1810
Langmuir 2006, 22, 1810-1816
Synthesis of Quantum Dot-Tagged Submicrometer Polystyrene Particles by Miniemulsion Polymerization Nancy Joumaa, Muriel Lansalot,* Alain The´retz, and Abdelhamid Elaissari UMR 2714 CNRS-bioMe´ rieux “Syste` mes Macromole´ culaires et Physiopathologie Humaine” - ENS Lyon 46, alle´ e d’Italie - 69364 Lyon Cedex 07, France
Alyona Sukhanova, Mikhail Artemyev, Igor Nabiev, and Jacques H. M. Cohen EA no. 3798 “De´ tection et Approches The´ rapeutiques Nanotechnologiques dans les Me´ canismes Biologiques de De´ fense”, IFR 53, UniVersite´ de Reims Champagne-Ardenne - 51, rue Cognacq Jay 51100 Reims, France ReceiVed August 12, 2005. In Final Form: NoVember 22, 2005 Submicrometer fluorescent polystyrene (PS) particles have been synthesized via miniemulsion polymerization using CdSe/ZnS core-shell quantum dots (QDs). The influence of QD concentration, QD coating (either trioctylphosphine oxide (TOPO)-coated or vinyl-functionalized), and surfactant concentration on the polymerization kinetics and the photoluminescence properties of the prepared particles has been analyzed. Polymerization kinetics were not altered by the presence of QDs, whatever their surface coating. Latexes exhibited particle sizes ranging from 100 to 350 nm, depending on surfactant concentration, and a narrow particle size distribution was obtained in all cases. The fluorescence signal of the particles increased with the number of incorporated TOPO-coated QDs. The slight red shift of the emission maximum was correlated with phase separation between PS and QDs, which occurred during the polymerization, locating the QDs in the vicinity of the particle/water interface. QD-tagged particles displayed higher fluorescence intensity with TOPO-coated QDs compared to those with the vinyl moiety. The obtained fluorescent particles open up new opportunities for a variety of applications in biotechnology.
Introduction Submicrometer fluorescent colloidal particles attract considerable attention in medical and biotechnological applications,1 where their use as fluorescent probes for diagnosis, imaging, and optical tracking makes them very valuable tools. Such particles are already commercially available, but their efficiency is limited by their large diameter and/or size heterogeneity. Moreover, traditional fluorescent latexes tagged with organic dyes suffer from photobleaching and technical difficulties with their simultaneous excitation for multiple target labeling. The production of fluorescent submicrometer polymer particles exhibiting a narrow particle size distribution, improved photostability and which would be adapted for multiplexing experiments is thus very challenging. Recently, luminescent semiconductor quantum dots (QDs) have attracted the interest of many research groups as fluorescent probes because of their unique properties.2 Regardless of the excitation wavelength, a narrow emission peak is observed, the emission wavelength being size-dependent (around 2-10 nm). Furthermore, QDs show excellent photostability, and different-sized QDs can be simultaneously excited by a single wavelength, because each type of QD exhibits a specific emission peak. The presence of an optically inactive shell preserves the optically active core from fluorescence quenching. Some of the most studied and documented QDs are composed of CdSe only, or of a CdSe core passivated with a ZnS shell. A few attempts have already been made to incorporate these outstanding fluorescent labels onto the surface or inside colloidal polymer particles. Fluorescent colloids have thus been obtained * Corresponding author. E-mail:
[email protected]. (1) Arshady, R. In Preparation and Chemical Applications; Arshady, R., Ed.; MML Series Citus Books: London, 1999; Vol. 2, pp 116-118. (2) Murphy, C. J. Anal. Chem. 2002, 520A-526A.
