Article pubs.acs.org/Macromolecules
Effects of Chain-Ends Tethering on the Crystallization Behavior of Poly(ε-caprolactone) Confined in Lamellar Nanodomains Shintaro Nakagawa,† Takashi Ishizone,† Shuichi Nojima,*,† Kohei Kamimura,‡ Kazuo Yamaguchi,‡,¶ and Seiichi Nakahama¶ †
Department of Organic and Polymeric Materials, Graduate School of Science and Engineering, Tokyo Institute of Technology, 2-12-1-H-125 Ookayama, Meguro-ku, Tokyo 152-8552, Japan ‡ Department of Chemistry, Faculty of Science, Kanagawa University, 2941 Tsuchiya, Hiratsuka-shi, Kanagawa 259-1293, Japan ¶ Research Institute for Photofunctionalized Materials, Kanagawa University, 2941 Tsuchiya, Hiratsuka-shi, Kanagawa 259-1293, Japan S Supporting Information *
ABSTRACT: We have investigated the effects of chain-ends tethering at nanodomain interfaces on the crystallization behavior of poly(ε-caprolactone) (PCL) chains spatially confined in lamellar nanodomains. The PCL chains tethered at both chain-ends (T2-PCL), one chain-end (T1-PCL), and no chain-end (i.e., PCL homopolymers) (T0-PCL) were prepared from polystyrene-block-PCL-blockpolystyrene (PS-b-PCL-b-PS) triblock copolymers having a photocleavable onitrobenzyl (ONB) group at either or both of block junctions. The lamellar nanodomain was provided by the microphase separation of PS-b-PCL-b-PS copolymers, in which T2-PCL was inherently confined. Subsequently, the ONB group was cleaved by the irradiation of ultraviolet light to yield T1-PCL or T0-PCL from T2-PCL without perturbing the lamellar nanodomain. T2-PCL showed an extremely low melting temperature and crystallinity as compared with T1-PCL and T0-PCL, indicating that the crystallization was significantly affected by chain mobility imposed by chain tethering at both ends to yield immature PCL crystallites. All the PCL chains exhibited a sigmoidal time evolution of crystallinity during isothermal crystallization, implying that their crystallization was driven by a conventional nucleation and growth mechanism. However, the crystallization rate at a same temperature depended largely on the state of chain tethering; T0-PCL crystallized moderately faster than T1-PCL, whereas T2PCL crystallized extraordinarily slower than T1-PCL. The crystallization kinetics of these PCL chains confined in lamellar nanodomains is qualitatively discussed.
1. INTRODUCTION
nanopore diameter D ranging from 15 to 110 nm and found that their crystallization behavior depended strongly on D. The influences of space confinement on the crystallization have also been investigated using crystalline−amorphous diblock copolymers with a glassy amorphous block,27−35 where the crystalline block is confined in nanodomains formed by the microphase separation of block copolymers. Sun et al., for example, reported that the nucleation mechanism and crystal orientation of poly(ε-caprolactone) (PCL) blocks confined in lamellar nanodomains of PCL-block-poly(4-vinylpyridine) depended significantly on the nanodomain size.30 However, the crystallization of block chains in this case is affected not only by the space confinement but also by chainend tethering at nanodomain interfaces, which we call chain conf inement. We have recently examined the effects of chain confinement using PCL-block-polystyrene (PCL-b-PS) diblock copolymers having a photocleavable o-nitrobenzyl (ONB)
Multicomponent polymeric materials having nanometer-scale domains (nanodomains) have been attracting much attention over the past years because their properties can be tuned with greater flexibility. Considering the fact that crystallization can drastically change the physical properties of materials, it is of great importance to understand the crystallization of polymer chains spatially confined in various nanodomains. Such crystallization is known to be unique as compared with that of bulk homopolymers and has been extensively reviewed.1−6 We term the spatial restriction imposed by nanodomains as space conf inement. Several experimental methods have been employed to prepare crystalline homopolymers confined in nanodomains, including droplets,7−10 micelles in solution,3,11,12 anodic aluminum oxide (AAO) templates,13−21 or nanolayered films.22−26 The crystallization of such confined homopolymers is known to be significantly affected by the shape and size of nanodomains. For example, Woo et al.15 investigated the crystallization behavior of polyethylene homopolymers infiltrated in nanopores of AAO templates as a function of © XXXX American Chemical Society
Received: August 6, 2015 Revised: September 9, 2015
A
DOI: 10.1021/acs.macromol.5b01744 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules
Figure 1. Schematic illustration of PCL chains tethered at both chain-ends (T2-PCL, a), one chain-end (T1-PCL, b), and no chain-end (T0-PCL, c) all confined in an identical lamellar nanodomain sandwiched between PS layers. The black circles represent the chain-ends tethering at nanodomain interfaces.
