Article pubs.acs.org/Langmuir
Efficient Pathway for Preparing Hollow Particles: Site-Specific Crosslinking of Spherical Polymer Particles with Photoresponsive Groups That Play a Dual Role in Shell Crosslinking and Core Shielding Yukiya Kitayama, Kazuki Yoshikawa, and Toshifumi Takeuchi* Graduate School of Engineering, Kobe University, 1-1, Rokkodai-cho, Nada-ku, Kobe 657-8501, Japan S Supporting Information *
ABSTRACT: Site-specific a posteriori photocrosslinking of homogeneous spherical polymer particles and subsequent removal of the particle corethe self-templating strategyhas been developed as an efficient pathway for hollow particle formation. In this approach, homogeneous polymer particles containing linear polymers bearing post-crosslinkable side-chain groups are first synthesized, and the photoinduced crosslinking occurred only at the shell region in the homogeneous polymer particles. Our fundamental studies clarified that the remaining non-crosslinked photoresponsive groups in the shell region played a crucial role in shielding the core region from photoirradiation. The shell-selective crosslinking was successfully applied to hollow polymer particle formation by core removal. This facile route to polymeric hollow particle formation via a self-templating strategy has great potential to be used as an alternative because the route has high mass productivity and high simplicity as a result of the nonuse of additional sacrificial template particles and highly toxic solvents.
1. INTRODUCTION Polymeric hollow particles have been attracting great attention owing to their widespread industrial/research applications such as organic white pigments,1 microreactors,2−4 cellular imaging agents,5−7 drug delivery vehicles,8−14 and self-healing materials.15−17 Remarkable progress has been made in the preparation of the hollow polymer particles, and versatile intelligent strategies for hollow polymer particle fabrication have been reported. Interfacial polymerization18−20 and polymersome formations via block copolymer self-assembly21−23 have been reported as routes to the fabrication of polymeric hollow particles. These strategies require a complex procedure including the prepreparation of amphiphilic block polymeric surfactants. Okubo and co-workers developed the self-assembling phase-separated polymer (SaPSeP) method as a powerful strategy for the preparation of hollow polymer particles.24,25 The driving force behind polymeric shell formation in this method is the selfassembly of phase-separated crosslinked polymers at the oil− water interface; therefore, the limitation of the methodology is © XXXX American Chemical Society
in the combination of polymer species and core materials, where a hydrophobic core material is necessary. In addition, it is often necessary to use an acceleration material for the phase separation of crosslinked polymers. Use of sacrificial template particles is another important strategy for producing polymeric hollow particles, and several intelligent routes have been reported based on templating methodology, such as seeded dispersion polymerization,26 surface-initiated polymerization,27,28 block copolypeptides assembly,29 and layer-by-layer processing.30−33 Although these approaches can prepare versatile hollow particles, the high complexity involved is a disadvantage because of the requirement of the use of different types of template particles (e.g., inorganic particles) from the shell-formed polymer species, leading to the necessity of using highly toxic solvents (e.g., hydrogen fluorite) for removing the sacrificial template Received: June 20, 2016 Revised: August 6, 2016
A
DOI: 10.1021/acs.langmuir.6b02295 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
Scheme 1. Synthetic Route to the Formation of Hollow Particles via Shell-Selective Photoinduced Crosslinking, Followed by the Removal of the Non-crosslinked Core Region