(1) by the swelling of polymer beads with an organic solvent in the presence of QDs,3-8 (2) via the layer-by-layer (LBL) deposition technique,9-11 controlled precipitation12 or controlled precipitation combined with LBL assembly,13-16 (3) heterocoagulation via electrostatic-driven interactions17 or thiol bonding,18,19 or (4) by entrapment of QDs inside microgels via pH modification.20 However, these different strategies often require multiple steps, notably the preliminary synthesis of a colloidal (3) Han, M.; Gao, X.; Su, J. Z.; Nie, S. Nat. Biotechnol. 2001, 19, 631-635. (4) Xu, H.; Sha, M. Y.; Wong, E. Y.; Uphoff, J.; Xu, Y.; Treadway, J. A.; Truong, A.; O’Brien, E.; Asquith, S.; Stubbins, M.; Spurr, N. K.; Lai, E. H.; Mahoney, W. Nucleic Acids Res. 2003, 31, 43-52. (5) Mulvaney, S. P.; Mattoussi, H. M.; Whitman, L. J. BioTechniques 2004, 36, 602-609. (6) Stsiapura, V.; Sukhanova, A.; Artemyev, M.; Pluot, M.; Cohen, J. H. M.; Baranov, A. V.; Oleinikov, V.; Nabiev, I. Anal. Biochem. 2004, 334, 257-265. (7) Gao, X.; Nie, S. Anal. Chem. 2004, 76, 2406-2410. (8) Bradley, M.; Bruno, N.; Vincent, B. Langmuir 2005, 21, 2750-2753. (9) Susha, A. S.; Caruso, F.; Rogach, A. L.; Sukhorukov, G. B.; Kornowski, A.; Mo¨hwald, H.; Giersig, M.; Eychmu¨ller, A.; Weller, H. Colloids. Surf., A 2000, 163, 39-44. (10) Rogach, A. L.; Susha, A. S.; Caruso, F.; Sukhorukov, G. B.; Kornowski, A.; Kershaw, S.; Mo¨hwald, H.; Eychmu¨ller, A.; Weller, H. AdV. Mater. 2000, 12, 333-337. (11) Wang, D.; Rogach, A. L.; Caruso, F. Nano Lett. 2002, 2, 857-861. (12) Radtchenko, I. L.; Sukhorukov, G. B.; Gaponik, N.; Kornowski, A.; Rogach, A. L.; Mo¨hwald, H. AdV. Mater. 2001, 13, 1684-1687. (13) Gaponik, N.; Radtchenko, I. L.; Sukhorukov, G. B.; Weller, H.; Rogach, A. L. AdV. Mater. 2002, 14, 879-882. (14) Gaponik, N.; Radtchenko, I. L.; Gerstenberger, M. R.; Fedutik, Y. A.; Sukhorukov, G. B.; Rogach, A. L. Nano Lett. 2003, 3, 369-372. (15) Gaponik, N.; Radtchenko, I. L.; Sukhorukov, G. B.; Rogach, A. L. Langmuir 2004, 20, 1449-1452. (16) Zebli, B.; Susha, A. S.; Sukhorukov, G. B.; Rogach, A. L.; Parak, W. J. Langmuir 2005, 21, 4262-4265. (17) Sherman, R. L., Jr.; Ford, W. T. Langmuir 2005, 21, 5218-5222. (18) Hirai, T.; Saito, T.; Komasawa, I. J. Phys. Chem. B 2000, 104, 1163911643. (19) Hirai, T.; Saito, T.; Komasawa, I. J. Phys. Chem. B 2001, 105, 97119714. (20) Kuang, M.; Wang, D.; Bao, H.; Gao, M.; Mo¨hwald, H.; Jiang, M. AdV. Mater. 2005, 17, 267-270.
10.1021/la052197k CCC: $33.50 © 2006 American Chemical Society Published on Web 01/11/2006
Synthesis of Quantum Dot-Tagged Polystyrene
support. To overcome this issue, some research groups have exploited in situ QD synthesis either inside preformed microgels particles21 or onto the surface of polymer particles.22,23 These techniques, however, do not allow the formation of high-quality QDs, which has a negative impact on the fluorescence properties of the resulting particles. Efforts have also focused on the incorporation of QDs into polymeric particles via polymerization in dispersed media. In one of the reported works, styrene suspension polymerization was carried out in the presence of CdSe QDs coated with polymerizable ligands, leading to 100 µm fluorescent beads.24 In the same way, trioctylphosphine oxide (TOPO)-coated CdSe/ ZnS QDs were successfully incorporated into the outer shell of micrometric polystyrene (PS) particles.8 Although really interesting for specific applications, these two examples deal only with micrometer-sized particles. Submicrometer particles were recently obtained by emulsion polymerization with effective incorporation of hydrophobic TOPO-coated CdSe into carboxylic functionalized and crosslinked PS particles, which can be subsequently coated with silica.25 Particle sizes ranging from 300 nm to 20 µm were obtained, depending on the recipe used for the synthesis. Nevertheless, the synthetic methodology used was quite complex and timeconsuming (24 h). In addition, the incorporation of hydrophobic QDs into the particles via a proper emulsion polymerization may be very challenging owing to the poor aqueous-phase transport of these species. Successful incorporation of hydrophilic poly(cysteine acrylamide)-stabilized QDs into 80 to 200 nm fluorescent latexes was achieved via emulsion polymerization, as reported by Sherman et al.17 using two different procedures. In the first one, a two-step shot growth emulsion polymerization of styrene and sodium 4-styrenesulfonate was performed in the presence of a solution of hydrophilic poly(cysteine acrylamide)-stabilized CdS or CdSe/ CdS QDs. In the second approach, the same QDs were first electrostatically modified by vinylbenzyl(trimethyl)-ammonium chloride and subsequently copolymerized with styrene. The resulting particles were effectively fluorescent, but less than particles labeled with TOPO-coated QDs. More recently, two groups took advantage of miniemulsion polymerization, which allows the direct encapsulation of hydrophobic species into polymeric particles. Fleischhaker et al.28 have incorporated CdSe-core/(CdS-ZnS)-shell QDs into PS particles. A poly(methyl methacrylate) shell was then formed by shot-growth emulsion polymerization. The particles obtained were indeed fluorescent, but no detailed information on experimental parameters affecting the first miniemulsion step of the process was provided. Esteves et al.26 showed effective encapsulation of (CdS or CdSe)-TOPO-coated QDs into either PS or poly(butyl acrylate) particles via miniemulsion. Homogeneous polymer nanocomposite particles were obtained, and the integrity of the nanocrystals and their photoluminescence (PL) properties were maintained. Nevertheless, two particle size (21) Zhang, J.; Xu, S.; Kumacheva, E. J. Am. Chem. Soc. 2004, 126, 79087914. (22) Zhang, J.; Coombs, N.; Kumacheva, E.; Lin, Y.; Sargent, E. H. AdV. Mater. 