Scheme 1. Schematic Illustration Showing the Sample Preparation Method Used in This Studya
a
SCS and SC′S form a lamellar microphase-separated structure in which both of PCL chain-ends are tethered at nanodomain interfaces (a and a′). After UV irradiation, both of PCL chain-ends in SCS are cleaved to yield T0-PCL (b), while one of PCL chain-ends in SC′S is cleaved to yield T1PCL (c).
group between two blocks.36−40 The most profound effect of chain confinement was observed when PCL chains (PCL blocks or PCL homopolymers) were confined in an identical lamellar nanodomain with the layer thickness dPCL of 8.7 nm.40 That is, the PCL homopolymer showed a sigmoidal time evolution of crystallinity Xc with a finite induction time, whereas the PCL block showed an abrupt increase in Xc with no induction time at the initial stage of crystallization. This result indicates that chain confinement practically controls the crystallization mechanism through changing the mobility of PCL chains. Our previous study described above focused on the difference in the crystallization behavior between PCL blocks (PCL chains tethered at one chain-end) and PCL homopolymers (PCL chains tethered at no chain-end). Considering the observed significant effects of chain confinement on the
crystallization behavior, more interesting crystallization behavior can be expected for PCL blocks with both chain-ends tethered at nanodomain interfaces, since the mobility of PCL chains would decrease further. On the basis of this idea, the present study aims to elucidate the effects of chain tethering at either or both of chain-ends on the crystallization behavior of PCL chains confined in an identical lamellar nanodomain. We hereafter denote PCL chains tethered at both chain-ends, one chain-end, and no chain-end as T2-PCL, T1-PCL, and T0-PCL, respectively (Figure 1). The crystallization behavior of crystalline blocks with both chain-ends tethered at nanodomain interfaces has been studied by several groups41−44 using ABA amorphous−crystalline−amorphous triblock copolymers. However, they did not consider the effect of chain mobility imposed by chain confinement on the crystallization. In this study, we designed novel model systems in which T2-PCL, T1-PCL, and B
DOI: 10.1021/acs.macromol.5b01744 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules Table 1. Characterization of the Samples Used in This Study Mn (g mol−1) a
sample code
PCL
SCS SC′S S/C/S S/C′S
27 400 27 400 27 400 27 400
PSb 13 500 13 500 13 500 13 500
× × × ×
2 2 2 2
totalc
Mw/Mnb
ϕPCLd
dPCL(nm)e
pf
54 400 54 400 − −
1.04 1.05 − −
0.50 0.50 0.50 0.50
10.9 10.9 11.0 11.0
− − 0.79 0.95
a Determined by 1H-NMR. bDetermined by GPC calibrated using PS standards. cCalculated from Mns of PCL and PS blocks. dVolume fraction of PCL chains at 100 °C calculated from specific volumes and Mns of PCL and PS chains. ePCL layer thickness calculated from SAXS long period and fPCL. fMaximum photocleavage yield obtained by GPC.
columns (Shodex K-803L, Showa Denko, Japan), a UV absorbance detector (Waters 2487, Waters, USA), and a RI detector (Waters 2410, Waters, USA), and the eluent was chloroform. The system was calibrated using polystyrene standards. The UV absorbance at 254 nm (mainly arising from absorption by phenyl rings of PS repeating units) was analyzed to evaluate the photocleavage yield as a function of total UV irradiation energy. 2.4. Small-Angle X-ray Scattering (SAXS). The microphaseseparated structure formed in the samples was examined by smallangle X-ray scattering (SAXS) using synchrotron radiation. The experiments were performed at Photon Factory in High Energy Accelerator Research Organization, Tsukuba, Japan, with a small-angle X-ray equipment for solution (SAXES) installed at beamline BL-10C. The X-ray wavelength λ used was 0.1488 nm, and the sample-todetector distance was ca. 2.6 m. 2-d scattering patterns were collected by a pixel-array detector (PILATUS3 2M, DECTRIS Ltd., Switzerland) consisting of 1475 × 1679 pixels with 172 × 172 μm in each size.49,50 Typically 10 sample films (total thickness: ca. 0.5 mm) were layered to obtain optimal scattering intensity. The concentric 2-d patterns obtained were first circularly averaged to get 1-d profiles and then corrected for the background scattering and absorption. The scattered intensity was plotted against wavenumber s (=2 sin θ/λ; 2θ = scattering angle). Silver behenate51 was used for the calibration of SAXES. 2.5. Differential Scanning Calorimetry (DSC). The melting temperature Tm and crystallinity Xc of PCL chains in the samples were determined by differential scanning calorimetry (DSC) using Diamond DSC (PerkinElmer, USA) equipped with a liquid nitrogen cooling system. Sample films (ca. 5 mg in weight) were sealed in an aluminum pan for measurements. The sample was first annealed at 60 °C for 3 min to melt PCL crystals and erase previous thermal histories, and then quenched to the selected crystallization temperature Tc at −500 °C min−1 and held there for the prescribed crystallization time t. Finally, it was heated to 60 °C at 10 °C min−1 to observe an endothermic peak due to the melting of PCL crystals formed. Xc was ° ), where ΔHm is the heat of fusion calculated as ΔHm/(wPCLΔHm ° that of a perfect PCL obtained from the endothermic peak area, ΔHm crystal (= 135 J g−147), and wPCL the weight fraction of PCL chains in the sample. Tm was defined as the temperature at the top of endothermic peaks. The time evolution of Xc in SCS and SC′S, in which T2-PCL crystallized below room temperature, was investigated by repeating the above temperature program with various Tc and t. The PCL chains in S/C/S and S/C′S crystallized above room temperature, so that the time evolution of normalized crystallinity was conveniently pursued by time-resolved IR measurements, as described in the next section, and only final Xc and Tm at various Tc were evaluated by DSC. 2.6. Infrared Spectroscopy (IR). The time evolution of normalized crystallinity X̃ c of PCL chains, defined as Xc at t divided by Xc at t = ∞, during isothermal crystallization was measured by timeresolved infrared spectroscopy (IR) using FT-IR 6200 spectrometer (JASCO, Japan) with a spectral resolution of 4 cm−1. Measurements were performed in transmission mode for the sample films placed on a silicon substrate. This technique was not available below room temperature due to instrumental limitations. Figure S1 in the Supporting Information shows representative time-resolved IR results
T0-PCL were spatially confined in an identical lamellar nanodomain. Two lamella-forming PS-b-PCL-b-PS triblock copolymers having ONB groups at either or both of block junctions were synthesized, and T2-PCL confined in the lamellar nanodomain formed by the microphase separation was converted to T1-PCL or T0-PCL using the photocleavage of ONB groups induced by the irradiation of ultraviolet (UV) light.