were observed using a scanning electron microscope (SEM) (VE9800, Keyence, Osaka, Japan). UV−vis spectral measurements were performed using a V-560 spectrophotometer (JASCO Ltd., Tokyo, Japan). MALDI-TOF mass spectra were recorded using Voyager-DE1000 (Ab Sciex, Tokyo, Japan). 1H NMR spectra were recorded using 300 MHz FT-NMR apparatus (JNM-LA300 FT NMR system, JEOL Ltd., Tokyo, Japan). Fluorescence spectra measurements were carried out using a F-2500 (Hitachi, Japan). MW and MWDs were analyzed via gel permeation chromatography (GPC) at 40 °C using two poly(styrene/divinylbenzene) gel columns (Tosoh Corp., TSK gel GMHHR-H, 7.8 mm i.d. × 30 cm) with THF as the eluant, a flow rate of 1.0 mL min−1, a refractive index (RI) detector (TOSOH RI-8020/ 21), an ultraviolet (UV) detector (TOSOH UV-8II). The columns were calibrated to the PS calibration standards (Mw = 1.05 × 103 to 5.48 × 106, Mw/Mn = 1.01−1.15). A C2si confocal laser-scanning microscope (CLSM, Nikon Corp.,Tokyo Japan) with excitation at 488 nm (for fluorescein) and an emission bandpass filter at 499−529 nm (for fluorescein) was used. 2.3. Cinnamoyloxyethyl Methacrylate (CEMA). HEMA (5.8 mL, 72 mmol) and pyridine (5.24 mL, 43.2 mmol) were dissolved in dry DCM (30 mL) in a round-bottom flask. Cinnamoyl chloride (11.8 g, 71.1 mmol) dissolved in dry DCM (30 mL) was slowly added to the solution in an ice bath. The mixture was stirred overnight at room temperature. The reaction mixture was washed with a saturated citric acid aqueous solution to remove pyridine compounds, Na2CO3 aqueous solution, and brine (saturated NaCl aqueous solution). The organic phase was dried in vacuo. Yield: 75%, 1H NMR (300 MHz, CDCl3): δ = 1.96 (s, 3H), 4.43 (m, 4H), 5.60 (s, 1H), 6.15 (s, H), 6.46 (d, 1H), 7.42 (m, 3H), 7.55 (m, 2H), 7.72 (d, 1H). MALDI-TOF-MS: 283.19 [M + Na]. 2.4. Fluorescein Acrylamide (FAm). 4-Amino fluorescein (0.125 g, 0.36 mmol) was dissolved in dry acetone (10 mL). The solution was stirred in an ice bath for 30 min in a nitrogen atmosphere. Acryloyl chloride (0.05 mL, 0.61 mmol) dissolved in dry acetone solution (2 mL) was slowly added to the solution in an ice bath. After overnight reaction at ambient temperature, the precipitant was corrected by filtration, and the solids were washed by diethyl ether. The obtained product was dried in vacuo. Yield: 98%, 1H NMR (300 MHz, DMSO): δ = 5.85 (d, 1H), 6.3− 6.7 (m, 9H), 7.23 (d, 1H), 7.89 (d, 1H), 8.41 (s, 1H), 10.1 (br, 1H), 10.6 (s, 1H). MALDI-TOF-MS: 402.16 [M + H], 424.28 [M + Na]. 2.5. PCEMA, PMMA, and Various P(CEMA-MMA). PCEMA, PMMA, and various P(CEMA-MMA) were prepared by conventional radical polymerization as follows. MMA (0, 1.0, 1.5, 2.0, or 3.0 mmol), CEMA (3.0, 2.0, 1.5, 1.0, or 0 mmol), and AIBN (12.4 mg) were dissolved in THF (5 mL) to prepare the prepolymerization mixture for PCEMA, P(CEMA-MMA)-67, P(CEMA-MMA)-50, P(CEMAMMA)-33, and PMMA, respectively. The solution was added to a Schlenk flask. After N2−vacuum cycles, conventional radical polymer-
particles. Therefore, the development of a simple and robust route to the synthesis of hollow polymer particles is considerably challenging. Here, we report a serendipitously discovered new route to the fabrication of polymeric hollow particles based on shellselective crosslinking of homogeneous polymer particles, a “self-templating” strategy. In this approach, homogeneous polymer particles containing linear polymers bearing photoresponsive cinnamoyl groups in their side chains were first synthesized, and photoirradiation is a trigger for crosslinking because of the [2 + 2] dimerization reaction between cinnamoyl groups. We discovered that the crosslinking occurred only at the shell region in the homogeneous polymer particles, and further removal of the non-crosslinked core region resulted in hollow particle formation (Scheme 1). The key for the shell-selective crosslinking is discussed in detail, and it is clarified that the remaining non-crosslinked cinnamoyl groups with a high absorption coefficient in the crosslinked shell region played a crucial role in shielding the core region from photoirradiation. To the best of our knowledge, this approach via the “self-templating” pathway for formation of hollow polymer particles has not been reported to date. It has great potential to be used as an efficient fabrication route in the synthesis of hollow polymer particles because it has high mass productivity and high simplicity attributable to the non-use of additional sacrificial template particles and highly toxic solvent, and it can be applied for the synthesis of advanced functional polymer particles because of its high robustness.
