Langmuir 2002, 18, 5023-5026
5023
Formation of BaSO4 Nanoparticles in Microemulsions with Polymerized Surfactant Shells Mark Summers, Julian Eastoe,* and Sean Davis School of Chemistry, University of Bristol, Bristol BS8 1TS, United Kingdom Received December 21, 2001. In Final Form: March 22, 2002
Introduction Reversed microemulsions have proved useful as reaction template media for nanoparticle formation, as shown recently by Pileni;1 the classic systems are Aerosol-OTstabilized water-in-oil (w/o) microemulsions. Since these surfactant films are stabilized by weak van der Waals forces, direct templating is difficult to achieve, and often the resulting particles are much larger than the initial microemulsion droplets. Instances of shape replication are rare, the best documented being the formation of copper nanorods produced with Cu(AOT)2.1 A new approach is tested here, employing microemulsion films composed of partially polymerized surfactant (surfmer) mixtures, which enclose nanometer-sized aqueous reaction domains.2,3 The enhanced interactions brought about by covalently linking surfactant molecules in the film give rise to improved templating effects, as compared with the initial unpolymerized system. BaSO4 nanoparticles were obtained by mixing two separate microemulsions containing BaCl2 and Na2SO4, and transmission electron microscopy (TEM) is ideal for sizing the resulting BaSO4 particles [e.g., refs 4-8]. Recently, these polymerizable surfactants A and B and microemulsions have been extensively studied and characterized by small-angle neutron scattering (SANS).2,3 Spherical w/o droplets can be readily formulated from mixtures of a single-chain surfmer, (11-methacrloyloxy)undecyltrimethylammonium bromide (A), and its doublechain analogue, dodecyl-(11-methacryloyloxy)dimethylammonium bromide (B) (molecular structures are shown in Figure 1a). To enhance surfmer-oil compatibility, ethyl nonanoate was chosen over a normal hydrocarbon. It was established2,3 that replacing surfmer B with surfmer A in a mixture induces a more planar film curvature, widening the microemulsion phase region, thereby increasing the maximum water content and the droplet radius. Therefore, by varying composition of the surfmer mixtures a series of spherical “nanoreactors” of varying size can be produced. NMR and SANS data demonstrated that these films can be partially polymerized in situ, while still preserving the original structure of the microemulsion and also avoiding * To whom correspondence should be addressed. E-mail: julian.
[email protected]. Tel: +117 9289180. Fax: +117 9250612. (1) Pileni, M. P. Langmuir 2001, 17, 7449. (2) Summers, M. J.; Eastoe, J.; Davis, S. A.; Du, Z.; Richardson, R. M.; Heenan, R. K.; Steytler, D.; Grillo, I. Langmuir 2001, 17, 5388. (3) Summers, M. J.; Eastoe, J.; Heenan, R. K.; Steytler, D.; Grillo, I. J. Dispersion Sci. Technol. 2001, 22, 597. (4) Rees, G. D.; Evans-Gowing, R.; Hammond, S. J.; Robinson, B. H. Langmuir 1999, 15, 1993. (5) Hopwood, J. D.; Mann, S. Chem. Mater. 1997, 9, 1819. (6) Li, M.; Mann, S. Langmuir 2000, 16, 7088. (7) Mann, S. Angew. Chem., Int. Ed. 2000, 39, 3392. (8) Qi, L.; Ma, J.; Cheng, H.; Zhao, Z. Colloids Surf., A 1996, 108, 117.
Figure 1. (a) Chemical structure of the surfactants. (b) TEM micrograph of the parent, reacant-free microemulsion composed of X ) 40 and w ) 10 made in ethyl nonanoate.