2. EXPERIMENTAL SECTION 2.1. Strategy for Preparing Model Systems. Scheme 1 illustrates the strategy to prepare the model systems used in this study. We start from two symmetric amorphous−crystalline− amorphous PS-b-PCL-b-PS triblock copolymers where the volume fraction of PCL blocks in each copolymer is nearly 0.5. The constituent PS (or PCL) block has a constant molecular weight in two PS-b-PCL-b-PS copolymers, so that they form an identical lamellar microphase-separated structure in the melt. Two triblock copolymers differ only in that one has photocleavable o-nitrobenzyl (ONB) groups45,46 in both of two block junctions (denoted SCS), whereas the other in either of two block junctions (SC′S). SCS and SC′S contain T2-PCL (a and a′ in Scheme 1) confined in a microphase separated structure. Then they are irradiated with ultraviolet (UV) light to cleave ONB groups, and consequently T2PCL in SCS and SC′S is transformed to T0-PCL (b) and T1-PCL (c), respectively. SCS and SC′S after a sufficient UV irradiation are designated as S/C/S and S/C′S, respectively. Since the glass transition temperature Tg of PS homopolymers (ca. 100 °C) is significantly higher than the melting temperature Tm of PCL homopolymers (ca. 60 °C), the microphase-separated structure formed in molten SCS and SC′S is expected to be frozen by the vitrification of PS chains at temperatures around Tm of PCL chains. 2.2. Samples and Sample Preparation. Synthesis methods, chemical structures, and photocleavage reactions of PS-b-PCL-b-PS triblock copolymers are illustrated in Supporting Information (section 1 and Scheme S1). The results of molecular characterization are summarized in Table 1. The volume fraction of PCL blocks ϕPCL, calculated from the weight fraction and specific volumes of amorphous PCL47 and PS homopolymers48 at 100 °C, is equal to 0.5, so that SCS and SC′S are expected to form a lamellar microphase-separated structure. The PS-b-PCL-b-PS copolymers dissolved in toluene were cast on an ethylene−tetrafluoroethylene copolymer sheet, and dried at 60 °C in atmospheric pressure overnight, followed by drying at 100 °C under reduced pressure for 3 h to ensure the complete removal of toluene. The films were circular disk in shape having a diameter of ca. 5 mm and thickness of ca. 50 μm. The UV light from high-pressure mercury lamp (USH-500SC, USHIO Inc., Japan) was filtered with aqueous solutions of CuSO4 in Pyrex cell to remove low-wavelength light, and irradiated on amorphous SCS and SC′S films at room temperature to prepare S/C/S and S/C′S. UV light with an intensity of 4.0 J s−1 cm−2 was applied for 2400 s. 2.3. Gel Permeation Chromatography (GPC). Gel permeation chromatography (GPC) was used to characterize the samples and examine the photocleavage process. The GPC system consisted of two C
DOI: 10.1021/acs.macromol.5b01744 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules for S/C/S during isothermal crystallization at Tc = 32.5 °C. Several absorption peaks increased with the progress of crystallization, which could be attributed to the formation of PCL crystals.52 An absorption band located at the wavenumber of 1295 cm−1 was chosen to evaluate X̃ c, since it was relatively strong and isolated from other PCL and PS absorption bands. X̃ c was calculated by dividing the peak area at t by that after the complete crystallization.
3. RESULTS AND DISCUSSION 3.1. Photocleavage Behavior. The photocleavage behavior of SCS and SC′S was investigated by GPC measurements as a function of total UV irradiation energy Et (proportional to irradiation time). Figure 2 shows the GPC chromatograms
Figure 3. Mole fractions of T2-PCL (solid squares), T1-PCL (open squares), and T0-PCL (open circles) in SCS (a) and SC′S (b) plotted against total UV energy irradiated. Dashed curves are just a guide to the eye.
