conference papers
Effect of lithium trifluoromethanesulfonate on the phase diagram of a liquid-crystalline amphiphilic diblock copolymer
aGraduate School of Engineering, Tokyo Metropolitan University, Japan, bChemical Resources Laboratory, Tokyo Institute of Technology, Japan, cCREST-JST, Japan, and dFaculty of Urban Environmental Science, Tokyo Metropolitan University, Japan
*Correspondence e-mail: yoshida-hirohisa@c.metro-u.ac.jp
Phase transitions and nanometre-scale ordered structures of a binary system of a liquid-crystalline amphiphilic diblock copolymer, poly(ethylene oxide)-b-poly{11-[4-(4-butylphenylazo)phenoxy]undecyl methacrylate} [PEOm-b-PMA(Az)n, where m and n are the degrees of polymerization of the PEO and PMA(Az) domains, respectively], and lithium trifluoromethanesulfonate (LiCF3SO3) were investigated by differential scanning calorimetry and small-angle X-ray scattering (SAXS). PEO114-b-PMA(Az)51 formed a highly ordered hexagonally packed PEO cylinder structure in the temperature range below 393 K and transformed to a body-centred-cubic structure in the isotropic state above 393 K. The PEO114-b-PMA(Az)51/LiCF3SO3 systems with various LiCF3SO3 concentrations (molar ratio 0 < LiCF3SO3/EO = fLi < 1) formed the hexagonally packed cylinder structure at room temperature. From the effects of LiCF3SO3 concentration on the phase transitions, the size and the order of the hexagonally packed cylinder structure, it was found that PEO114-b-PMA(Az)51 and LiCF3SO3 formed a complex efficiently at a molar equivalent of three ethylene oxide repeating units per LiCF3SO3 unit. The ordering of the hexagonally packed cylinder structure decreased with increasing LiCF3SO3 concentration and the radius of the PEO cylinder evaluated by SAXS profile fitting increased from 2.7 to 8.3 nm. For the PEO114-b-PMA(Az)51/LiCF3SO3 system with fLi = 1, the hexagonally packed cylinder structure remained even in the isotropic state because the PEO volume fraction (ϕPEO) increased from ϕPEO = 0.06 (fLi = 0) to ϕPEO = 0.23 (fLi = 1) on the formation of the LiCF3SO3/PEO complex.
Keywords: amphiphilic diblock copolymers; small-angle X-ray scattering; lithium trifluoromethanesulfonate; nanocomposites.
1. Introduction
Nanometre-scale electron-conducting materials have been paid much attention because they are expected to provide a breakthrough in lithography limits. These materials are also interesting in terms of special physical properties like quantum size effects. Block copolymers are used as templates for making nanoscale objects in many studies (Lazzari & López-Quintela, 2003) because they can provide various nanoscale ordered structures which are easily modified by changing their fractions, degree of polymerization and chemical structures (Hamley, 1998).
There are various methods for analyzing nanostructures, e.g. atom-force microscopy (AFM), transmission and scanning electron microscopies (TEM and SEM) and small-angle scattering (SAS), which have different advantages. SAS is a powerful tool for analyzing average structures of materials because the irradiated volume is much larger than the measurement area for AFM or TEM (Hamley & Castelletto, 2004). Other advantages of SAS are that it can be used for various samples such as liquids, solutions, films, powders or supercritical fluids (Arai et al., 2003) and can be coupled with various fields, for instance, temperature (Yoshida et al., 1995), pressure (Kawabata et al., 2004), magnetic, mechanical and shear fields (Kato et al., 2004). Many numerical analysis methods for analyzing SAS data for micelle solutions (Lemmich et al., 1996) and block copolymers (Hashimoto et al., 1994) have also been reported. The combination of the use of brilliant synchrotron X-radiation with thermal analysis makes it possible to analyze the dynamics of phase transitions for organic and polymeric materials (Yamada et al., 2005).
