Preparation of Raspberry-like Polymer Particles by a

Dec 18, 2012 - Esther U. Ikhuoria , Stanley O. Omorogbe , Bertrand T. Sone , Zebib Y. Nuru , Saleh Khamlich , Malik Maaza. Journal of Modern Optics 20...
0 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/Langmuir

Preparation of Raspberry-like Polymer Particles by a Heterocoagulation Technique Utilizing Hydrogen Bonding Interactions between Steric Stabilizers Hideto Minami,* Yusuke Mizuta, and Toyoko Suzuki Department of Chemical Science and Engineering, Graduate School of Engineering, Kobe University, Kobe 657-8501, Japan ABSTRACT: Large polystyrene particles stabilized by poly(acrylic acid) (PAA) (L-PSPAA) (as the core) and small polystyrene particles stabilized by poly(vinyl pyrrolidone) (PVP) (S-PSPVP) (as the corona) were successfully used to prepare raspberry-like particles by a heterocoagulation technique utilizing the hydrogen bonding interaction between PAA and PVP. The coverage of L-PSPAA by S-PSPVP could be controlled by adding PVP homopolymer to the L-PSPAA dispersion and by changing the molecular weight of the stabilizers. Moreover, the heterocoagulation of large poly(methyl methacrylate) particles stabilized by PAA (LPMMAPAA) and S-PSPVP particles was also accomplished, resulting in the formation of L-PMMAPAA-core/S-PSPVPcorona raspberry-like composite particles. These results suggested that the raspberry-like particles composed of various polymer particles could be formed by the heterocoagulation technique utilizing the hydrogen bonding interaction.

■

INTRODUCTION Raspberry-like particles, each consisting of one large particle and many small particles on the surface of the large particle, have been investigated for their potential in applications such as wetting-controlled films,1−5 core−shell composite particles,6−8 hollow particles,6,9−12 and Janus particles.13,14 Furthermore, they exhibit optical and electrical properties when combined with other materials.15 Raspberry-like particles have been prepared by Pickering emulsion,11,13,14,16−18 seed polymerization,19 silica synthesis,20,21 and heterocoagulation techniques. The heterocoagulation technique, which involves the adsorption of small particles onto large particles by the use of electrostatic interaction,9,10,12,22−24 hydrophobic interaction,25 and covalent bonding,1,4 has been reported. Okubo et al. reported the preparation of a raspberry-like particle by a “stepwise” heterocoagulation technique in which many small cationic polymer particles were adsorbed onto one large anionic polymer particle prepared by emulsifier-free emulsion polymerization.26−33 Recently, Yang et al. reported the preparation of a polymer composite with raspberry-like morphology by selfassembled heterocoagulation based on a hydrogen bonding interaction between cross-linked poly(acrylic acid) (PAA) particles and cross-linked poly(4-vinylpyridine) (PVPy) in organic solvents.34,35 The hydrogen bonding interaction between the hydrogen bond donor polymer and the hydrogen bond acceptor polymer, such as PAA and PVPy, respectively, often induces the formation of precipitates in aqueous solution, although both polymers are water-soluble; the formed © XXXX American Chemical Society

precipitates are hydrogen bonding interpolymer complexes. However, in this case, the kinds of polymer particles should be limited because the interaction is derived from the particles comprising polymers. Moreover, it would seem to be difficult to control the morphology of a number of small particles on one large particle. In this article, we demonstrated the facile preparation and development of raspberry-like polystyrene (PS) particles by utilizing the hydrogen bonding interaction between PAA and poly(vinylpyrrolidone) (PVP, a hydrogen bond acceptor polymer) as stabilizers in which large PS particles and small PS particles stabilized by PAA and PVP, respectively, were simply mixed. This method is expected to be applicable to the preparation of raspberry-like particles consisting either of various polymers or different types of particles. Moreover, the morphology of the raspberry-like particles was controlled.

■

EXPERIMENTAL SECTION

Materials. Styrene (S) and methyl methacrylate (MMA, Nacalai Tesque, Inc., Kyoto, Japan) were purified by distillation under reduced pressure in a nitrogen atmosphere. Reagent-grade 2,2′-azobis(isobutyronitrile) (AIBN, Wako Pure Chemical Industries, Osaka, Japan) was purified by recrystallization in methanol. Reagent grades of methanol, ethanol, standard buffer solution, poly(vinylpyrrolidone) Received: November 2, 2012 Revised: December 17, 2012

A

dx.doi.org/10.1021/la3043507 | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