respectively, after a sufficient UV irradiation (>2.5 kJ cm−2). On the other hand, the mole fraction of T2-PCL decreases and that of T1-PCL increases monotonically with increasing Et in the case of SC′S, and the presence of T0-PCL cannot be observed during the photocleavage process (Figure 3b). The final mole fractions of T0-PCL, T1-PCL, and T2-PCL in SC′S are 0, 0.95, and 0.05, respectively. These results agree well with those expected from the molecular design of SCS and SC′S. On the basis of Figure 3, we prepared S/C/S and S/C′S from SCS and SC′S by irradiating UV with the intensity of 4.0 J s−1 cm−2 for 2400 s (i.e., 9.6 kJ cm−2). 3.2. Microphase-Separated Structure. Figure 4 shows the SAXS curves of crystallized SCS, S/C/S, SC′S, and S/C′S. The SAXS curves of crystallized SCS and SC′S (thick solid curves in Figure 4) exhibit several higher order peaks, the positions of which exactly correspond to a ratio of 1:2:3, indicating the formation of a well-defined lamellar microphaseseparated structure consisting of alternating layers of PS and PCL chains. The same primary peak position and similar shape of the SAXS curves suggest that an identical lamellar microphase-separated structure is formed in SCS and SC′S. Moreover, the SAXS curves of amorphous SCS and SC′S at 60 °C (dotted curves in Figure 4) also show first and second order peaks at exactly the same positions as the crystallized samples. These results indicate that the lamellar microphase-separated structure formed in amorphous states is completely preserved after the crystallization of PCL chains. The only noticeable change caused by the melting of PCL crystals is the disappearance of the third order peak. This is because the electron density difference between PS48 (332 e cm−3) and amorphous PCL47 (337 e cm−3) at 100 °C is significantly smaller than that between PS and crystallized PCL (393 e cm−3 for a perfect crystal). It is confirmed from above results that all the PCL chains (i.e., T0-PCL, T1-PCL, and T2-PCL) in SCS, SC′S, S/C/S, and S/C′S are confined in the same lamellar nanodomain. The PCL layer thickness dPCL was calculated from the primary peak position and the volume fraction of PCL ϕPCL at 60 °C to be 10.9 nm for both SCS and SC′S (Table 1). It is necessary to examine whether the melting and/or crystallization of PCL chains could lead to a destruction of the lamellar nanodomain formed in S/C/S and S/C′S, since these
Figure 2. GPC chromatograms of SCS (a) and SC′S (b) after the irradiation of UV light with 4.0 J s−1 cm−2 in intensity for 2400 s (thick solid curves). The chromatograms before UV irradiation are also shown with dotted curves for comparison. Thin solid curves in each panel are the decomposed peaks corresponding to PS-b-PCL-b-PS triblocks, PCL-b-PS diblocks, dimerized PS homopolymers, and PS homopolymers, respectively (in increasing order of elution time).
measured by a UV absorption detector at the wavelength of 254 nm for SCS and SC′S before and after UV irradiation. The unimodal and narrow GPC peak of SCS and SC′S is observed before UV irradiation (dotted curves), whereas it splits into several peaks after UV irradiation (thick solid curves). These peaks can be assigned as unreacted PS-b-PCL-b-PS, PCL-b-PS, dimerized PS homopolymers, and PS homopolymers in the increasing order of elution time, the last three of which have been generated by the photocleavage of SCS or SC′S. The dimerization of PS homopolymers has also been observed in our previous study,39 probably due to a coupling reaction between nitroso moieties produced by the photocleavage of ONB groups.53 PCL homopolymers appearing after the photocleavage are not detected because PCL chains are insensitive to the wavelength used in the GPC measurement. For a quantitative evaluation of the photocleavage reaction during UV irradiation, we calculated the mole fractions of T0PCL, T1-PCL, and T2-PCL in SCS and SC′S from the area ratio of four peaks assigned above, the method of which is precisely described in the Supporting Information (section 2). Figure 3 shows the mole fractions of T0-PCL, T1-PCL, and T2-PCL for SCS (a) and SC′S (b) plotted against Et. The mole fraction of T2-PCL decreases and that of T0-PCL increases monotonically with increasing Et irradiated to SCS, whereas the mole fraction of T1-PCL first increases from 0 but decreases asymptotically after taking a maximum at Et ∼ 0.3 kJ cm−2 (Figure 3a). This behavior can be qualitatively explained by a two-step reaction kinetics (i.e., PS-b-PCL-b-PS → PS + PCL-bPS and PCL-b-PS → PCL + PS), where the rate constants of each step are comparable. The final mole fractions of T0-PCL, T1-PCL, and T2-PCL in SCS reach 0.79, 0.16, and 0.05, D
DOI: 10.1021/acs.macromol.5b01744 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules
Figure 5. Melting temperature (a) and crystallinity (b) of PCL chains in SCS (closed circles) and SC′S (closed squares), S/C/S (open circles), and S/C′S (open squares) plotted against crystallization temperature. Broken lines are just a guide for the eye.