2. EXPERIMENTAL SECTION 2.1. Materials. Methyl methacrylate (MMA), 2-hydroxyethyl methacrylate (HEMA), citric acid, sodium carbonate (Na2CO3), sodium chloride (NaCl), tetrahydrofuran (THF), diethyl ether, dichloromethane (DCM), and sodium dodecyl sulfate (SDS) were purchased from Nacalai Tesque Co. (Kyoto, Japan). Acryloyl chloride, methacryloyl chloride, and 4-amino fluorescein were purchased from Tokyo Chemical Industries (Tokyo, Japan). 2,2′-Azobisisobutyronitrile (AIBN), cinnamoyl chloride, and 4-cyano-4(phenylcarbonothioylthio)pentanoic acid were purchased from Sigma-Aldrich (USA). Methanol (MeOH), acetone, hexane (Hex), toluene, pyridine, and poly(vinyl alcohol) (PVA; degree of polymerization 1000, degree of saponification 88%) were purchased from Wako Pure Chemical Co. Ltd (Osaka, Japan). Deionized water used was obtained from a Millipore Milli-Q purification system. 2.2. Characterization. FT-IR measurements were carried out by the KBr method using an infrared spectrograph (Varian 660 KU-IR, Agilent Inc., California, USA). The particle sizes and morphologies B
DOI: 10.1021/acs.langmuir.6b02295 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir ization was performed at 60 °C for 16 h. The polymers were obtained by reprecipitation using Hex and dried in vacuo. In reversible addition−fragmentation chain transfer (RAFT) polymerization, 4-cyano-4-(phenylcarbonothioylthio)pentanoic acid was used as the RAFT agent. Typically, MMA (500 mg, 5.0 mmol), CEMA (1300 mg, 5.0 mmol), AIBN (1.64 mg, 10 μmol), and 4-cyano4-(phenylcarbonothioylthio)pentanoic acid (14 mg, 50 μmol) were dissolved in THF (5 mL), and RAFT polymerization was carried out at 60 °C for 12 h. 2.6. PCEMA, PMMA, and P(CEMA-MMA) Particles by Solvent Evaporation. The typical procedure is outlined below. P(CEMAMMA)-50 (50 mg) was dissolved in toluene (1 mL), and the polymer solution was mixed with PVA aqueous solution (0.0067 wt %, 25 mL). Toluene droplets containing P(CEMA-MMA)-50 were prepared by a homogenizer (POLYTRON PT 1600 E, Kinematica Inc.) with 10 000 rpm for 1 min. After the preparation of droplets, toluene was evaporated at ambient temperature with a gentle stirring. PCEMA, PMMA, and other P(CEMA-MMA) particles were prepared by the same procedure. Submicrometer-sized P(CEMA-MMA)-50 particles were also prepared by the solvent evaporation method, where the droplets were prepared using a homogenizer at 20 000 rpm for 2 min, followed by ultrasonication for 30 min. The average particle size and the coefficient of variation (CV) were measured from SEM images (the number of measured particles: ca. 200). 2.7. Hollow Polymer Particles Fabricated by Photoinduced Crosslinking and Subsequent Removal of Non-crosslinking Polymers. The typical photoinduced crosslinking and capsule formation procedures are as follows. P(CEMA-MMA)-50 particles (solid contents: 2 mg/mL) dispersed in PVA aqueous solution (3 mL) were placed in a 5 mL vial. Photoirradiation (7 cm from the dispersion top surface, 254 nm, 2000 μW/cm2, GL 15, Toshiba) was performed for the P(CEMA-MMA)-50 dispersion for 16 h at ambient temperature. After photoinduced crosslinking, the P(CEMA-MMA)50 particles were washed with water (once) and with THF (thrice) to remove PVA and non-crosslinked polymers, respectively. The gel fraction was calculated using the remaining crosslinked polymers by gravimetry. 2.8. Determination of Gel Fraction in the Polymer Film States. P(CEMA-MMA)-50 films were prepared on a glass Petri dish using the solvent cast method. We dissolved P(CEMA-MMA)-50 (100, 200, or 300 mg) in DCM, and the solvent was placed on a glass Petri dish (area of base: 18, 51, and 127 cm2) and was slowly evaporated. Photoinduced crosslinking for the P(CEMA-MMA)-50 films prepared on the glass Petri dish was carried out. To determine the time for achieving the maximum gel fraction, the irradiation time was set from 15 to 200 min. The crosslinked polymer was corrected by filtration, where the non-crosslinked polymer was removed by washing with DCM. The gel fraction was calculated from the initial and crosslinked polymer weights, and the effective film thickness was calculated from the initial film thickness and the gel fraction.
