Effects of Sonication on the Size and Crystallinity of Stable Zwitterionic ...
7990
Langmuir 2005, 21, 7990-7994
Effects of Sonication on the Size and Crystallinity of Stable Zwitterionic Organic Nanoparticles Formed by Reprecipitation in Water Rabih O. Al-Kaysi, Astrid M. Mu¨ller, Tai-Sang Ahn, Soohyun Lee, and Christopher J. Bardeen* Department of Chemistry, University of California, Riverside, California 92521 Received May 2, 2005. In Final Form: June 17, 2005 Nanoparticles of a novel organic zwitterionic Meisenheimer complex, N′,N′′,N′′′-tri(isopropyl)-4-oxo-6(isopropyliminio)-2-s-(2H)triazinespiro-1′-2′,4′,6′-trinitrocyclohexadienylide, were synthesized by reprecipitation in water under different conditions. While reprecipitation alone resulted in a suspension of amorphous particles that fell out of solution within hours, sonication for different periods of time resulted in the formation of crystalline particles that were stable in solution over the course of weeks. The diskshaped particles had an average diameter of 140 nm and a thickness of 70 nm. Comparison of the optical spectroscopy of these particles with the monomer indicates that they possess delocalized excitonic states and enhanced radiative decay rates. The use of zwitterionic molecules in conjunction with sonication provides a way to exert some level of control over particle size and morphology, as well as increased colloidal stability.
Introduction In the past decade, several methods have been developed that yield exquisite control over both the size and shape of inorganic and metal nanoparticles. These inorganic nanoparticles have been shown to have a variety of novel electronic properties, with potential applications in fields as diverse as biological labeling,1,2 lasers,3 and solar energy conversion.5,6 A natural question is whether nanoparticles composed of organic semiconductors can also yield interesting electronic properties. Organic nanoparticles can exhibit interesting electronic phenomena, including photocatalytic activity,7 enhanced photoinduced charge separation,8,9 and size effects on the electronic properties.10-12 Nevertheless, progress in this area has lagged behind the inorganic field. One problem is the lack of general methods for controlling the average particle size and dispersity of organic nanoparticles. While optimized methods for making inorganic nanocrystals are now well-established,13 comparable guidelines for fabricating organic nanoparticles have yet to be formulated. Understanding how * Corresponding author: phone 951-827-2723; fax 951-827-4713; e-mail [email protected]. (1) Bruchez, M.; Moronne, M.; Gin, P.; Weiss, S.; Alivisatos, A. P. Science 1998, 281, 2013-2016. (2) Chan, W. C. W.; Nie, S. Science 1998, 281, 2016-2019. (3) Klimov, V. I.; Mikhailovsky, A. A.; Xu, S.; Malko, A.; Hollingsworth, J. A.; Leatherdale, C. A.; Eisler, H.-J.; Bawendi, M. G. Science 2000, 290, 314-317. (4) Kamat, P. V. J. Phys. Chem. B 2002, 106, 7729-7744. (5) Gratzel, M. J. Photochem. Photobiol., A 2004, 164, 3-14. (6) Schaller, R. D.; Klimov, V. I. Phys. Rev. Lett. 2004, 92, 186601/ 1-4. (7) Kim, H. Y.; Bjorklund, T. G.; Lim, S.-H.; Bardeen, C. J. Langmuir 2003, 19, 3941-3946. (8) Biju, V.; Sudeep, P. K.; Thomas, K. G.; George, M. V.; Barazzouk, S.; Kamat, P. V. Langmuir 2002, 18, 1831-1839. (9) Hasobe, T.; Imahori, H.; Fukuzumi, S.; Kamat, P. V. J. Phys. Chem. B 2003, 107, 12105-12112. (10) Matsui, A. H.; Mizuno, K.; Nishi, O.; Matsushima, Y.; Shimizu, M.; Goto, T.; Takeshima, M. Chem. Phys. 1995, 194, 167-174. (11) Kasai, H.; Kamatani, H.; Okada, S.; Oidawa, H.; Matsuda, H.; Nakanishi, H. Jpn. J. Appl. Phys. 1996, 35, L221-L223. (12) Fu, H.-B.; Yao, J.-N. J. Am. Chem. Soc. 2001, 123, 1434-1439. (13) Cushing, B. L.; Kolesnichenko, V. L.; O’Connor, C. J. Chem. Rev. 2004, 104, 3893-3946.
various types of noncovalent interactions allow for nucleation and growth, and then stabilize the aggregate, is a complex challenge in colloidal science.14 A second problem is that organic nanoparticles tend to be metastable in solution, often aggregating and falling out of solution over the course of a few days. Despite these challenges, some progress has been made. In the simplest synthetic approach, a dilute solution of the monomer is rapidly injected into a solvent in which the monomer is insoluble.15,16 The subsequent reprecipitation of the monomer molecules forms a colloidal solution of particles whose size can range from nanometers to micrometers. To gain better control over the reprecipitation process, workers have varied experimental parameters such as initial concentration and temperature11,12 and have added surfactants.17 All these methods have met with limited success, and often it is difficult to quantify the size of the effect. Physical agitation or milling has also been used to make organic nanoparticles, usually in the form of mechanical crushing or sonication.18 In one case, a combination of reprecipitation and sonication was used to make anthracene nanoparticles,19 although there was no evidence that this resulted in smaller or more monodisperse nanoparticles. The empirical observation that molecules which can carry charge on one or more moieties seem to provide more control over nanoparticle size and stability12 led us to examine nanoparticle formation using a recently synthesized zwitterionic compound, N′,N′′,N′′′-tri(isopropyl)-4-oxo-6-(isopropyliminio)-2-s-(2H)(14) Keuren, E. V.; Georgieva, E.; Adrian, J. Nanoletters 2001, 1, 141-144. (15) Kasai, H.; Nalwa, H. S.; Oikawa, H.; Okada, S.; Matsuda, H.; Minami, N.; Kakuta, A.; Ono, K.; Mukoh, A.; Nakanishi, H. Jpn. J. Appl. Phys. 1992, 31, L1132-L1134. (16) LaMer, V. K.; Dinegar, R. H. J. Am. Chem. Soc. 1950, 72, 48474854. (17) Debuigne, F.; Jeunieau, L.; Wiame, M.; Nagy, J. B. Langmuir 2000, 16, 7605-7611. (18) Horn, D.; Rieger, J. Angew. Chem., Int. Ed. 2001, 40, 43304361. (19) Kang, P.; Chen, C.; Hao, L.; Zhu, C.; Hu, Y.; Chen, Z. Mater. Res. Bull. 2004, 39, 545-551.