Figure 4. SAXS curves of SCS, S/C/S, SC′S, and S/C′S. The dotted curves were measured when PCL chains were amorphous and the solid curves after sufficient crystallization of PCL chains to achieve better contrast between PCL and PS layers. The vertical broken line indicates primary peak positions.
which is usually observed for the crystallization of bulk homopolymers. Xc of T1-PCL and T0-PCL in S/C′S and S/ C/S also increases steadily with increasing Tc in a way similar to Tm. On the other hand, Tm and Xc of T2-PCL in SCS and SC′S are significantly lower than those of T1-PCL and T0-PCL and independent of Tc in the temperature range investigated. Furthermore, Tm and Xc of T2-PCL in SCS are almost the same as those in SC′S, indicating that both T2-PCLs exist under the similar environment. It is reasonable to assume that the crystallites of PCL chains confined in lamellar nanodomains cannot grow thicker than the layer thickness of nanodomains if the crystal stems take a random or perpendicular orientation with respect to nanodomain interfaces. Even if the crystal stems are oriented parallel to nanodomain interfaces, the crystallite thickness would be limited since the nanodomain interfaces are not infinitely straight. Therefore, a limiting thickness of PCL crystallites, which is mainly determined by the layer thickness of lamellar nanodomains, can be assumed. Moreover, it is plausible that the PCL repeating units near tethered chain-ends cannot participate in crystallization, which further reduces the limiting thickness of crystallites. Therefore, the maximum crystallite thickness of T2-PCL confined in lamellar nanodomains may be extremely small to yield low Tm and Xc, as shown in Figure 5. In addition, the crystallite thickness of T2-PCL already reaches its limiting value even at the lowest Tc investigated (∼0 °C), which would render Tm of T2-PCL in SCS and SC′S constant at higher Tc. If this speculation is true, T0-PCL in S/C/S and T1PCL in S/C′S should also have the Tc range in which Tm is constant and independent of Tc, but the crystallization at such high Tc would require a long time to complete. 3.4. Crystallization Behavior of PCL Chains. The normalized crystallinity of PCL chains X̃ c during isothermal crystallization was pursued using either time-resolved IR (for Tc > 20 °C) or DSC (for Tc < 10 °C) measurements. Representative X̃ c curves are shown in Figure 6 as a function of t, where X̃ c changes sigmoidally in all samples, implying that the crystallization process is controlled by a conventional
two samples are in fact a polymer blend mainly consisting of two components. Figure S2 in the Supporting Information shows the change in the primary peak intensity and long period after repeated crystallization at 5 °C and melting at 60 °C as a function of repeating cycles. The long period of S/C/S and S/ C′S does not change at all after 5 crystallization/melting cycles, indicating the preservation of the lamellar nanodomains by the crystallization and melting of PCL chains. However, a small decrease in the primary peak intensity of S/C/S can be detected with increasing the crystallization/melting cycles, suggesting that the initial lamellar nanodomain in S/C/S might be slightly deformed after several crystallization/melting cycles. In fact, when the same isothermal crystallization program is repeatedly applied to S/C/S, the crystallinity evaluated by DSC gradually increases, probably reflecting some deformation of lamellar nanodomains. On the other hand, S/C′S shows almost no reduction in primary peak intensity, indicating that the initial lamellar nanodomain in S/C′S is stable against the crystallization and melting of PCL chains. This is also supported by a good reproducibility of crystallinity in repeated DSC measurements. A possible reason for the difference in the thermal stability between lamellar nanodomains of S/C/S and S/C′S is that T0-PCL in S/C/S can move rather freely while T1-PCL in S/C′S is not allowed to diffuse because one of chain-ends is always tethered at nanodomain interfaces. Considering the above results, fresh S/C/S was used for every isothermal crystallization experiment by DSC and IR. 3.3. Melting Behavior of PCL Chains. Figure 5 shows the melting temperature Tm and crystallinity Xc of PCL chains in SCS, SC′S, S/C′S, and S/C/S after isothermal crystallization as a function of crystallization temperature Tc. Tm and Xc of T1PCL and T0-PCL in S/C′S and S/C/S increase linearly with increasing Tc. Since Tm reflects the crystallite thickness as indicated by the Gibbs−Thomson equation, Figure 5a suggests that T1-PCL and T0-PCL form thicker crystallites at higher Tc, E
DOI: 10.1021/acs.macromol.5b01744 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules
Figure 6. Time evolution of normalized crystallinity of PCL chains in SCS (solid circles) and SC′S (solid squares) during isothermal crystallization at 5.0 °C (a), S/C′S at 21.5 °C (b), and S/C/S at 32.5 °C (c). Data in panel a are obtained by DSC measurements, whereas those in panels b and c by time-resolved IR measurements. Dashed curves are just a guide to the eye.