2002, 14, 1756-1759. (23) Zhang, J.; Coombs, N.; Kumacheva, E. J. Am. Chem. Soc. 2002, 124, 14512-14513. (24) O’Brien, P.; Cummins, S. S.; Darcy, D.; Dearden, A.; Masala, O.; Pickett, N. L.; Ryley, S.; Sutherland, A. J. Chem. Commun. 2003, 2532-2533. (25) Yang, X.; Zhang, Y. Langmuir 2004, 20, 6071-6073. (26) Esteves, A. C. C.; Barros-Timmons, A.; Monteiro, T.; Trindade, T. J. Nanosci. Nanotechnol. 2005, 5, 766-771. (27) Esteves, A. C. C.; Bombalski, L.; Cusick, B.; Barros-Timmons, A.; Matyjaszewski, K.; Trindade, T. Polym. Prepr. 2005, 46, 134-135. (28) Fleischhaker, F.; Zentel, R. Chem. Mater. 2005, 17, 1346-1351.
Langmuir, Vol. 22, No. 4, 2006 1811
distributions were observed, with the lower one in size being nontagged by QDs, which implied necessary successive centrifugations to extract the QD-tagged particles. Moreover, there was no data on QD concentration. The same group then applied controlled radical polymerization to miniemulsion using recently reported AGET (activators generated by electron transfer) ATRP (Atom Transfer Radical Polymerization).27 CdS QDs functionalized with a trialkylphosphine modified with an ATRP initiator (i.e., 2-chloropropionate) were successfully used in butyl acrylate miniemulsion polymerization. Complete conversion was attained in 24 h, and 100 nm particles with good control over polymers chain length and polydispersity were obtained with a low proportion of aggregated particles observed. (No data on particle size distribution was presented.) Neither the initial QD concentration nor the PL signals before and after incorporation of the QDs were displayed, but the integrity of the QDs seemed to be preserved. With the aim of synthesizing submicrometer and monodisperse fluorescent particles, we have carried out an extensive study of the incorporation of hydrophobic QDs into polystyrene particles via polymerization in dispersed media. Styrene emulsion and miniemulsion polymerizations have been performed in the presence of either TOPO-coated or vinyl-functionalized CdSe/ ZnS QDs. Both emulsion and miniemulsion processes were first applied to the incorporation of TOPO-coated QDs. Then, the concentration and type of QDs as well as the surfactant concentration have been varied in order to investigate the influence of these parameters on the miniemulsion polymerization kinetics and PL properties of the final particles. The obtained latexes have been characterized with respect to conversion, solid content, particle size, particle size distribution, number of incorporated QDs, and fluorescence properties. Experimental Section Materials. Water was deionized before use (Millipore Milli-Q purification system). Styrene (St, 99%, Aldrich) was distilled under vacuum before use. Potassium persulfate (KPS, >98%, Prolabo), sodium dodecyl sulfate (SDS, >99%, Fluka), hexadecane (HD, 99%, Aldrich), sodium hydrogencarbonate (NaHCO3, >99.5%, Merck), hydroquinone (99%, Janssen), trioctylphosphine oxide (TOPO, Aldrich), hexadecylamine (Fluka), hexamethyldisilthiane (Fluka), trioctylphosphine (Fluka), diethylzinc (Strem), and dimethylcadmium (97%, Strem) were used without further purification. Quantum Dot Synthesis. Batch 1 of TOPO-coated CdSe/ZnS QDs with moderate quantum yield (QY) was purchased from Evident Technologies. Another batch of highly luminescent CdSe/ZnS coreshell nanocrystals of ca. 3.5-4 nm diameter was synthesized according to a published procedure29 (batch 2). The nanocrystal surface was modified with 4-mercaptovinylbenzene (batches 3 and 4). Briefly, 1 µmole of 4-mercaptovinylbenzene in diethyl ether was added to a colloidal solution of nanocrystals in benzene (ca. 10 nmole in 1 mL). The solution was stirred for 30 min at room temperature, and the modified nanocrystals were recovered by evaporation of the solvent. The excess mercaptovinylbenzene was removed by washing with methanol and centrifugation. The precipitate was redissolved in 2-3 mL of benzene, followed by the addition of 5-10 mg of 1-hexadecanol as a nanocrystal surface protector, and the benzene was then allowed to evaporate. Prior to further preparation of the fluorescent latex, the hexadecanol was removed by washing with methanol. The characteristics of the different QDs used in this study are reported in Table 1. Emulsion and Miniemulsion Polymerization Procedures. Batch emulsion and miniemulsion polymerizations of styrene were performed at 75 °C in a conventional thermostated 100 mL reactor (29) Sukhanova, A.; Venteo, L.; Devy, J.; Artemyev, M.; Oleinikov, V.; Pluot, M.; Nabiev, I. Lab. InVest. 2002, 82, 1259-1261.