Figure 1. 1H NMR (a), FT-IR (b), and UV−vis spectra (c) for P(CEMA) (red), P(CEMA-MMA)-67 (orange), P(CEMA-MMA)-50 (green), P(CEMA-MMA)-33 (blue), and PMMA (black). Time course of gel fraction via photoirradiation for P(CEMA) (red), P(CEMA-MMA)-67 (orange), P(CEMA-MMA)-50 (green), P(CEMA-MMA)-33 (blue), and PMMA (black) (d).
in P(CEMA-MMA)-50 were also confirmed by 1H NMR (δ: 7−8 ppm derived from the benzene ring, as shown in Figure 1a) and FT-IR (750 and 1650 nm−1 derived from the benzene ring and CC bonds, respectively, as shown in Figure 1b), and the 280 nm UV−vis absorption derived from cinnamoyl groups was also confirmed (Figure 1c). P(CEMA-MMA)-50 particles were prepared by solvent evaporation (Scheme S1),35−38 where the organic solvent (toluene) dissolving the polymers was homogenized in an aqueous solution containing PVA as a stabilizer. This resulted in polymer-dissolving toluene droplets, followed by slow organic solvent evaporation, allowing the formation of spherical P(CEMA-MMA)-50 particles. The synthesis of P(CEMAMMA)-50 particles was confirmed by an optical micrograph and SEM. Micrometer-sized spherical particles were obtained [number-average particle size (dn): 8.05 μm, coefficient of variation (CV): 32% calculated from SEM images], as shown in Figure 2. For a posteriori crosslinking of cinnamoyl groups, photoirradiation (254 nm) of P(CEMA-MMA)-50 dispersion in water was performed for 16 h at room temperature, and the particle morphology of P(CEMA-MMA)-50 was spherical as well as that before photoirradiation with slightly smaller particle size because of the density increase after the crosslinking (dn: 7.79 μm, CV: 28% by SEM). The particle crosslinking which was qualitatively checked from their insolubility in THF (THF can completely dissolve P(CEMA-MMA)-50 particles before photoirradiation). The time course of the a posteriori crosslinking of P(CEMA-MMA)-50 was investigated by gravimetry after washing out non-crosslinked polymers with THF, resulting in an increased gel fraction as time progressed. It is worth noting that the thermal stimulus was not effective on the crosslinking of the polymers, which was confirmed by the complete dissolution of the heat-treated P(CEMA-MMA)-50 film at 90 °C for 16 h. The ratio finally reached ca. 68% after 16
3. RESULTS AND DISCUSSION 3.1. Hollow Polymer Particle Formation via “SelfTemplating” Shell-Selective Crosslinking. The cinnamoyl group was selected as a posteriori photoinduced crosslinking moiety, and CEMA was prepared as a posteriori crosslinking monomer by a conjugation reaction between HEMA and cinnamoyl chloride (Figure S1). CEMA exhibits UV−vis absorption with a maximum absorption at 280 nm, which is derived from the cinnamoyl moiety, and electron excitation due to light absorption is the trigger for the a posteriori crosslinking (Figure 1).34 CEMA was copolymerized with MMA [initial ratio of MMA and CEMA was 50:50 (mol/mol)] by free radical polymerization with AIBN to obtain linear P(CEMAMMA)-50 [Mn: 12 000; Mw/Mn: 2.8 measured by GPC; CEMA content measured by 1H NMR was ca. 46% (Figure S2)], wherein we selected nonliving radical polymerization because of the high simplicity of this procedure. The cinnamoyl groups C
DOI: 10.1021/acs.langmuir.6b02295 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
Figure 2. Chemical structure of P(CEMA-MMA) (a) and SEM images of P(CEMA-MMA)-50 particles before photoirradiation (b), after photoirradiation (c), and after THF treatment [high magnification (d), low magnification (e)].