reduction in mobility of PCL chains caused by chain-end tethering, though block chains tethered at nanodomain interfaces is known to take a slightly stretched conformation which may facilitate the crystallization. In order to quantitatively evaluate the effects of chain tethering at either or both of chain-ends on the crystallization behavior, Tc at t1/2 = 10 min was defined as T10 and estimated by interpolating the data in Figure 7 using an Arrhenius-type expression ln t1/2 = ln A + B/Tc, where A and B are constants.28 T10 can be regarded as the temperature with a constant crystallization rate. Figure 8 shows T10 plotted against dPCL,
nucleation and growth mechanism. This result is in contrast with our previous study,40 where T1-PCL confined in the lamellar nanodomain with dPCL = 8.9 nm shows an abrupt increase in X̃ c at the initial stage of crystallization followed by an asymptotic increase, while T0-PCL in the identical lamellar nanodomain shows a sigmoidal increase in X̃ c. This difference in the crystallization behavior is attributed to a change in nucleation mechanism from homogeneous to heterogeneous nucleation by the disappearance of chain-end tethering. A similar phenomenon is not observed for T2-PCL that should experience much more chain confinement than T1-PCL. The strong effects of chain confinement on the nucleation mechanism in lamellar nanodomains with dPCL = 8.9 nm might be significantly reduced in larger nanodomains with dPCL = 10.9 nm. It should also be noted that Figure 6a shows one master curve, again confirming that T2-PCL chains in SCS and SC′S are confined in an identical lamellar nanodomain. For a quantitative analysis on the crystallization rate, the crystallization half-time t1/2 was determined by fitting X̃ c data to the Avrami equation log(− ln (1 − X̃ c)) = n log t + log K, where n is the Avrami exponent and K is a rate constant. The Avrami equation is known to hold only in the initial stage of crystallization, therefore only the data with X̃ c < 1/2 were used for this analysis. t1/2 was defined as the crystallization time at X̃ c = 1/2 and calculated from n and K by t1/2 = (ln 2/K) 1/n. Figure 7 shows t1/2−1, a measure for the crystallization rate, plotted against Tc. It is clear that T2-PCL needs fairly lower Tc than T1-PCL and T0-PCL to crystallize in an experimentally accessible time scale. This fact arises from a significant
Figure 8. T10, defined as the temperature at which the crystallization half-time is equal to 10 min, plotted against the PCL layer thickness dPCL for T2-PCL (solid square), T1-PCL (solid circles), and T0-PCL (open circles). The values of T10 at dPCL = 8.7 and 15.8 nm are taken from the data in our previous study.40 Dashed curves are just a guide to the eye.
together with the data from our previous study.40 For PCL chains confined in lamellar nanodomains with 10.9 nm ≤ dPCL ≤ 15.8 nm, T10 of T1-PCL is 8−10 °C lower than that of T0PCL. On the other hand, T10 of T2-PCL confined in lamellar nanodomains with dPCL = 10.9 nm decreases by 27 °C from that of T0-PCL, which is interesting by considering that T10 of T0-PCL decreases by 13 °C as dPCL is reduced from 15.8 to 8.7 nm. Therefore, it is concluded that chain tethering at both ends is more effective on decreasing the crystallization rate than the sharp reduction of lamellar nanodomain size (15.8 nm →8.7 nm). The Tc dependence of crystallization rates for bulk homopolymers is often interpreted using the Hoffman− Lauritzen theory,54,55 and the spherulitic growth rate G is written as
Figure 7. Inverse of crystallization half-time of T2-PCL in SCS (solid circles) and SC′S (solid squares), T1-PCL in S/C′S (open squares), and T0-PCL in S/C/S (open circles) plotted against crystallization temperature. Dashed curves are just a guide to the eye.
⎛ −U * ⎞ ⎛ −K *Tm° ⎞ G ∝ exp⎜ ⎟ exp⎜ ⎟ ⎝ Tc − T∞ ⎠ ⎝ TcΔT ⎠ F
(1)
DOI: 10.1021/acs.macromol.5b01744 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules where U* is the constant associated with the energy barrier to a short-range transport of polymer chains across the phase boundary, T∞ the temperature at which the transport ceases (usually T∞ = Tg − 30 °C), K* the constant reflecting the difficulty in forming a critical nucleus in the secondary nucleation process, Tm ° the equilibrium melting temperature (= 78.9 °C for PCL homopolymers56), and ΔT = Tm ° − Tc. The Tc dependence of t1/2−1 shown in Figure 7 can be qualitatively discussed on the basis of eq 1, assuming that t1/2−1 is proportional to G. The increase in exp(−U*/(Tc − T∞)) with increasing Tc is insignificant when Tc − T∞ is sufficiently large, hence the difference in the Tc dependence of t1/2−1 among T0-PCL, T1-PCL, and T2-PCL will mainly arise from the variation of K*. Figure 9 shows the Tc dependence of
crystallization of T1-PCL was moderately slower than that of T0-PCL, whereas T2-PCL crystallized extremely slow as compared with T0-PCL. This difference in the crystallization rate was qualitatively discussed from the viewpoint of energy barrier to form a critical nucleus in the secondary nucleation process.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.macromol.5b01744. Sample synthesis and calculation of mole fractions (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*(S.No.) E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The SAXS measurement was performed under the approval of Photon Factory Advisory Committee (No. 2014G011). This work was supported by JSPS Grant-in-Aid for Scientific Research (Grant No. 15J10502). S. Nakagawa gratefully acknowledges a JSPS Research Fellowship for Young Scientists.
Figure 9. exp(−K*T°m/(TcΔT)) in eq 1 plotted against crystallization ° = 78.9 °C and K* = 730, 870, or 1060 K were used. temperature. Tm
■
° /(TcΔT)) calculated using K* = 730, 870, or 1060 exp(−K*Tm K. The values of K* and the range of vertical axis were arbitrarily chosen so that the horizontal positions of resulting curves roughly coincide with those of t1/2−1 vs Tc curves shown in Figure 7. Since higher K* implies the deceleration of secondary nucleation, we speculate that the chain confinement makes the formation of secondary nuclei more difficult (i.e., increasing K*) by reducing the mobility of PCL chains to decrease the overall crystallization rate.
