pubs.acs.org/Langmuir © 2010 American Chemical Society
Narrowly Size-Distributed Cobalt Salt Containing Poly(2-hydroxyethyl methacrylate) Particles by Inverse Miniemulsion Zhihai Cao,† Zhuo Wang,† Christine Herrmann,† Ulrich Ziener,*,† and Katharina Landfester*,†,‡ Institute of Organic Chemistry III - Macromolecular Chemistry and Organic Materials, University of Ulm, Albert-Einstein-Allee 11, 89081 Ulm, Germany, and ‡Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany Received November 19, 2009. Revised Manuscript Received December 22, 2009
†
Cobalt-containing hybrid particles have been prepared through the encapsulation of cobalt tetrafluoroborate hexahydrate (CoTFB) via inverse miniemulsion polymerization of 2-hydroxyethyl methacrylate (HEMA). We systematically varied the amount and type of cosolvent (water, methanol, ethanol, ethylene glycol), apolar continuous phase (cyclohexane, isooctane, isopar M, hexadecane), amount of cobalt salt, and molecular weight of the polymeric surfactant. The influence of those parameters on the particle size, size distribution, and particle morphology were investigated. Narrowly size-distributed hybrid particles with good colloidal stability could be obtained in a wide range of cobalt content between 5.7 and 22.6 wt % salt relative to the monomer. The addition of a cosolvent such as water not only promotes the loading of metal salt but also has a positive influence on narrowing the particle size distribution. We assume that generally narrowly size-distributed particles can be obtained for a large variety of combinations of polar/apolar phase by adjusting the balance between osmotic and Laplace pressure via the solubility of the metal salt in the continuous phase and lowering the interfacial tension by adjusting the hydrophilic-lipophilic balance (HLB) value of the surfactant. The results show a significant advantage of the inverse miniemulsion over the direct system with respect to the variability and total amount of metal salt without losing the narrow particle size distribution and colloidal stability.
Introduction Metal nanoparticles have many applications in various fields because of their semiconducting, optical, electronic, catalytic, and sensing properties.1,2 However, often they need to be stabilized or supported by a matrix due to their strong tendency to agglomerate. Polymer particles are one of the most commonly used matrixes. The metal-containing moiety can be immobilized on the surface of polymer particles,3,4 homogenously distributed in the matrix of particles,5-9 or as a separate phase dispersed in the particles.10,11 A versatile heterophase system for the preparation of those materials is offered by miniemulsions. Miniemulsion polymerization as a promising technique for preparing polymer nanoparticles has achieved significant progress in recent years.12-14 A miniemulsion *To whom correspondence should be addressed. E-mail: ulrich.ziener@ uni-ulm.de (U.Z.);
[email protected] (K.L.).
(1) Daniel, M.-C.; Astruc, D. Chem. Rev. 2004, 104, 293–346. (2) Burda, C.; Chen, X.; Narayanan, R.; El-Sayed, M. A. Chem. Rev. 2005, 105, 1025–1102. (3) Tamai, T.; Watanabe, M.; Hatanaka, Y.; Tsujiwaki, H.; Nishioka, N.; Matsukawa, K. Langmuir 2008, 24, 14203–14208. (4) Wen, F.; Zhang, W.; Wei, G.; Wang, Y.; Zhang, J.; Zhang, M.; Shi, L. Chem. Mater. 2008, 20, 2144–2150. (5) Ramirez, L. P.; Antonietti, M.; Landfester, K. Macromol. Chem. Phys. 2006, 207, 160–165. (6) Manzke, A.; Pfahler, C.; Dubbers, O.; Plettl, A.; Ziemann, P.; Crespy, D.; Schreiber, E.; Ziener, U.; Landfester, K. Adv. Mater. 2007, 19, 1337–1341. (7) Schreiber, E.; Ziener, U.; Manzke, A.; Plettl, A.; Ziemann, P.; Landfester, K. Chem. Mater. 2009, 21, 1750–1760. (8) Brieger, S.; Dubbers, O.; Fricker, S.; Manzke, A.; Pfahler, C.; Plettl, A.; Ziemann, P. Nanotechnology 2006, 17, 4991–4994. (9) K€astle, G.; Boyen, H.-G.; Weigl, F.; Lengl, G.; Herzog, T.; Ziemann, P.; Riethm€uller, S.; Mayer, O.; Hartmann, C.; Spatz, J. P.; M€oller, M.; Ozawa, M.; Banhart, F.; Garnier, M. G.; Oelhafen, P. Adv. Funct. Mater. 2003, 13, 853–861. (10) Erdem, B.; Sudol, E. D.; Dimonie, V. L.; El-Aasser, M. S. J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 4419–4430. (11) Fleischhaker, F.; Zentel, R. Chem. Mater. 2005, 17, 1346–1351. (12) Asua, J. M. Prog. Polym. Sci. 2002, 27, 1283–1346. (13) Landfester, K. Angew. Chem., Int. Ed. 2009, 48, 4488–4507. (14) Schork, F. J.; Luo, Y.; Smulders, W.; Russum, J. P.; Butte, A.; Fontenot, K. Adv. Polym. Sci. 2005, 175, 129–256.