1812 Langmuir, Vol. 22, No. 4, 2006
Joumaa et al.
Table 1. Characteristics of the QDs Used in This Study batch
type
diameter (nm)
1a 2 3 4
TOPO TOPO vinyl vinyl
4.2 4 4.5 4
λem (nm)
fwhmb (nm)
quantum yield (%)
583 586 602 582
32 29 30 35
∼25 ∼60 ∼20 ∼20
a Specifications from Evident Technologies, Nanomaterials Catalog v5.0, August 2004. b fwhm: full-width half-maximum.
equipped with a condenser and a nitrogen inlet. In the emulsion process, water, SDS, and NaHCO3 were first introduced into the reactor. The mixture of styrene and QDs was then added under stirring. The reactor content was deoxygenated by purging with nitrogen for 45 min while increasing the temperature to 75 °C. The addition of KPS corresponded to zero time of the polymerization process. Miniemulsion polymerizations followed a different procedure. Styrene was first mixed with hexadecane and QD nanocrystals (when used). This organic phase was then added to the aqueous phase (water, SDS, and NaHCO3) under vigorous stirring. After 10 min, the resulting mixture was ultrasonicated (HD 2200 Bandelin Sonoplus ultrasonic homogenizer, flat tip TT13, amplitude 30%, HF-generator power: 200 W) for 7 min. The obtained stable miniemulsion was then transferred to the reactor and was deoxygenated by purging with nitrogen for 45 min while the temperature was raised to 75 °C. Finally, the addition of KPS dissolved in water gave the zero time of the polymerization. The experimental conditions of all of the polymerizations performed in this study are displayed in Table 2. Characterization of the Latexes. Monomer consumption was determined by the gravimetric analysis of samples withdrawn from the polymerization medium (and quenched with hydroquinone) at different times. The particle size and particle size distribution were obtained by dynamic light scattering (DLS) (Zetasizer 3000HS, from Malvern Instruments). The particle morphologies and particle size distribution were further examined by transmission electron microscopy at an accelerating voltage of either 80 or 120 kV (Philips CM 120, Centre Technologique des Microstructures (CTµ), Claude Bernard University, Lyon, France and Philips CM 200-Cryo CERMAV, Joseph Fourier University, Grenoble, France). To prepare ultrathin sections of the particles (ca. 90 nm), the latexes were first dehydrated by successive washing in an ethanol bath. The particles were then embedded in an Epon resin that was cut with an ultramicrotome Leica Ultracut E equipped with a diamond knife. The composition of some samples was determined by elemental analysis (SCA CNRS, Solaize, France). Fluorescence Measurements. Emission spectra of TOPO-coated QDs, vinyl-functionalized QDs, and QD-tagged PS particles were recorded using a fluorescence spectrophotometer (LS 50 system, Perkin-Elmer). Particles (either raw or washed) were dried under
vacuum and then dissolved in chloroform (0.15 g/2 mL). When used, the washing procedure was as follows: An aliquot of the latex was placed in a centrifuge Eppendorf tube equipped with a 0.1 µm filter (commercial device from Millipore). The particles were subsequently separated from the supernatant by centrifugation. QDs that were not incorporated into the particles should be in the supernatant. Fluorescence imaging was performed with a Zeiss Axioplan 2 Imaging microscope equipped with a camera and a 100× infinitycorrected 1.3 numerical aperture oil objective. For fluorescence observations, a Hg lamp was used in combination with a filter set for fluorescein. A drop of diluted latex was deposited onto a glass slide and allowed to dry before analysis.
