retention capacity. The modified filtration model, which accounts for first-order attachment and a maximum retention capacity, successfully captured the delayed nC60 breakthrough, as well as the gradual ascent and steep descent of the effluent BTCs and relatively flat profile of the nC60 solid-phase concentrations (Figure 1A). The fitted attachment rate (katt) and maximum retention capacity (Smax) for the baseline experiment were 6.45 1/h and 1.8 μg/g (Table 1), respectively, which were in close agreement with results of previous studies conducted under similar conditions (katt = 7.23 1/h; Smax = 1.13 ug/g).2 Effects of THF on nC60 Mobility. Effluent BTCs and retention profiles obtained for nC60 in the presence of 44.5 mg/ L THF and at background levels (≤0.8 mg/L, baseline experiment) are shown in Figure 1. These data indicate that, even at a relatively high concentration, THF had no discernible effect on nC60 transport or retention behavior in watersaturated 40−50 mesh Ottawa sand. The effluent BTC and retention profile, as well as the model-fitted values katt and Smax, were nearly identical to those determined in the baseline nC60 column experiment (Table 1). Therefore, in all subsequent nC60 experiments, where the residual background THF concentration was ≤0.8 mg/L, the influence of THF on nC60 mobility was considered to be negligible. Effects of NOM on nC60 Mobility. In the presence of either 20 mg C/L SRHA or 20 mg C/L SRFA, nC60 appeared in the column effluent after 1 PV and rapidly increased to the applied concentration (i.e., C/Co = 1) (Figure 1C). Once the pulse injection ceased, nC60 concentrations decreased sharply to below the detection limit, yielding a BTC that nearly coincided with that measured for the nonreactive tracer. As anticipated from the effluent BTCs, the amount of nC60 deposited in the column represented less than 2% of the injected mass when the nC60 influent suspension contained either SRHA or SRFA. A maximum nC60 solid-phase concentration of ca. 0.2 μg/g was observed nearest the column inlet, and concentrations decreased gradually to nondetectable levels with increasing distance from the inlet (Figure 1D). In comparison to the baseline experiment (Figure 1), which was characterized by delayed breakthrough and a gradual increase in effluent concentration, nC60 breakthrough was rapid and retained mass decreased more than 7-fold, which clearly demonstrates the ability of relatively low concentrations of NOM (20 mg C/L) to dramatically enhance nC60 mobility in porous media. As shown in Figure 1, the modified filtration model was able to simulate the measured nC60 effluent BTC and the general shape of the retention profile in the presence of SRHA or SRFA. The observed discrepancies between the model fits and the retention profiles can be attributed, in large part, to the very 11766



Article

Figure 2. Measured and simulated effluent breakthrough curves (A) and retention profiles (B) for nC60 transport in water-saturated 40−50 mesh Ottawa sand that was pretreated with Tween 80. The mathematical model accounted for surfactant adsorption−desorption dynamics, surfactant− nanoparticle interactions, and first-order nanoparticle attachment kinetics.

(Figure S4, Supporting Information). In addition, the large confidence intervals estimated for Smax (Table 1) suggest that the breakthrough curve and retention profile are relatively insensitive to this parameter. The declining plateau behavior is suggestive of filter ripening,49 a process where attached particles serve as collectors, resulting in greater particle deposition over time. Modification of the mathematical model to incorporate filter ripening50 yielded improved simulations of the effluent BTCs (i.e., declining plateau); however, the model was unable to reproduce observed nC60 retention profiles (Figure S4). To explore the potential influence of surfactant sorption− desorption on nC60 mobility, the nanoparticle filtration model was coupled with a surfactant transport module. Here, a ratelimited Langmuir sorption model51 was used to describe interactions between Tween 80 and Ottawa sand