7054 DOI: 10.1021/la904380k
consists of a dispersed phase containing the monomer(s) and osmotic pressure agent, and a continuous phase containing a surfactant. The dispersed phase is composed of droplets in the size range of 30-500 nm formed by a homogenization process, and the polymerization can be initiated either in the dispersed or continuous phase. Since the effective mass transfer is suppressed by the addition of an osmotic agent, each nanodroplet can be regarded as a nanoreactor in an ideal state. The reaction mechanism of miniemulsion polymerization displays the inherent advantage of encapsulating different materials to elaborate hybrid or composite nanoparticles.13 As reported in previous papers, the narrowly distributed metalcontaining particles have been prepared by the encapsulation of a hydrophobic metal salt via the direct miniemulsion or emulsion polymerization.6,7 These metal-containing particles could be applied in the field of nanolithography as an etching mask to create nanopillars or nanoholes. However, the low solubility of metal salts in the hydrophobic dispersed phase displays a severe limitation of the system with respect to the amount and type of metal (compound) in the hybrid particles. Hence, the transfer of the technique to an inverse system would allow one to increase the encapsulated amount of metal compound in the hybrid particles, as most metal compounds as salts are highly polar and hydrophilic. Actually, the inverse miniemulsion (co)polymerization has been successfully utilized to synthesize water-soluble or block (co)polymers.15-21 In addition, it has been proven that latex (15) Landfester, K.; Willert, M.; Antonietti, M. Macromolecules 2000, 33, 2370– 2376. (16) Willert, M.; Landfester, K. Macromol. Chem. Phys. 2002, 203, 825–836. (17) Oh, J. K.; Dong, H.; Zhang, R.; Matyjaszewski, K.; Schlaad, H. J. Polym. Sci., Part A: Polym. Chem. 2007, 45, 4764–4772. (18) Oh, J. K.; Tang, C.; Gao, H.; Tsarevsky, N. V.; Matyjaszewski, K. J. Am. Chem. Soc. 2006, 128, 5578–5584. (19) Qi, G.; Jones, C. W.; Schork, F. J. Macromol. Rapid Commun. 2007, 28, 1010–1016. (20) Blagodatskikh, I.; Tikhonov, V.; Ivanova, E.; Landfester, K.; Khokhlov, A. Macromol. Rapid Commun. 2006, 27, 1900–1905.
Published on Web 01/29/2010
Langmuir 2010, 26(10), 7054–7061
Cao et al.
Article
particles with a narrow particle size distribution could be obtained in inverse miniemulsion systems, and moreover the particle size could be tuned by using different types and amounts of surfactant.15 In the present paper, cobalt-containing hybrid particles with a narrow size distribution which is of importance for some applications, for example, for a new approach in nanolithography,6 have been synthesized via the encapsulation of cobalt tetrafluoroborate hexahydrate (CoTFB) by the inverse miniemulsion polymerization of 2-hydroxyethyl methacrylate. Compared to the direct miniemulsion polymerization, narrowly size-distributed hybrid particles containing a much higher weight metal content in form of a metal salt could be produced in inverse miniemulsion polymerization. The addition of cosolvents not only promotes the dissolution of CoTFB in the dispersed phase but also has a positive influence on narrowing the particle size distribution. A similar effect is found for the introduction of a suitable amount of CoTFB which in addition improves the colloidal stability. Particle properties including size, size distribution, and morphology depend greatly on the choice of the types of cosolvent, continuous phase, and surfactant.
Experimental Section Materials. The monomer, 2-hydroxyethyl methacrylate (HEMA, Aldrich, 97.0%), was purified by passing through a column filled with alumina, stored in the refrigerator (8 °C), and used up in 2 weeks. The block copolymer surfactant, poly(ethylene-cobutylene)-b-poly(ethylene oxide) (P(E/B)-PEO), was synthesized according to the literature.22,23 The molecular weight of the P(E/B)-PEOs used in this paper includes 9800 g/mol (P(E/B)PEO1), 6700 g/mol (P(E/B)-PEO2), and 6200 g/mol (P(E/B)PEO3), determined by 1H NMR spectroscopy. The constant block length of the hydrophobic block (4000 g/mol) results in hydrophilic-lipophilic balance (HLB) values for P(E/B)-PEO1, P(E/B)PEO2, and P(E/B)-PEO3 of 12, 8, and 7, respectively.24 The cosolvents including methanol (Merck, 99.9%), ethanol (Merck, 99.5%), and ethylene glycol (EG, Aldrich, 99.0%), the initiator, R,R0 -azoisobutyronitrile (AIBN, Merck, 98.0%), and the apolar solvents cyclohexane (CH, 99.5%, VWR Prolabo), isooctane (Merck, 99.0%), isopar M (a C12-C14 isoparaffinic mixture, Caldic Deutschland), and hexadecane (HD, Merck, 99.0%), and the metal salt cobalt tetrafluoroborate hexahydrate (CoTFB, Aldrich, 99.0%) were used as received. Demineralized water with Milli-Q grade (resistivity: 18 MΩ) was used.
Preparation of Inverse Miniemulsion and Miniemulsion Polymerization. A suitable amount of the surfactant P(E/B)PEO was dissolved in the corresponding apolar solvent under magnetic stirring. For the polar phase, CoTFB was first dissolved in the cosolvent, and then 1.5 g of HEMA was added to the CoTFB solution. The polar phase was mixed with the surfactant solution. After 15 min pre-emulsification under strong magnetic stirring, the mixture was treated with 120 s ultrasound with a Branson 450W digital sonifier at 90% amplitude in an ice bath to prepare a miniemulsion. The initial miniemulsion and AIBN were introduced to the reactor and then purged with argon for 3 min under magnetic stirring. The argon protected reaction mixture was placed in a preheated oil bath at 65 °C and stirred for 4 h. The respective amounts and detailed conditions are given in Table 1. Characterization. Dynamic Light Scattering (DLS). The particle size and size distribution (as PDI and intensity particle (21) Oh, J. K.; Perineau, F.; Matyjaszewski, K. Macromolecules 2006, 39, 8003–8010. (22) Schlaad, H.; Kukula, H.; Runloff, J.; Below, I. Macromolecules 2001, 34, 4302–4304. (23) Thomas, A.; Schlaad, H.; Smarsly, B.; Antonietti, M. Langmuir 2003, 19, 4455–4459. (24) Rosen, M. J. Surfactants and interfacial phenomena; Wiley: New York, 1978; p 304.