10.1021/la051183b CCC: $30.25 © 2005 American Chemical Society Published on Web 07/22/2005
Sonication of Zwitterionic Organic Nanoparticles
Figure 1. Normalized ground-state absorption (s) and fluorescence (---) spectra of the monomeric MHC in THF at room temperature. The structure of MHC is shown in the inset.
triazinespiro-1′-2′,4′,6 ′-trinitrocyclohexadienylide.20 This molecule, whose structure is shown in an inset to Figure 1, belongs to a family of molecules known as Meisenheimer complexes21 and will be denoted by the abbreviation MHC. This compound, whose crystal structure is known, has a fluorescence quantum yield in tetrahydrofuran of Φf(THF) ) 0.50 and a sizable transition dipole moment as judged from its peak absorption coefficient of 19 000 cm-1 M-1 at 528 nm.22 Interactions between the transition dipoles of neighboring molecules in the nanoparticle would be expected to form new excitonic states, with potentially useful emissive or nonlinear optical properties. In this paper, we present data on the formation and the basic photophysical properties of MHC nanoparticles. Using controlled sonication in conjunction with reprecipitation in water, we show how both the size and crystallinity of the nanoparticles can be controlled. By careful adjustment of the experimental conditions, we obtain a dilute solution of nanocrystals that is stable for weeks when stored at room temperature. The crystalline nanoparticles exhibit the clear spectroscopic signatures of J-aggregate formation, including a change in vibronic line shape and an enhanced radiative decay rate. The combination of reprecipitation and sonication reported here provides an effective way to produce stable organic nanoparticles with interesting optical properties and to guide further advances in the preparation of these potentially useful semiconductor colloids.
Langmuir, Vol. 21, No. 17, 2005 7991 starting the next sonication period. The resulting solution is clear and pink-colored. Filtering the solution through a 0.8 µm nitrocellulose Millipore disk filter removed the largest remaining MHC particles, and the filtered solution was used for further study. By use of this procedure, the particle sizes and electronic spectra were identical between different preparations, so the reproducibility was quite good for a colloidal synthesis. Nanoparticle Characterization and Photophysics. A drop of the nanoparticle solution was deposited on a freshly cleaved 1 × 1 cm mica surface (SPI, V1 mica) and dried in a Petri dish filled with Dryerite covered with a Kimwipe. Atomic force microscopy (Novascan, ESPM II) was done in intermittent tapping mode to minimize surface perturbation. The nanoparticles were also characterized by their steady-state UV-vis absorption (on an Ocean Optics SD2000 absorption spectrometer with a 10 cm path length cell). Fluorescence measurements and quantum yields were determined using Spex Fluorolog Tau-3 fluorescence spectrophotometer. Samples were prepared with maximum optical density of 0.1 at the excitation wavelength, to minimize self-absorption and quenching. A solution of anthracene (Aldrich, zone refined) in ethanol was used as a standard for the determination of Φf of the MHC/THF solution.23 Five concentrations of anthracene in ethanol were prepared to give optical densities of 0.02, 0.014, 0.01, 0.006, and 0.001 at the excitation wavelength of 356 nm. These samples were used as calibration standards. Similar optical densities of MHC solution in THF were used. The area of the fluorescence emission was plotted against the corresponding optical density to give a straight line. The slopes of both straight lines and the fluorescence quantum yield of anthracene in ethanol (Φf ) 0.27) were used to determine Φf of MHC.24 For determination of the nanoparticle quantum yield, we used the Φf value obtained for the MHC/THF solution and excited the samples at 500 nm to reduce the effect of light scattering from the nanoparticle sample. Fluorescence lifetimes were measured by exciting the samples with pulses centered at 400 nm derived from a 40 kHz regeneratively amplified Ti: sapphire laser system with a pulse width of 150 fs and an instrument response of 15 ps. When the data for this paper were taken, a replica pulse at a delay of about 50 ps resulted in an apparent broadening of the instrument response near zero delay. This can be seen in the rounded appearance in the signal peak in Figure 4 but does not affect the decay times extracted from the data. The sample was in a 1 cm path length quartz cuvette, and the emission was collected at 90° relative to the excitation. The emission was directed into a monochromator attached to a picosecond streak camera (Hamamatsu C4334 Streakscope), which provides both time- and wavelength-resolved fluorescence data, with resolutions of 15 ps and 2.5 nm, respectively.
Experimental Section
Results and Discussion
Synthesis of MHC. The zwitterionic Meisenheimer complex was synthesized and purified according to a previously reported method.20 The solid sample is indefinitely stable at room temperature. Synthesis of MHC Nanoparticles. Ultrapure water (20.0 mL) from a Milli-Q filtration system (Millipore) was added to a 25 mL vial and chilled to 5 °C in a beaker of ice-water slush. A solution of MHC (4.3 × 10-3 M) in tetrahydrofuran (THF; Aldrich, Optima grade), was prepared and stored in the dark. A titanium probe sonicator (horn diameter 0.5 cm) was inserted (1 in. from the bottom) into the chilled vial of water. The sonicator (S-250/450A Analogue sonifier with a 400 W output) was set at an amplitude of 4 with an 80% duty cycle. While the water was being sonicated, 47 µL of the MHC solution was rapidly injected, giving an effective concentration of MHC in water of 1 × 10-5 M. The solution is then sonicated for a total of 12 min, making sure the temperature of the water/MHC nanoparticles does not rise above 10 °C. This is accomplished by sonicating in 4 min intervals, then pausing and letting the solution cool to 5 °C before
Figure 1 shows the normalized absorption and fluorescence spectra of monomeric MHC in room-temperature THF. The molecular structure is shown in the inset. The absorption has two peaks at 407 and 528 nm (typical of a σ complex ground-state absorption spectrum) and shows a pronounced vibronic shoulder at 504 nm whose height is about 70% of the peak at 528 nm. The MHC studied in this work is unique in that it is not a salt but a zwitterion, which is soluble in most organic solvents but not in water. Its fluorescence, peaked at 558 nm in THF, has a slight charge-transfer character but originates from the same state as the absorption, as surmised from its small Stokes shift (1.0 × 103 cm-1) and the fact that it mirrors the absorption line shape. The fluorescence quantum yield Φf ) 0.50 is also significantly greater than that of most Meisenheimer complex salts, which are typically 10% or less.25,26 The fluorescence decay in THF is singleexponential with a lifetime of 9.7 ns, very similar to the
(20) Al-Kaysi, R. O.; Guirado, G.; Valente, E. J. Eur. J. Org. Chem. 2004, 3408-3411. (21) Terrier, F. Nucleophilic Aromatic Displacement; VCH: New York, 1991. (22) Al-Kaysi, R. O.; Creed, D.; Valente, E. J. J. Chem. Crystallogr. 2004, 34, 685-692.
