Environ. Sci. Technol. 2009, 43, 330–336
Formation of Aqueous Suspensions of Fullerenes XIN MA* AND DERMONT BOUCHARD U.S. Environmental Protection Agency, National Exposure Research Laboratory, Athens Georgia 30605
Received July 7, 2008. Revised manuscript received November 5, 2008. Accepted November 6, 2008.
Colloidal suspensions of C60, C70, and a derivative of C60, PCBM ([6,6]-Phenyl C61-butyric acid methyl ester) were produced by extended mixing in water. We examined the contribution of background solution chemistry (pH, ionic strength) on the formation of colloidal suspensions in terms of mass, aggregate size, and zeta potential. These parameters were also compared between free-settling and filtered treatments. Results indicated that all three fullerenes were highly negatively charged in aqueous systems, that it took a finite time to reach kinetically stable suspensions, and that suspension formation was pH and ionic strength dependent. With isoelectric points approaching zero, the suspensions were generally stable at pH > 3. The results indicate that it is critical to state the condition under which the formation of aqueous fullerene suspensions occurs when employing such suspensions to evaluate environmental toxicity or fate and transport of fullerenes.
Introduction Due to their high strength, electrical conductivity, and electron affinity, fullerenes and fullerene derivatives have found a wide range of applications and created significant commercial interest. C60 is utilized in fuel cells, drug delivery, imaging agents in medicine, and cosmetics (1, 2). After C60, C70 is the most abundant higher fullerene in fullerene production. C70 has been shown to have higher photoinduced charge separation capability than C60 and may be used in artificial photosynthesis (3). The tendency of C60 to crystallize during film formation limits its use in the high concentration blends used to improve photoinduced electron transfer efficiencies in the fabrication of photovoltaic devices. To overcome these problems, more soluble C60 derivatives, such as PCBM (Supporting Information (SI) Figure S1), which is also a strong electron acceptor, have been used in polymer/ fullerene blends (4). Most fullerene environmental studies to date have focused on C60 (5-11); much less is known about the behavior of C70 and PCBM in the environment. Though C60 is very insoluble in water (12, 13), colloidal suspensions of C60 (nC60) are commonly created using solvent exchange (11, 14), or prolonged stirring techniques (5, 6). In solvent exchange methods, C60 is first dissolved in an organic solvent, usually tetrahydrofuran (THF). The C60-THF solution is then dispersed in water followed by THF removal by distillation (11, 14). The result is an aqueous suspension of C60 aggregates (THF/nC60). However, THF may be retained in the cluster structure of THF/nC60, which may influence colloidal suspension surface properties (5, 8) and contribute to charge transfer from THF to C60 molecules (5, 14). Extended * Corresponding author phone: (706) 355-8330; fax: (706) 3558160; e-mail:
[email protected]. 330 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 2, 2009
stirring in water without organic solvents can also produce colloidal suspensions of C60 (aqu/nC60) that contain fewer preparation artifacts, and therefore are more environmentally relevant (5). The use of these different fullerene suspension preparation techniques has led to a problem in comparing and interpreting data across studies. For example, THF/nC60 and aqu/ nC60 aggregates differ with respect to size, shape, polydispersity, and charge characteristics (5-7) and also in the toxic effects they elicit (15). Also, when solid fullerenes are dispersed in aqueous systems, the establishment of the final equilibrium structure at the particle/solution interface takes a finite time. It is unknown how aqu/C60 aggregates in suspension change over time with respect to size, zeta potential (ζ potential), and concentration. Clearly, fullerene suspension preparation method is a major determinant of the results observed in transport and toxicity studies, and quantifying suspension properties is a sine qua non for environmental studies with fullerenes. The objective of this work was to measure the changes in important fullerene aggregate parameters during suspension formation by extended stirring in water, and to interpret these changes in light of potential mechanisms for fullerene charge acquisition in aqueous systems. To this end, we examined the effect of pH and ionic strength (I) on the formation of fullerene colloidal suspensions in terms of mass/ concentration, particle size, and ζ potential. These parameters are also compared between free-settling and filtered treatments. This investigation is the first detailed quantitative investigation of solution chemistry effects on fullerene colloid formation for a suite of fullerenes in aqueous systems and provides insight into the potential behavior of these materials when released into the environment.