Langmuir 2010, 26(10), 7054–7061
size distribution) were measured by DLS (Nano-Zetasizer, Malvern Instruments) at 20 °C under the scattering angle of 173° at a wavelength of 633 nm. The dispersions were diluted with the corresponding apolar cosolvent in a glass cuvette before the measurement. Particle sizes and PDIs are given as the average of three measurements. The size and size distribution of the particles were calculated from the following equation: lnðG1Þ ¼ a þ bt þ ct2 þ dt3
ð1Þ
where G1 is a correlation function. The second order cumulant b is converted to a size using the dispersant viscosity and instrumental constants. The coefficient of the squared term c, when scaled as 2c/b2, is defined as the polydispersity index (PDI). The PDI is a measure of the particle size distribution, and the PDI is a dimensionless number that describes the heterogeneity of the sample; it can range from 0 (monodisperse) to 1 (polydisperse). Transmission Electron Microscopy (TEM). TEM measurements were performed on a Philips EM 400 microscope. A amount of 1.5 μL of dispersion was diluted with 3 mL of the corresponding apolar solvent, and then 1.5 μL of the diluted sample was placed on a 400-mesh carbon-coated copper grid and dried at 40 °C for at least 4 days. To enhance the contrast, an additional carbon-coating was performed before the measurement. Polymer Characterization. The glass transition temperature (Tg) was determined by differential scanning calorimetry (DSC) on a Perkin-Elmer DSC 7 differential scanning calorimeter. Three scanning cycles of heating-cooling were performed for each sample with a 10 °C/min heating rate for the first two cycles and a 20 °C/min heating rate for the third cycle. The last heating run was evaluated to calculate the Tg. The DSC samples were prepared by freeze-drying the dispersions. Coagulum. The coagulums produced during the polymerization accumulated on the reactor wall. They were collected, rinsed several times with CH to remove the residual polymer dispersion, and dried at 90 °C to constant weight. Viscosity. The viscosities of the polar mixture and the oil solution of P(E/B)-PEO were measured on a Haake MARS RheoStress 6000 rheometer in a thermostatically controlled measuring unit “plate/plate” (PP35/Ti). The temperature was kept constant at 20 °C, and the shear rate was set to 400 s-1. Interfacial Tension. The interfacial tensions were measured by the Du No€ uy ring method on a Dataphysics DCAT 11 tensiometer at 20 °C. Fourier Transform Infrared (FTIR) Measurement. FTIR spectra of HEMA, solutions of HEMA/CoTFB, and HEMA/ CoTFB/water were recorded by using a Bruker IFS113v FTIR spectrometer (Rheinstetten/Karlsruhe).
Results Cyclohexane as Continuous Phase. All syntheses were performed as radical polymerizations of the monomer HEMA initiated by AIBN in inverse miniemulsion with varying continuous phases and composition of the dispersed polar phase. As stabilizer, the block copolymer poly(ethylene-co-butylene)-bpoly(ethylene oxide) (P(E/B)-PEO) was employed. Cyclohexane (CH) is a commonly used continuous phase in inverse miniemulsion because it can be easily removed from the system by evaporation.17-21,25 Hence, in the first set of experiments, we concentrated on CH as continuous phase. a. Variation of Cosolvent. Four different cosolvents for CoTFB were employed, namely, water, ethylene glycol, methanol, (25) Luo, Y.-D.; Dai, C.-A.; Chiu, W.-Y. J. Polym. Sci., Part A: Polym. Chem. 2008, 46, 8081–8090.
DOI: 10.1021/la904380k
7055
Article
Cao et al. Table 1. Formulations, Particle Sizes, PDIs, and Amounts of Coagulum for All Experimentsa
run
P(E/B)-PEO type
surfactant mass (mg)
surfactant content (wt %)b
apolar solventc
1 2 3 4 5
P(E/B)-PEO1 P(E/B)-PEO1 P(E/B)-PEO1 P(E/B)-PEO1 P(E/B)-PEO1
64.6 64.3 64.8 64.5 64.3
3.37 3.36 3.38 3.37 3.36
CH CH CH CH CH
6
P(E/B)-PEO1
64.5
3.37
CH
7
P(E/B)-PEO1
64.7
3.38
CH
8 9 10 11 12 13 14
P(E/B)-PEO2 P(E/B)-PEO3 P(E/B)-PEO2 P(E/B)-PEO3 P(E/B)-PEO2 P(E/B)-PEO3 P(E/B)-PEO1
64.5 64.5 64.4 64.1 64.5 64.7 79.9
3.37 3.37 3.36 3.35 3.37 3.38 4.00
CH CH CH CH CH CH CH
15
P(E/B)-PEO1
86.4
4.00
CH
16
P(E/B)-PEO1
93.2
4.00
CH
