pubs.acs.org/Langmuir © 2009 American Chemical Society
Photochemical Gluing of Colloidal Particles by a Simple Interparticle Polymerization Route D. Marksteiner, S. Wasser, and W. Sch€artl* Institut f€ ur Physikalische Chemie, Welderweg 11, 55099 Mainz, Germany Received July 17, 2009 We have developed a simple strategy to chemically cross-link polystyrene latex spheres by UV-light-induced polymerization after the particles, which themselves were internally cross-linked with divinylbenzene (DVB), have been carefully swollen in aqueous dispersion with a small amount of the monomer styrene. As a proof of concept, aggregates were preformed by salt-induced aggregation, and successful fixation of the aggregate structure was quantitatively confirmed by dynamic light scattering after the sample was diluted with distilled water and treated with ultrasound. Under this treatment, non-UV-irradiated aggregates dispersed and the single particle size was recovered. This concept was successfully applied also to commercial latex particles. For the future, arranging the colloids in a well-defined way by laser traps before photocross-linking opens up a facile method to prepare microscopic devices.
*To whom correspondence should be addressed. E-mail: schaertl@ uni-mainz.de.
dispersion to screen the interparticle Coulomb repulsion;11,12 (ii) adding small polymer chains to a particle dispersion to introduce depletion attraction, an effectively entropic force;13-15 and (iii) using laser traps to assemble a few colloidal particles in a defined way and keeping them in close contact during the photocross-linking.16-18 Here, we use approach (i) to obtain aggregates which are growing so slow that photo-cross-linking by UVirradiation for several hours leads to a stable dispersion of chemically fixed aggregates. Our studies were mainly motivated by looking for a much simpler and also more general concept for photo-cross-linking of colloidal particles compared to our previous approach of surface functionalization: we decided to swell cross-linked polymer particles in aqueous dispersion with a small amount of the respective monomer and then initiate the polymerization of the “glue” by UV-irradiation. As a first example to test this concept, we have chosen polystyrene latex spheres cross-linked with divinylbenzene, and swollen with styrene. The intraparticle cross-linking is necessary, otherwise the particles may lose their spherical shape during swelling with the monomer. The idea is to polymerize the styrene while the swollen particles are touching, thereby forming interpenetrating polymer chains and gluing the particles together. Note that these clusters are not chemically cross-linked, but if the polymer has a glass transition temperature higher than room temperature the interpenetrating chains will become kinetically trapped within the particle network after either all styrene has polymerized or the nonreacted styrene has evaporated. The big advantage of this simple approach is that it can be applied to virtually any type of cross-linked polymer particles, the reaction itself takes place in aqueous solution, and even many commercial particles widely available these days can be directly glued together
(1) Terray, A.; Oakey, J.; Marr, D. W. M. Science 2002, 296, 1841–1844. (2) Lee, S. K.; Kim, S. H.; Kang, J. H.; Park, S. G.; Jung, W. J.; Yi, G. R.; Yang, S. M. Microfluid. Nanofluid. 2008, 4, 129–144. (3) Chen, Y.; Geh, J. L. Polymer 1996, 37, 4481–4486. (4) Chen, Y.; Jean, C. S. J. Appl. Polym. Sci. 1997, 64, 1759–1768. (5) Chen, Y.; Tsai, C. H. J. Appl. Polym. Sci. 1998, 70, 605–611. (6) Yuan, X.; Fischer, K.; Sch€artl, W. Adv. Funct. Mater. 2004, 14, 457. (7) Yuan, X.; Schnell, M.; Muth, S.; Schartl, W. Langmuir 2008, 24, 5299–5305. (8) Meakin, P. Phys. Rev. Lett. 1983, 51, 1119–1122. (9) Sander, L. M.; Cheng, Z. M.; Richter, R. Phys. Rev. B 1983, 28, 6394–6396. (10) Witten, T. A.; Sander, L. M. Phys. Rev. B 1983, 27, 5686–5697. (11) Broide, M. L.; Cohen, R. J. J. Colloid Interface Sci. 1992, 153, 493–508. (12) Adachi, Y. Adv. Colloid Interface Sci. 1995, 56, 1–31.