Materials and Methods Creation of Aqueous Fullerene Suspensions (aqu/C60, aqu/ C70, and aqu/PCBM). The C60 (purity 99.9%) and C70 (purity 99.0%) fullerenes were purchased from MER Corp. (Tucson, AZ) and the PCBM (purity 99.5%) from MTR Ltd. (Cleveland, OH). Seven treatments of aqueous suspensions were prepared in background solutions of double deionized water (DDI) (resistivity >18 MΩ/cm), or at pH values of 4, 7, and 10 with I equivalent to 10 mM NaCl, respectively. An amount of 100 mg of C60, C70 or PCBM was added to 400 mL of aqueous solution, the mixtures were stirred using magnetic plates, and sampling was performed at least weekly initially and later biweekly. Before each sampling event, the mixtures were allowed to free-settle for an hour. One hour was chosen because it was a relatively short time during which the larger particles were settled while smaller particles were still in suspension. A 20 mL aliquot was collected 2 cm below the suspension surface, and 20 mL of background solution was replenished after sampling. The samples collected were not representative of the total suspension volume (where particle size and concentration will vary with depth); but rather, provided a consistent sampling point for very heterogeneous samples so that changes in suspension properties over time could be measured with acceptable precision. A 10 mL aliquot was filtered through a 0.45 µm cellulose acetate filter (to allow comparison to prior studies which usually used filtered suspensions) while the other half-remained unfiltered (to represent an environmental sample collected in the field). Characterization of Aqueous Suspensions. The mass quantification method for aqu/C60, aqu/C70, and aqu/PCBM is described in a separate publication (16). Briefly, the aqueous fullerene samples were extracted into toluene and 10.1021/es801833p
Not subject to U.S. Copyright. Publ. 2009 Am. Chem. Soc.
Published on Web 12/15/2008
FIGURE 1. Variation of particle size (a) (b), ζ potential (c) (d), and mass (e) (f) of aqu/C60 in unfiltered (a) (c) (e) and filtered (b) (d) (f) samples during 111-day extended stirring. The four treatments were C60 at pH 4, 7, and 10 in 10 mM NaCl and C60 in DDI water. Solutions with different pH were buffered with 0.01 M acetate (pH 4.00 ( 0.01), HEPES (pH 7.01 ( 0.04), or Tris (pH 10.03 ( 0.06). The measurement temperature was set at 25 °C. then analyzed by HPLC-UV (16). ζ-potentials of the suspensions were monitored using a ZetaSizer Nano ZS instrument (Malvern Instruments, Worcestershire, UK). This instrument uses phase analysis light scattering (PALS) to measure the electrophoretic mobility of charged particles. The Smoluchowski equation was used to calculate ζ-potential from electrophoretic mobility. Three measurements (12 runs per
measurement) were acquired from each unfiltered sample while six measurements were acquired from each filtered sample. Instrument performance was verified using NISTtraceable polystyrene microsphere standards. Changes in fullerene aggregate size were examined by dynamic light scattering (DLS) (ZetaSizer Nano ZS). Six measurements (12 runs per measurement) were acquired VOL. 43, NO. 2, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
331
FIGURE 2. Intensity-weighted (a), volume-weighted (b), and number-weighted (c) size distribution changes of filtered aqu/ C60 in pH 4 over time. Solid, long dash, dotted lines represent day 20, 48, and 96, respectively. The peaks shifted toward smaller sizes over time. The polydispersity index (PDI) values decreased over time (0.762, 0.652, and 0.218, respectively), indicating the formation of less dispersed suspensions. Refractive index: 1.96 (29). from each sample. The autocorrelation function was analyzed by the cumulant method to obtain the moments of the aggregate size distribution. The fluctuations of scattered light from particles can be mathematically correlated to the diffusion coefficient. The intensity average (Z-average) hydrodynamic diameter was calculated from measured diffusivities using the Stokes-Einstein equation.