Figure 3. CLSM images of P(CEMA-MMA)-50 particles before photoirradiation (a,d), P(CEMA-MMA)-50 particles after photoirradiation (b,e), and P(CEMA-MMA)-50 particles treated by THF after photoirradiation (c,f). Optical image (a−c); fluorescent image (d−f).
50:50:1 (mol/mol/mol)], P(CEMA-MMA-FAm)-50), was synthesized by conventional free radical copolymerization. P(CEMA-MMA-FAm)-50 particles also had a spherical morphology (Figure S4), and the particles had sufficient FL intensity to observe the particle morphology using CLSM. Before photoirradiation, the particles had a solid morphology, which was not changed even after photoinduced crosslinking. On the other hand, the particle was changed to a hollow morphology after the removal of non-crosslinked polymer from the crosslinked P(CEMA-MMA-FAm)-50 particles (Figure 3). As a result, spherical hollow particles were prepared in a dispersed system by photoinduced crosslinking only in the particle shell region. 3.2. Effect of Photocrosslinking Group Content. To investigate the effect of CEMA content on photoinduced crosslinking, we prepared four additional polymer mixtures of P(CEMA-MMA) with different CEMA contents [initial molar ratio of CEMA and MMA, 67:33 (P(CEMA-MMA)-67) or 33:67 (P(CEMA-MMA)-33) (mol/mol)] and PCEMA or
h of irradiation (Figure 1d). The washed P(CEMA-MMA)-50 particles were observed by SEM, where the SEM sample was prepared from particle dispersion in THF, yielding nonspherical particles with a dimple, indicating hollow morphology (Figure 2). It is worth noting that the nonspherical morphology was formed during the SEM sample preparation because the spherical P(CEMA-MMA)-50 particles can be observed from the same sample as when the SEM sample was prepared after redispersion in water of THF-washed P(CEMA-MMA)-50 particles (Figure S3). To obtain further evidence of the hollow morphology formation after a posteriori crosslinking and removal of noncrosslinked polymers, we observed particle morphology by CLSM. To give a fluorescent property to the particles, we synthesized fluorescein acrylamide (FAm) (maximum emission wavelength: 514 nm; excitation wavelength: 490 nm) by a coupling reaction of 4-amino-fluorescenin and acryloyl chloride. The synthesis was confirmed by 1H NMR. Poly(CEMA-coMMA-co-FAm) [initial ratio of CEMA, MMA, and FAm: D
DOI: 10.1021/acs.langmuir.6b02295 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
Figure 4. SEM images of P(CEMA) (a−c), P(CEMA-MMA)-67 (d−f), P(CEMA-MMA)-50 (g−i), P(CEMA-MMA)-33 particles (j−l), after a posteriori photoinduced crosslinking for 1 h (a, d, g, and j), 2 h (b, e, h, and k), and 16 h (c, f, i, and l), followed by THF washing.