(23) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of Photochemistry, 2nd ed.; Marcel Dekker: New York, 1993. (24) Melhuish, W. H. J. Phys. Chem. 1961, 65, 229-235. (25) Farnham, S.; Taylor, R. J. Org. Chem. 1974, 39, 2446-2448. (26) Hiratsuka, T. Eur. J. Biochem. 2003, 270, 3479-3485.
7992
Langmuir, Vol. 21, No. 17, 2005
Al-Kaysi et al.
Figure 2. (a) Normalized absorption spectrum of the MHC reprecipitated in stirred water (s) and normalized spectrum of a film of MHC spin cast from a THF solution (---). (b) Normalized absorption spectra of MHC (47 µL of 4.3 × 10-3 M THF solution) injected in water and sonicated for 1 min (upper), 12 min (3 × 4 min intervals) (middle), and then passed through a 0.8 µm filter (lower). As the average particle size becomes smaller, the spectrum narrows and the low energy wing decreases. (c) Normalized absorption spectrum for a sample of MHC that undergoes uninterrupted sonication for 20 min, reaching a final temperature of 40 °C. The elevated baseline is the result of increased light scattering.
value of 9.1 ns previously obtained in CH2Cl2.20 In short, the MHC shown in the inset of Figure 1 combines the relatively simple photophysics of an organic dye with a highly polarized zwitterionic structure. When a concentrated MHC/THF solution is injected into highly pure H2O, which is stirred but not otherwise perturbed, the MHC does not crystallize. Figure 2a shows the absorption spectrum of such a solution, which looks qualitatively similar to that of the monomer in THF. The main differences are a slight red shift (2 nm) and broadening of the line shape. If nothing further is done to this sample, within 2 h the MHC falls out of solution and coats the sides and bottom of the sample flask with a thin red layer. The absorption line shape in Figure 2a does not change at all but just diminishes with time. The same spectrum is also seen when the MHC is spin-cast on a surface directly from THF, as shown in Figure 2a, which indicates that this spectrum is characteristic of amorphous MHC and basically reflects the monomer in a highly disordered environment. Although the similar spectra are suggestive, we cannot say with certainty if the unsonicated MHC in water exists in amorphous aggregates or in a monomeric form. If, on the other hand, the MHC solution is injected into a water solution that is sonicated, within 1 min, crystallites are formed, as judged by the dramatic changes in the absorption line shape. Relative to the monomer spectrum, the new absorption spectrum is narrower and red-shifts to 548 nm, with a decrease in the intensity of the high-energy vibronic shoulder from 70% to 35% of the absorption peak. Also, there is a pronounced low-energy wing that extends from 550 nm out past 700 nm. This wing is the result of scattering due to large (on the order of a wavelength of light) particles in the sample. This conclusion is supported by the fact that this wing is not observed in the fluorescence
excitation spectrum, and the observation of large particles in AFM measurements on this sample. If sonication is continued for another set of three 4-min periods with 2-min cooling intervals, the low-energy wing decreases. Passing the sample through a 0.8 µm filter results in an additional decrease in the low-energy wing, as well as a 30% reduction of the peak absorption. The normalized absorption of the filtered sample is the lowest curve shown in Figure 2b. Passing the previously filtered sample through a 0.2 µm filter reduces the peak absorption of the sample without affecting the overall shape of the spectrum. The 0.8 µm filtered solution is quite stable. After an initial drop of about 9% in OD at λmax during the first 15 h, the peak absorbance remains constant for a period of weeks while stored at constant room temperature. This is in contrast to aqueous solutions of tetracene nanoparticles, whose absorption decreases by 50% within about 5 days after formation.7 This increased stability is consistent with previous observations of surfaces and nanoparticles that have been stabilized with zwitterionic ligands27,28 and is also consistent with the empirical observation that molecules which can acquire a partial charge, for example, that contain a carbonyl or nitro group, can form stable, relatively monodisperse solutions. The formation of a negative ζ charge on the surface layer of these particles has been postulated to be responsible for this phenomenon but has not been conclusively proven.12 The duration of sonication is critical in the formation of this type of sample. If sonication is continued without interruption for longer periods of time, larger particles are re-formed, as shown in Figure 2c, where there is now (27) Tatumi, R.; Fujihara, H. Chem. Commun. 2005, 83-85. (28) Kitano, H.; Kawaski, A.; Kawaski, H.; Morokoshi, S. J. Colloid Interface Sci. 2005, 282, 340-348.
Sonication of Zwitterionic Organic Nanoparticles
Langmuir, Vol. 21, No. 17, 2005 7993
Figure 3. (a) AFM image of the 12 min (3 × 4 min intervals) sonicated sample after passage through a 0.8 µm filter. This sample corresponds to the lowest spectrum in Figure 2b. (b) AFM image of MHC nanoparticles collected after 20 min of uninterrupted sonication. This sample corresponds to the spectrum in Figure 2c. Note the different scale.
a very pronounced scattering background. This reaggregation of the particles does not seem to be due to the sonication itself but is instead an artifact of the heating of the solution, which induces reaggregation of the particles. If the 12-min sonication period is broken up into 4-min intervals followed by 2-min cooling periods, the temperature of the solution never rises above 10 °C, and the absorption of the middle curve in Figure 2b is obtained. On the other hand, 20 min of uninterrupted sonication leads to a final solution temperature of 40 °C and the absorption spectrum shown in Figure 2c. Higher temperatures have been observed to lead to larger crystallites in other systems,11,15 most likely due to more rapid aggregation kinetics. We have also seen changes indicative of larger crystallites when we heat a solution of small nanocrystals up to 40 °C for a period of 10 min or more. The key observation is that solution heating means that there is an optimum duration for sonication to initiate growth and refine the particle size. Ideally, we would be able to control the temperature of the solution at the point of sonication, for example, by using a rapidly recirculating sample cell as opposed to an external ice bath. Such a device was unavailable during these experiments. From the absorption spectra in Figure 2, we surmise that the formation of MHC nanoparticles in water proceeds in the following steps: (1) supersaturation of the MHC monomer in water, (2) sonication-induced nucleation and crystal growth, (3) sonication milling of the larger particles into smaller, more uniform nanocrystals, and (4) reaggregation of the nanocrystals as the temperature of the sonicated solution rises. Further support for this sequence of events is given by AFM data taken at stages 3 and 4 in the particle development. After 12 min of intermittent sonication and subsequent filtration, the sample consists almost entirely of the disk-shaped nanocrystals shown in Figure 3a. Their average dimensions, about 140 nm in diameter and 70 nm in height, are smaller than visible light wavelengths, explaining the reduced scattering seen in the absorption spectrum in Figure 2b. As the period of sonication is increased, large micrometer-sized aggregates reappear, as shown in Figure 3b. In this scenario, sonication fulfills several functions. First, at the molecular scale, it seems to provide a uniform start for the nucleation process. Sonication is not necessary for nanocrystal formation, which can also be initiated in solutions of lower purity water or water that has been seeded with nanocrystals from a sonication preparation. But these particles are typically larger, with a very broad size distribution,
Figure 4. Normalized fluorescence decay of the MHC nanoparticle sample from Figures 2b (lowest curve) and 3a. The signal is integrated over all wavelengths and plotted on a logarithmic scale to show both the fast (135 ps) and slow (1.2 ns) decay components.