with enhanced nC60 mobility (and negligible attachment), suggesting that factors other than electrostatic interactions and van der Waals forces govern nC60 attachment in the presence of Tween 80. On the basis of the measured specific surface area of 40−50 mesh Ottawa sand (0.0125 m2/g), the mass of Tween 80 required for monolayer coverage was 10.7 μg/g, assuming a Tween 80 surface coverage of 245 Å2/molecule.46,47 At the conclusion of the nC60 transport experiment, the retained Tween 80 mass was 15.9 μg/g corresponding to a surfactant coverage of ca. 1.5 monolayer. Additionally, the nC60 zeta potential decreased (became less negative) upon addition of 1000 mg/L Tween 80 to the influent suspension, indicating surfactant adsorption onto nC60 (Table 1). Thus, the observed mobility enhancement and negligible attachment of nC60 in the presence of Tween 80 was attributed to steric repulsion arising from surfactant coatings on nC60.48 Effects of Adsorbed Surfactant on nC60 Mobility. Due to the ability of Tween 80 to substantially increase nC60 mobility, an additional set of experiments was undertaken to examine the impact of adsorbed-phase surfactant on nC60 transport and retention. In these experiments, the 40−50 mesh Ottawa sand columns were flushed with 10 PVs of nC60free Tween 80 solution (1000 mg/L) prior to the introduction of a 3 PV pulse of surfactant-free nC60 suspension. The resulting nC60 effluent BTCs and retention profiles obtained for replicate column experiments are shown in Figure 2A,B, respectively. The ascending and descending portions of nC60 BTCs matched those of the nonreactive tracer, but the relative concentration plateau did not reach unity and declined over time. Retention of nC60 was greatest near the column inlet, where solid-phase concentrations reached 1.9 μg/g, and decreased hyper-exponentially along the column, approaching the detection limit at the column outlet. These findings demonstrate that adsorbed Tween 80 enhanced nC60 mobility relative to the baseline (surfactant-free) experiment, although the degree of enhancement was less than that observed when the surfactant was mixed with the nC60 influent suspension prior to injection into the column (Figure 1E,F). In contrast to simulations of nC60 transport and retention in the baseline and premixed (Tween 80) column experiments, the modified filtration model was unable to capture the declining plateau of the BTCs and the hyper-exponential decay in retention profiles observed after pretreatment with Tween 80

ρ ∂S ∂C TW ∂ 2C TW ∂C + b TW = DH − vp TW 2 ∂t θw ∂t ∂x ∂x

(5)

⎛ QbC TW ⎞ ∂STW = k TW ⎜ − STW ⎟ ∂t ⎝ 1 + bC TW ⎠

(6)

where CTW and STW are concentrations of Tween 80 in the aqueous and solid phases, respectively, and kTW is the Tween 80 adsorption rate coefficient. The maximum sorption capacity of Ottawa sand for Tween 80, Q (49.9 μg/g), and sorption constant, b (0.0036 L/mg), were obtained by fitting to Tween 80 batch adsorption data (Figure S5, Supporting Information). The transport/sorption model (eqs 5 and 6) was verified through comparisons to measured effluent BTC for Tween 80 performed in the absence of nanoparticles (Figure S6, Supporting Information), yielding a sorption rate coefficient (kTW) of 0.3 1/h with a 95% confidence interval of (0.14, 0.46) 1/h. Transport and retention of nC60 was then dynamically coupled to surfactant transport and sorption by relating effective attachment rate (k) to the aqueous phase Tween 80 concentration. Here, it was assumed that aqueous phase Tween 80 is in dynamic equilibrium with Tween 80 adsorbed on the sand, which in turn may modify surface properties and block adsorption sites. Aqueous phase Tween 80 may also adsorb onto nC60 surfaces. Both of these factors may contribute to changes in the maximum retention capacity and attachment 11767



■

rate coefficient of nC60. These effects were modeled by modifying the effective nC60 attachment rate, k ⎧ 0, C TW > Tw ⎪ k=⎨ ⎛ C ⎞ S ⎞⎛ ⎟⎜1 − TW ⎟ , C TW < Tw ⎪ katt⎜1 − Smax ⎠⎝ Tw ⎠ ⎩ ⎝

AUTHOR INFORMATION

Corresponding Author

*Phone: 617-627-3099; fax: 617-627-3994; e-mail: kurt. [email protected]. Notes

The work has not been subject to EPA or NSF review, and, therefore, does not necessarily reflect the views of either Agency, and no official endorsement should be inferred. The authors declare no competing financial interest.