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53
type of cosolvent water methanol ethanol EG water/ethanol (E=9.2 wt %) water/ethanol (E=22.5 wt %) water/ethanol (E=44.1 wt %) water water methanol methanol ethanol ethanol water/ethanol (E=22.5 wt %) water/ethanol (E=22.5 wt %) water/ethanol (E=22.5 wt %) water water water methanol methanol methanol water water water methanol methanol methanol water water water methanol methanol methanol water water water water water water water water water water water water
cosolvent (wt %)d
CoTFB (wt %)e
particle size (nm)f
PDI
coagulum (wt %)e
22.0 22.0 22.0 22.0 22.0
5.65 5.65 5.65 5.65 5.65
127 105 101 152 114
0.118 0.007 0.026 0.046 0.076
3.1 4.5 1.8 2.2 4.1
22.0
5.65
114
0.061
4.5
22.0
5.65
114
0.062
7.1
22.0 22.0 22.0 22.0 22.0 22.0 22.0
5.65 5.65 5.65 5.65 5.65 5.65 11.3
121 88 79 75 68 71 103
0.123 0.130 0.036 0.073 0.019 0.077 0.065
1.1 2.8 1.6 2.2 2.4 2.6 2.5
22.0
22.6
102
0.091
10.6
22.0
34.0
113
0.115
15.6
P(E/B)-PEO1 64.8 3.38 isooctane 22.0 5.65 145 0.142 6.7 P(E/B)-PEO2 64.4 3.36 isooctane 22.0 5.65 142 0.129 3.6 P(E/B)-PEO3 64.5 3.37 isooctane 22.0 5.65 142 0.039 4.1 P(E/B)-PEO1 64.7 3.38 isooctane 22.0 5.65 128 0.033 1.9 P(E/B)-PEO2 64.7 3.38 isooctane 22.0 5.65 98 0.117 1.2 P(E/B)-PEO3 64.5 3.37 isooctane 22.0 5.65 119 0.052 1.4 P(E/B)-PEO1 64.3 3.36 isopar M 22.0 5.65 164 0.056 17.4 P(E/B)-PEO2 64.3 3.36 isopar M 22.0 5.65 151 0.155 20.2 P(E/B)-PEO3 64.6 3.37 isopar M 22.0 5.65 158 0.032 20.3 P(E/B)-PEO1 64.6 3.37 isopar M 22.0 5.65 136 0.082 2.9 P(E/B)-PEO2 64.6 3.37 isopar M 22.0 5.65 143 0.221 9.5 P(E/B)-PEO3 64.6 3.37 isopar M 22.0 5.65 154 0.116 10.1 P(E/B)-PEO1 64.8 3.38 HD 22.0 5.65 134 0.057 34.8 P(E/B)-PEO2 64.8 3.38 HD 22.0 5.65 121 0.028 38.6 P(E/B)-PEO3 64.8 3.38 HD 22.0 5.65 113 0.022 30.2 P(E/B)-PEO1 64.6 3.37 HD 22.0 5.65 130 0.121 3.8 P(E/B)-PEO2 64.6 3.37 HD 22.0 5.65 161 0.235 15.1 P(E/B)-PEO3 64.6 3.37 HD 22.0 5.65 188 0.136 12.9 P(E/B)-PEO3 36.0 1.88 HD 22.0 5.65 141 0.037 54.2 P(E/B)-PEO3 46.0 2.40 HD 22.0 5.65 128 0.030 41.0 P(E/B)-PEO3 55.0 2.87 HD 22.0 5.65 123 0.032 38.9 P(E/B)-PEO3 75.7 3.92 HD 22.0 5.65 113 0.032 26.4 P(E/B)-PEO3 84.0 4.41 HD 22.0 5.65 106 0.021 42.3 P(E/B)-PEO3 71.7 3.92 HD 22.0 0 79 0.141 56.4 P(E/B)-PEO3 73.6 3.92 HD 22.0 2.86 134 0.057 27.5 P(E/B)-PEO3 78.4 3.92 HD 22.0 11.3 108 0.023 10.2 P(E/B)-PEO3 81.7 3.92 HD 22.0 17.0 103 0.022 6.1 P(E/B)-PEO3 84.7 3.92 HD 22.0 22.6 104 0.030 3.0 P(E/B)-PEO3 88.4 3.92 HD 22.0 28.4 106 0.048 6.6 P(E/B)-PEO3 91.7 3.92 HD 22.0 34.0 111 0.063 5.3 P(E/B)-PEO3 65.6 3.92 HD 11.3 137 0.183 3.3 P(E/B)-PEO3 69.4 3.92 HD water 6.7 11.3 139 0.058 2.7 P(E/B)-PEO3 72.1 3.92 HD water 11.3 11.3 136 0.060 7.9 P(E/B)-PEO3 75.7 3.92 HD water 16.7 11.3 117 0.014 6.5 P(E/B)-PEO3 81.6 3.92 HD water 27.3 11.3 102 0.040 11.6 P(E/B)-PEO3 85.1 3.92 HD water 33.3 11.3 106 0.020 7.5 P(E/B)-PEO3 88.5 3.92 HD water 40.0 11.3 110 0.053 3.9 a The monomer mass for all the experiments was 1.5 g. b Based on the mass of the polar phase. c The mass of the apolar solvent was 12.5 g for each experiment. d The amount of cosolvent does not include the crystal water in the cobalt salt, and the weight percentage is based on the monomer mass. e Based on the monomer mass. f z-average particle size determined by DLS.
and ethanol. In the same order, the solubility of CoTFB is decreasing because of a decreasing dielectric constant (Supporting Information, Table S1). For all systems, CoTFB-containing spherical particles could be prepared (Figure 1 and Supporting Information, Figure S7) where the average particle size obtained by TEM and DLS of about 130 nm decreased to around 100 nm and the PDI also 7056 DOI: 10.1021/la904380k
decreased significantly from 0.118 to 0.007 by going from water to the more hydrophobic cosolvents (Table 1, runs 1-4). It has to be noted that the polar phase still contained water arising from the crystal water of CoTFB (31.7 wt %). A similar trend could be observed when pure water as cosolvent was exchanged by water/ethanol mixtures (9.2, 22.5, and 44.1 wt % relative to the overall mass of cosolvent, Table 1, runs 5-7) Langmuir 2010, 26(10), 7054–7061
Cao et al.