Recently we reported the syntheses of liquid-crystalline amphiphilic diblock copolymers, poly(ethylene oxide)-b-poly{11-[4-(4-butylphenylazo)phenoxy]undecyl methacrylate} [PEOm-b-PMA(Az)n, where m and n are the degrees of polymerization of the PEO and PMA(Az) domains, respectively] (Tian et al., 2002). PEOm-b-PMA(Az)n copolymers (Fig. 1) form a highly ordered hexagonally packed PEO cylinder structure selectively over a wide range of volume fraction because the liquid crystallinity of the PMA(Az) domains stabilizes a cylinder structure rather than a sphere structure (Yoshida et al., 2004). We also reported the nanometre-size control of the hexagonally packed cylinder structure of PEOm-b-PMA(Az)n by blending a PEO homopolymer (Jung et al., 2005) and PEOm-b-PMA(Az)n (Jung & Yoshida, 2006).
PEO has potential as an electron-conducting material because it forms complexes with alkali metal ions (Rhodes & Frech, 2001). Since PEO–metal ion complexes have ionic conductivity in the solid state, they have been investigated as polymer electrodes in rechargeable batteries (Tominaga et al., 2003). The combination of an anisotropic nanometre-scale structure of block copolymers and PEO–metal ion complexes is expected to produce new ionic conductive materials. We have reported the anisotropic ionic conductivity of the PEOm-b-PMA(Az)n complex with lithium trifluoromethanesulfonate (LiCF3SO3) (Li et al., 2005). The structure and the phase transitions of the PEOm-b-PMA(Az)n complex with LiCF3SO3 are also important in considering the anisotropic conductivity. In this study, the effects of LiCF3SO3 on the phase diagram of PEO114-b-PMA(Az)51 were investigated by differential scanning calorimetry (DSC) and small-angle X-ray scattering (SAXS).
2. Samples and experiments
2.1. Samples
PEO114-b-PMA(Az)51 copolymer was synthesized by atomic transfer radical polymerization as reported elsewhere (Tian et al., 2002). The molecular weight dispersion was 1.23 determined by gel permeation chromatography. LiCF3SO3 supplied from Aldrich was used without further purification. LiCF3SO3 and PEO114-b-PMA(Az)51 were dissolved in distilled tetrahydrofuran separately. Then an appropriate amount of LiCF3SO3 solution was added to the PEO114-b-PMA(Az)51 solution. The mixed solution was stirred at room temperature for 6 h. After removing the solvent, the complex samples were annealed at 413 K for 16 h and cooled down to room temperature. The LiCF3SO3 concentration is indicated by the molar ratio of LiCF3SO3 to one repeating unit of PEO, LiCF3SO3/EO = fLi. The fLi range in this study was from 0 to 1 (Table 1).
|
2.2. Experiments
2.2.1. Differential scanning calorimetry
DSC measurements were performed using a DSC 6200 calorimeter (Seiko Instruments Inc.) equipped with an electric cooling control apparatus (Haake EK90/MT) over the temperature range between 223 and 473 K. The scanning rate was 10 K min−1 in a nitrogen flow atmosphere (40 ml min−1) and the sample weight used for the DSC was about 3 mg.
2.2.2. Small-angle X-ray scattering
SAXS measurements were performed using the synchrotron X-radiation facility of the 2.5 GeV storage ring at BL-10C at the Photon Factory, High Energy Accelerator Research Organization (Tsukuba, Japan). Monochromatic X-rays with a wavelength λ of 0.1488 nm selected by double Si crystals were used for SAXS measurements. Two kinds of optics which covered 0.1 < q < 3 nm−1 and 0.06 < q < 1.5 nm−1 were used, where q = (4π/λ)sin θ and θ is half of the diffraction angle. SAXS profiles at various temperatures were obtained using the simultaneous DSC instrument (Yoshida et al., 1995). Powder samples were covered with thin aluminium foil and put in an aluminium vessel. The exposure time was 60–300 s.