(13) Poon, W. C. K. J. Phys.: Condens. Matter 2002, 14, R859–R880. (14) Ramakrishnan, S.; Fuchs, M.; Schweizer, K. S.; Zukoski, C. F. J. Chem. Phys. 2002, 116, 2201–2212. (15) Tuinier, R.; Rieger, J.; de Kruif, C. G. Adv. Colloid Interface Sci. 2003, 103, 1–31. (16) Klajn, R.; Bishop, K. J. M.; Grzybowski, B. A. Proc. Natl. Acad. Sci. U.S. A. 2007, 104, 10305–10309. (17) Misawa, H.; Sasaki, K.; Koshioka, M.; Kitamura, N.; Masuhara, H. Appl. Phys. Lett. 1992, 60, 310–312. (18) Misawa, H.; Sasaki, K.; Koshioka, M.; Kitamura, N.; Masuhara, H. Macromolecules 1993, 26, 282–286.
Introduction Colloidal clusters are of imminent technological relevance in food science, paper science, and so on. Also, assemblies of colloidal nano- or microspheres have been used recently as welldefined microdevices in miniaturized setups,1,2 for instance, as microscopic pumps. To explore the structure-property relationship in detail, it is desirable to freeze the clusters during the aggregation process. This of course is also necessary if one wants to permanently fix a microscopic device based on a colloidal aggregate. One possible strategy is to functionalize the surface of the colloidal particles with reactive groups, which then can undergo photo-cross-linking, leading to a chemical coupling between neighboring particles. Whereas this concept has been successfully applied to polymer chains,3-5 to our knowledge, we were the first and, so far, only ones who prepared permanently fixed colloidal particle aggregates by photo-cross-linking of functionalized colloidal particles.6,7 In these previous studies, we still used the random collisions between the diffusing particles to form the aggregates in a diffusion-controlled aggregation mechanism,8-10 which was possible since our reactive groups, nitrocinnamate, were highly efficient and reactive on a very short time scale. If the functional groups are less reactive, however, it is necessary to prearrange the colloidal particles into aggregates before the photofixation, which then may take a few hours. Such preaggregation could be achieved in several ways, and the most common ways are (i) adding salt to charge-stabilized colloidal particles in aqueous
Langmuir 2009, 25(22), 12843–12846
Published on Web 10/26/2009
DOI: 10.1021/la903733b
12843
Letter
without difficult additional chemical modification. We should not forget to mention that Misawa et al. previously had developed a similar approach, the photo-cross-linking of colloidal microparticles within an organic solvent containing a photoinitiator and vinylic monomers.17,18 They aligned their particles with laser traps to well-defined aggregates and finally fixed these superstructures by photo-cross-linking. The major difference compared to our concept is that we swell only the particles themselves with the vinylic monomer and that our cross-linking therefore may take place in an aqueous dispersion in a welldefined way although we are irradiating the whole sample, with the “glue” being present at the particle surface only. In contrast, Misawa et al. had to irradiate exactly the touching particle surfaces with a laser beam otherwise their whole sample would have been polymerized.
Experimental Section A. Particle Synthesis. 1. Polystyrene (PS) Latex Spheres with 1 mol % Cross-Linker Divinylbenzene (DVB). Following Thomson et al.,19,20 80 mL of distilled Milli-Q water and 90 mL of methanol were mixed in a 250 mL flask equipped with a stirrer and reflux condenser and then flushed with nitrogen for 15 min to remove oxygen. Next, a mixture of 28 mL of freshly distilled styrene and 0.321 g of divinylbenzene was added, and the reaction mixture was again flushed with nitrogen for 15 min. Finally, 0.216 g of potassium peroxodisulfate was added, the apparatus was sealed under a nitrogen atmosphere, and the reaction mixture was heated to 80 °C. The polymerization took place for 48 h under rapid stirring. After the reaction mixture cooled down slowly, it was purified first by filtration through ordinary paper filters and then further purified by dialysis versus pure distilled water for 9 days, using dialysis membranes with a molecular weight cutoff of 50 kDa. During dialysis, the outer aqueous phase was replaced by fresh distilled water five times daily. The purified latex dispersion was kept at room temperature.