Article
Figure 1. (a) Representative TEM image of the CoTFB-containing particles with ethanol as cosolvent and (b) intensity particle size distribution with the different cosolvents (see Table 1, runs 1-4).
with a basically invariant PDI between 0.076 and 0.062 but an increased solubility of the salt relative to pure ethanol (Supporting Information, Figure S8). Regarding the effect of ethylene glycol as cosolvent on particle size and distribution, the higher viscosity (Supporting Information, Figure S1) has to be taken into account hampering an efficient homogenization by ultrasound and leading to a relatively larger z-average size of 152 nm but a still low PDI of 0.046 by DLS. The TEM images of all the particles presented in this contribution show a homogeneous contrast suggesting that the different compounds in the particles are fully miscible although the limited TEM resolution does not allow a precise analysis. This suggestion is further supported by the fact that methanol, ethanol, and ethylene glycol are all good solvents for poly(HEMA) whereas in water the polymer (at high molecular weight) is only partly soluble and forms gels. Experimental proof for the miscibility of the components and its dependence on the presence of the cobalt salt is hampered by the low content of cosolvent in the polar phase. b. Variation of Surfactant Composition. In the present system, the block copolymer poly(ethylene-co-butylene)-b-poly(ethylene oxide) (P(E/B)-PEO) was employed as an efficient surfactant in inverse miniemulsion.15,26 Three P(E/B)-PEOs with different molecular weights (9800, 6700, and 6200 g/mol) by varying length of the hydrophilic block were used. The particle sizes decreased with decreasing molecular weight of P(E/B)-PEO, which is attributed to the increasing protection capability of P(E/B)-PEO for the inverse miniemulsion system with decreasing molecular weight (decrease of the HLB value). The size distribution (PDI) of the spherical particles became a little worse with decreasing molecular weight (Table 1, runs 1, 8, and 9; Figure 2 and Supporting Information, Figure S7a). All these results show that the molecular weight of P(E/B)PEO is an effective parameter to adjust the particle size, but using the P(E/B)-PEO with a lower molecular weight cannot effectively narrow the particle size distribution with CH as continuous phase. The overall tendency that the sizes and the PDIs of latex particles with water are larger than those of the systems with methanol or ethanol as cosolvent with the same P(E/B)-PEO as surfactant is ascribed to the differences in the interfacial tension (see below). c. Variation of CoTFB Content. The influence of the CoTFB content on the particle size, particle size distribution, and particle morphology were investigated in the systems with ethanol-water mixtures as cosolvent, CH as continuous phase, and 4 wt % P(E/ B)-PEO1 as surfactant. Based on the PDI (Table 1), TEM image (Figure 3a), and intensity particle size distribution (Figure 3b), a (26) Crespy, D.; Landfester, K. Macromolecules 2005, 38, 6882–6887.
Langmuir 2010, 26(10), 7054–7061
Figure 2. (a,b) TEM images and (c) intensity particle size distribution of the CoTFB-containing particles with water as cosolvent with different molecular weights P(E/B)-PEOs: (a) P(E/B)-PEO2, Mn=6700 g/mol and (b) P(E/B)-PEO3, Mn=6200 g/mol; P(E/B)PEO1, Mn=9800 g/mol (see Supporting Information, Figure S7a and Table 1, runs 1, 8, and 9).
Figure 3. (a) Exemplary TEM image (CoTFB = 11.3 wt %) and (b) intensity particle size distribution of the CoTFB-containing particles with a mixture of ethanol and water as cosolvent with different CoTFB weight contents (see Table 1, runs 14-16).
narrow particle size distribution could be obtained with 11.3 wt % CoTFB with respect to the monomer mass. In addition, this system showed a high colloidal stability with a low amount of coagulum. Increasing the weight content of CoTFB led to a destabilization and a broadening of the size distribution observable by the PDI (Table 1) and the intensity particle size distribution (Figure 3b). The poor colloidal stability is ascribed to the insufficient protection by P(E/B)-PEO1 in the presence of a high weight content of CoTFB. Depending on the continuous phase, a high colloidal stability can be obtained for 34 wt % CoTFB with P(E/B)-PEO3 as surfactant (see below). Isooctane as Continuous Phase. By going from CH to isooctane as continuous phase with water as cosolvent, the particle DOI: 10.1021/la904380k
7057
Article
Cao et al.
Figure 4. (1) Representative TEM images and (2) intensity particle size distribution of the CoTFB-containing particles with isooctane as continuous phase, (a) water and (b) methanol as cosolvent, and P(E/B)-PEO3 (Mn = 6200 g/mol) as surfactant.
Figure 5. (1) Representative TEM images and (2) intensity particle size distribution of the CoTFB-containing particles with isopar M as apolar solvent, (a) water and (b) methanol as cosolvent, and P(E/B)-PEO3 (Mn = 6200 g/mol) as surfactant.