Results and Discussion Fullerene Aggregate Size Change over Time. The arc process for producing C60 and its derivatives generates a totally dehydrated hydrophobic solid (17). It takes a finite time to establish the final equilibrium structure when solid fullerenes are dispersed in aqueous systems. The mixing of fullerenes in water provides shear force energy to disperse the solids, mimicking the turbulence in a natural water system. In our study, shear forces were maintained as constant as practicable across treatments using magnetic stirring plates. After 3 days of stirring, the Z-average hydrodynamic diameters for unfiltered aqu/C60 suspensions in pH 7, pH 10, and DDI reached kinetically stable sizes of 514 ( 17 nm (mean ( 95% confidence limit), 479 ( 31 nm, and 535 ( 9 nm, respectively, and these sizes remained relatively constant throughout the 111-day experimental period (Figure 1a), i.e., the fullerene particles did not aggregate further to create larger clusters. However, the unfiltered aqu/C60 suspensions formed in pH 4 did not reach a stable aggregate size until day 33 (Figure 1a). In addition, the Z-average values of the aqu/C60 in pH 4 treatments (767 ( 63 nm) were larger than those at the higher pH’s. This kinetic effect was also observed in a size distribution shift toward smaller size and less polydispersed suspensions over time (Figure 2a), with the polydispersity index (PDI) stabilizing near 0.20 by day 69. Particles are 332
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 2, 2009
kinetically stable when they remain dispersed over a long period of time; they may collide by Brownian motion or shear flow, but do not stick together after the collision. The result is dispersed particles in media where size and zeta potential do not change over time. A limitation of DLS is the reliance on a single intensityaveraged value, the Z-average diameter, for representation of the entire distribution of hydrodynamic diameters. Inspection of scattering intensity distributions revealed that most fullerene aggregate size distributions were multimodal (Figure 2a). Although derivation of volume/mass-average hydrodynamic diameter (dh,v), and especially numberaverage hydrodynamic diameter (dh,n), from dynamic light scattering data involves numerous assumptions, the resulting volume and number distributions can be useful for making comparisons among samples, bearing in mind that confidence in the absolute magnitude of dh,v and dh,n is less than for Z-average. Figure 2 shows a comparison of the change of intensity-average, volume-average, and number-average distribution in C60 in pH 4 filtered samples over time. The intensity-average and volume-average distributions showed the multimodal populations of the aggregates while numberaverage distributions showed monomodal peaks. This indicated that the presence of very few large aggregates contributed a large amount of the scattering intensity. We reported the changes in first-order Z-average values with time as there was less uncertainty in intensity-average distributions. Aggregate size (725 ( 52 nm) of the unfiltered aqu/C70 suspension stabilized after day 16 (SI Figure S2a). Among all the unfiltered fullerene treatments, aqu/PCBM exhibited the largest hydrodynamic diameters (Figure 3a). The sizes of aqu/ PCBM aggregates were stable after day 33 with Z-averages of 1082 ( 106 nm and 817 ( 92 nm for aqu/PCBM in pH 7 and DDI, respectively. Our results indicated that the phenylester moiety of PCBM did not interfere sterically with the formation of PCBM aggregates and implied that other fullerene derivatives may also form stable colloidal suspensions. To our knowledge, these are the first investigations of aqu/C70 suspension without solvent aids and the first report on the formation of