Environ. Sci. Technol. 2009, 43, 6597–6603
Colloidal Properties of Aqueous Fullerenes: Isoelectric Points and Aggregation Kinetics of C60 and C60 Derivatives D E R M O N T B O U C H A R D , * ,† X I N M A , † A N D CARL ISAACSON‡ U.S. Environmental Protection Agency, National Exposure Research Laboratory, and National Research Council Research Associate, Athens Georgia 30605
Received May 6, 2009. Revised manuscript received June 29, 2009. Accepted July 14, 2009.
Aqueous colloidal suspensions of C60 (aqu/C60) and the C60 derivatives PCBM ([6,6]-phenyl C61-butyric acid methyl ester) and the corresponding butyl and octyl esters, PCBB and PCBO (aqu/PCB-R, where R is an alkyl group), were produced by stirring in double deionized water for 5 months. Kinetically stable fullerene aggregates were formed using this procedure that ranged in intensity-averaged hydrodynamic diameter (Dh) from 193 ( 2 nm (95% C.L.) for aqu/C60 to 259 ( 6 nm for aqu/ PCBO. Measured zeta potentials (ζ) were < -50 mV, and the isoelectric points (pI) were 99%), PCBB ([6,6]-phenyl C61-butyric acid butyl ester, purity >97%), and PCBO ([6,6]-phenyl C61-butyric acid octyl ester, purity >99%) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Aqueous suspensions were prepared in double deionized water (DDI) (resistivity >18 MΩ · cm) by adding 100 mg of fullerene to 400 mL of DDI and then stirring the suspensions under ambient laboratory lighting on magnetic plates for 5 months. All of the PCB-R fullerenes were added to water as received from the manufacturer; however, the C60 was ground with a glass mortar and pestle prior to mixing with water due to its more granular composition. To obtain working suspensions, after the extended stirring period the suspensions were allowed to free-settle for an hour. Then, a 20 mL aliquot was collected from 1 to 2 cm below the suspension surface and filtered through a 0.45 µm cellulose acetate filter. The aliquots for each fullerene suspension were then combined to form 100 mL working suspension volumes. Characterization of Aqueous Fullerene Suspensions. The method used for mass quantification of the fullerenes in aqueous suspension was based on toluene extraction with HPLC-DAD and is similar to a method described in an earlier publication (24). Fullerene aggregate size in aqueous suspension was measured using dynamic light scattering (DLS) with a ZetaSizer Nano ZS (Malvern Instruments, Worcestershire, UK). The intensity average (Z-average) hydrodynamic diameter (Dh) was calculated from measured diffusivities using the Stokes-Einstein equation. For initial fullerene aggregate size characterization, six measurements (12 runs per measurement) were acquired from each sample. ζ-Potentials of the suspensions were also measured using a ZetaSizer Nano ZS instrument which employs 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. Six measurements (12 runs per measurement) were acquired for each sample. Instrument performance was verified using NIST-traceable latex microsphere and polystyrene microsphere standards. pH Titrations/Isoelectric Point (pI) Determination. The titrants used were 0.50-1.0 M HCl, 0.1-0.4 M NaOH, and 0.01-0.05 M HCl. All titrations were performed in triplicate, and all titrants were degassed prior to titration. 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 6598
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 17, 2009
before and after each titration session with DDI. Fullerene aggregate size and ζ-potential were determined using a ZetaSizer Nano ZS as described above. Since the pI’s of the fullerenes were slightly lower than the lowest titrated pH, pI’s were estimated by fitting a regression line to data points in the linear region between pH 1 and 2 and extrapolating to ζ-potential of 0. Fullerene Aggregation Kinetics. Time-resolved dynamic light scattering (TRDLS) was used to measure fullerene aggregate intensity-averaged hydrodynamic diameter (Dh) change with time and as a function of background solution ionic strength. In these experiments, 600 µL of aqueous fullerene suspension was pipetted into a quartz cuvette, and an additional 600 µL aliquot of either DDI or of a stock NaCl solution was added to yield a specific NaCl and fullerene concentration. The final background solutions included DDI and solutions with NaCl concentrations ranging from 25 to 1000 mM NaCl. The capped cuvette was then immediately vortexed for 1 s and then placed in the DLS instrument. Autocorrelation functions were accumulated for 15 s, and since the different fullerenes exhibited a wide range in aggregation rate, measurements were conducted over time periods ranging from 25 to 300 min. The initial aggregation period was defined as the time period from experiment initiation (t0) to the time when measured Dh values exceeded 1.25Dh,initial. The initial aggregation rate constants (ka) for the fullerenes are proportional to the initial rate of increase of Dh with time (13, 25) ka ∝
(
1 dDh(t) N0 dt
)
(1)
tf0
where N0 is the initial particle concentration. The particle attachment efficiency R (or inverse stability ratio, 1/W) is used to quantify fullerene particle aggregation kinetics; it is defined as the initial aggregation rate constant (ka) normalized by the aggregation rate constant measured under diffusionlimited (fast) conditions (13)
ka 1 R) ) ) W ka,fast
( ) ( )
1 Dh(t) N0 dt Dh(t) 1 N0,fast dt
tf0
(2)
tf0,fast
Using the sample preparation technique described above, the concentration of each fullerene suspension (half that reported in Table 1 due to 1:1 dilution) remained constant for each measurement made at varying NaCl concentration. This simplifies eq 2 (i.e., No drops out) so that R can be determined directly by normalizing the initial slope of the aggregation profile for a specific background solution chemistry by the initial slope under diffusion-limited (fast) conditions.
FIGURE 2. Representative TEM and high-resolution TEM (insets) of (a) aqu/PCBO and (b) aqu/C60 aggregates. Sonicated fullerene suspensions were dropped onto 300 mesh Formvar coated copper grids. Images were acquired on a Hitachi-7600 at 120 kV and on a Hitachi-9500 at 300 kV.
Results and Discussion Characterization of Fullerenes. The 190-800 nm UV scans for each PCB-R fullerene in 70/30 v/v toluene-acetonitrile are nearly identical; the scan for C60 is very similar to those of the PCB-R fullerenes, only it is shifted to slightly longer wavelengths (Figure SI-1 of the Supporting Information). The best analytical wavelengths for the fullerenes were determined to be 330 nm for the PCB-R fullerenes and 333 nm for C60 with the log molar absorption coefficient (log ε) values g4.5 at these wavelengths (Table SI-1 of the Supporting Information). Although it can be seen in the UV scans in Figure SI-1 that the PCB-R fullerenes have marginally higher absorbances near 280 nm, baseline noise is also significantly greater at these lower wavelengths due to toluene absorbance. These observations are consistent with prior studies on C60 (26) and PCBM (24). TEM images indicate that the aqu/fullerene aggregate population is polydisperse and characterized by both irregular and spherical structures (Figure 2a, b). Large (>200 nm) particles that had partially eroded under stirring and small particles (