2. PS Latex Spheres with 2 mol % Cross-Linker DVB. Commercial polystyrene microbeads (from Bangs Laboratories Inc.) were investigated for comparison. The particle size provided by the manufacturer was 4.2 μm, and the solid content of the aqueous dispersion, which also contained a surfactant for stabilization, was 1 wt %. B. Characterization Methods. 1. Gravimetry. To determine the reaction yield and therefore the solid content of the latex dispersions, two drops of the dispersion were put onto a microscope glass slide whose weight was previously measured, and the mass of this sample was determined. All water was evaporated in a vacuum oven at T=60 °C until the total weight remained constant. The difference between the weight of the glass slide and the sample weight before and after drying then yielded the solid content. 2. Dynamic Light Scattering (DLS). Our light scattering setup consisted of a goniometer Sp-125 equipped with an ALV/ SO-SIPD single photon detector, ALV-5000/EPP60X Multitau realtime digital correlator, and argon ion laser (Spectra Physics Stabilite 2060-4S, operating at wavelength 514 nm and laser power 400 mW). The power of the incident laser beam was adjusted with an attenuator. Our sample cells are cylindrical Suprasil glass cuvettes of diameter 1 cm from Hellma. All sample cells were purified from dust by flushing with freshly distilled acetone, and the solvent (water) was filtered with Millex W filters (pore size 0.1 μm, Duropore PVDF membrane, Millipore), using 90 drops of filtered water per drop of original latex particle (19) Thomson, B.; Rudin, A.; Lajoie, G. J. Polym. Sci., Part A: Polym. Chem. 1995, 33, 345–357. (20) Thomson, B.; Rudin, A.; Lajoie, G. J. Appl. Polym. Sci. 1996, 59, 2009– 2028.
12844 DOI: 10.1021/la903733b
Marksteiner et al. Table 1. Particle Characteristics of Cross-Linked Latex Spheres, Determined by DLS and TEM sample
RH (DLS)/nm
R (TEM)/nm
PS-DVB 1% PS-DVB 2%
375 ( 15
420 ( 25 2350 ( 25
dispersion. Note that the particle dispersion itself could not be filtered due to the large size of the latex particles. During the measurement, the sample was kept in a toluene bath at T=20 °C. DLS measurements were conducted at several scattering angles, and particle sizes were either determined, if possible, from the angular dependence of the correlation times by interpolation toward zero scattering vector (= z-average) or given as the apparent particle radius at a given scattering angle.22 3. Optical Microscopy. We used a Zeiss Axioskop instrument equipped with a Colorview 12 camera (Olympus). Samples were prepared by putting directly one to two drops of the dispersion on a microscopy slide and putting a cover glass on top. 4. Transmission Electron Microscopy (TEM). For size determination of single latex particles by TEM, we used an EM 420 (Philips) instrument. The sample dispersion was first diluted 1:100, and then a few drops were put into an ultrasonic spray evaporator, used to spray the sample dispersion onto carbon grids. The grids were dried overnight before the measurements.
C. Photo-Cross-Linking of Colloidal Aggregates. 1. Swelling of PS-DVB Latex Spheres with Styrene. First, the latex particles were swollen with a small amount of freshly distilled styrene. The dispersion was rapidly stirred while styrene was added carefully with an Eppendorf pipet dipping into the dispersion. Small portions of only 3 μL of styrene were added at once, and the sample was stirred for 30 min before the next addition of styrene. After all the styrene was added, the sample cell was sealed and stirred for another 6 h.
2. Preaggregation of Swollen PS-DVB Latex Spheres by Adding Salt. To destabilize the latex dispersion and induce particle aggregation, 2 mL of an aqueous magnesium salt solution of various concentrations was added to 2 mL of the PS-DVB latex particle dispersion. The mixture was rapidly stirred for 2 h, and the size of the aggregates was determined by dynamic light scattering or optical microscopy. To confirm that the salt-induced aggregation was reversible, the sample was diluted with distilled water, treated by 15 min ultrasonication, and characterized by DLS or optical microscopy.