when compared with the sonicated preparation. Second, at the nanometer scale, it appears to prevent the formation of very large particles, possibly through a milling type of process or due to small-scale turbulence in the solution. Similar experiments in our lab on sonication of reprecipitated anthracene showed little or no effect of reprecipitation on either nucleation or particle size, so the success of this method will likely depend on the chemical nature of the molecule used. The ability to form size-controlled nanoparticles of the MHC permits the study of their photophysical properties. In addition to the steady-state absorption spectra in Figure 2, we have also measured the steady-state and timeresolved fluorescence of these nanocrystals in water at room temperature. The fluorescence quantum yield of the nanocrystals is 0.04, more than an order of magnitude less than that of the monomer in THF. The reason for this decrease is the very short lifetime of the nanocrystal emission, as shown in Figure 4. The bulk of the biexponential fluorescence decay (92%) occurs with a time constant of 135 ps, with a smaller, long-lived component that decays with a time constant of 1.2 ns. We examine the spectra of the short- and long-lived components in Figure 5. The fluorescence within the first 1 ns, shown in Figure 5a, is dominated by a sharp spectrum that looks very similar to the absorption and has a small Stokes shift of 430 cm-1. After this initial decay, which accounts for 92% of the integrated fluorescence, a broader, redshifted spectrum is observed, which is very similar to that of MHC injected into water without sonication. The measured decay times of 1.3 ns for the unsonicated MHC in water and 1.2 ns for the long-lived component in the
7994
Langmuir, Vol. 21, No. 17, 2005
Al-Kaysi et al.
Figure 5. (a) Normalized fluorescence MHC nanocrystals in water collected during the interval 0.0-0.8 ns after excitation, showing the shape of the short-lived component that accounts for 92% of the total fluorescence. The integrated fluorescence is very similar to this spectrum. (b) Normalized fluorescence of the long-lived component, collected during the interval 0.8-1.8 ns after excitation (s), compared with the fluorescence spectrum of amorphous MHC in water (---).
nanoparticle sample are very similar. Figure 5b shows these two spectra normalized and overlapped. These data show how sonication does not completely remove the noncrystalline fraction of the MHC in water, which may be either associated with the nanocrystals or suspended independently in solution. It is expected that both reprecipitation and sonication can lead to nonequilibrium structures that might act as defects in these particles. The time-resolved fluorescence data did not indicate any energy transfer between the short-lived nanocrystal peak and the longer-lived amorphous peak, however, so there is no direct spectroscopic evidence that the noncrystalline MHC acts as a low-energy trap for the initially excited exciton. The final question concerns the origin of the new absorption and emission line shapes. The combination of shift to lower energy, sharpening, and the decrease in the strength of the higher energy vibronic peaks are all qualitatively consistent with J-aggregate formation, as has been surmised for other organic nanocrystal systems.29 The complexity of the MHC crystal structure, along with its complex molecular geometry, precludes easy prediction of whether the transition dipole orientations in the crystal favor J- or H-aggregates.30 Most likely, it forms a J-like aggregate, as in the case of tetracene,31 where there is an enhanced oscillator strength at the bottom of the exciton band, but that enhancement is not necessarily linearly proportional to the number of monomers contained in the aggregate, as would be the case in an ideal J-aggregate. Nevertheless, we can show that the new states seen in the nanocrystals do possess the enhanced radiative decay rate (superradiance) characteristic of J-like aggregates. If we consider the fraction of the total quantum yield originating from the short-lived component, we can calculate the radiative rate for this component according to
krad )
(fraction of fluorescence) × (quantum yield) (fluorescence decay time) (1)
With a fluorescence decay time of 135 ps, a quantum yield of 0.04, and a fraction of fluorescence of 0.92, we obtain krad ) 2.7 × 108 s-1. For the monomer, we use the usual relationship (i.e., the fraction of the fluorescence from the state of interest is assumed to be 1.0), and using a (29) An, B.-K.; Kwon, S.-K.; Jung, S.-D.; Park, S. Y. J. Am. Chem. Soc. 2004, 124, 14410-14415. (30) Kasha, M.; Rawls, H. R.; El-Bayoumi, M. A. J. Pure. Appl. Chem. 1965, 11, 371-392. (31) Lim, S.-H.; Bjorklund, T. G.; Spano, F. C.; Bardeen, C. J. Phys. Rev. Lett. 2004, 92, 107402/1-4.
fluorescence decay time of 9.7 ns and a quantum yield of 0.50, we obtain krad ) 5.2 × 107 s-1. Thus the nanocrystals have a radiative decay rate that is enhanced by roughly a factor of 5 relative to the monomer. As is often the case in solid-state systems, this enhanced radiative decay rate is more than compensated by enhanced nonradiative decay channels that are present in the nanoparticles. Thus the total quantum yield decreases, despite the increase in krad. Nevertheless, the enhancement in krad, along with the other spectral changes observed as the MHC forms nanocrystals, provides clear evidence for the formation of a delocalized excitonic state within the nanoparticles. Conclusion We have shown how reprecipitation in conjunction with sonication provides an efficient way to synthesize organic nanoparticles. By controlling the duration of sonication, we were able to control both the size distribution and crystallinity of the particles. We hypothesize that this control, along with the enhanced stability of the MHC nanoparticles, results from the unique surface charge properties of the zwitterion, which slow the reaggregation process. These nanoparticles exhibit a red shift, a sharpening of the spectrum, and a decrease in vibronic line intensity, all consistent with the formation of a superradiant J-type aggregate. The enhanced radiative decay rate (krad), as compared to that of the monomer, confirms this. Although this work is preliminary, it clearly justifies the further investigation of highly charged organic compounds for use in producing stable organic colloids with interesting electronic properties. While the generality of the sonication method is an open question, we hope that by controlling the sonication conditions more precisely, especially the solution temperature and the ultrasound intensity, we can achieve more systematic control over the nanoparticle size distribution. The extension of this work to different types of zwitterionic compounds with different photophysical and photochemical properties should yield more information about the role of surface charges and crystal structure on nanoparticle size and stability. Acknowledgment. This material is based upon work supported by the National Science Foundation under Grant 0517095. We gratefully acknowledge the use of the sonicator in Professor Q. Cheng’s lab at UCR. We also acknowledge Professor Edward J. Valente of Mississippi College for providing the X-ray structure of the MHC. LA051183B
Langmuir 2005, 21, 7990-7994
Effects of Sonication on the Size and Crystallinity of Stable Zwitterionic Organic Nanoparticles Formed by Reprecipitation in Water Rabih O. Al-Kaysi, Astrid M. Mu¨ller, Tai-Sang Ahn, Soohyun Lee, and Christopher J. Bardeen* Department of Chemistry, University of California, Riverside, California 92521 Received May 2, 2005. In Final Form: June 17, 2005 Nanoparticles of a novel organic zwitterionic Meisenheimer complex, N′,N′′,N′′′-tri(isopropyl)-4-oxo-6(isopropyliminio)-2-s-(2H)triazinespiro-1′-2′,4′,6′-trinitrocyclohexadienylide, were synthesized by reprecipitation in water under different conditions. While reprecipitation alone resulted in a suspension of amorphous particles that fell out of solution within hours, sonication for different periods of time resulted in the formation of crystalline particles that were stable in solution over the course of weeks. The diskshaped particles had an average diameter of 140 nm and a thickness of 70 nm. Comparison of the optical spectroscopy of these particles with the monomer indicates that they possess delocalized excitonic states and enhanced radiative decay rates. The use of zwitterionic molecules in conjunction with sonication provides a way to exert some level of control over particle size and morphology, as well as increased colloidal stability.