The UV−vis absorption at 280 nm also increased gradually as the CEMA molar ratio increased (Figure 1c). Micrometersized spherical PMMA (dn: 8.91 μm, CV: 31%), P(CEMAMMA)-33 (dn: 8.58 μm, CV: 29%), P(CEMA-MMA)-67 (dn: 8.96 μm, CV: 30%), and PCEMA (dn: 8.24 μm, CV: 38%) particles were successfully synthesized (Figure S5). After photoirradiation, all polymer particles kept their spherical morphology (Figure S6). In the case of PMMA particles, the particles were completely dissolved in THF even after 16 h of photoirradiation (Figure 1d); on the other hand, the other particles [PCEMA and the three P(CEMA-MMA) particles] were crosslinked by photoirradiation, and the dimpled spherical particles were obtained after THF washing (Figure 4). These results indicate that a posteriori crosslinking for the polymers was caused by photoinduced crosslinking of cinnamoyl groups in the polymer side chains, and the final morphology after photoirradiation and subsequent solvent treatment for PCEMA and P(CEMA-MMA) particles was that of hollow spheres. Moreover, an SEM image of broken P(CEMA-MMA) particles after a posteriori crosslinking followed by THF treatment supported that the morphology was hollow (Figure S7). Time courses of the gel fraction with photoirradiation were performed for all particles, and the gel fraction increased as the CEMA molar content increased (Figure 1d). Important differences on the gel fractions for PCEMA and the three
PMMA homopolymers. The incorporated CEMA ratio increased as the initial CEMA molar ratio in the radical (co)polymerizations increased. The peak intensities at 750 and 1650 cm−1 derived from the benzene ring and the CC bond in the cinnamoyl group increased as the CEMA content increased as indicated by FT-IR, and the ratio of the CEMA content in the polymers was estimated from the peak intensities derived from the CEMA (δ = 4.2−4.5 ppm derived from −CHCH− in the cinnamoyl group) and MMA (δ = 3.6 ppm derived from C(O)OCH3) content in the 1H NMR spectra, resulting in 0, 29, 46, 62, and 100%, respectively, for PMMA, P(CEMA-MMA)-33, P(CEMA-MMA)-50, P(CEMA-MMA)67, and PCEMA (Figure 1a,b). From the GPC measurements, the Mn values of PCEMA, P(CEMA-MMA)-67, P(CEMAMMA)-50, P(CEMA-MMA)-33, and PMMA were 17 900 (Mw/Mn: 5.4), 14 400 (Mw/Mn: 3.4), 12 000 (Mw/Mn: 2.8), 10 400 (Mw/Mn: 2.6), and 6400 (Mw/Mn: 1.9). The degrees of polymerization of these polymers calculated from the molecular weights and the CEMA/MMA molar ratio were similar for these polymers [PCEMA: 69, P(CEMA-MMA)-67: 72, P(CEMA-MMA)-50: 69, P(CEMA-MMA)-33: 71, and PMMA: 64]. The high Mw/Mn value with the high CEMA content is unclear, but these results may be caused by a side reaction (chain transfer reaction) and decreasing termination reaction constant (kt) because of its diffusion control property. E
DOI: 10.1021/acs.langmuir.6b02295 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
Figure 5. Molecular weight distributions of P(CEMA-MMA)-50 bearing different molecular weights prepared by RAFT polymerization at 60 °C using 4-cyano-4-(phenylcarbonothioylthio)pentanoic acid and AIBN as the chain transfer agent and initiator, respectively. Mn value: 4300 (blue) and 8000 (red) (a). Time course of the gel fraction of P(CEMA-MMA)-50 prepared by RAFT polymerization: Mn value: 4300 (close) and 8000 (open) (b).