(7)

where Tw is a threshold concentration of aqueous Tween 80, above which nanoparticle attachment is completely prevented as a result of steric repulsion. This expression also assumes that the influence of CTW on effective attachment is linear as its value decreases below Tw. Note that the Tween 80 mass required for monolayer coverage of the solid phase was negligible (0.5%) compared to the initial Tween 80 mass applied to the column (10 PVs at 1000 mg/L). For the coupled model simulations, the aqueous and solid phase Tween 80 concentrations immediately before nC60 introduction were predicted using the surfactant transport model (eqs 5 and 6; Figure S7, Supporting Information). Then, eqs 1, 2, and 7were solved using the value of CTW calculated from eqs 5 and 6 in a previous time step. Tw was the only fitting parameter, obtained by minimizing the sum of squares residuals between measured and simulated nC60 effluent concentration and retention profile data using a nonlinear least-squares optimization function in MATLAB ver. R2009b. The resulting mathematical model, which accounted for coupled Tween 80 adsorption−desorption dynamics, surfactant−nC60 interactions, and nanoparticle attachment kinetics, captured the general trends of both the nC60 effluent BTCs and retention profiles (Figure 2). The fitted threshold aqueous Tween 80 concentration (Tw) was only 1.9 mg/L with a 95% confidence interval of (1.62, 2.17) mg/L, which is more than 6fold smaller than the critical micelle concentration (CMC) of Tween 80.52 However, this value was more than 10 times larger than the aqueous phase Tween 80 concentration (0.16 mg/L) required for monolayer coverage of nC60. As shown in Figure S7, the concentrations of both aqueous and solid phase Tween 80 decreased over the course of the nC60 pulse injection. Thus, desorption of Tween 80 from the quartz sand would (a) increase the solid surface area available for nC60 attachment over time, consistent with the temporal decline in the plateau of the nC60 BTCs, and (b) be sufficient to coat nC60 thereby enhancing mobility. These findings demonstrate the ability of natural and commercial stabilizing agents to dramatically enhance the mobility of nC60 in water-saturated porous media, and the importance of accounting for dynamic interactions between stabilizing agents and both solid and nanoparticle surfaces to accurately simulate nC60 behavior in complex systems.

■

Article

■

ACKNOWLEDGMENTS We thank Dr. Nils Kroger, Georgia Institute of Technology, for providing access to a Malvern Zetasizer instrument. This research was supported by Grant RD-832535 from the U.S. Environmental Protection Agency (EPA) Science to Achieve Results (STAR) program and Grant CBET-0854136 from the National Science Foundation (NSF).