Introduction In the past decade, several methods have been developed that yield exquisite control over both the size and shape of inorganic and metal nanoparticles. These inorganic nanoparticles have been shown to have a variety of novel electronic properties, with potential applications in fields as diverse as biological labeling,1,2 lasers,3 and solar energy conversion.5,6 A natural question is whether nanoparticles composed of organic semiconductors can also yield interesting electronic properties. Organic nanoparticles can exhibit interesting electronic phenomena, including photocatalytic activity,7 enhanced photoinduced charge separation,8,9 and size effects on the electronic properties.10-12 Nevertheless, progress in this area has lagged behind the inorganic field. One problem is the lack of general methods for controlling the average particle size and dispersity of organic nanoparticles. While optimized methods for making inorganic nanocrystals are now well-established,13 comparable guidelines for fabricating organic nanoparticles have yet to be formulated. Understanding how * Corresponding author: phone 951-827-2723; fax 951-827-4713; e-mail [email protected]. (1) Bruchez, M.; Moronne, M.; Gin, P.; Weiss, S.; Alivisatos, A. P. Science 1998, 281, 2013-2016. (2) Chan, W. C. W.; Nie, S. Science 1998, 281, 2016-2019. (3) Klimov, V. I.; Mikhailovsky, A. A.; Xu, S.; Malko, A.; Hollingsworth, J. A.; Leatherdale, C. A.; Eisler, H.-J.; Bawendi, M. G. Science 2000, 290, 314-317. (4) Kamat, P. V. J. Phys. Chem. B 2002, 106, 7729-7744. (5) Gratzel, M. J. Photochem. Photobiol., A 2004, 164, 3-14. (6) Schaller, R. D.; Klimov, V. I. Phys. Rev. Lett. 2004, 92, 186601/ 1-4. (7) Kim, H. Y.; Bjorklund, T. G.; Lim, S.-H.; Bardeen, C. J. Langmuir 2003, 19, 3941-3946. (8) Biju, V.; Sudeep, P. K.; Thomas, K. G.; George, M. V.; Barazzouk, S.; Kamat, P. V. Langmuir 2002, 18, 1831-1839. (9) Hasobe, T.; Imahori, H.; Fukuzumi, S.; Kamat, P. V. J. Phys. Chem. B 2003, 107, 12105-12112. (10) Matsui, A. H.; Mizuno, K.; Nishi, O.; Matsushima, Y.; Shimizu, M.; Goto, T.; Takeshima, M. Chem. Phys. 1995, 194, 167-174. (11) Kasai, H.; Kamatani, H.; Okada, S.; Oidawa, H.; Matsuda, H.; Nakanishi, H. Jpn. J. Appl. Phys. 1996, 35, L221-L223. (12) Fu, H.-B.; Yao, J.-N. J. Am. Chem. Soc. 2001, 123, 1434-1439. (13) Cushing, B. L.; Kolesnichenko, V. L.; O’Connor, C. J. Chem. Rev. 2004, 104, 3893-3946.
various types of noncovalent interactions allow for nucleation and growth, and then stabilize the aggregate, is a complex challenge in colloidal science.14 A second problem is that organic nanoparticles tend to be metastable in solution, often aggregating and falling out of solution over the course of a few days. Despite these challenges, some progress has been made. In the simplest synthetic approach, a dilute solution of the monomer is rapidly injected into a solvent in which the monomer is insoluble.15,16 The subsequent reprecipitation of the monomer molecules forms a colloidal solution of particles whose size can range from nanometers to micrometers. To gain better control over the reprecipitation process, workers have varied experimental parameters such as initial concentration and temperature11,12 and have added surfactants.17 All these methods have met with limited success, and often it is difficult to quantify the size of the effect. Physical agitation or milling has also been used to make organic nanoparticles, usually in the form of mechanical crushing or sonication.18 In one case, a combination of reprecipitation and sonication was used to make anthracene nanoparticles,19 although there was no evidence that this resulted in smaller or more monodisperse nanoparticles. The empirical observation that molecules which can carry charge on one or more moieties seem to provide more control over nanoparticle size and stability12 led us to examine nanoparticle formation using a recently synthesized zwitterionic compound, N′,N′′,N′′′-tri(isopropyl)-4-oxo-6-(isopropyliminio)-2-s-(2H)(14) Keuren, E. V.; Georgieva, E.; Adrian, J. Nanoletters 2001, 1, 141-144. (15) Kasai, H.; Nalwa, H. S.; Oikawa, H.; Okada, S.; Matsuda, H.; Minami, N.; Kakuta, A.; Ono, K.; Mukoh, A.; Nakanishi, H. Jpn. J. Appl. Phys. 1992, 31, L1132-L1134. (16) LaMer, V. K.; Dinegar, R. H. J. Am. Chem. Soc. 1950, 72, 48474854. (17) Debuigne, F.; Jeunieau, L.; Wiame, M.; Nagy, J. B. Langmuir 2000, 16, 7605-7611. (18) Horn, D.; Rieger, J. Angew. Chem., Int. Ed. 2001, 40, 43304361. (19) Kang, P.; Chen, C.; Hao, L.; Zhu, C.; Hu, Y.; Chen, Z. Mater. Res. Bull. 2004, 39, 545-551.