Results and Discussion Emulsion Polymerization with TOPO-Coated QDs. As previously mentioned, the aim of the present work was to synthesize fluorescent polystyrene particles by taking advantage of the outstanding properties of QDs. Our first attempt at TOPOcoated CdSe/ZnS QD (batch 1) incorporation into polystyrene particles was performed by batch emulsion polymerization using a conventional synthetic recipe (see latex 1, Table 2). However, as evidenced by the pinkish layer observed on the inner wall of the reactor as well as by the very white latex obtained, QD incorporation was not successful. This was likely due to the very hydrophobic character of the QDs, which prevented efficient transport through the water phase. Indeed, to be effectively located inside the polystyrene particles, the QDs need to be transported from the monomer reservoir droplets to the polymerization site (i.e., the particles). The rate of diffusion through the water phase was either completely prevented or too slow in comparison with the rate of polymerization. These limitations led us to carry out the next experiments by batch miniemulsion polymerization. Miniemulsion Polymerization. Miniemulsion polymerization provides an efficient means of incorporating hydrophobic compounds into polymer particles.30,31 Hence, (1) QDs should be in the reaction loci (i.e., the monomer droplets) from the beginning of the polymerization, avoiding any loss in the water phase, and (2) they should be theoretically homogeneously distributed among the particles so that all of the particles exhibit the same fluorescence properties. During the preparation of this article, a similar approach based on a miniemulsion process was reported by Fleischhaker et al.,28 which confirms the potential of this approach. Influence of Initial QD Concentration. Various initial amounts of TOPO-coated QDs (batch 1) were first used in miniemulsion polymerization (latexes 2-5, Table 2). In contrast with what
Table 2. Detailed Experimental Conditions of All Latexes Synthesized in This Studya expt latex 1 latex 2 latex 3 latex 4 latex 5 latex 6 latex 7 latex 8 latex 9 latex 10 latex 11 latex 12
type of QDs
[SDS]b (mol‚L-1water)
batch 1 none batch 1 batch 1 batch 1 batch 2 batch 3 batch 1 batch 1 batch 4 batch 4 batch 4
3.4 × 1.8 × 10-3 1.8 × 10-3 1.8 × 10-3 1.8 × 10-3 1.8 × 10-3 1.8 × 10-3 1.3 × 10-3 3.4 × 10-3 1.3 × 10-3 3.4 × 10-3 9.1 × 10-3 10-3
[QDs]0 (mol‚L-1St) 4.9 ×
10-6
4.1 × 10-6 1.2 × 10-5 2.2 × 10-5 4.9 × 10-6 4.9 × 10-6 4.1 × 10-6 4.1 × 10-6 4.9 × 10-6 4.9 × 10-6 4.9 × 10-6
t final (h) 3.5 5 5 5 5 4.2 4 6.5 3.5 6.5 2.5 3.5
Dhc (nm)
Np (L-1emulsion)
NQDd
115 259 242 256 255 265 268 320 168 343 178 100
2.40 × 2.14 × 1016 2.59 × 1016 2.20 × 1016 2.20 × 1016 1.94 × 1016 1.88 × 1016 1.09 × 1016 7.40 × 1016 9.10 × 1015 6.56 × 1016 3.70 × 1017
21 74 132 34 34 49 7 71 10 2
1017
a All experiments were carried out with [KPS]0 ) 5.6 × 10-3 mol‚L-1water; [NaHCO3]0 ) 1.1 × 10-2 mol‚L-1water; 20 wt % styrene; hexadecane: 3.5 wt %/styrene (except latex 1); T ) 75 °C. b Critical micelle concentration (cmc) ) 7.6 × 10-3 mol‚L-1. c Obtained by dynamic light scattering. d NQD ) Number of QDs per particle; estimated values taking into account initial number of QDs and experimental values of particle numbers, under the assumption there are no free QDs left in the continuous phase.
Synthesis of Quantum Dot-Tagged Polystyrene
Langmuir, Vol. 22, No. 4, 2006 1813
Figure 1. Photographs of latexes prepared with increasing concentrations of QDs, without and under irradiation (λexc ) 365 nm). See Table 2 for detailed experimental conditions. Figure 3. TEM photos of latex 3 ([QDs]0 ) 4.1 × 10-6 mol/LSt).