■

REFERENCES

(1) Koruga, D.; Hameroff, S.; James, W.; Loutfy, R.; Sundareshan, M. Fullerene C60: History, Physics, Nanobiology, Nanotechnology; NorthHolland: Amsterdam, 1993. (2) Li, Y. S.; Wang, Y. G.; Pennell, K. D.; Abriola, L. M. Investigation of the transport and deposition of fullerene (C60) nanoparticles in quartz sands under varying flow conditions. Environ. Sci. Technol. 2008, 42, 7174−7180. (3) Colvin, V. L. The potential environmental impact of engineered nanomaterials. Nat. Biotechnol. 2003, 21, 1166−1170. (4) Wiesner, M. R.; Lowry, G. V.; Alvarez, P.; Dionysiou, D.; Biswas, P. Assessing the risks of manufactured nanomaterials. Environ. Sci. Technol. 2006, 40, 4336−4345. (5) Pelley, A. J.; Tufenkji, N. Effect of particle size and natural organic matter on the migration of nano- and microscale latex particles in saturated porous media. J. Colloid Interface Sci. 2008, 321, 74−83. (6) Wang, Y. G.; Li, Y. S.; Pennell, K. D. Influence of electrolyte species and concentration on the aggregation and transport of fullerene nanoparticles in quartz sands. Environ. Toxicol. Chem. 2008, 27, 1860−1867. (7) Wang, Y. G.; Li, Y. S.; Kim, H.; Walker, S. L.; Abriola, L. M.; Pennell, K. D. Transport and retention of fullerene nanoparticles in natural soils. J. Environ. Qual. 2010, 39, 1925−1933. (8) Wang, Y. G.; Li, Y. S.; Fortner, J. D.; Hughes, J. B.; Abriola, L. M.; Pennell, K. D. Transport and retention of nanoscale C-60 aggregates in water-saturated porous media. Environ. Sci. Technol. 2008, 42, 3588− 3594. (9) Zhang, L.; Hou, L.; Wang, L.; Kan, A. T.; Chen, W.; Tomson, M. B. Transport of fullerene nanoparticles (nC60) in saturated sand and sandy soil: Controlling factors and modeling. Environ. Sci. Technol. 2012, 46, 7230−7238. (10) Chen, K. L.; Elimelech, M. Aggregation and deposition kinetics of fullerene (C-60) nanoparticles. Langmuir 2006, 22, 10994−11001. (11) Espinasse, B.; Hotze, E. M.; Wiesner, M. R. Transport and retention of colloidal aggregates of C-60 in porous media: Effects of organic macromolecules, ionic composition, and preparation method. Environ. Sci. Technol. 2007, 41, 7396−7402. (12) Lecoanet, H. F.; Wiesner, M. R. Velocity effects on fullerene and oxide nanoparticle deposition in porous media. Environ. Sci. Technol. 2004, 38, 4377−4382. (13) Isaacson, C.; Zhang, W.; Powell, T.; Ma, X.; Bouchard, D. Temporal changes in aqu/c(60) physical-chemical, deposition, and transport characteristics in aqueous systems. Environ. Sci. Technol. 2011, 45, 5170−5177. (14) Gu, B. H.; Schmitt, J.; Chen, Z. H.; Liang, L. Y.; McCarthy, J. F. Adsorption and desorption of natural organic-matter on iron-oxide mechanisms and models. Environ. Sci. Technol. 1994, 28, 38−46.

ASSOCIATED CONTENT

S Supporting Information *

Additional data and descriptions of nonreactive tracer test, XPS analysis of the sand surface, transport of stabilizing agents, and surfactant sorption. Simulations of nC60 column data using clean-bed filtration theory, modified filtration theory, and filter ripening model. Modeling of surfactant transport in 40−50 mesh Ottawa sand. Calculations of Hamaker constants and interaction energy of nC60 with sand. This material is available free of charge via the Internet at http://pubs.acs.org. 11768