10.1021/la051183b CCC: $30.25 © 2005 American Chemical Society Published on Web 07/22/2005
Sonication of Zwitterionic Organic Nanoparticles
Figure 1. Normalized ground-state absorption (s) and fluorescence (---) spectra of the monomeric MHC in THF at room temperature. The structure of MHC is shown in the inset.
triazinespiro-1′-2′,4′,6 ′-trinitrocyclohexadienylide.20 This molecule, whose structure is shown in an inset to Figure 1, belongs to a family of molecules known as Meisenheimer complexes21 and will be denoted by the abbreviation MHC. This compound, whose crystal structure is known, has a fluorescence quantum yield in tetrahydrofuran of Φf(THF) ) 0.50 and a sizable transition dipole moment as judged from its peak absorption coefficient of 19 000 cm-1 M-1 at 528 nm.22 Interactions between the transition dipoles of neighboring molecules in the nanoparticle would be expected to form new excitonic states, with potentially useful emissive or nonlinear optical properties. In this paper, we present data on the formation and the basic photophysical properties of MHC nanoparticles. Using controlled sonication in conjunction with reprecipitation in water, we show how both the size and crystallinity of the nanoparticles can be controlled. By careful adjustment of the experimental conditions, we obtain a dilute solution of nanocrystals that is stable for weeks when stored at room temperature. The crystalline nanoparticles exhibit the clear spectroscopic signatures of J-aggregate formation, including a change in vibronic line shape and an enhanced radiative decay rate. The combination of reprecipitation and sonication reported here provides an effective way to produce stable organic nanoparticles with interesting optical properties and to guide further advances in the preparation of these potentially useful semiconductor colloids.
Langmuir, Vol. 21, No. 17, 2005 7991 starting the next sonication period. The resulting solution is clear and pink-colored. Filtering the solution through a 0.8 µm nitrocellulose Millipore disk filter removed the largest remaining MHC particles, and the filtered solution was used for further study. By use of this procedure, the particle sizes and electronic spectra were identical between different preparations, so the reproducibility was quite good for a colloidal synthesis. Nanoparticle Characterization and Photophysics. A drop of the nanoparticle solution was deposited on a freshly cleaved 1 × 1 cm mica surface (SPI, V1 mica) and dried in a Petri dish filled with Dryerite covered with a Kimwipe. Atomic force microscopy (Novascan, ESPM II) was done in intermittent tapping mode to minimize surface perturbation. The nanoparticles were also characterized by their steady-state UV-vis absorption (on an Ocean Optics SD2000 absorption spectrometer with a 10 cm path length cell). Fluorescence measurements and quantum yields were determined using Spex Fluorolog Tau-3 fluorescence spectrophotometer. Samples were prepared with maximum optical density of 0.1 at the excitation wavelength, to minimize self-absorption and quenching. A solution of anthracene (Aldrich, zone refined) in ethanol was used as a standard for the determination of Φf of the MHC/THF solution.23 Five concentrations of anthracene in ethanol were prepared to give optical densities of 0.02, 0.014, 0.01, 0.006, and 0.001 at the excitation wavelength of 356 nm. These samples were used as calibration standards. Similar optical densities of MHC solution in THF were used. The area of the fluorescence emission was plotted against the corresponding optical density to give a straight line. The slopes of both straight lines and the fluorescence quantum yield of anthracene in ethanol (Φf ) 0.27) were used to determine Φf of MHC.24 For determination of the nanoparticle quantum yield, we used the Φf value obtained for the MHC/THF solution and excited the samples at 500 nm to reduce the effect of light scattering from the nanoparticle sample. Fluorescence lifetimes were measured by exciting the samples with pulses centered at 400 nm derived from a 40 kHz regeneratively amplified Ti: sapphire laser system with a pulse width of 150 fs and an instrument response of 15 ps. When the data for this paper were taken, a replica pulse at a delay of about 50 ps resulted in an apparent broadening of the instrument response near zero delay. This can be seen in the rounded appearance in the signal peak in Figure 4 but does not affect the decay times extracted from the data. The sample was in a 1 cm path length quartz cuvette, and the emission was collected at 90° relative to the excitation. The emission was directed into a monochromator attached to a picosecond streak camera (Hamamatsu C4334 Streakscope), which provides both time- and wavelength-resolved fluorescence data, with resolutions of 15 ps and 2.5 nm, respectively.
Experimental Section
Results and Discussion
Synthesis of MHC. The zwitterionic Meisenheimer complex was synthesized and purified according to a previously reported method.20 The solid sample is indefinitely stable at room temperature. Synthesis of MHC Nanoparticles. Ultrapure water (20.0 mL) from a Milli-Q filtration system (Millipore) was added to a 25 mL vial and chilled to 5 °C in a beaker of ice-water slush. A solution of MHC (4.3 × 10-3 M) in tetrahydrofuran (THF; Aldrich, Optima grade), was prepared and stored in the dark. A titanium probe sonicator (horn diameter 0.5 cm) was inserted (1 in. from the bottom) into the chilled vial of water. The sonicator (S-250/450A Analogue sonifier with a 400 W output) was set at an amplitude of 4 with an 80% duty cycle. While the water was being sonicated, 47 µL of the MHC solution was rapidly injected, giving an effective concentration of MHC in water of 1 × 10-5 M. The solution is then sonicated for a total of 12 min, making sure the temperature of the water/MHC nanoparticles does not rise above 10 °C. This is accomplished by sonicating in 4 min intervals, then pausing and letting the solution cool to 5 °C before
Figure 1 shows the normalized absorption and fluorescence spectra of monomeric MHC in room-temperature THF. The molecular structure is shown in the inset. The absorption has two peaks at 407 and 528 nm (typical of a σ complex ground-state absorption spectrum) and shows a pronounced vibronic shoulder at 504 nm whose height is about 70% of the peak at 528 nm. The MHC studied in this work is unique in that it is not a salt but a zwitterion, which is soluble in most organic solvents but not in water. Its fluorescence, peaked at 558 nm in THF, has a slight charge-transfer character but originates from the same state as the absorption, as surmised from its small Stokes shift (1.0 × 103 cm-1) and the fact that it mirrors the absorption line shape. The fluorescence quantum yield Φf ) 0.50 is also significantly greater than that of most Meisenheimer complex salts, which are typically 10% or less.25,26 The fluorescence decay in THF is singleexponential with a lifetime of 9.7 ns, very similar to the
(20) Al-Kaysi, R. O.; Guirado, G.; Valente, E. J. Eur. J. Org. Chem. 2004, 3408-3411. (21) Terrier, F. Nucleophilic Aromatic Displacement; VCH: New York, 1991. (22) Al-Kaysi, R. O.; Creed, D.; Valente, E. J. J. Chem. Crystallogr. 2004, 34, 685-692.