3. Results and discussion
3.1. Phase transitions
Fig. 2 shows DSC heating curves for the LiCF3SO3/PEO114-b-PMA(Az)51 systems with various fLi. In the case of PEO114-b-PMA(Az)51 (fLi = 0), four endothermic phase transitions were observed around 313, 333, 373 and 393 K. These phase transitions correspond to melting of the PEO domain (~313 K), melting of the azobenzene moieties (~333 K), the liquid-crystalline transition from smectic C to smectic A (~373 K, indicated by an arrow in the inset) and the isotropic transition (~393 K) (Yoshida et al., 2004; Watanabe et al., 2006). With increasing LiCF3SO3 concentration the endothermic peak of the PEO melting decreased and disappeared above fLi = 1.3 × 10−1. LiCF3SO3/PEO114-b-PMA(Az)51 above fLi = 2.5 × 10−1 had a new endothermic peak around 433 K. This temperature is similar to the melting temperature of complexes of PEO with LiCF3SO3 (Rhodes & Frech, 2001). According to their report, one LiCF3SO3 forms a complex with three ethylene oxide (EO) units in the PEO/LiCF3SO3 complex. LiCF3SO3/PEO114-b-PMA(Az)51 with fLi = 2.5 × 10−1, in which the number of PEO repeating units per LiCF3SO3 was four, gave the largest fusion enthalpy of all the LiCF3SO3/PEO114-b-PMA(Az)51 complexes, therefore the endothermic peak of LiCF3SO3/PEO114-b-PMA(Az)51 with fLi = 2.5 × 10−1 and 1.0 observed around 433 K was assigned as melting of the LiCF3SO3/PEO complex in the LiCF3SO3/PEO114-b-PMA(Az)51 system. The melting peak of the LiCF3SO3/PEO complex in LiCF3SO3/PEO114-b-PMA(Az)51 with fLi = 1 was broad and small. This fact suggested that the excess amount of LiCF3SO3 disturbed the crystallization of the PEO/LiCF3SO3 complex. Below fLi = 1.3 × 10−1, in which the number of PEO repeating units per LiCF3SO3 was more than 20, the melting peak of PEO in the LiCF3SO3/PEO114-b-PMA(Az)51 systems became broad and small with increasing LiCF3SO3 concentration. The existence of a small amount of LiCF3SO3 relative to the PEO repeating unit inhibited the crystallization of the PEO domain in PEO114-b-PMA(Az)51. The molar equivalent was important for forming crystals of PEO/LiCF3SO3 complex.
On the other hand, phase transitions concerning the hydrophobic PMA(Az) domain in PEO114-b-PMA(Az)51 showed different tendencies with LiCF3SO3 concentration. Both the melting of azobenzene moieties and the isotropic transition were scarcely influenced by the addition of LiCF3SO3 for the LiCF3SO3/PEO114-b-PMA(Az)51 systems below fLi = 5.0 × 10−2. However, in the case of fLi = 1.3 and 2.5 × 10−1, the isotropic transition temperature (TIso) and enthalpy (ΔHIso) shifted slightly to higher temperatures and decreased compared with the isotropic transition of PEO114-b-PMA(Az)51 (fLi = 0). The solid PEO/LiCF3SO3 complex in the PEO domain, having a higher melting temperature than TIso, reduced the molecular mobility of the PMA(Az) domain in the isotropic state, therefore TIso increased. The decrease of ΔHIso indicated the restricted molecular motion of the liquid-crystalline parts by the solid PE/LiCF3SO3 complex and disorder in the smectic layer of the PMA(Az) domain.