size increased obviously to 145 nm, and the particle size distribution became broader (PDI = 0.142). The interfacial tensions increase with the increase of hydrophobicity of the apolar solvent (Supporting Information, Figure S2). In addition, the amount of coagulum increased (6.7 wt % of coagulum relative to the monomer was formed, Table 1). The coagulum amount decreased in the systems with P(E/B)-PEO2 and P(E/B)-PEO3 as surfactant. For P(E/B)-PEO3, additionally a low PDI (0.039) with relatively narrowly size-distributed spherical particles has been found (Figure 4a and Supporting Information Figure S9). Please note that the nonspherical appearance of some particles is caused by a degradation of poly(HEMA) under the electron beam. Compared to CH, the relatively lower viscosity of isooctane may be another possible reason for the broader particle size distribution with isooctane as continuous phase. If the polar cosolvent was changed to methanol (Figure 4b and Supporting Information Figure S9), the particle sizes of this system were generally lower than those in the case with water as cosolvent and isooctane as continuous phase, but larger than those with methanol as cosolvent and CH as continuous phase. The quantity of coagulum is negligible and amounted to less than 2 wt % with respect to monomer mass. Comparing the systems with the same P(E/B)-PEO as surfactant and water as cosolvent reveals that particles synthesized with methanol as cosolvent display a narrower particle size distribution when isooctane was used as continuous phase. Isopar M as Continuous Phase. If the more hydrophobic apolar solvent, isopar M, is used as continuous phase, particles with a significantly small PDI of 0.032 are only obtained when P(E/B)-PEO3 with the lowest HLB value is used as surfactant (run 25, Table 1). This is also confirmed by TEM (Figure 5a and Supporting Information Figure S10) which shows for the other surfactants the appearance of some small particles which we
attribute to homogeneous secondary nucleation. It should be noted that about 20 wt % of coagulum relative to the monomer mass was produced with isopar M as the continuous phase. The PDIs for the systems with methanol as cosolvent (0.082 (P(E/B)-PEO1), 0.221 (P(E/B)-PEO2), and 0.116 (P(E/B)-PEO3)) indicate that only relatively broadly size-distributed particles could be synthesized. This finding is also supported by the TEM results (Figure 5b1 and Supporting Information Figure S10) and the intensity particle size distribution (Figure 5b2). HD as Continuous Phase. Switching to HD as continuous phase and water as cosolvent significantly improved the particle size distribution with both surfactants P(E/B)-PEO2 and P(E/B)PEO3 (Figure 6a) showing a narrow size distribution of welldefined spherical particles with a high homogeneity in highly organized arrays. On the contrary, for the system with P(E/B)PEO1 as surfactant, some small particles could be clearly detected in the TEM image (Figure 6a1), resulting in a relatively broad particle size distribution. It should be mentioned that the amount of coagulum increased to about 39 wt % for P(E/B)-PEO2 and about 30 wt % for P(E/B)-PEO3 as surfactant. We attribute the worse colloidal stability in the systems using HD and isopar M compared to CH as continuous phase to the increase of the differences in the hydrophobicity of the dispersed polar phase and the continuous phase. Similar to the systems with isopar M as continuous phase, only particles with a relatively high polydispersity could be obtained (0.121 (P(E/B)-PEO1), 0.235 (P(E/B)-PEO2), 0.136 (P(E/B)PEO3)) with HD as continuous phase and methanol as cosolvent. These results were confirmed by TEM (Figure 6b). Further experiments were carried out with HD as continuous phase, water as cosolvent, and P(E/B)-PEO3 as surfactant, as they led to narrowly size-distributed hybrid particles in highly ordered arrays (see Figure 6a2 and a3) although the amount of coagulum was relatively high. As expected, the particle size decreased when
7058 DOI: 10.1021/la904380k
Langmuir 2010, 26(10), 7054–7061
Cao et al.
Article
Figure 7. (a) Particle sizes and PDIs, (b) amounts of coagulum of the CoTFB-containing particles with water as cosolvent, HD as continuous phase, and P(E/B)-PEO3 as surfactant (runs 40-46, Table 1).
Figure 8. (a) Particle sizes and PDIs, (b) amount of coagulum, and (c,d) representative TEM images of the CoTFB-containing particles with water as cosolvent, HD as continuous phase, and P(E/B)PEO3 as the surfactant with (c) only crystal, no additional water, and (d) 16.7 wt % water.
Figure 6. (1-3) TEM images and (4) intensity particle size distribution of the CoTFB-containing particles with HD as apolar solvent with (a) water and (b) methanol as cosolvent and (1) P(E/ B)-PEO1 (Mn = 9800 g/mol), (2) P(E/B)-PEO2 (Mn = 6700 g/ mol), and (3) P(E/B)-PEO3 (Mn = 6200 g/mol) as surfactant; runs 29-34 (Table 1).
the surfactant amount was increased (Supporting Information, Figure S11). The PDIs were basically insensitive to the variation of surfactant amount in the range of 1.9-4.4 wt % with respect to the mass of the polar phase and were in the range of 0.02-0.04. The amount of coagulum first decreased and then increased with increasing amount of surfactant (Supporting Information, Figure S11b). When the amount of CoTFB was increased from 0 to 34 wt % with a constant surfactant content of 3.9 wt %, the amount of coagulum decreased from 56% significantly to below 10%. In parallel, the particle size distribution decreases also Langmuir 2010, 26(10), 7054–7061
strongly with an optimum range of cobalt salt between 5.6 and 22.6 wt % (Figure 7 and Supporting Information Figure S12). The influence of the weight content of cosolvent on the particle properties was investigated by varying the water content in the system with HD as continuous phase, water as cosolvent, and P(E/B)-PEO3 as surfactant. The weight content of P(E/B)-PEO3 was kept constant at 3.9 wt % with respect to the mass of the polar phase; the weight content of CoTFB relative to HEMA was maintained at 11.3 wt %. A broad particle size distribution was obtained when besides the crystal water no further water was added (PDI = 0.183, Figure 8a and c). With the increase of the water content, the PDIs decreased significantly; that is, the addition of water not only improves the loading capacity with CoTFB but also has a positive influence on narrowing the particle size distribution (Figure 8 and Supporting Information, Figure S13). According to the FTIR results, we believe that the dissociation of CoTFB can be promoted by the introduction of water due to its stronger interaction compared to HEMA. Compared to the undissociated CoTFB, the dissociated CoTFB may more effectively work as lipophobe to improve the droplet stability by suppressing the molecular diffusion. This is supported by the fact the particle size distribution is significantly narrowed in the systems containing water in contrast to the system in the absence DOI: 10.1021/la904380k
7059
Article
of water. The optimum range of water content for obtaining narrowly size-distributed particles lies in the range of 16.7-33.3 wt % with respect to the monomer mass. The particle size distribution deteriorates a little with a relatively small or large amount of water. As shown in the Supporting Information (Figure S5), the interfacial tensions between the apolar and the polar phases increased with the increase of the water content. Therefore, the stability of droplets may become a little worse with a large amount of water, and as a result more monomer diffusion may take place in these cases, resulting in the occurrence of homogeneous nucleation to produce the particles with a small size. On the other hand, the osmotic pressure may not effectively establish to counteract the Laplace pressure due to the low extent of dissociation of CoTFB in the system with a low weight content of water. As shown in Figure 8a, not only the distribution but the particle size itself could be adjusted by using different amounts of water. It decreased with the increase of the water content in the range of 6.7-27.3 wt % water, and then slightly increased in the range of 27.3-40 wt %. Referring to the dependence of particle size on the weight content of CoTFB, the enhancement of the dissociation of CoTFB in polar mixtures seems to result in decreasing particle size. In addition, the viscosities of the polar phase decreased with the increase of water content in the polar mixture (Supporting Information, Figure S6). This can be another reason for the decrease of particle size with the introduction of water. The slight increase of particle size with a high water content can be ascribed to the increase of the interfacial tensions between the apolar and the polar phases with the increase of the water content.