Figure 2. Influence of the initial concentration of TOPO-coated QDs (batch 1) on the evolution of monomer conversion for styrene miniemulsion polymerization. See Table 2 for detailed experimental conditions.
was observed in conventional emulsion polymerization, stable and slightly pinkish latexes were obtained, which became fluorescent under UV light (Figure 1). As evidenced by Figure 2, the addition of QDs actually had no marked effect on the polymerization kinetics, and complete conversion was reached in ca. 4 h. Final particle diameters were in each case close to 250 nm, indicating that the initial QD concentration had no significant influence on the final particle size. All of the latexes exhibited a narrow particle size distribution as evidenced by TEM analysis and shown for latex 3 in Figure 3 (polydispersity index ) 1.024). The number-average diameter Dn ) 246 nm calculated from the TEM photos was found to be in good agreement with the hydrodynamic diameter Dh ) 242 nm obtained from DLS measurement. Fluorescence spectrometry was used as a powerful tool for determining the presence of QDs inside the particles. The aqueous supernatant never showed a detectable fluorescence signal. Fluorescence analysis of the washed and dried particles dissolved in chloroform (Experimental Section) followed the expected trend (i.e., the fluorescence signal increased with the number of incorporated QDs, Figure 4). All of the latexes exhibited narrow emission peaks (fwhm ≈ 40 nm) with an emission wavelength of ca. 589 nm, a slightly red-shifted value compared with that of the QDs alone (583 nm), which indicated an alteration of the optical properties of the QDs. This may be correlated with changes induced by the polymerization in the vicinity of the QDs surface. It should be noted that the simple mixing of QDs and polystyrene (30) Asua, J. M. Prog. Polym. Sci. 2002, 27, 1283-1346. (31) Antonietti, M.; Landfester, K. Prog. Polym. Sci. 2002, 27, 689-757.
Figure 4. Fluorescence spectra of washed and dried particles prepared with increasing concentrations of QDs (λexc ) 450 nm). See Table 2 for detailed experimental conditions.
in chloroform did not induce any emission maximum red shift, showing that the presence of individual PS chains had no effect on the fluorescence analysis. In addition, holding the QDs at 75 °C for 3 h in the presence of styrene did not induce either an emission maximum red shift or a broadening of the emission peak, even if a noticeable decay of the fluorescence signal was observed. (A loss of 75% of the initial intensity was measured, and it is postulated that this is the result of the oxidation of the TOPO molecules on the QD’s surface, followed by partial QD aggregation.) To explain the slight red shift of the emission maximum, the latexes were further analyzed by electron microscopy. As shown for latex 5 in Figure 5, QDs were found to be essentially located in the outer shell of the particles. To determine if any QDs were present in the core of the particles, the dehydrated latexes were then embedded in a resin and sliced. The cross sections of latex 5 in Figure 6A confirmed the previous observations and provided clear evidence that phase separation between PS and QDs occurred during the polymerization, driving the QDs toward the particle/ water interface. TOPO-coated QDs and PS may not be compatible with each other, thus inducing phase separation. It was, however, difficult to determine if the phase separation induced (partial) aggregation of the QDs or if the QDs were just excluded from the PS phase without aggregation. In any case, these observations may explain the slight red shift of the emission observed for all of the tagged particles. Taking into account that the ultrathin cut of the embedded samples randomly sliced the particles, the fact that a great number of them contained QDs (Figure 6B) indicated that the majority of the particles were tagged with QDs. Besides,
1814 Langmuir, Vol. 22, No. 4, 2006
Joumaa et al.
Figure 7. Fluorescence imaging for latex 5 ([QDs]0 ) 2.2 × 10-5 mol/LSt).
Figure 5. TEM photos of latex 5 ([QDs]0 ) 2.2 × 10-5 mol/LSt) (diluted and dried sample).
Figure 8. Evolution of fluorescence intensity as a function of [QDs]0 for latexes synthesized with varying concentrations of TOPO-coated QDs (batch 1).
Figure 6. TEM photos of ultrathin sections of latex 5 ([QDs]0 ) 2.2 × 10-5 mol/LSt).
these analyses consisted of the first successful visualization of QDs incorporated into submicrometer particles. Phase separation between QDs and PS has been observed for the bulk32 and the suspension polymerization of styrene in the presence of TOPO-coated CdSe.8 In both cases, aggregates of QDs were effectively present either in the bulk matrix or in the particles, and the resulting product was not fluorescent. The (32) Erskine, L. E.; Emrick, T.; Alivisatos, A. P.; Fre´chet, J. M. J. Polym. Prepr. 2000, 41, 593-594.