REFERENCES
(1) Lin, M.-C.; Nandan, B.; Chen, H.-L. Soft Matter 2012, 8, 7306− 7322. (2) He, W.-N.; Xu, J.-T. Prog. Polym. Sci. 2012, 37, 1350−1400. (3) Michell, R. M.; Blaszczyk-Lezak, I.; Mijangos, C.; Müller, A. J. Polymer 2013, 54, 4059−4077. (4) Takeshita, H.; Shiomi, T.; Takenaka, K.; Arai, F. Polymer 2013, 54, 4776−4789. (5) Huang, S.; Jiang, S. RSC Adv. 2014, 4, 24566−24583. (6) Nakagawa, S.; Marubayashi, H.; Nojima, S. Eur. Polym. J. 2015, 70, 262−275. (7) Massa, M. V.; Dalnoki-Veress, K. Phys. Rev. Lett. 2004, 92, 255509. (8) Massa, M. V.; Lee, M. S. M.; Dalnoki-Veress, K. J. Polym. Sci., Part B: Polym. Phys. 2005, 43, 3438−3443. (9) Massa, M. V.; Carvalho, J. L.; Dalnoki-Veress, K. Phys. Rev. Lett. 2006, 97, 247802. (10) Carvalho, J. L.; Dalnoki-Veress, K. Phys. Rev. Lett. 2010, 105, 237801. (11) Zhang, Q.; Wang, M.; Wooley, K. L. Curr. Org. Chem. 2005, 9, 1053−1066. (12) Su, Y.; Liu, G.; Xie, B.; Fu, D.; Wang, D. Acc. Chem. Res. 2014, 47, 192−201. (13) Steinhart, M.; Goering, P.; Dernaika, H.; Prabhukaran, M.; Goesele, U.; Hempel, E.; Thurn-Albrecht, T. Phys. Rev. Lett. 2006, 97, 027801. (14) Wu, H.; Wang, W.; Yang, H.; Su, Z. Macromolecules 2007, 40, 4244−4249. (15) Woo, E.; Huh, J.; Jeong, Y. G.; Shin, K. Phys. Rev. Lett. 2007, 98, 136103. (16) Shin, K.; Woo, E.; Jeong, Y. G.; Kim, C.; Huh, J.; Kim, K.-W. Macromolecules 2007, 40, 6617−6623. (17) Duran, H.; Steinhart, M.; Butt, H.-J.; Floudas, G. Nano Lett. 2011, 11, 1671−1675. (18) Maiz, J.; Martin, J.; Mijangos, C. Langmuir 2012, 28, 12296− 12303.
4. CONCLUSIONS The crystallization behavior of poly(ε-caprolactone) (PCL) chains tethered at both chain-ends (T2-PCL), one chain-end (T1-PCL), and no chain-end (T0-PCL) all confined in an identical lamellar nanodomain was investigated using DSC or IR to clarify the effects of chain confinement on the crystallization. Polystyrene-block-PCL-block-polystyrene (PS-bPCL-b-PS) triblock copolymers having o-nitrobenzyl (ONB) groups at either or both of block junctions were employed to prepare T0-PCL, T1-PCL, and T2-PCL. The melting temperature Tm and crystallinity Xc of T0-PCL and T1-PCL after isothermal crystallization at crystallization temperature Tc increased steadily with increasing Tc, as is usually observed in bulk homopolymers, whereas those of T2-PCL were constant irrespective of Tc, suggesting that the crystallite size was significantly restricted by chain tethering at both ends. The time evolution of normalized crystallinity for T0-PCL, T1-PCL, and T2-PCL during isothermal crystallization showed a sigmoidal change with a finite induction time, indicating that the crystallization was controlled by a conventional nucleation and growth mechanism. Furthermore, the chain confinement significantly affected the crystallization rate. That is, the G
DOI: 10.1021/acs.macromol.5b01744 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules (19) Noirez, L.; Stillings, C.; Bardeau, J.-F.; Steinhart, M.; Schlitt, S.; Wendorff, J. H.; Pepy, G. Macromolecules 2013, 46, 4932−4936. (20) Martin, J.; Nogales, A.; Mijangos, C. Macromolecules 2013, 46, 7415−7422. (21) Suzuki, Y.; Duran, H.; Akram, W.; Steinhart, M.; Floudas, G.; Butt, H.-J. Soft Matter 2013, 9, 9189−9198. (22) Wang, H.; Keum, J. K.; Hiltner, A.; Baer, E.; Freeman, B.; Rozanski, A.; Galeski, A. Science 2009, 323, 757−760. (23) Wang, H.; Keum, J. K.; Hiltner, A.; Baer, E. Macromolecules 2010, 43, 3359−3364. (24) Flores, A.; Arribas, C.; Fauth, F.; Khariwala, D.; Hiltner, A.; Baer, E.; Baltá-Calleja, F. J.; Ania, F. Polymer 2010, 51, 4530−4539. (25) Lai, C.; Ayyer, R.; Hiltner, A.; Baer, E. Polymer 2010, 51, 1820− 1829. (26) Wang, H.; Keum, J. K.; Hiltner, A.; Baer, E. Macromol. Rapid Commun. 