any catastrophic phase separation or loss of interfacial activity of the polysurfactants.2,3 By following NMR intensities of the vinyl protons, it was found that the microemulsion interfaces contained approximately 35% of the polymerized surfactant.3 Furthermore, detailed contrast variation SANS experiments showed that the droplet and interfacial structures were essentially identical before and after polymerization. Adding cyclohexane as a cosolvent induced the distinct formation of cylindrical microemulsion droplets.2,3 This structural change in the background microemulsion is a key aspect of the present work here, in that well-defined nanorods of BaSO4 can be made inside the aqueous channels of partially polymerized surfmer shells, whereas spherical inorganic particles dominate with “weaker” unpolymerized interfaces. This approach is unique, and the results demonstrate potential uses of these new microemulsions with modified surfactant interfaces. Experimental Section (11-Methacrloyloxy)undecyltrimethylammonium bromide (A) and dodecyl-(11-methacryloyloxy)dimethylammonium bromide (B) were prepared as previously described.2 Water-in-oil (w/o or L2) microemulsions were composed of xA + (1 - x)B (where x ) [A]/([A] + [B])), with ethyl nonanoate (97% Aldrich) or mixtures of this solvent with cyclohexane (HPLC grade) as the oil component. Samples of desired composition were made at a constant surfactant concentration of 0.1 mol dm-3 and w value () [water]/[surfactant]). BaCl2 (99.9% Aldrich) and Na2SO4 (GPR BDH) were prepared as 0.2 mol dm-3 aqueous solutions and used to formulate transparent w/o phases. The preparation of BaSO4 was achieved by rapidly mixing equal volumes of the two
10.1021/la015755d CCC: $22.00 © 2002 American Chemical Society Published on Web 05/15/2002
5024
Langmuir, Vol. 18, No. 12, 2002
Notes
Table 1. TEM Particle Sizing Analysisa
Figure
X
w
1b 0.40 10 2a 2b 2c 1 week aged as in Figure 2b nonpolymerized 0.40 5 polymerized 3a 0.45 20 3b 3c
Dsphere/ nm
Dcylinder/ nm
e
e
f
f
2.8 2.7 2.8 18.7 2.8 25.8 2.8 3.8 2.8 23.5
L/nm e
f
p 0.20 0.29 0.22 0.26 0.25
1.4 14.0 1.4 14.3 35.0
0.19 0.20 0.39 2.7 32.6 0.22 62.5 59.6 0.23 22.9 0.20 2.7 28.8 0.22 62.5 52.8 0.32 29.7 0.23 2.7 36.1 0.19 62.5 66.7 0.27
a D sphere is the mean diameter of spherical particles, Dcylinder is the mean cross-sectional diameter of cylindrical particles, L is cylinder length, and p is the polydispersity. e is the expected size based on the background microemulsion, calculated from eq 1 for spherical particles and determined by SANS for the cylinders (refs 2 and 3). f represents the particle dimension found by TEM analysis.
separate microemulsions containing the respective reactants. Dynamic exchange between aqueous droplets is rapid,9 although particle growth is typically longer than this (minutes to hours) [e.g., ref 10]. Directly after mixing, polymerization was initiated using an external UV-light source.2,3 The morphology and size distribution of BaSO4 nanoparticles was determined by transmission electron microscopy (JEOL 1200EX TEM, operated at 120 kV up to magnifications of 500k). Samples were prepared by depositing a drop of solution onto Parafilm and then placing over a 3 mm copper grid. After 1 min, the grid was removed and excess solution was wicked away using filter paper. Images were recorded on a SIS MegaView II digital camera. Particle sizes were determined using Soft Imaging Systems GmbH analySIS 3.0 software. The polydispersity index is given by p ) (D/σ(D)) where D is the mean particle diameter and σ(D) is the standard deviation. Reproducibility was checked with freshly prepared and repeat samples.
Results and Discussion (a) Background Microemulsion. Figure 1b is a reference image of the pure background microemulsion, with no added reagents. The mean particle diameter and polydispersity are given in Table 1; the expected diameter can be estimated using eq 1:
Rc )
3wν Ah
(1)
Rc is the core radius, w is defined above, ν is the molecular volume of water, and Ah is the effective interfacial molecular area. As described elsewhere,2,3 SANS gave a mean value for Ah ) 64 ( 2 Å2, in good agreement with similar results for structurally related but chemically inert quaternary ammonium surfactants.11 Hence, the calculated diameter (2.8 nm) agrees well with that determined by TEM. (b) Spherical BaSO4 Nanoparticles. BaSO4 nanoparticles were prepared in microemulsions composed of (9) Fletcher, P. D. I.; Howe, A. M.; Robinson, B. H. J. Chem. Soc., Faraday Trans. 1 1987, 83, 985. (10) Eastoe, J.; Stebbing, S.; Dalton, J.; Heenan, R. K. Colloids Surf., A 1996, 119, 123. (11) Warr, G. G.; Sen, R.; Evans, D. F.; Trend, J. E. J. Phys. Chem. 1988, 92, 774.
Figure 2. (a) BaSO4 nanoparticles prepared in a microemulsion stabilized by adsorbed surfactant monomers with X ) 40, w ) 10, and ethyl nonanoate. (b) BaSO4 nanoparticles prepared from a microemulsion stabilized by a partially polymerized surfactant film; X ) 40 and w ) 10 with ethyl nonanoate. The surfmer polymerization was UV initiated. (c) The image in (b) magnified to a 50 nm scale.