Discussion We have shown that in the present inverse miniemulsion system a subtle interplay between the type and amount of polar cosolvent and surfactant and the amount of cobalt salt effectively controls the droplet stability. We assume that an increased droplet stability not only decreases the amount of coagulum but also narrows the size distribution of the resulting hybrid nanoparticles. A low interfacial tension and a high osmotic pressure guarantee the high colloidal droplet stability. The interfacial tension is lowered (i) if for a given nonpolar continuous phase the polarity of the cosolvent or (ii) the HLB value of the surfactant decreases. Furthermore, the osmotic pressure is increased if (i) the dissociation of the cobalt salt as lipophobe is promoted and (ii) the partition of the salt between the continuous and the polar phase is minimized. Here, the interfacial tensions between a specific apolar solvent and the polar mixture with different cosolvents are expected to decrease in the order of water > ethylene glycol > methanol > ethanol according to the dielectric constant of these cosolvents (Supporting Information, Table S1). On the other side, the interfacial tensions between a polar mixture with a specific cosolvent and different continuous phases decrease in the order of HD > isopar M > isooctane > CH (Supporting Information, Figure S2). Furthermore, based on the dielectric constants of the cosolvents, it is supposed that the solubility of the cosolvents in a specific apolar solvent increases in the order of water < ethylene glycol < methanol < ethanol. The (slight) solubility of the polar cosolvents in the hydrocarbons promotes also the solubility of CoTFB in the continuous phase; that is, with a specific cosolvent, the solubility of CoTFB increases in the order of HD < isopar M < isooctane < CH.27 On the other hand, the solubility of CoTFB in a specific continuous phase increases for the polar cosolvent in the order of water < ethylene glycol < methanol < (27) Ruelle, P.; Kesselring, U. W. J. Solution Chem. 1996, 25, 657–665.
7060 DOI: 10.1021/la904380k
Cao et al.
Figure 9. FTIR spectra of HEMA and the solutions of HEMA/ water, HEMA/CoTFB, and HEMA/CoTFB/water.
ethanol. Additionally, the dissociation of CoTFB will be mostly promoted with water as cosolvent which subsequently will (i) lower the solubility of the salt in the continuous phase and (ii) increase the osmotic pressure of the droplets as the molar amount of components inside the droplets is increased. For the systems with CH as continuous phase, in the order ethanol < methanol < ethylene glycol < water, the solubility of CoTFB in the continuous phase decreases (osmotic pressure increases) but the interfacial tension increases. Experimentally, a broadening of the particle size distribution in the same order is found. Thus, we assume that the droplet stability in this system is governed by the change of the interfacial tension and not by the osmotic pressure. For the systems with isooctane as continuous phase, the interfacial tension between the dispersed and the continuous phases increases relatively to CH as continuous phase (Supporting Information, Figure S2), presumably resulting in a decrease of the droplet stability. On the other hand, the solubility of CoTFB or its ions in the continuous phase is decreased, causing a higher osmotic pressure compared to CH as continuous phase. Here, a significant influence of the HLB value of the surfactant can be found. Only P(E/B)-PEO3 delivers a satisfyingly low PDI value. We assume that the osmotic pressure gains more influence by counteracting the Laplace pressure due to the decrease of interfacial tension by the surfactant with the lowest HLB value. Even with methanol as cosolvent, the particle size distributions are still relatively narrow due to the relatively lower interfacial tension compared to the systems using water. For the systems using isopar M and HD as continuous phase, the above trend is continued; that is, the influence of interfacial tension through the HLB value of the surfactant becomes more important than the effect of decreasing solubility of the salt in the continuous phase and thus increasing osmotic pressure. To get more insight into the interaction between the cobalt salt and the polymer/cosolvent and further support our mechanistic suggestion, FTIR spectroscopy was performed. The FTIR spectra of pure HEMA and the solutions of HEMA/water (run 40), HEMA/water/CoTFB (run 42), and HEMA/CoTFB (run 47) are shown in Figure 9. The characteristic band for the CdO stretching vibration in HEMA displays two minima of the transmittance at 1724 and 1699 cm-1 which can be attributed to free CdO groups and groups with hydrogen bonding between some HEMA molecules, respectively.28 With the introduction of CoTFB, we find that the band position of the free CdO groups is nearly unaffected but the band position belonging to the hydrogen bonded CdO groups is shifted by 9 cm-1 to lower wavenumbers (1690 cm-1). These results support the existence of an interaction between HEMA and CoTFB. The band position of free CdO in (28) Perova, T. S.; Vij, J. K.; Xu, H. Colloid Polym. Sci. 1997, 275, 323–332.
Langmuir 2010, 26(10), 7054–7061
Cao et al.