(PVP, weight-average molecular weights 4.0 × 104 and 9.0 × 105, Nacalai Tesque, Inc., Kyoto, Japan), and PAA (weight-average molecular weight 2.5 × 105, Sigma-Aldrich Chemical Co., USA) were used as received. Preparation of PS Particles. The PS particles stabilized by PVP (L-PSPVP and S-PSPVP) were prepared by dispersion polymerization in a sealed glass tube as follows: S (1.0 g), AIBN (0.01 g), PVP (0.2 g), and the medium were mixed in the reactor and held at 60 °C for 24 h under a nitrogen atmosphere. The tube was shaken horizontally at 80 cycle/min (3 cm strokes). The PS particles stabilized by PAA (L-PSPAA and S-PSPAA) were prepared under the same conditions, except for the use of PAA (0.2 g) as the stabilizer. Ethanol (6.0 g) and methanol/ water (4.8 g/1.2 g) were used as the medium in the case of the large PS particles (L-PSPVP and L-PSPAA) and the small PS particles (S-PSPVP and S-PSPAA), respectively. These particles were used after centrifugal washing with methanol to remove ungrafted PAA and PVP, and then the medium was replaced with water. Preparation of Raspberry-like Particles. The L-PSPAA and SPSPVP aqueous dispersions after centrifugal washing were separately diluted with buffer solutions to 0.5 and 2.0 wt %, respectively. Aqueous buffer solutions (0.01 mol/L) of tetraoxalate, hydrogen phthalate, dihydrogenphosphate, and carbonate-bicarbonate potassium salts were used for pH values of 1.7, 4.0, 6.9, and 10.0, respectively. The L-PSPAA dispersion (0.5 wt %, 0.8 mL) was added to the S-PSPVP dispersion (2.0 wt %, 2.0 mL) by dropwise addition with magnetic stirring. After the heterocoagulation process, free S-PSPVP particles were removed by centrifugation, and the raspberry-like particles were placed in a sealed glass tube and held for 1 h at 100 °C in buffer solution. Characterization. The PS and the raspberry-like particles were observed with a scanning electron microscope (SEM, JSM-6510, JEOL Ltd., Tokyo, Japan) and an optical microscope (ECLIPSE 80i, Nikon). The number of average-diameter (Dn) particles and the coefficient of variation (Cv) of the particles were measured with image analysis software (WinROOF, Mitani Co., Ltd.). Monomer conversions were determined by gas chromatography (GC, GC-18A, Shimadzu Corporation, Kyoto, Japan) employing helium as the carrier gas, N,N-dimethylformamide as the solvent, and p-xylene as the internal standard.

medium (ungrafted) was measured by gravimetry after centrifugal washing of these particles. According to this measurement, the L-PSPAA and S-PSPVP particles contained 2.2 wt % PAA and 8.9 wt % PVP, respectively. Figure 2 shows optical micrographs of PS particles after mixing and the visual appearance of the mixture systems of the

Figure 2. Optical micrographs of the L-PSPAA dispersion (a) after the addition of the S-PSPVP dispersion and the S-PSPVP dispersion after batch addition (b) and the dropwise addition (c) of the L-PSPAA dispersion in buffer solution at pH 4.0 (insets: dispersion appearance).

L-PSPAA and S-PSPVP dispersions. When the S-PSPVP dispersion was poured into the L-PSPAA dispersion (Figure 2a), the system immediately coagulated and sedimentation was observed. In this case, S-PSPVP particles should work as bridging particles between L-PSPAA particles to result in coagulation because the number of S-PSPVP particles was considerably smaller than the number of L-PSPAA particles at the beginning of the addition of the S-PSPVP dispersion. As shown in the optical micrograph in Figure 2a, one S-PSPVP particle was sandwiched between two LPSPAA particles. On the contrary, when the L-PSPAA dispersion was poured into the S-PSPVP dispersion (Figure 2b), raspberrylike particles in which the L-PSPAA particle was surrounded by a monolayer of S-PSPVP particles were observed. However, the corona S-PSPVP particles were likely attached to another LPSPAA particle, and sedimentation occurred. Generally, the formation of a hydrogen bonding complex between polymers is very fast,36 but when the L-PSPAA dispersion was added to the S-PSPVP dispersion by dropwise addition, well-dispersed raspberry-like particles were observed, as shown in Figure 2c. Furthermore, the ratio between the large and small particles (w/w) was varied in a series of experiments. In the case of fewer small particles, the system became unstable and coagulated. However, above the optimum ratio (small/large 1/0.3 w/w), the stable raspberry-like particles were obtained without coagulum. In this heterocoagulation technique, heterocoagulation should occur under kinetically controlled conditions. Thus, the method of mixing the dispersion was very important in preventing the formation of bridging. To examine whether the driving force of the heterocoagulation was the hydrogen bonding interaction between the stabilizers, the same mixing procedure was performed with a large particle dispersion and a small particle dispersion using the same kind of stabilizer. The L-PSPVP particles (1.8 μm, Cv = 6.5%) and the S-PSPAA particles (0.91 μm, Cv = 7.4%) were also prepared by the dispersion polymerization of styrene using PVP and PAA as the stabilizers, respectively. As shown in Figure 3, heterocoagulation was not observed in the case of mixing either L-PSPVP and S-PSPVP particles or L-PSPAA and S-PSPAA particles. In both cases, the large and small particles remained colloidally stable after the mixing process, indicating that heterocoagulation did not occur and is probably caused by the hydrogen bonding interaction between the hydrogen bond donor (PAA) and the hydrogen bond acceptor stabilizer (PVP).