X ) 0.40, w ) 10, and ethyl nonanoate oil. In general, mixing the two separate microemulsions produced a slightly turbid phase that separated after polymerization to give a transparent solution coexisting with a fine layer of white precipitate. TEM analysis was performed on the upper liquid phase, immediately after polymerization and after fixed aging periods. Shown in Figure 2 are particles obtained with nonpolymerized (Figure 2a) and polymerized (Figure 2b) microemulsion films; particle sizes are
Notes
given in Table 1. The initial “weak” shell system yields relatively polydisperse (p ) 0.29) spherical particles of average diameter 18.7 nm, that is, significantly larger than the background microemulsion droplets. A similar diameter was found for the polymerized systems, but there was a clear decrease in polydispersity (Table 1). An interesting result, common to all samples, is the coexistence of a population of much smaller particles, which are just visible in Figure 2b. Figure 2c is an enlargement, showing relatively polydisperse spherical particles of mean diameter of 3.8 nm. This could represent a bimodal particle distribution or precursor microemulsion droplets loaded with reactant ions. Obviously, the mean particle diameter is larger than that found with the equivalent “reactantfree” microemulsion. Polymerization may affect droplet coalescence and fission, thereby altering the intercompartment exchange process. Similar particle sizes were observed in BaNaAOT reverse micelles:5 oval-shaped particles, 70150 nm long and 50-100 nm wide, composed of smaller amorphous, roughly spherical BaSO4 nanoparticles between 2 and 4 nm in diameter were reported. Over a period of 1 week, no obvious changes in macroscopic appearance of the systems were observed, and the results of aging experiments are given in Table 1. Well-defined spherical particles were maintained only with the partially polymerized shells. However, as previously found,7 some of the crystals appeared rectangular or rhombic. The population of smaller particles (such as shown in Figure 2c) with a mean diameter of 3.1 nm was always observed. Reducing w from 10 to 5 gave rise to a reduction in particle size by approximately the same factor (0.6, see Table 1). (c) Reactions in Cylindrical Microemulsion Droplets: Effect of Cosolvent. As shown by SANS,2,3 cylinder microemulsion droplets are formed with a solvent mixture of 50:50 ethyl nonanoate and cyclohexane. The systems studied here were with X ) 0.45 and w ) 20. BaSO4 particles prepared in unreacted and UV-polymerized shell systems are shown in parts a and b of Figure 3, respectively, and the sizes are given in Table 1. The monomeric system shows little direct templating effect, and spherical particles, of mean diameter of 35.0 nm, dominate the picture. Nonetheless, a small fraction of cylindrical particles were observed, with a mean crosssectional diameter of 32.6 nm and length of 60 nm (axial ratio ∼ 1.8). Although the radial dimension is somewhat larger than for the starting droplets, the length is of similar size [see results in ref 2]. On the other hand, the polymeric system gave rise to a large majority of cylindrical particles, clearly indicating an improved templating effect. The mean cross-sectional diameter was found to be 28.8 nm, and the length was 52.8 nm (axial ratio of 1.8). A small minority of spherical particles exist with a mean diameter of 23 nm. The effect of time was investigated, and Figure 3c is an image for the polysurfmer shell system taken 1 week after preparation. Although particle size increased (Table 1), a large majority of cylindrical particles were still present, clearly demonstrating that the templating effect is preserved. In comparison with other nonreacted shell systems described in ref 1, some success has been demonstrated in obtaining cylindrical Cu nanoparticles in Cu(AOT)2 reversed micelles. The majority of particles remained spherical with typically 13% cylinder formation. However, it was found that using interconnected cylinders as templates leads to a higher fraction (42%) of rod-shaped particles.
Langmuir, Vol. 18, No. 12, 2002 5025
Figure 3. (a) BaSO4 nanoparticles prepared from a microemulsion stabilized by adsorbed surfactant monomers with X ) 45, w ) 20, and a 50:50 ethyl nonanoate/cyclohexane oil mixture. (b) BaSO4 nanoparticles prepared from a microemulsion stabilized by a partially polymerized surfactant film; X ) 45, w ) 20, and a 50:50 ethyl nonanoate/cyclohexane oil mixture. (c) BaSO4 nanoparticles prepared as in (b), but aged for 1 week.
Conclusions BaSO4 nanoparticles were prepared in water-in-oil microemulsions, formulated with mixtures of single- and double-chain polymerizable cationic surfactants (surfmers). Particle formation was studied before and after initiating a polymerization reaction to covalently link surfactant monomers at the oil-water interface. Starting with spherical microemulsion droplets (ethyl nonanoate
5026
Langmuir, Vol. 18, No. 12, 2002
solvent), the unpolymerized shell systems gave rise to polydisperse nanospheres, of mean diameter significantly larger than the starting droplets. Spherical particles were also formed starting with cylindrical microemulsion droplets (50:50 ethyl nonanoate/cyclohexane mixed solvent), implying that the initial surfactant film was too weak to limit particle growth. However, enhanced templating effects were observed with partially polymerized surfactant films, especially for cylindrical microemulsion droplets. For the spherical systems, a relatively mono-
Notes
disperse distribution was obtained that remained stable after a week, accompanied with a slight increase in polydispersity. Applied to the cylindrical system, it was found the majority of particles formed were rodlike with an axial ratio of 1.8. These results demonstrate clear advantages of partially polymerized surfactant shells for templating anisotropic inorganic particles from waterin-oil microemulsions. LA015755D