aqueous colloidal aggregates for a fullerene derivative. For the filtered samples, aggregate size in most C60 treatments was stable after day 8. Similar to the unfiltered samples, aqu/C60 aggregate size in pH 4 filtered treatments did not stabilize as quickly as the other C60 treatments (Figure 1a and b). The data indicated that the 0.45 µm filters removed particles larger than 450 nm and that little postfiltration aggregation occurred. Pooling data from day 69-111 when aggregate size was stable, the Z-averages for the filtered C60 samples were 191 ( 23 nm, 171 ( 13 nm, 165 ( 4 nm, and 164 ( 7 nm for pH in 4, 7, 10, and DDI, respectively, which were significantly smaller than the unfiltered treatments. The aqu/C60 Z-average of the filtered samples at the end of our experiment were similar to the result (163 nm) of Duncan and co-workers using pulverized C60 mixed in water for 14 days (6). The Z-averages measured near day 14 in our study were not stable (Figure 1b), perhaps indicating the role pregrinding or pulverization can play in increasing the rate to reach stable fullerene aggregates. Background solution pH affected fullerene aggregate size in both the unfiltered and filtered samples. In aqu/C60 unfiltered and filtered samples, larger aggregates were observed at pH 4 than at pH 7, pH 10 or in DDI. At pH 4, increased proton activity may neutralize the negative charges at the aqu/C60 surfaces. With less electrostatic double layer repulsion, van der Waals attractive forces led to the formation of larger aggregates. It is generally thought that aqu/C60 will quickly aggregate in natural waters, settle out of suspension, and not pose an environment risk (9, 10, 17). However, the
FIGURE 3. Variation of particle size (a) (b), ζ potential (c) (d), and mass (e) (f) of aqu/PCBM in unfiltered (a) (c) (e) and filtered (b) (d) (f) samples during 111-day extended stirring. The two treatments were PCBM at pH 7 in 10 mM NaCl and in DDI water. The solution at pH 7 was buffered with HEPES (pH 7.01 ( 0.04). The measurement temperature was set at 25 °C. results from this study indicated that aqu/C60 will form aggregates in water at I common in natural waters (18) that are stable over a long period of time. Examination of fullerene aggregates by HRTEM revealed that many smaller aggregates formed larger, irregular clusters (SI Figure S3a-c). AFM imaging also indicated a wide range of aggregate sizes in the suspensions (SI Figure S3d). Measurements with both microscope techniques were generally consistent with DLS measurements given the inherent differences in the techniques since DLS
Z-average is influenced by hydration or solvation effects while TEM is a number-weighted average size of a dehydrated hard sphere. Mass Change over Time. The unfiltered aqu/C60 treatments reached relatively stable concentrations after about day 26 and an inverse relationship between the mass of aqu/ C60 in the unfiltered suspensions and the solution pH was observed (Figure 1e): aqu/C60 mass concentration decreased from 107 to 84 to 43 mg/L as pH increased from 4 to 7 to 10 at the end of the experiment on day 111. The dependence VOL. 43, NO. 2, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
333
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 2, 2009
0.61 ( 0.21 0.59 ( 0.16 0.84 ( 0.34 0.91 ( 0.10 0.45 ( 0.34 0.71 ( 0.48 0.61 ( 0.43 0.52 ( 0.11 0.94 ( 0.04 1.07 ( 0.04 0.77 ( 0.11 0.58 ( 0.16 0.72 ( 0.06
0.96 ( 0.13
unfiltered filtered unfiltered filtered unfiltered filtered unfiltered filtered
Titration was performed over a range from pH 1-10 and pI was measured or extrapolated at the point where the ζ-potential ) 0. The aqueous suspensions at the end of day 111 were used for the titration measurements. At least three titrations were measured for each treatment.