(23) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of Photochemistry, 2nd ed.; Marcel Dekker: New York, 1993. (24) Melhuish, W. H. J. Phys. Chem. 1961, 65, 229-235. (25) Farnham, S.; Taylor, R. J. Org. Chem. 1974, 39, 2446-2448. (26) Hiratsuka, T. Eur. J. Biochem. 2003, 270, 3479-3485.
7992
Langmuir, Vol. 21, No. 17, 2005
Al-Kaysi et al.
Figure 2. (a) Normalized absorption spectrum of the MHC reprecipitated in stirred water (s) and normalized spectrum of a film of MHC spin cast from a THF solution (---). (b) Normalized absorption spectra of MHC (47 µL of 4.3 × 10-3 M THF solution) injected in water and sonicated for 1 min (upper), 12 min (3 × 4 min intervals) (middle), and then passed through a 0.8 µm filter (lower). As the average particle size becomes smaller, the spectrum narrows and the low energy wing decreases. (c) Normalized absorption spectrum for a sample of MHC that undergoes uninterrupted sonication for 20 min, reaching a final temperature of 40 °C. The elevated baseline is the result of increased light scattering.
value of 9.1 ns previously obtained in CH2Cl2.20 In short, the MHC shown in the inset of Figure 1 combines the relatively simple photophysics of an organic dye with a highly polarized zwitterionic structure. When a concentrated MHC/THF solution is injected into highly pure H2O, which is stirred but not otherwise perturbed, the MHC does not crystallize. Figure 2a shows the absorption spectrum of such a solution, which looks qualitatively similar to that of the monomer in THF. The main differences are a slight red shift (2 nm) and broadening of the line shape. If nothing further is done to this sample, within 2 h the MHC falls out of solution and coats the sides and bottom of the sample flask with a thin red layer. The absorption line shape in Figure 2a does not change at all but just diminishes with time. The same spectrum is also seen when the MHC is spin-cast on a surface directly from THF, as shown in Figure 2a, which indicates that this spectrum is characteristic of amorphous MHC and basically reflects the monomer in a highly disordered environment. Although the similar spectra are suggestive, we cannot say with certainty if the unsonicated MHC in water exists in amorphous aggregates or in a monomeric form. If, on the other hand, the MHC solution is injected into a water solution that is sonicated, within 1 min, crystallites are formed, as judged by the dramatic changes in the absorption line shape. Relative to the monomer spectrum, the new absorption spectrum is narrower and red-shifts to 548 nm, with a decrease in the intensity of the high-energy vibronic shoulder from 70% to 35% of the absorption peak. Also, there is a pronounced low-energy wing that extends from 550 nm out past 700 nm. This wing is the result of scattering due to large (on the order of a wavelength of light) particles in the sample. This conclusion is supported by the fact that this wing is not observed in the fluorescence
excitation spectrum, and the observation of large particles in AFM measurements on this sample. If sonication is continued for another set of three 4-min periods with 2-min cooling intervals, the low-energy wing decreases. Passing the sample through a 0.8 µm filter results in an additional decrease in the low-energy wing, as well as a 30% reduction of the peak absorption. The normalized absorption of the filtered sample is the lowest curve shown in Figure 2b. Passing the previously filtered sample through a 0.2 µm filter reduces the peak absorption of the sample without affecting the overall shape of the spectrum. The 0.8 µm filtered solution is quite stable. After an initial drop of about 9% in OD at λmax during the first 15 h, the peak absorbance remains constant for a period of weeks while stored at constant room temperature. This is in contrast to aqueous solutions of tetracene nanoparticles, whose absorption decreases by 50% within about 5 days after formation.7 This increased stability is consistent with previous observations of surfaces and nanoparticles that have been stabilized with zwitterionic ligands27,28 and is also consistent with the empirical observation that molecules which can acquire a partial charge, for example, that contain a carbonyl or nitro group, can form stable, relatively monodisperse solutions. The formation of a negative ζ charge on the surface layer of these particles has been postulated to be responsible for this phenomenon but has not been conclusively proven.12 The duration of sonication is critical in the formation of this type of sample. If sonication is continued without interruption for longer periods of time, larger particles are re-formed, as shown in Figure 2c, where there is now (27) Tatumi, R.; Fujihara, H. Chem. Commun. 2005, 83-85. (28) Kitano, H.; Kawaski, A.; Kawaski, H.; Morokoshi, S. J. Colloid Interface Sci. 2005, 282, 340-348.
Sonication of Zwitterionic Organic Nanoparticles
Langmuir, Vol. 21, No. 17, 2005 7993
Figure 3. (a) AFM image of the 12 min (3 × 4 min intervals) sonicated sample after passage through a 0.8 µm filter. This sample corresponds to the lowest spectrum in Figure 2b. (b) AFM image of MHC nanoparticles collected after 20 min of uninterrupted sonication. This sample corresponds to the spectrum in Figure 2c. Note the different scale.
a very pronounced scattering background. This reaggregation of the particles does not seem to be due to the sonication itself but is instead an artifact of the heating of the solution, which induces reaggregation of the particles. If the 12-min sonication period is broken up into 4-min intervals followed by 2-min cooling periods, the temperature of the solution never rises above 10 °C, and the absorption of the middle curve in Figure 2b is obtained. On the other hand, 20 min of uninterrupted sonication leads to a final solution temperature of 40 °C and the absorption spectrum shown in Figure 2c. Higher temperatures have been observed to lead to larger crystallites in other systems,11,15 most likely due to more rapid aggregation kinetics. We have also seen changes indicative of larger crystallites when we heat a solution of small nanocrystals up to 40 °C for a period of 10 min or more. The key observation is that solution heating means that there is an optimum duration for sonication to initiate growth and refine the particle size. Ideally, we would be able to control the temperature of the solution at the point of sonication, for example, by using a rapidly recirculating sample cell as opposed to an external ice bath. Such a device was unavailable during these experiments. From the absorption spectra in Figure 2, we surmise that the formation of MHC nanoparticles in water proceeds in the following steps: (1) supersaturation of the MHC monomer in water, (2) sonication-induced nucleation and crystal growth, (3) sonication milling of the larger particles into smaller, more uniform nanocrystals, and (4) reaggregation of the nanocrystals as the temperature of the sonicated solution rises. Further support for this sequence of events is given by AFM data taken at stages 3 and 4 in the particle development. After 12 min of intermittent sonication and subsequent filtration, the sample consists almost entirely of the disk-shaped nanocrystals shown in Figure 3a. Their average dimensions, about 140 nm in diameter and 70 nm in height, are smaller than visible light wavelengths, explaining the reduced scattering seen in the absorption spectrum in Figure 2b. As the period of sonication is increased, large micrometer-sized aggregates reappear, as shown in Figure 3b. In this scenario, sonication fulfills several functions. First, at the molecular scale, it seems to provide a uniform start for the nucleation process. Sonication is not necessary for nanocrystal formation, which can also be initiated in solutions of lower purity water or water that has been seeded with nanocrystals from a sonication preparation. But these particles are typically larger, with a very broad size distribution,
Figure 4. Normalized fluorescence decay of the MHC nanoparticle sample from Figures 2b (lowest curve) and 3a. The signal is integrated over all wavelengths and plotted on a logarithmic scale to show both the fast (135 ps) and slow (1.2 ns) decay components.