2010, 31, 356−361. (27) Zhu, L.; Cheng, S. Z. D.; Calhoun, B. H.; Ge, Q.; Quirk, R. P.; Thomas, E. L.; Hsiao, B. S.; Yeh, F.; Lotz, B. J. Am. Chem. Soc. 2000, 122, 5957−5967. (28) Shiomi, T.; Tsukada, H.; Takeshita, H.; Takenaka, K.; Tezuka, Y. Polymer 2001, 42, 4997−5004. (29) Loo, Y.-L.; Register, R. A.; Ryan, A. J.; Dee, G. T. Macromolecules 2001, 34, 8968−8977. (30) Sun, Y.-S.; Chung, T.-M.; Li, Y.-J.; Ho, R.-M.; Ko, B.-T.; Jeng, U.-S.; Lotz, B. Macromolecules 2006, 39, 5782−5788. (31) Vasilev, C.; Reiter, G.; Pispas, S.; Hadjichristidis, N. Polymer 2006, 47, 330−340. (32) Cai, T.; Qian, Y.; Ma, Y.; Ren, Y.; Hu, W. Macromolecules 2009, 42, 3381−3385. (33) Chung, T.-M.; Wang, T.-C.; Ho, R.-M.; Sun, Y.-S.; Ko, B.-T. Macromolecules 2010, 43, 6237−6240. (34) Lin, M.-C.; Wang, Y.-C.; Chen, H.-L.; Müller, A. J.; Su, C.-J.; Jeng, U.-S. J. Phys. Chem. B 2011, 115, 2494−2502. (35) Chen, L.; Jiang, J.; Wei, L.; Wang, X.; Xue, G.; Zhou, D. Macromolecules 2015, 48, 1804−1812. (36) Nojima, S.; Ohguma, Y.; Namiki, S.; Ishizone, T.; Yamaguchi, K. Macromolecules 2008, 41, 1915−1918. (37) Nojima, S.; Ohguma, Y.; Kadena, K.-i.; Ishizone, T.; Iwasaki, Y.; Yamaguchi, K. Macromolecules 2010, 43, 3916−3923. (38) Nakagawa, S.; Kadena, K.-i.; Ishizone, T.; Nojima, S.; Shimizu, T.; Yamaguchi, K.; Nakahama, S. Macromolecules 2012, 45, 1892− 1900. (39) Nakagawa, S.; Tanaka, T.; Ishizone, T.; Nojima, S.; Kakiuchi, Y.; Yamaguchi, K.; Nakahama, S. Macromolecules 2013, 46, 2199−2205. (40) Nakagawa, S.; Tanaka, T.; Ishizone, T.; Nojima, S.; Kamimura, K.; Yamaguchi, K.; Nakahama, S. Polymer 2014, 55, 4394−4400. (41) Weimann, P. A.; Hajduk, D. A.; Chu, C.; Chaffin, K. A.; Brodil, J. C.; Bates, F. S. J. Polym. Sci., Part B: Polym. Phys. 1999, 37, 2053− 2068. (42) Takeshita, H.; Ishii, N.; Araki, C.; Miya, M.; Takenaka, K.; Shiomi, T. J. Polym. Sci., Part B: Polym. Phys. 2004, 42, 4199−4206. (43) Xu, J.-T.; Yuan, J.-J.; Cheng, S.-Y. Eur. Polym. J. 2003, 39, 2091− 2098. (44) Yu, P.-Q.; Yan, L.-T.; Chen, N.; Xie, X.-M. Polymer 2012, 53, 4727−4736. (45) Yamaguchi, K.; Kitabatake, T.; Izawa, M.; Fujiwara, T.; Nishimura, H.; Futami, T. Chem. Lett. 2000, 29, 228−229. (46) Zhao, H.; Sterner, E. S.; Coughlin, E. B.; Theato, P. Macromolecules 2012, 45, 1723−1736. (47) Crescenzi, V.; Manzini, G.; Calzolari, G.; Borri, C. Eur. Polym. J. 1972, 8, 449−463. (48) Richardson, M. J.; Savill, N. G. Polymer 1977, 18, 3−9. (49) Hijikawa, R.; Huang, L.; Kiyofuji, G.; Marubayashi, H.; Nojima, S. Polymer 2014, 55, 6960−6966. (50) Gondo, S.; Osawa, S.; Marubayashi, H.; Nojima, S. Polym. J. 2015, 47, 556−563. (51) Huang, T. C.; Toraya, H.; Blanton, T. N.; Wu, Y. J. Appl. Crystallogr. 1993, 26, 180−184.
(52) Duan, Y.; Zhang, J.; Chang, H.; Yan, S.; Yang, C.; Takahashi, I.; Ozaki, Y. Macromolecules 2010, 43, 5315−5322. (53) Guillier, F.; Orain, D.; Bradley, M. Chem. Rev. 2000, 100, 2091− 2157. (54) Hoffman, J. D.; Lauritzen, J. I. J. Res. Natl. Bur. Stand., Sect. A 1961, 65A, 297−336. (55) Mandelkern, L. Crystallization of Polymers, 2nd ed., vol. 2; Cambridge University Press: Cambridge, U.K., 1976. (56) Chen, H.-L.; Li, L.-J.; Ou-Yang, W.-C.; Hwang, J. C.; Wong, W.Y. Macromolecules 1997, 30, 1718−1722.
H
DOI: 10.1021/acs.macromol.5b01744 Macromolecules XXXX, XXX, XXX−XXX