main difference between the QDs used in those studies and the ones used in our work is the presence of a ZnS shell, which seems to play a critical role in the final fluorescence properties. This shell may not prevent QDs from phase separation but may at least preserve their optical properties, explaining why our latexes were fluorescent despite the observed phase separation. The fluorescence of the particles was determined by the comparison of the transmitted light signal and fluorescence (Figure 7), which showed that most if not all of the particles were fluorescent. Finally, a further indication of effective incorporation of the QDs was given by elemental analysis: cadmium, selenium, zinc, and phosphorus were found in the samples of washed and dried particles. The number average of QDs per polystyrene particle NQD was found to be between 21 and 132, depending on the initial QD concentration. NQD stands in the range of values recently reported for emulsion polymerization,17 or miniemulsion.28 Moreover, linear evolution of the fluorescence intensity with the number of QDs introduced is observed (Figure 8). Surprisingly, our results did not corroborate the observations reported by Fleischhaker et al.,28 who studied the miniemulsion polymerization of styrene in the presence of different QDs. They observed that the incorporation of CdSe/ZnS QDs into PS particles did not lead to the formation of fluorescent latexes. The authors attributed this quenching effect to an irregular coating of CdSe by ZnS, owing to the abrupt lattice constant change from CdSe to ZnS.28 The quality of the ZnS shell is known to play a crucial role in ensuring the high QD quantum yield and the preservation of the CdSe core fluorescence from quenching.33 The ZnS shell changes its structure from amorphous (irregular) to crystalline as its thickness increases from 0.5 to 3 monolayers. As far as we have shown before, the highest quantum yield and stability of PL of CdSe/ZnS QDs is achieved when the CdSe core is (33) Baranov, A. V.; Rakovich, Yu. P.; Donegan, J. F.; Perova, T. S.; Moore, R. A.; Talapin, D. V.; Rogach, A. L.; Masumoto, Y.; Nabiev, I. Phys. ReV. B 2003, 68, 165306.
Synthesis of Quantum Dot-Tagged Polystyrene
Figure 9. Influence of the type of QDs (either TOPO-coated QDs, latex 6 or vinylfunctionalized QDs, latex 7) on the evolution of monomer conversion. See Table 2 for detailed experimental conditions.
protected with 2.0-2.5 monolayers of the ZnS shell. The more likely explanation accounting for the different results obtained from Fleischhaker’s work and ours is the better coverage of the QDs used in our study (either commercial or self-made). Influence of the Type of QDs. The results in the previous section seem to indicate that TOPO-coated QDs may be subject to phase separation during polymerization. In an attempt to better disperse the QDs throughout the PS phase, vinyl-functionalized QDs (batch 3) were used and compared with TOPO-coated ones (batch 2). The vinyl function was expected to copolymerize with styrene and to allow a homogeneous dispersion of the nanocrystals through the polymer particles, as well as irreversible incorporation. The properties of the obtained latexes (latexes 6 and 7, respectively) are displayed in Table 2. No influence of the type of QDs on either the polymerization kinetics or the final diameter and particle size distribution was observed (Figure 9 and Table 2). However, vinyl-functionalized QDs were intrinsically less fluorescent than TOPO-coated ones (which may be related to the modification of their surface), and the same trend was observed for the final particles: the fluorescence signal was 1.5 times more intense for particles tagged with TOPO-coated QDs. A slight red shift of the emission maximum was noticeable (from 602 to 605 nm). Ultrathin sections of the particles observed by TEM revealed that vinyl-functionalized QDs were located at the interface, as in the case of TOPO-coated QDs. The expected homogeneous dispersion of the QDs throughout the whole particle was never observed. Despite the presence of the vinyl function, polymerization also led to the phase separation of the QDs, which indicates that copolymerization did not occur. This would suggest that (i) the reactivity ratios of styrene and vinyl-functionalized QDs were not in favor of copolymerization and/or (ii) the incomplete coverage of the QD surfaces with vinyl functions. More work is currently in progress to understand these observations. Influence of Surfactant Concentration. To assess the influence of the number of QDs per particle, batch 1 of TOPO-coated QDs was used in miniemulsion polymerization, with varying amounts of SDS (latexes 3, 8, and 9, Table 2) and a constant concentration of QDs (4.1 × 10-6 mol‚L-1St). As expected, increasing SDS concentration induced a higher polymerization rate, which was directly correlated with an increase in the final number of particles Np (Figure 10 and Table 2). Complete conversion was achieved in each case, and all of the latexes were isodispersed in size with a final particle diameter of between 168 and 320 nm. The number of QDs per particle NQD logically decreased when Np increased, and NQD could be correlated with the initial amount of SDS for a given QD concentration (Figure 11A). The same observation
Langmuir, Vol. 22, No. 4, 2006 1815
Figure 10. Influence of the initial concentration of SDS on the evolution of monomer conversion. See Table 2 for detailed experimental conditions.