3.2. Nanostructures
3.2.1. PEO114-b-PMA(Az)51
Fig. 3 shows SAXS profiles of PEO114-b-PMA(Az)51 (fLi = 0) at room temperature and 423 K (isotropic state). The nanostructure at room temperature was assigned as the hexagonally packed PEO cylinder structure, because the ratio of each diffraction peak to the first-order peak (q*) was q*: (3q*)0.5: (4q*)0.5: (7q*)0.5. At q = 2.0 nm−1, another peak corresponding to smectic layers of PMA(Az) domains was observed at room temperature. At 423 K, where the PMA(Az) domain was in the isotropic state, the peak coming from the smectic layers disappeared. On the other hand, three peaks, shown by arrows in Fig. 3, were observed, although these peak intensities were much weaker than those at room temperature. Since the ratios of the diffraction peaks were q*: (2q*)0.5: (3q*)0.5, the nanostructure at 423 K was assigned as having a body-centered-cubic (b.c.c.) structure. The hexagonally packed PEO cylinder structure of PEO114-b-PMA(Az)51 changed to the b.c.c. structure at the isotropic transition.
The radius of the PEO cylinder (R) was estimated by the paracrystal model (Hashimoto et al., 1994). In this model, the observed scattering intensity from a hexagonally packed cylinder structure having a random orientation [I(q)] is given by
where is a scattering function having uniaxial orientation. is given by
where f and Z indicate the form factor and the structure factor, respectively. The form factor f(q) is given by
where C is a constant including the electron density difference and scattering volume, J1 is the first-order Bessel function, R is the radius of the cylinder, is the mean radius, σR is the standard deviation of R and σs is the characteristic thickness of a interface. σs is related to the thickness of the interface between each domain (t) by t/(2π)0.5 (Shibayama & Hashimoto, 1986). The structure factor Z is given by
where a and Δa are the lattice constant of the hexagonal structure and the paracrystal distortion factor, respectively. For the profile-fitting calculation, t was fixed to 0.5 nm because the interface thickness was estimated as two or three repeating units of PMA(Az) from the thermodynamic relationship between the isotropic transition entropy and the degree of polymerization of PMA(Az) (Yamada et al., 2004). Since the effect of σR on the profile-fitting calculation was small, σR/R was also fixed to 0.05. Four fitting parameters, a, Δa, R and C, were used for the profile -fitting calculation. The best-fit results for PEO114-b-PMA(Az)51 are superimposed on the SAXS profile observed at room temperature in Fig. 3. The fitting results are listed in Table 1. The calculated result did not match the experimental data because the scattering from large grains overlapped in the low-q region and the intensity was not high enough in the high-q region. The PEO volume fraction (ϕPEO) was estimated using
ϕPEO of PEO114-b-PMA(Az)51 was 0.06. In the case of a linear diblock copolymer having a step-function-type electron density profile, the border between the hexagonally packed cylinder structure and the b.c.c. structure is ϕA = 0.12, where ϕA is the minor component (Lodge & Muthukumar, 1996). Although the ϕPEO value of PEO114-b-PMA(Az)51 (0.06) was lower than the border ϕPEO value between the hexagonally packed cylinder structure and the b.c.c. structure (0.12) (Lodge & Muthukumar, 1996), the hexagonally packed cylinder structure was stable at room temperature for PEO114-b-PMA(Az)51. For PEO114-b-PMA(Az)n, the hexagonally packed cylinder structure is stable when the PMA(Az) domain is in the smectic phase to compensate for the conformational entropy loss of the smectic phase formation by the increase of freedom in the interface between the hydrophilic and hydrophobic domains (Yoshida et al., 2004). This assumption conformed to the observation of the transformation from the hexagonally packed cylinder structure to the b.c.c. structure at the isotropic transition. Anthamatten & Hammond determined phase diagrams of side-chain liquid-crystalline copolymers by theoretical calculation. When the side-chain liquid crystal took a planar anchoring on the interface between each block, the side-chain liquid-crystal copolymer preferred to form the cylinder structure over a wide range of volume fractions compared with linear block copolymers (Anthamatten & Hammond, 2001).