■

RESULTS AND DISCUSSION Figure 1 shows optical micrographs and SEM photographs of the large PS particles stabilized by PAA (L-PSPAA) and the small

Figure 1. Optical micrographs (a, b) and SEM photographs (a′, b′) of L-PSPAA particles (a, a′) and S-PSPVP particles(b, b′).

PS particles stabilized by PVP (S-PSPVP) prepared by dispersion polymerization. In both cases, the obtained particles were colloidally stable and monodisperse. The particle size/ coefficient of variation of L-PSPAA and S-PSPVP were 1.8 μm/ 16% and 0.57 μm/4.8%, respectively. To estimate the quantity of the grafted stabilizer, the amount of dissolved stabilizer in the B

dx.doi.org/10.1021/la3043507 | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

Figure 3. Optical micrographs of a mixed L-PSPAA and S-PSPAA dispersion (a) and a mixed L-PSPVP and S-PSPVP dispersion (b) in buffer solution at pH 4.0.

Additionally, to clarify the interaction between PAA and PVP, mixing of the S-PSPVP and L-PSPAA dispersions was carried out in buffer solutions of various pH values (Figure 4). Figure 5. SEM photographs of L-PSPAA/S-PSPVP heterocoagulated particles prepared by mixing dispersions in buffer solution at pH 4.0 after drying in the absence (a, b) and presence (c) of Emulgen 909. Optical micrograph (d) and SEM photographs (e, f) of heterocoagulated particles after heat treatment at 100 °C for 1 h.

dispersion before drying. As a result, the desorption of S-PSPVP particles from the L-PSPAA surface was clearly reduced (Figure 5c). To prevent the desorption of all S-PSPVP particles from the L-PSPAA surface without the addition of the nonionic emulsifier, the heterocoagulated particles in the buffer solution at pH 4.0 were held at 100 °C for 1 h (Tg,PS ≈ 100 °C) to promote sintering between the small and large particles. As a result of the heat treatment, the heterocoagulated particles were colloidally stable (Figure 5d), and the objective of creating raspberry-like particles without the desorption of S-PSPVP particles (Figure 5e,f) was successfully achieved. Hereafter, the heat treatment of the obtained heterocoagulated particles was conducted before the preparation of the dry sample. To control the particle morphology, that is, to control the number of S-PSPVP particles adsorbed on an L-PSPAA particle, the amount of valid PAA stabilizer on the surface of L-PSPAA particles was varied. The amount of PAA on the surface was changed by adding free PVP to the L-PSPAA dispersion prior to mixing in the S-PSPVP particles. Hydrogen bonds should be formed between the free PVP and PAA stabilizers on the surface of L-PSPAA particles, resulting in the reduction of valid PAA on the L-PSPAA particles. The ratios of the free PVP added were determined on the basis of the amount of PAA contained in the L-PSPAA particles (2.2 wt % based on the L-PSPAA particle), for which the molar ratio of the pyrrolidone group of the free PVP added to the carboxyl group of the PAA contained in L-PSPAA particles was calculated. At a ratio of 100%, L-PSPAA particles would be fully covered with free PVP and would not be adsorbed by S-PSPVP particles. Figure 6 shows SEM photographs of the heterocoagulated particles after heat treatment prepared by adding various amounts of free PVP before mixing the S-PSPVP and L-PSPAA dispersions. From the SEM observation, the number of adsorbed S-PSPVP particles on the L-PSPAA clearly decreased with an increase in the amount of free PVP up to the ratio of 30%. However, above a 30% ratio of free PVP the number of S-PSPVP particles adsorbed on L-PSPAA particles was observed to increase as shown in Figure 7f−h. To quantify this result, the degree of coverage (Pc %), which is expressed by eq 1, was estimated by counting the number of adsorbed S-PSPVP particles in the SEM data. To obtain an accurate measurement, the heterocoagulated particles located in isolated areas were observed in SEM photographs.