3. UV-Fixation of Swollen and Aggregated PS-DVB Latex Particles. Light scattering cuvettes containing aggregates were irradiated with a 200 W HgXe lamp (Amko, Germany) for 3 h, and then the samples were characterized by DLS or optical microscopy. During irradiation, the sample was put onto a stage slowly rotating with 0.03 turns per second. To prove successful photo-cross-linking of the latex aggregates, the irradiated dispersions were diluted with distilled water and ultrasonicated for 15 min (see also section C.2 above). If DLS optical microscopy showed no change in aggregate size after this treatment, then the chemical cross-linking is considered successful.
Results and Discussion A. Characterization of PS-DVB Latex Spheres. Since the size polydispersity of our homemade PS-DVB1% latex particles was 10-15%, the DLS correlation functions were fitted with a double-exponential decay, with comparable amplitudes for each mode. Average apparent hydrodynamic radii as well as particle (21) Min, K.; Hu, J. H.; Wang, C. C.; Elaissari, A. J. Polym. Sci., Part A: Polym. Chem. 2002, 40, 892–900. (22) Sch€artl, W. Light Scattering from Polymer Solutions and Nanoparticle Dispersions; Springer: Berlin, Heidelberg, 2007.
Langmuir 2009, 25(22), 12843–12846
Marksteiner et al.
Figure 1. Apparent radius of aggregates of PS-DVB1% latex particles at various salt concentrations, determined by DLS at scattering angle 70°. Lines are guides to the eye.
radii determined by TEM (number-average) are summarized in Table 1. Note that the particle diameter of 750 nm is so large that the angular regime qR < 1 (= scattering angle < 10°) could not be accessed experimentally, so in this case the average still contains significant particle form factor contributions and therefore is not a well-defined average but an apparent hydrodynamic radius. However, the size polydispersity of our particles is so small that the angular dependence of the apparent particle size could be ignored with an experimental error of less than 5%. TEM confirms the spherical shape of the particles, but the size seems to be slightly larger than that predicted by DLS. Note that sample dilution did not effect our DLS result, and therefore, particle interactions as a potential reason for this deviation can be ruled out. This phenomenon of a significant deviation of particle sizes determined by optical techniques and TEM has been reported before, but without explanation.2 One possible reason is that the particles still contain some nonpolymerized styrene, which is evaporating in the vacuum of the TEM chamber and thereby leading to an increase in particle size, a little bit like blowing up a balloon. For our commercial particles, characterization by TEM also gave a larger value than the size provided by the manufacturer, that is, 4.7 μm compared to 4.2 μm, respectively. B. Salt-Induced Aggregation of Nonswollen PS-DVB Latex Spheres. We first investigated the salt-induced aggregation behavior of PS latex spheres not swollen with the monomer styrene. We identified MgSO4 as the best-suited salt to induce the aggregation of our aqueous latex particle dispersions. Note that the goal of these aggregation tests was to identify a regime where the aggregates are definitely larger than single particles but not yet growing so rapidly that they might precipitate within 1 day. This ensures that the aggregate dispersion will remain stable during UV-irradiation for several hours, as needed for successfully “gluing together” the styrene-swollen latex spheres. Figure 1 shows the average apparent hydrodynamic radius as a function of aggregation time for PS-DVB 1% particles at three different salt concentrations, determined by DLS at scattering angle 70°. At the lowest salt concentration 0.01 mol/L, the latex particles did not aggregate at all. At salt concentration 0.08 mol/L, the aggregates grew so fast that their apparent radius reached 1 μm within a few minutes, thereby exceeding the reasonable regime of characterization by DLS. At the intermediate salt concentration of 0.05 mol/L, aggregates grew but very slowly, so their size stayed nearly constant for several hours and beyond the sedimentation limit even after 1-2 days. Although these aggregates are Langmuir 2009, 25(22), 12843–12846
Letter
Figure 2. Optical micrograph of fixed aggregates of PS-DVB2% latex particles.