Article
mirror the influence of the varying salt content on the colloidal stability and thus the particle size distribution which indeed is found (see Figure 7a). Therefore, we believe that the osmotic pressure established in the droplets increases at first with the increase of ion amount in the polar mixture thus narrowing the particle size distribution and decreasing the particle size. A further increase of salt concentration leads to the formation of ionic aggregates31 which subsequently causes a lower osmotic pressure. Additionally, we suppose that the strong interactions in the droplets may also help to suppress kinetically the molecular diffusion to the continuous phase to narrow the particle size distribution further. Figure 10. Influence of the CoTFB weight content on the glass transition temperature of the CoTFB-containing polymers in inverse miniemulsion polymerization of HEMA using water as cosolvent, isooctane as continuous phase, and P(E/B)-PEO3 as surfactant.
the solution of HEMA/water/CoTFB is still unchanged with respect to pure HEMA, whereas the position of the minimum for the bonded CdO lies between the one for pure HEMA and the solution of HEMA and CoTFB. The shift of the minimum for the bonded CdO to higher wavenumbers, compared to the solution of HEMA/CoTFB, indicates that the interaction between HEMA and CoTFB was weakened by the introduction of water. This can be attributed to a competition between water and HEMA molecules as ligands for the cobalt ions. This is further supported by the spectrum of HEMA/water showing a further increase of the band position for the hydrogen bonded species to 1704 cm-1 indicating a stronger interaction between water and HEMA molecules. A more precise view at the two bands reveals that also the intensity of the lower energy band exceeds the one at higher wavenumbers which is presumably responsible for the red shift of the higher energy band to 1718 cm-1. The interaction between the cobalt ions and the carbonyl groups of poly(HEMA) should also have an effect on the thermal properties of the polymer. Thus, the glass transition temperature (Tg) of the dried particles and its dependence on the CoTFB content were investigated in the systems with isooctane as continuous phase, water as cosolvent, and P(E/B)-PEO3 as surfactant (Figure 10). For all samples, only one Tg could be found in the thermogram. A sample of polymer particles without CoTFB shows a Tg at about 105 °C. With an increasing weight content of salt, the Tg’s run through a maximum at about 123 °C in the range of 11.3-17.0 wt % of CoTFB. This kind of dependence of the Tg’s on the weight content of metal salt is frequently observed in the composites of polymer and metal salt.29-31 The Tg’s increase with the increase of weight content of metal salts in the range of low weight content which can be attributed to an increase of ion-polymer interactions.29 In our case, it means the enhancement of interactions between poly(HEMA) and Co2þ presumably leads to a weak cross-linking. However, the metal ion cannot work as cross-linker or effectively establish the interaction between polymer chains and salts at a higher metal salt due to the formation of ion pairs or higher order ionic aggregates.31 The interaction between polymer molecules and metal salts can be significantly weakened due to the presence of ionic aggregates, resulting in the decrease of Tg.31 As the trend of the Tg’s reflects the change of interactions between the metal ions and the polymer molecules it should (29) Kim, Y. J.; Hong, S. U.; Won, J.; Kang, Y. S. Macromolecules 2000, 33, 3161–3165. (30) Angell, C. A.; Liu, C.; Sanchez, E. Nature 1993, 362, 137–139. (31) Kim, J. K.; Min, B. R.; Won, J.; Kang, Y. S. J. Phys. Chem. B 2003, 107, 5901–5905.
Langmuir 2010, 26(10), 7054–7061
Conclusion Narrowly size-distributed cobalt-containing hybrid particles have been successfully synthesized by the encapsulation of cobalt tetrafluoroborate hexahydrate in poly(2-hydroxyethyl methacrylate) via inverse miniemulsion polymerization. In the combination of cosolvent, apolar solvent, and surfactant, the hydrophobicity of the cosolvent which has a significant impact on the solubility of CoTFB in the continuous phase and the interfacial tension between the dispersed and the continuous phase are critical for the droplet stability and thus particle size distribution. In the systems with cyclohexane as continuous phase, monodisperse hybrid particles could be obtained when methanol and ethanol were employed as cosolvent, and a high molecular weight surfactant (P(E/B)-PEO1) due to the relatively lower interfacial tension between the dispersed and the continuous phases compared to the system using water as cosolvent. In addition, the particle size distribution can be significantly narrowed by decreasing the solubility of CoTFB in the continuous phase by the appropriate choice of cosolvent and thus effectively increasing the osmotic pressure in the droplets accompanied with the decrease of the interfacial tension by using surfactants with a low HLB value. We believe that it is possible to synthesize narrowly size-distributed particles for a broad variety of compositions of the polar phase and continuous phase by fine-tuning the balance of osmotic and Laplace pressure to an optimum. For the systems containing HD, water, and P(E/B)-PEO3, the particle size decreased with the increase of the surfactant weight content with low polydispersity indexes. The introduction of a suitable amount of CoTFB has also a positive influence on the decrease of the polydispersity index and a low amount of coagulum. The amount of encapsulated CoTFB exceeds by far more than what is possible for a corresponding direct miniemulsion or emulsion system because of the poor solubility of the metal salts in the organic dispersed phase. The addition of water to the dispersed phase in the inverse system not only promotes the loading of CoTFB but also has a positive influence on narrowing the particle size distribution. Consequently, large amounts of coagulum were produced in the systems with the combination of a high interfacial tension between dispersed and continuous phase and a small amount of CoTFB. Although this work focused on the encapsulation of a cobalt salt, we believe that narrowly size-distributed hybrid particles with a broad variety of compounds from the vast field of hydrophilic metal salts can also be obtained by the reported technique. Supporting Information Available: Additional tables, plots, and TEM images. This material is available free of charge via the Internet at http://pubs.acs.org. DOI: 10.1021/la904380k
7061