filtered
unfiltered
pH 7 C70
DDI DDI pH 10
C60 pH 7 pH 4 unfiltered
a
filtered unfiltered
DDI 9
filtered
PCBM 334
TABLE 1. Mean ± 95% Confidence Limits for Isoelectric Points (pI) of C60, C70, and PCBM treatmentsa
FIGURE 4. ζ-potential and Z-average of unfiltered aqu/C60 in pH 10 as a function of pH. The pH titrations were performed using an autotitrator in conjunction with real time monitoring of aggregate size and ζ-potential changes. The titrants used were 1-0.50 M HCl, 0.100.4 M NaOH, and 0.01-0.05 M HCl. All titrants were pumped through a vacuum degasser prior to titration to prevent any absorption of oxygen that could cause pH drift. A pH probe calibration in the autotitrator was performed at the beginning of each titration session, the titrants were primed to exclude any bubbles in the tubes, and the sample loop was cleaned before and after each titration session with DDI. of aqu/C60 mass in suspension on solution pH clearly has implications for environmental transport and fate processes. The mass of aqu/C70 in the unfiltered suspensions increased at a much slower rate than aqu/C60 or aqu/PCBM (SI Figure S2e). At the end of the 111-day sampling period the mass of aqu/C70 in DDI was only 35 and 52% that of the DDI treatments of aqu/C60 and aqu/PCBM, respectively. Since aqu/C60 and aqu/PCBM reached their maximum concentrations by day 20, the mass differences with aqu/C70 were even greater at that point in time: suspended C70 mass was only 11% of aqu/C60 and aqu/PCBM. Clearly, C70, which has been the object of much less environmental study than C60, forms aqueous suspensions far more slowly than C60. In contrast to C60 and C70, the unfiltered aqu/PCBM formed high mass colloidal suspensions within a day, but the PCBM treatments also exhibited the largest variations in mass throughout the experimental period (Figure 3e). For most of the data, the mass in the filtered samples was generally pH 7> DDI > pH 4. The average unfiltered sample ζ potentials over the entire course
were -62.2 ( 1.1, -55.0 ( 1.1, -47.0 ( 4.9, -45.8 ( 1.8 mV, respectively; the ζ potentials for the filtered samples were -59.1 ( 3.4, -45.7 ( 11.5, -39.0 ( 7.2, and -33.6 ( 11.5 mV, respectively. As pH decreased, the increased proton activity in solution appeared to hinder the acquisition of negative charge by the fullerenes. The pH effect observed in this study differs from that reported by Brant and co-workers (5) who found that THF/aqueous samples at pH 10.4 and pH 5.6 were similarly charged. In our study, different ζ potentials were observed for aqu/C60 at pH 4 and pH 10 over the entire course of the study. Unfiltered aqu/C70 samples had a ζ potential of -51 ( 2.4 mV while aqu/PCBM in pH 7 and in DDI had more negative ζ potentials of -61.6 ( 3.0 and -63.3 ( 2.6 mV. The fullerene’s high negative charges will increase aggregate stability and mobility in aqueous media and decrease binding to other negatively charged species in natural waters. The ζ potential values of the filtered samples were much more variable and a trend of decreasing ζ potential over time was observed in all filtered treatments (Figures 1-3d). This increase in the negative charge was concomitant with the increase of the colloidal concentration in the filtered samples. The reason behind such a kinetic process with respect to the ζ potential is unclear, but this observation again underscores the importance of kinetics in fullerene suspension formation and stability. However, by the end of the study, the ζ potential values for the filtered and unfiltered samples had converged and were not significantly different by paired t-tests (data not shown). Isoelectric Point (pI). The ζ potential changes were measured as pH of the suspensions was titrated at the end of the experiment. Figure 4 shows a representative ζ/size-pH plot for aqu/C60 in pH 10. The titrations of all other treatments exhibited a similar trend and the pIs are tabulated in Table 1. It is clear from Figure 4 that the ζ potential increased as pH decreased, with ζ potential increasing slowly from pH 10 to pH 2, then increasing rapidly as the isoelectric point (pI ) 1.07) was approached. Concurrently, the Z-average increased when the pH was near pI. The increased proton activity in solution appeared to neutralize the aggregates’ negative charges and greater aggregation would be expected at the lower pH, although the nature of the charge was unclear. Paired t-tests and 95% confidence limits were calculated to evaluate whether there were significant differences in isoelectric points between treatments (Table 1). Significant differences between unfiltered and filtered samples were found in aqu/C60 in pH 4 (p ) 0.0261) and aqu/C60 in pH 10 (p ) 0.00097). The pI of aqu/C60 in DDI was significantly different from the C60 treatments in pH 4 (p ) 0.0287), pH 7 (p ) 0.0187) and pH 10 (p ) 0.00015), indicating the effect of ionic strength on the formation of the aqu/C60 suspensions. A similar ionic strength effect was observed in aqu/PCBM in pH 7 and in DDI (p ) 0.0355). A significant difference was also found between aqu/C60 in pH 4 and aqu/C60 in pH 10 (p ) 0.0071). It is important to understand that the observed differences in pI were a function of the background solution parameters (pH, I) during fullerene suspension formation, not during pI measurement, thus indicating that solution characteristics during suspension formation have apparently irreversible effects on aggregrate electrokinetic properties. The high negative ζ potentials (∼60 mV) observed in all three fullerene species and different solution treatments pose a question on the nature of the charge. The surface charge of most colloidal dispersions arises from the ionization of surface groups, differential loss of ions from the crystal lattice, or adsorption of charged species (ions and ionic surfactants) (21). The C60, C70, and PCBM molecules have no readily ionizable functional groups or lattice ions; therefore, the first two mechanisms do not apply to these fullerenes. The