Article

zene in micellar solutions of ethoxylated nonionic surfactants. Environ. Sci. Technol. 1997, 31, 1382−1389. (36) 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. (37) Toride, N.; Leij, F. J.; van Genuchten, M. T.The CXTFIT code for estimating transport parameters f rom laboratory or field tracer experiments; U.S. Salinity Laboratory, Agricultural Research Service, USDA: Riverside, CA; 1999, Version 2.1, Research report No. 137; http://www.ars.usda.gov/services/software/92/cxtfit.pdf. (38) Hubaux, A.; Vos, G. Decision and detection limits for linear calibration curves. Anal. Chem. 1970, 42, 849−855. (39) Deshiikan, S. R.; Papadopoulos, K. D. Modified Booth equation for the calculation of zeta potential. Colloid Polym. Sci. 1998, 276, 117−124. (40) Yao, K. M.; Habibian, M. M.; Omelia, C. R. Water and waste water filtrationConcepts and applications. Environ. Sci. Technol. 1971, 5, 1105−1112. (41) Tufenkji, N.; Elimelech, M. Correlation equation for predicting single-collector efficiency in physicochemical filtration in saturated porous media. Environ. Sci. Technol. 2004, 38, 529−536. (42) Chin, Y. P.; Aiken, G.; Oloughlin, E. Molecular-weight, polydispersity, and spectroscopic properties of aquatic humic substances. Environ. Sci. Technol. 1994, 28, 1853−1858. (43) 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. (44) Guzman, K. A. D.; Finnegan, M. P.; Banfield, J. F. Influence of surface potential on aggregation and transport of titania nanoparticles. Environ. Sci. Technol. 2006, 40, 7688−7693. (45) Min, Y. J.; Akbulut, M.; Kristiansen, K.; Golan, Y.; Israelachvili, J. The role of interparticle and external forces in nanoparticle assembly. Nat. Mater. 2008, 7, 527−538. (46) Edwards, D. A.; Adeel, Z.; Luthy, R. G. Distribution of nonionic surfactant and phenanthrene in a sediment aqueous system. Environ. Sci. Technol. 1994, 28, 1550−1560. (47) Somasundaran, P.; Snell, E. D.; Xu, Q. Adsorption behavior of alkylarylethoxylated alcohols on silica. J. Colloid Interface Sci. 1991, 144, 165−173. (48) Fritz, G.; Schadler, V.; Willenbacher, N.; Wagner, N. J. Electrosteric stabilization of colloidal dispersions. Langmuir 2002, 18, 6381−6390. (49) Ginn, T. R.; Wood, B. D.; Nelson, K. E.; Scheibe, T. D.; Murphy, E. M.; Clement, T. P. Processes in microbial transport in the natural subsurface. Adv. Water Resour. 2002, 25, 1017−1042. (50) Simunek, J.; van Genuchten, M. T.; Sejna, M. The Hydrus-1d Software Package for Simulating the One-Dimensional Movement of Water, Heat, and Multiple Solutes in Variably-Saturated Media; University of California-Riverside Research Reports, Riverside, CA; 2005, http://www.ars.usda.gov/sp2UserFiles/Place/53102000/ pdf_pubs/P2119.pdf . (51) Abriola, L. M.; Dekker, T. J.; Pennell, K. D. Surfactant-enhanced solubilization of residual dodecane in soil columns. 2. Mathematical modeling. Environ. Sci. Technol. 1993, 27, 2341−2351. (52) Pennell, K. D.; Abriola, L. M.; Weber, W. J., Jr. Surfactantenhanced solubilization of residual dodecane in soil columns. 1. Experimental investigation. Environ. Sci. Technol. 1993, 27, 2332− 2340.