3.2.2. LiCF3SO3/PEO114-b-PMA(Az)51 systems
Fig. 4 shows SAXS profiles of LiCF3SO3/PEO114-b-PMA(Az)51 systems having various fLi values at room temperature. In the q range between 0.1 and 1 nm−1, the diffraction peaks corresponding to the hexagonally packed cylinder structure were observed for all systems. The diffraction peak of the smectic layers also appeared at q = 2 nm−1. Although the peak position of the smectic layers scarcely changed below fLi = 2.5 × 10−1, the half bandwidth increased with increasing LiCF3SO3 concentration. In particular, above fLi = 1.3 × 10−1 the smectic layer peak became broad and the position of the peak shifted to lower q for the system with fLi = 1.0. When the LiCF3SO3 concentration was close to the molar equivalent value, the formation of the complex between three EO units and one LiCF3SO3 occurred efficiently. The complex of PEO/LiCF3SO3 acted as a cross linkage between the PEO chains and decreased the molecular mobility of the PEO. Then the hydrophilic PEO cylinder changed to a viscous cylinder and induced disordering of the smectic layers in the hydrophobic PMA(Az) domain. The SAXS experiment results showed a good agreement with the DSC results shown in Fig. 2.
The nanostructures of LiCF3SO3/PEO114-b-PMA(Az)51 were evaluated by SAXS profile fitting using equation (1). The fitted results for LiCF3SO3/PEO114-b-PMA(Az)51 systems with fLi = 2.5 × 10−1 and 1.0 are shown in Fig. 5, and the values obtained are listed in Table 1. These fitted results also did not match the experimental results in the low-q region for the same reason as for PEO114-b-PMA(Az)51 (fLi = 0). Since the PEO complexes with LiCF3SO3 had a larger volume than PEO, the radius of the PEO cylinder (R) increased with increasing LiCF3SO3 concentration. Concurrently, LiCF3SO3 acted as a cross-linker between the PEO chains and restricted the molecular motion of the PEO, therefore the disorder (Δa/a) also increased with increasing LiCF3SO3 concentration.
Fig. 6 shows SAXS profiles for the LiCF3SO3/PEO114-b-PMA(Az)51 systems with fLi = 2.5 × 10−2 (ϕPEO = 0.09) and fLi = 1 (ϕPEO = 0.23) at 423 and 473 K, respectively, where both systems are in the isotropic state. The nanostructure of the system with fLi = 2.5 × 10−2 in the isotropic state was the b.c.c. structure, the same as that of PEO114-b-PMA(Az)51 at 423 K (Fig. 2). The ϕPEO values of the LiCF3SO3/PEO114-b-PMA(Az)51 systems used in this study, from 0.06 to 0.23 as shown in Table 1, covered the border ϕPEO value (0.12) between the hexagonally packed cylinder structure and the b.c.c. structure expected by theoretical calculations for linear block copolymers (Lodge & Muthukumar, 1996). Therefore, the nanostructure of the system with fLi = 1.0 was the hexagonally packed cylinder structure in the isotropic state, because ϕPEO = 0.23 was larger than the border value. These results also indicated that LiCF3SO3 increased ϕPEO of the LiCF3SO3/PEO114-b-PMA(Az)51 systems by the formation of an LiCF3SO3/PEO complex.
4. Conclusion
Phase transitions and nanostructures of LiCF3SO3/PEO114-b-PMA(Az)51 systems were investigated by DSC and SAXS. The LiCF3SO3/PEO114-b-PMA(Az)51 systems above fLi = 2.5 × 10−1, which corresponds to four PEO repeating units per LiCF3SO3 molecule, formed a complex of LiCF3SO3/(EO)3. The complex of LiCF3SO3/PEO melted at 433 K, which was higher than the isotropic transition of the PMA(Az) domain. The addition of LiCF3SO3 induced disordering in the PEO114-b-PMA(Az)51 nanostructure and the liquid-crystal structure due to the molecular motion of the PEO domains being restricted by the complex formation. On the other hand, the addition of LiCF3SO3 increased the radius of the PEO cylinder and the volume fraction of the PEO domain. The structure transition from the hexagonally packed cylinder to the b.c.c. structure occurred for LiCF3SO3/PEO114-b-PMA(Az)51 with fLi = 2.5 × 10−2 (ϕPEO = 0.09) in the isotropic state, but not for LiCF3SO3/PEO114-b-PMA(Az)51 with fLi = 1.0 (ϕPEO = 0.23).