Figure 11. Evolution of NQD as a function of SDS concentration. (A) [TOPO-coated QDs]0 ) 4.1 × 10-6 mol‚L-1St. (B) [Vinylfunctionalized QDs]0 ) 4.9 × 10-6 mol‚L-1St.
stands true for vinyl-functionalized QDs (Figure 11B, Latexes 7 and 10-12). According to the results presented above, the nature and/or concentration of QDs influences neither the polymerization kinetics nor Np. Assuming this, the evolution of Np as a function of SDS concentration (Log-Log plot, Figure 12) was plotted, taking into account all of the experiments performed in this study, and Np was found to be proportional to [SDS]R ) 1.84. This R value is in the range of those reported (R ) 1.3834) or extracted from literature data (R ) 0.9235 or 2.1936) for styrene (and hexadecane) miniemulsion polymerization. The variation ob(34) Anderson, C. D.; Sudol, E. D.; El-Aasser, M. S. J. Appl. Polym. Sci. 2003, 90, 3987-3993. (35) Landfester, K.; Bechthold, N.; Tiarks, F.; Antonietti, M. Macromolecules 1999, 32, 5222-5228. (36) Hansen, F. K.; Ugelstad, J. J. Polym. Sci., Polym. Chem. Ed. 1979, 17, 3047-3067.
1816 Langmuir, Vol. 22, No. 4, 2006
Figure 12. Evolution of the final number of particles Np as a function of SDS concentration for styrene miniemulsion polymerizations performed in this study (Log-Log plot).
served in the R values may be ascribed to differences in the formulation and/or operating conditions (by means of emulsification and temperature) from one system to another. Smith and Ewart’s theory for conventional emulsion polymerization37 predicts that Np should vary with [SDS]R)0.6 for styrene polymerization. The same R value cannot be expected for miniemulsion polymerization, which is based on different mechanisms. In emulsion polymerization, surfactant is mostly present in the water phase as free molecules or as micelles. These monomer-swollen micelles are the main locus of nucleation, and the large monomer droplets, which offer a small interfacial area, simply act as reservoirs for the growing particles. In contrast, in miniemulsion systems the synthesis recipe, as well as the means of emulsification, provides small monomer droplets exhibiting a large interfacial area. The surfactant is mainly located on the droplet surface, and only a few molecules are in the aqueous phase (concentration below the cmc). Under these conditions, the probability of homogeneous nucleation is reduced, and micellar nucleation is suppressed, leading to effective droplet nucleation.
Conclusions In this article, the synthesis of fluorescent polystyrene particles tagged with luminescent quantum dots was investigated. The (37) Smith, W. V.; Ewart, R. H. J. Chem. Phys. 1948, 16, 592-599.
Joumaa et al.
first experiments performed in emulsion polymerization showed unsuccessful encapsulation of the QDs by polystyrene particles. In contrast, the use of either TOPO-coated or vinyl-functionalized QDs in batch miniemulsion polymerization of styrene resulted in submicrometer and isodispersed fluorescent latexes. The final particle size could be tuned between 100 and 350 nm by varying the initial surfactant (SDS) concentration. Polymerization kinetics were not altered by the presence of QDs, whatever their surface properties. The fluorescence signal increased with the number of incorporated TOPO-coated QDs, and the slight red shift of the emission maximum, induced by the polymerization, was correlated with the modification of the medium surrounding the QDs. All of the results undoubtedly showed effective incorporation of the QDs, whatever the type of surface coating, even though latexes labeled with TOPO-coated QDs displayed higher fluorescence intensity than vinyl-QD labeled ones. In both cases, phase separation between PS and QDs occurred during the polymerization, and QDs were clearly visible at the particle/ water interface. This was essentially ascribed to a strong incompatibility between PS and QDs. In conclusion, the results reported herein open up new opportunities for the development of biofunctionalized submicrometer particles suitable for a variety of applications in biotechnology. Modifications of the surface of the particles for further binding with biomolecules and multiplexed coding are now under investigation.
Acknowledgment. N.J. acknowledges bioMe´rieux S.A. for a Ph.D. fellowship, and A.S. is grateful to FEBS for providing a postdoctoral fellowship. This work was supported in part by bioMe´rieux S.A., INTAS and NATO grants, and Ligue Nationale contre le Cancer. H. Mouaziz is greatly acknowledged for TEM analyses. We also thank J.-L. Putaux (CERMAV, Grenoble) as well as A. Rivoire and C. Boule´ (CTµ, Lyon) for their helpful advice on TEM analyses and C. Place (Laboratoire de PhysiqueENS Lyon, UMR 5672, Lyon) for his help with fluorescence optical microscopy. M.L. gratefully acknowledges C. Pichot and D. Morsley for their relevant comments on this manuscript. LA052197K