(15) Johnson, W. P.; Logan, B. E. Enhanced transport of bacteria in porous media by sediment-phase and aqueous-phase natural organic matter. Water Res. 1996, 30, 923. (16) Avena, M. J.; Koopal, L. K. Desorption of humic acids from an iron oxide surface. Environ. Sci. Technol. 1998, 32, 2572−2577. (17) Amirbahman, A.; Olson, T. M. Transport of humic mattercoated hematite in packed-beds. Environ. Sci. Technol. 1993, 27, 2807− 2813. (18) Deshiikan, S. R.; Eschenazi, E.; Papadopoulos, K. D. Transport of colloids through porous beds in the presence of natural organic matter. Colloids Surf., A 1998, 145, 93−100. (19) Franchi, A.; O’Melia, C. R. Effects of natural organic matter and solution chemistry on the deposition and reentrainment of colloids in porous media. Environ. Sci. Technol. 2003, 37, 1122−1129. (20) Chen, G.; Liu, X.; Su, C. Distinct effects of humic acid on transport and retention of TiO2 rutile nanoparticles in saturated sand columns. Environ. Sci. Technol. 2012, 46, 7142−7150. (21) Chen, K. L.; Elimelech, M. Interaction of fullerene (C-60) nanoparticles with humic acid and alginate coated silica surfaces: Measurements, mechanisms, and environmental implications. Environ. Sci. Technol. 2008, 42, 7607−7614, DOI: 10.1021/es8012062. (22) Wang, L. L.; Huang, Y.; Kan, A. T.; Tomson, M. B.; Chen, W. Enhanced transport of 2,2′,5,5′-polychlorinated biphenyl by natural organic matter (NOM) and surfactant-modified fullerene nanoparticles (nC(60)). Environ. Sci. Technol. 2012, 46, 5422−5429. (23) Tadros, T.; Kessell, L. Stabilizing nanodispersions in personal care and cosmetic applications. Cosmet. Toiletries 2004, 119, 41−46. (24) De Jong, W. H.; Borm, P. J. A. Drug delivery and nanoparticles: Applications and hazards. Int. J. Nanomed. 2008, 3, 133−149. (25) McCoy, M. Soaps & detergents: Specialty chemical makers seek business both with cleaning product innovators and the companies that follow them. Chem. Eng. News 2006, 84, 13−19. (26) Phenrat, T.; Cihan, A.; Kim, H. J.; Mital, M.; Illangasekare, T.; Lowry, G. V. Transport and deposition of polymer-modified Fe-0 nanoparticles in 2-D heterogeneous porous media: Effects of particle concentration, Fe-0 content, and coatings. Environ. Sci. Technol. 2010, 44, 9086−9093. (27) Kim, H. J.; Phenrat, T.; Tilton, R. D.; Lowry, G. V. Effect of kaolinite, silica fines and pH on transport of polymer-modified zero valent iron nano-particles in heterogeneous porous media. J. Colloid Interface Sci. 2012, 370, 1−10. (28) Jeong, S. W.; Kim, S. D. Aggregation and transport of copper oxide nanoparticles in porous media. J. Environ. Monit. 2009, 11, 1595−1600. (29) Song, J. E.; Phenrat, T.; Marinakos, S.; Xiao, Y.; Liu, J.; Wiesner, M. R.; Tilton, R. D.; Lowry, G. V. Hydrophobic interactions increase attachment of gum Arabic- and PVP-coated Ag nanoparticles to hydrophobic surfaces. Environ. Sci. Technol. 2011, 45, 5988−5995. (30) Tian, Y. A.; Gao, B.; Silvera-Batista, C.; Ziegler, K. J. Transport of engineered nanoparticles in saturated porous media. J. Nanopart. Res. 2010, 12, 2371−2380. (31) Bouchard, D.; Zhang, W.; Powell, T.; Rattanaudompol, U. S. Aggregation kinetics and transport of single-walled carbon nanotubes at low surfactant concentrations. Environ. Sci. Technol. 2012, 46, 4458− 4465. (32) Bensasson, R. V.; Bienvenue, E.; Dellinger, M.; Leach, S.; Seta, P. C60 in model biological-systemsA visible-UV absorption study of solvent-dependent parameters and solute aggregation. J. Phys. Chem. 1994, 98, 3492−3500. (33) Henry, T. B.; Menn, F. M.; Fleming, J. T.; Wilgus, J.; Compton, R. N.; Sayler, G. S. Attributing effects of aqueous C-60 nanoaggregates to tetrahydrofuran decomposition products in larval zebrafish by assessment of gene expression. Environ. Health Perspect. 2007, 115, 1059−1065. (34) Andrievsky, G.; Klcohkov, V.; Drevyanchenko, L. Is the C60 fullerene molecule toxic?! Fullerenes, Nanotubes, Carbon Nanostruct. 2005, 13, 363−376. (35) Pennell, K. D.; Adinolfi, A. M.; Abriola, L. M.; Diallo, M. S. Solubilization of dodecane, tetrachloroethylene, and 1,2-dichloroben11769