Figure 4. Optical micrographs of mixed L-PSPAA and S-PSPVP dispersions in buffer solutions at various pH values: (a) 1.7, (b) 4.0, (c) 6.9, and (d) 10.0.

Hydrogen bonding complexes can be formed in the protonation state of a carboxyl group involving a polyacid such as PAA.37,38 In the buffer solution at pH 1.7 and 4.0, heterocoagulated particles were observed in which most of the carboxyl groups of PAA (pKa 4.8) should be protonated. However, in the buffer solution at pH 6.9 and 10.0, heterocoagulation was not observed, and PAA was protonated to only a very small extent. In addition, when the pH value increases to 10.0 in the heterocoagulated particle system (pH 4.0), the desorption of small particles was observed. Also, when the heterocoagulated particles were exposed to mechanical shear such as sonication, some the small particles that adsorbed on the large particles were desorbed. These results also support the premise that the driving force of heterocoagulation in this work is a hydrogen bonding interaction between the stabilizers. Figure 5a,b shows SEM photographs of the heterocoagulated particles prepared in mixed L-PSPAA and S-PSPVP dispersions at pH 4.0. The coverage of L-PSPAA particles by S-PSPVP particles seemed to be small compared to that of the heterocoagulated particles shown in the optical photograph in Figure 2b. The SEM observation also indicated that the interaction of the LPSPAA and S-PSPVP particles was relatively weak on the basis of the hydrogen bonding interaction. This desorption of S-PSPVP particles seemed to be induced by the capillary force exerted as the dispersion was dried. The capillary force is in proportion to the surface tension of their media. To decrease the capillary force, that is, the water surface tension, a nonionic emulsifier, Emulgen 909, was added to the heterocoagulated particle C

dx.doi.org/10.1021/la3043507 | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

Figure 6. SEM photographs of heterocoagulated particles prepared by mixing dispersions of S-PSPVP and L-PSPAA in buffer solutions at pH 4.0, to which various concentrations of free PVP were added before mixing the dispersions after heat treatment at 100 °C for 1 h. Added PVP (%): (a) 0, (b) 8, (c) 15, (d) 22, (e) 30, (f) 37, (g) 45, and (h) 75.

Q max ≈ 10 × Q min =

Nmax =

N × 100 Nmax 2π 3

⎛ RLP + R SP ⎞2 ⎜ ⎟ R SP ⎠ ⎝

(3)

where Qmax and Qmin are the maximum (saturated) and minimum (required to stabilize particles against same-sized coalescence) graft coverages (chain/nm2), respectively. Rg is the radius of gyration of stabilizer chains in solution (nm). The usual expression is Rg = A × MWb for an ungrafted stabilizer, in which A and b are constants. The expression becomes Rg = A(0.75 × MW)b for the grafted stabilizer because the average chain length of the long stabilizing end will be 75% of the molecular weight of the ungrafted stabilizer.39 MW is the molecular weight of the stabilizer. In the present work, the MW of PAA is 2.5 × 105, and the constants are taken to be A = 0.53 nm mol g−1 and b = 0.32.40 Assuming that the graft coverage of PAA on the PS particles in this work was saturated (Qmax), the ratio of the effective stabilizer could be expressed by the following equation

Figure 7. Degree of coverage of L-PSPAA particles by S-PSPVP particles as a function of free PVP added.

Pc(%) =

10 10 = 2 2 πRg πA (0.75 × MW)2b

(1)

wPAA(surface) = 4πr 2 × Q max ×

MW NA

4

wPAA(all) = 3 πr 3d × 2.2(wt%) (2)

(4)

where r and d are the radius and density of an L-PSPAA particle, respectively, and NA is Avogadro’s number. Equation 4 was calculated to be 30.8 wt %, meaning that the remaining ∼70 wt % PAA would be buried inside the particles. The behavior of Pc (Figure 7) in which the minimum value occurred at ∼30% PVP addition was in very good agreement with these considerations even though some experimental error should be included. That is, the Pc value could be controlled by the addition of PVP to the L-PSPAA dispersion by considering the amount of PAA on the surface of the particles. However, beyond 30% the leftover PVP might induce depletion flocculation because the surface of the L-PSPAA particles was almost covered with free PVP and leftover PVP acted as unadsorbed polymer in the medium. The depletion attraction increases with increasing concentration of unadsorbed polymer,41 and thus the number of adsorbed SPSPVP particles might increase with the amount of free PVP added beyond 30%. However, when a large amount PVP is added (