dramatic effect that pH exerts on the ζ potentials (and, by inference, the surface charge) indicates that H+ and OH- are potential-determining ions in the system. The fact that most of the pI < 1 indicated the relative high affinity of OH- for fullerenes and the significance of the OH- ion in determining the state of surface charge (Table 1). The electronic properties of fullerenes may lead to the possibility of donor-acceptor and charge-transfer interactions. C60 has a high electron affinity of 2.65 eV as measured in the gas phase (22) which makes C60 reactive with nucleophiles. Since suspension formation was in an aqueous system, the interaction between C60 and water has to be considered. The shearing force during the mixing may facilitate the donor-acceptor interaction of fullerenes with a polar solvent such as water. One possibility for charge acquisition is simply hydroxyl ion adsorption, which has been proposed as a mechanism of charge acquisition for hydrophobic surfaces in water (23). In this mechanism, the unpaired electrons of oxygen atoms in H2O may become more polarized near the electron-accepting fullerene surface. The interfacial water molecules are preferentially oriented with the oxygen atoms toward the hydrophobic phase to maximize the number of hydrogen bonds formed (24, 25). The strong dipole or hydrogen bonding between the OHions and the hydrogen atoms of the interfacial water molecules could result in the adsorption of hydroxyl ions (14, 23). Brant et al. rejected this hypothesis, reasoning that hydroxyl ion adsorption was not the major contributor to surface charge because of unchanged pH after cluster formation and no observed decrease in ζ potential at the higher pH (5). In our study, ζ potential was observed to decrease with increasing pH both during pH titration in the pI measurements and during suspension formation under different pH conditions. However, like Brant et al. (5), we did not observe a decrease in background solution pH. It is possible that any pH changes may be localized at the vicinity of the hydrated surface and therefore are not reflected in bulk solution measurements. While our data provides some evidence for the donor-acceptor hypothesis, it does not provide definitive proof that this is the mechanism behind fullerene charge acquisition. Another hypothesis for charge acquisition is that C60 is surrounded by an icosahedral water cluster (26, 27). Such clathrate-like icosahedral water clusters minimize the loss of hydrogen bonds at hydrophobic surfaces (27). As an electronegative molecule, C60 shows some aromatic behavior in the 26-membered rings with the π-orbital electron density biased outward (27). The fullerene may sit ideally in an icosahedral water cluster missing its inner water dodecahedron because the size of the C60 molecule and a (H2O)20 dodecahedral cage are similar (28). The second hydration shell of the icosahedral is composed of 60 fully hydrogen bonded water molecules. Such a hypothetical cluster would be of the C60@{H2O}80 type. In such a structure, the electrondeficient carbon atoms are capable of interacting with lone pair electrons donated by extra water molecules (27). The ionization of carbon-linked water increases the negative charge on the hydrated C60 surface. Up to six hydroxide ions would increase the aromatic electron density, resulting in strengthening the HOsH · · · π hydrogen bonding in the surrounding inner aqueous shell (27). Therefore, the interaction between fullerene and water via weak hydrogen bonds leads to C60’s high negative surface charge. Geometrical matching is another critical factor in charge acquisition. C70 has 1 order of magnitude higher estimated water solubility than C60 (13). It also has higher electron affinity (2.73 eV) than C60 (22). Yet C70 had lower mass concentrations in suspension than C60 in this study. The shape of C70 is more ovoid than the classic C60 soccer ball and this geometry does not fit in the water cluster very well. In fact, VOL. 43, NO. 2, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
335
it has been proposed that the water cluster is split into two halves and five extra water molecules are hydrogen-bonded to the five six-member rings of C70 as compared to the 20 in C60 (28). PCBM seemed to fit well within the proposed cage structure, as evidenced by the higher suspension concentrations of both filtered and unfiltered aqu/PCBM than for aqu/ C70. It may be that the phenyl and methyl-ester moiety of PCBM are small enough to fit into the interstitial spaces in the water shells, creating multiple van der Waals attractions.