phenyl groups formed after photoinduced crosslinking of cinnamoyl groups in the shell region absorbed the photoirradiation (254 nm). To confirm this hypothesis, we prepared poly(CEMA-co-phenyl methacrylate) [P(CEMA-PhMA)] comprising phenyl groups in its side chains by conventional radical copolymerization of CEMA and phenyl methacrylate (PhMA) with AIBN, where the initial ratio of CEMA to PhMA was 50:50 (mol/mol) (P(CEMA-PhMA)-50) (Mn: 12 900; Mw/Mn: 2.6 measured by GPC), and its UV absorption was derived from the cinnamoyl groups (Figure S8). P(CEMA-PhMA)-50 particles (dn: 7.07 μm, CV: 26%) were also prepared by solvent evaporation, and we investigated the gel fractions of the P(CEMA-PhMA)-50 particles by the same procedure. The gel fractions were measured by gravimetry after THF washing of non-crosslinked polymers, where the nonspherical particles were obtained (Figure S9), and the trends in the gel fraction time course of the P(CEMA-PhMA)-50 particles were similar to the case of P(CEMA-MMA)-50 particles (Figure S10). These results indicate that the light absorption derived from the phenyl groups remaining after photoinduced crosslinking of cinnamoyl groups in polymer side chains was not important for the formation of the non-crosslinked core region. From the results, we hypothesize that the possible UV (280 nm) absorption functional groups in the crosslinked shell region were non-crosslinked cinnamoyl groups. To confirm that cinnamoyl groups remained after photoirradiation, FT-IR spectra were recorded for the crosslinked P(CEMA-MMA) particles by photoirradiation after THF washing of noncrosslinked polymers. The peak of the CC bond derived from the remaining non-crosslinked cinnamoyl group is shown in Figure 6. The result is understandable because the dimerization of the cinnamoyl groups can happen if two cinnamoyl groups are placed in close proximity. Moreover, the UV−vis spectra of CEMA and PhMA at the same concentration indicate that the phenyl groups have a significantly smaller absorption coefficient when compared with that of the cinnamoyl groups. The molar absorption coefficients of CEMA at 280 nm (ε280 nm) was 17 000 [L mol−1 cm−1], and the value was clearly higher than that for PhMA (ε280 nm = 250 [L mol−1 cm−1]). From these results, the remaining cinnamoyl group had a crucial role in inhibiting the UV proceeding toward the particle core region. 3.5. Effect of Particle Size. Our proposed method for the formation of hollow capsule particles by photoirradiation has cinnamoyl groups remaining on the polymer side chains even
P(CEMA-MMA) particles were observed in the early stage of crosslinking; that is, the gel fraction increased rapidly with the photoirradiation time as the CEMA content increased in the obtained polymers. The phenomenon was comprehensible because the crosslinking between the polymer chains occurred with a higher probability as the CEMA content increased. The particulate morphology was not confirmed after 1 h of photoirradiation for P(CEMA-MMA)-33 and P(CEMAMMA)-50 particles because of their low gel fraction. On the other hand, the dimpled particulate morphology remained unchanged for PCEMA and P(CEMA-MMA)-67 (Figure 4). It is worth noting that the results for PCEMA particles also indicate that shell-selective crosslinking was not caused by the selective distribution of PCEMA-rich polymers at the particle surface in the case of P(CEMA-MMA) particles. 3.3. Effect of Chain Length. Polymer chain length is another possible parameter in the fabrication of hollow polymer particles. If the chain length is short, the probability of intermolecular crosslinking in the polymer particles should be very small, leading to a slow crosslinking rate. To investigate the effect of the molecular weight of P(CEMA-MMA)-50 on the gel fraction, we prepared two kinds of P(CEMA-MMA)-50 with different molecular weights with a narrow molecular weight distribution by RAFT polymerization39,40 and reversible-deactivation radical polymerization (RDRP),41−45 with 4cyano-4-(phenylcarbonothioylthio)pentanoic acid and AIBN as the chain transfer agent and initiator, respectively (Figure 5). For the low-molecular-weight P(CEMA-MMA)-50 (Mn: 4300, Mw/Mn: 1.4), the gel fraction was 0% until 9 h (Figure 5). After 16 h of photoirradiation, the ratio reached 33%. On the other hand, use of the higher-molecular-weight P(CEMA-MMA)-50 (Mn: 8000, Mw/Mn: 1.4) led to a higher gel fraction; that is, the gel fraction reached 83% after 16 h of irradiation. These results indicate that the molecular weight of P(CEMA-MMA)-50 greatly affected the gel fraction, and the greater gel fraction in the higher molecular weight polymer should be caused by the high probability of the intermolecular crosslinking between the different polymer chains. In addition, it is clarified that the crosslinked polymer particles could be obtained in