when compared with the sonicated preparation. Second, at the nanometer scale, it appears to prevent the formation of very large particles, possibly through a milling type of process or due to small-scale turbulence in the solution. Similar experiments in our lab on sonication of reprecipitated anthracene showed little or no effect of reprecipitation on either nucleation or particle size, so the success of this method will likely depend on the chemical nature of the molecule used. The ability to form size-controlled nanoparticles of the MHC permits the study of their photophysical properties. In addition to the steady-state absorption spectra in Figure 2, we have also measured the steady-state and timeresolved fluorescence of these nanocrystals in water at room temperature. The fluorescence quantum yield of the nanocrystals is 0.04, more than an order of magnitude less than that of the monomer in THF. The reason for this decrease is the very short lifetime of the nanocrystal emission, as shown in Figure 4. The bulk of the biexponential fluorescence decay (92%) occurs with a time constant of 135 ps, with a smaller, long-lived component that decays with a time constant of 1.2 ns. We examine the spectra of the short- and long-lived components in Figure 5. The fluorescence within the first 1 ns, shown in Figure 5a, is dominated by a sharp spectrum that looks very similar to the absorption and has a small Stokes shift of 430 cm-1. After this initial decay, which accounts for 92% of the integrated fluorescence, a broader, redshifted spectrum is observed, which is very similar to that of MHC injected into water without sonication. The measured decay times of 1.3 ns for the unsonicated MHC in water and 1.2 ns for the long-lived component in the
7994
Langmuir, Vol. 21, No. 17, 2005
Al-Kaysi et al.
Figure 5. (a) Normalized fluorescence MHC nanocrystals in water collected during the interval 0.0-0.8 ns after excitation, showing the shape of the short-lived component that accounts for 92% of the total fluorescence. The integrated fluorescence is very similar to this spectrum. (b) Normalized fluorescence of the long-lived component, collected during the interval 0.8-1.8 ns after excitation (s), compared with the fluorescence spectrum of amorphous MHC in water (---).
nanoparticle sample are very similar. Figure 5b shows these two spectra normalized and overlapped. These data show how sonication does not completely remove the noncrystalline fraction of the MHC in water, which may be either associated with the nanocrystals or suspended independently in solution. It is expected that both reprecipitation and sonication can lead to nonequilibrium structures that might act as defects in these particles. The time-resolved fluorescence data did not indicate any energy transfer between the short-lived nanocrystal peak and the longer-lived amorphous peak, however, so there is no direct spectroscopic evidence that the noncrystalline MHC acts as a low-energy trap for the initially excited exciton. The final question concerns the origin of the new absorption and emission line shapes. The combination of shift to lower energy, sharpening, and the decrease in the strength of the higher energy vibronic peaks are all qualitatively consistent with J-aggregate formation, as has been surmised for other organic nanocrystal systems.29 The complexity of the MHC crystal structure, along with its complex molecular geometry, precludes easy prediction of whether the transition dipole orientations in the crystal favor J- or H-aggregates.30 Most likely, it forms a J-like aggregate, as in the case of tetracene,31 where there is an enhanced oscillator strength at the bottom of the exciton band, but that enhancement is not necessarily linearly proportional to the number of monomers contained in the aggregate, as would be the case in an ideal J-aggregate. Nevertheless, we can show that the new states seen in the nanocrystals do possess the enhanced radiative decay rate (superradiance) characteristic of J-like aggregates. If we consider the fraction of the total quantum yield originating from the short-lived component, we can calculate the radiative rate for this component according to
krad )
(fraction of fluorescence) × (quantum yield) (fluorescence decay time) (1)
With a fluorescence decay time of 135 ps, a quantum yield of 0.04, and a fraction of fluorescence of 0.92, we obtain krad ) 2.7 × 108 s-1. For the monomer, we use the usual relationship (i.e., the fraction of the fluorescence from the state of interest is assumed to be 1.0), and using a (29) An, B.-K.; Kwon, S.-K.; Jung, S.-D.; Park, S. Y. J. Am. Chem. Soc. 2004, 124, 14410-14415. (30) Kasha, M.; Rawls, H. R.; El-Bayoumi, M. A. J. Pure. Appl. Chem. 1965, 11, 371-392. (31) Lim, S.-H.; Bjorklund, T. G.; Spano, F. C.; Bardeen, C. J. Phys. Rev. Lett. 2004, 92, 107402/1-4.
fluorescence decay time of 9.7 ns and a quantum yield of 0.50, we obtain krad ) 5.2 × 107 s-1. Thus the nanocrystals have a radiative decay rate that is enhanced by roughly a factor of 5 relative to the monomer. As is often the case in solid-state systems, this enhanced radiative decay rate is more than compensated by enhanced nonradiative decay channels that are present in the nanoparticles. Thus the total quantum yield decreases, despite the increase in krad. Nevertheless, the enhancement in krad, along with the other spectral changes observed as the MHC forms nanocrystals, provides clear evidence for the formation of a delocalized excitonic state within the nanoparticles. Conclusion We have shown how reprecipitation in conjunction with sonication provides an efficient way to synthesize organic nanoparticles. By controlling the duration of sonication, we were able to control both the size distribution and crystallinity of the particles. We hypothesize that this control, along with the enhanced stability of the MHC nanoparticles, results from the unique surface charge properties of the zwitterion, which slow the reaggregation process. These nanoparticles exhibit a red shift, a sharpening of the spectrum, and a decrease in vibronic line intensity, all consistent with the formation of a superradiant J-type aggregate. The enhanced radiative decay rate (krad), as compared to that of the monomer, confirms this. Although this work is preliminary, it clearly justifies the further investigation of highly charged organic compounds for use in producing stable organic colloids with interesting electronic properties. While the generality of the sonication method is an open question, we hope that by controlling the sonication conditions more precisely, especially the solution temperature and the ultrasound intensity, we can achieve more systematic control over the nanoparticle size distribution. The extension of this work to different types of zwitterionic compounds with different photophysical and photochemical properties should yield more information about the role of surface charges and crystal structure on nanoparticle size and stability. Acknowledgment. This material is based upon work supported by the National Science Foundation under Grant 0517095. We gratefully acknowledge the use of the sonicator in Professor Q. Cheng’s lab at UCR. We also acknowledge Professor Edward J. Valente of Mississippi College for providing the X-ray structure of the MHC. LA051183B