Acknowledgments We thank Nick Marshall (University of Georgia) for assistance with the AFM analysis, Amar Kumbhar (Electron Microscope Facility at Clemson University) for help with the TEM analysis, and Carl Isaacson and Jason Locklin for helpful discussions. This paper has been reviewed in accordance with the U.S. Environmental Protection Agency’s peer and administrative review policies and approved for publication. Mention of trade names or commercial products does not constitute endorsement or recommendation for use. Disclaimer: This paper has been reviewed in accordance with the U.S. Environmental Protection Agency’s peer and administrative review policies and approved for publication. Mention of trade names or commercial products does not constitute endorsement or recommendation for use.
Supporting Information Available PCBM structure; detailed characterization of C70; and TEM, AFM, and photo images of suspensions. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Colvin, V. L. The potential environmental impact of engineered nanomaterials. Nat. Biotechnol. 2003, 21 (10), 1166–1170. (2) Oberdo¨rster, E. Manufactured Nanomaterials (Fullerenes, C60) Induce oxidative stress in the brain of juvenile largemouth bass. Environ. Health Perspect. 2004, 112 (10), 1058–1062. (3) Koichi, T.; Hiroshi, I.; Yoshiteru, S.; Yoshinobu, N.; Iwao, Y.; Akihisa, S.; Tadashi, O. Photoinduced electron transfer in porphyrin-linked C70 and C60. Kokagaku Toronkai Koen Yoshishu 1999, 1999, 124. (4) Yu, G.; Gao, J.; Hummelen, J. C.; Wudl, F.; Heeger, A. J. Polymer photovoltaic cells: enhanced efficiencies via a network of internal donor-acceptor heterojunctions. Science 1995, 270, 1789–1791. (5) 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 (17), 6343– 6351. (6) 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 (1), 173–178. (7) Brant, J. A.; Labille, J.; Bottero, J. Y.; Wiesner, M. R. Characterizing the impact of preparation method on fullerene cluster structure and chemistry. Langmuir 2006, 22 (8), 3878–3885. (8) Andrievsky, G. V.; Klochkov, V. K.; Bordyuh, A. B.; Dovbeshko, G. I. Comparative analysis of two aqueous-colloidal solutions of C60 fullerene with help of FTIR reflectance and UV-Vis spectroscopy. Chem. Phys. Lett. 2002, 364 (1-2), 8–17.