Figure 3. Optical micrographs of UV-cross-linked PS-DVB2% latex particle aggregates diluted 1:5 with Milli-Q-water and additionally ultrasonicated for 15 min.
polydisperse, according to their average size, they comprise a limited number of single latex spheres (between 4 and 10) even after an aggregation time as long as 10 days. Therefore, a salt concentration of 0.05 M MgSO4 seemed to be a promising regime for our photofixation, assuming that the swelling with styrene does not change the aggregation kinetics. Next, we addressed the issue of reversibility of aggregate formation without UV-photo-cross-linking with these nonswollen particles: a dispersion of PS-DVB1% latex in 0.05 M MgSO4 with average aggregate radius 740 nm after aggregation time 30 h was diluted 1:50 with pure Milli-Q distilled water and treated for 15 min by ultrasonication. By DLS, we obtained a particle radius of 340 nm, which is comparable to the size of the original latex spheres, proving that the salt-induced aggregation without photocross-linking indeed is perfectly reversible as expected. The minor deviation (340 nm compared to 375 nm) is probably caused by partial particle destruction and/or dissolution of non-cross-linked polymer chains from the particles. C. UV-Photo-Cross-Linking of Preformed Aggregates. 1. PS-DVB1% Latex Spheres þ 2% Styrene. The original dispersion was diluted to a solid content of 5.9% before a total amount of 0.01 g of styrene was carefully added with an Eppendorf pipet in four portions. DLS at scattering angle 70° yielded a hydrodynamic radius of 402 nm for these swollen particles, compared to 375 nm before swelling. This unexpected large increase in hydrodynamic particle size is attributed to enhanced surface roughening during the swelling procedure. DOI: 10.1021/la903733b
12845
Letter
Marksteiner et al.
As a cross-check of successful photo-cross-linking, the dispersion was diluted 1:5 with Milli-Q water and ultrasonicated for 15 min. The aggregates in this case are still present but seem to have become slightly smaller (see Figure 3). We conclude that the swelling with styrene here was less efficient (compared to the PSDVB1% particles) due to the very low concentration of the latex dispersion. Our control experiment for nonirradiated swollen particle aggregates shows single latex particles, underlining the success of our photo-cross-linking approach also for the commercial DVB2%-PS spheres (see Figure 4).
Conclusions
Figure 4. Optical micrographs of nonirradiated PS-DVB2% latex particle aggregates diluted 1:5 with Milli-Q-water and treated with ultrasound for 15 min.
The swollen particle dispersion was aggregated by adding an identical volume of 0.05 mol/L MgSO4, diluted with 0.05 mol/L salt solution after 2 days of aggregation time as described for the preparation of DLS examples, and irradiated with UV-light for 3 h. DLS yielded a hydrodynamic radius of 742 nm for these fixed aggregates. After diluting 1:50 with pure Milli-Q water and 15 min of ultrasonication, the aggregate radius was still 672 nm, showing that the aggregates did not break up upon dissolution and ultrasonication and, therefore, the photofixation was successful. 2. PS-DVB2% Latex Spheres. A 0.1% solid content latex dispersion (1 mL, the low concentration was chosen for practical reasons only due to the small amount of commercial sample available at that time) was swollen with 200 wt % (with respect to the latex particles) styrene for 14 h. This large excess amount was necessary because of the very low latex particle concentration. Then 7 mL of 1 mol/L MgSO4 was added, and the mixture was left for an aggregation time of 7 days before UV-irradiation for 3 h. Figure 2 shows the optical micrograph of the resulting fixed aggregates.
12846 DOI: 10.1021/la903733b
We studied the photochemical cross-linking of colloidal polystyrene nanoparticles by UV-light-induced polymerization after the polystyrene particles, which were internally cross-linked with divinylbenzene, were carefully swollen in aqueous dispersion with a small amount of styrene and arranged into aggregates by saltinduced aggregation. Successful fixation of the aggregate structure was demonstrated by dynamic light scattering and by optical microscopy. Our new procedure is a very simple method to fix aggregates of cross-linked polymer particles by simply swelling them with a small amount of the respective monomer, and subsequent photopolymerization. It takes place in aqueous dispersion, does not effort a subsequent functionalization of latex particles as proven by its application to a commercial PS-DVB2% latex, and may easily be transferred to other kinds of polymers such as methacrylates. In addition, it is a very simple way to prepare well-defined colloidal aggregates if the single particles are not arranged in a nondefined random way by salt-induced aggregation but, for example, by laser traps. Here, fixation of well-defined colloidal aggregates leads to a facile fabrication of microscopic devices to be employed, for instance, in microfluidic systems. Acknowledgment. We thank R.W€urfel for the TEM characterization.
Langmuir 2009, 25(22), 12843–12846