Acknowledgements
The authors are grateful to the Japan Science and Technology Agency for support of this research through the CREST project.
References
Anthamatten, M. & Hammond, P. T. (2001). J. Polym. Sci. B, 39, 2671–2691. Web of Science CrossRef CAS
Arai, A.A., Morita, T. & Nishikawa, K. (2003). J. Chem. Phys. 119, 1502–1509. Web of Science CrossRef CAS
Hamley, I. W. (1998). The Physics of Block Copolymers. New York: Oxford University Press.
Hamley, I. W. & Castelletto, V. (2004). Prog. Polym. Sci. 29, 909–948. CAS
Hashimoto, T., Kawamura, T., Harada, M. & Tanaka, H. (1994). Macromolecules, 27, 3063–3072. CrossRef CAS Web of Science
Jung, S., Yamada, T., Iyoda, T. & Yoshida, H. (2005). J. Therm. Anal. Cal. 81, 563–567. Web of Science CrossRef CAS
Jung, S. & Yoshida, H. (2006). J. Therm. Anal. Cal. 85, 719–724. Web of Science CrossRef CAS
Kato, T., Minewaki, K., Kawabata, Y., Imai, M. & Takahashi, Y. (2004). Langmuir, 20, 3504–3508. Web of Science CrossRef PubMed CAS
Kawabata, Y., Nagao, M., Seto, H., Komura, S., Takeda, T., Schwahn, D., Yamada, N. L. & Nobutou, H. (2004). Phys. Rev. Lett. 92, 056103–056106. Web of Science CrossRef PubMed CAS
Lazzari, M. & López-Quintela, M. A. (2003). Adv. Mater. 15, 1583–1594. Web of Science CrossRef CAS
Lemmich, J., Mortensen, K., Ipsen, J. H., Hønger, T., Bauer, R. & Mouritsen, O. G. (1996). Phys. Rev. E, 53, 5169–5180. CrossRef CAS Web of Science
Li, J., Kamata, K., Yamada, T., Yoshida, H. & Iyoda, T. (2005). Polym. Prepr. Jpn, 53, 3N03.
Lodge, T. P. & Muthukumar, M. (1996). J. Phys. Chem. 100, 13275–13292. CrossRef CAS Web of Science
Rhodes, C. P. & Frech, R. (2001). Macromolecules, 34, 2660–2666. Web of Science CSD CrossRef CAS
Shibayama, M. & Hashimoto, T. (1986). Macromolecules, 19, 740–749. CrossRef CAS Web of Science
Tian, Y., Watanabe, K., Kong, X., Abe, J. & Iyoda, T. (2002). Macromolecules, 35, 3739–3747. Web of Science CrossRef CAS
Tominaga, Y., Izumi, Y., Kwak, G., Asai, S. & Sumita, M. (2003). Macromolecules, 36, 8766–8772. Web of Science CrossRef CAS
Watanabe, R., Iyoda, T., Yamada, T. & Yoshida, H. (2006). J. Therm. Anal. Cal. 85, 713–717. Web of Science CrossRef CAS
Yamada, T., Watanabe, R., Watanabe, K., Koyanagi, C., Yoshida, H. & Iyoda, T. (2004). Polym. Prepr. Jpn, 53, 1089.
Yamada, T., Yoshii, T., Ozawa, N., Kawabata, Y., Kato, T. & Yoshida, H. (2005). Trans. Mater. Res. Sci. Jpn, 30, 675–678. CAS
Yoshida, H., Kinoshita, R. & Teramoto, Y. (1995). Thermochim. Acta, 264, 173–183. CrossRef CAS Web of Science
Yoshida, H., Watabane, K., Watanabe, R. & Iyoda, T. (2004). Trans. Mater. Res. Sci. Jpn, 29, 861–864. CAS
© International Union of Crystallography. Prior permission is not required to reproduce short quotations, tables and figures from this article, provided the original authors and source are cited. For more information, click here.