336
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 2, 2009
(9) Lecoanet, H. F.; Wiesner, M. R. Velocity effects on fullerene and oxide nanoparticle deposition in porous media. Environ. Sci. Technol. 2004, 38 (16), 4377–4382. (10) 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 (21), 7396–7402. (11) 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. C-60 in water: Nanocrystal formation and microbial response. Environ. Sci. Technol. 2005, 39 (11), 4307–4316. (12) Jafvert, C. T.; Kulkarni, P. P. Buckminsterfullerene’s (C60) octanolwater partition coefficient (Kow) and aqueous solubility. Environ. Sci. Technol. 2008, 42, 5945–5950. (13) Heymann, D. Solubility of fullerenes C60 and C70 in seven normal alcohols and their deduced solubility in water. Fullerenes, Nanotubes and Carbon Nanostruct. 1996, 4 (3), 509–515. (14) Deguchi, S.; Alargova, R. G.; Tsujii, K. Stable dispersions of fullerenes, C60 and C70, in water. Preparation and characterization. Langmuir 2001, 17 (19), 6013–6017. (15) Zhu, S.; Oberdorster, E.; Haasch, M. L. Toxicity of an engineered nanoparticle (fullerene, C60) in two aquatic species, Daphnia and fathead minnow. Mar. Environ. Res. 2006, 62, S5–S9. (16) Bouchard, D.; Ma, X. Extraction and HPLC-UV analysis of C60, C70, and PCBM in synthetic and natural waters. J. chromatogr., A 2008, 1203, 153–159. (17) Wiesner, M. R.; Lowry, G. V.; Alvarez, P.; Dionysiou, D.; Biswas, P. Assessing the risks of manufactured nanomaterials. Environ. Sci. Technol. 2006, 40 (14), 4336–4345. (18) Pankow, J. F., Aquatic Chemistry Concepts; Lewis Publishers: Chelsea, MI, 1991. (19) Wiesner, M. R.; Hortze, E. M.; Brant, J. A.; Espinasse, B. Nanomaterials as possible contaminants: the fullerene example. Water Sci. Technol. 2008, 57 (3), 305–310. (20) Elimelech, M.; Jia, X.; Gregory, J.; Williams, R., Particle Deposition & Aggregation: Measurement, Modelling and Simulation; Butterworth-Heinemann: Woburn, WA, 1998. (21) Hunter, R. J., Zeta Potential in Colloid Science; Academic Press Inc.: Orlando, FL, 1981. (22) Boltalina, O. V.; Dashkova, E. V.; Sidorov, L. N. Gibbs energies of gas-phase electron transfer reactions involving the larger fullerene anions. Chem. Phys. Lett. 1996, 256 (3), 253–260. (23) Marinova, K. G.; Alargova, R. G.; Denkov, N. D.; Velev, O. D.; Petsev, D. N.; Ivanov, I. B.; Borwankar, R. P. Charging of oilwater interfaces due to spontaneous adsorption of hydroxyl Ions. Langmuir 1996, 12 (8), 2045–2051. (24) Israelachvili, J., Intermolecular and Surface Forces. 2nd ed.; Elsevier Academic Press: San Diego, CA, 1991. (25) van Oss, C. J., Interfacial Forces in Aqueous Media. 2nd ed.; CRC Press: Taylor & Francis Group: Boca Raton, FL, 2006. (26) Andrievsky, G. V.; Klochkov, V. K.; Bordyuh, A. B.; Dovbeshko, G. I. Comparative analysis of two aqueous-colloidal solutions of C60 fullerene with help of FTIR reflectance and UV-Vis spectroscopy. Chem. Phys. Lett. 2002, 364, 8–17. (27) Chaplin, M. F. Water structuring at colloidal surfaces. In Surface Chemistry in Biomedical and Environmental Science, NATO Security through Science Series; Blitz, J., Gun’ko, V., Eds.; Springer: Dordrecht, The Netherlands, 2006; Vol. 228, pp 1-10. (28) Chaplin, M. F. 2008 Aqueous C 60 and C 70 Fullerene Solutions.; South Bank University: London, UK; http://www.lsbu.ac.uk/ water/buckmin.html (accsessed May 2008). (29) Ruoff, R. S.; Tse, D. S.; Malhotra, R.; Lorents, D. C. Solubility of C60 in a variety of solvents. J. Phys. Chem. 1993, 97, 3379–3383.
ES801833P
