- Email: [email protected]
Interlayer spacing effect of alkylammonium-modified montmorillonite on conducting and mechanical behaviors of polymer composite electrolytes
Journal of Colloid and Interface Science 332 (2009) 145–150
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Interlayer spacing effect of alkylammonium-modified montmorillonite on conducting and mechanical behaviors of polymer composite electrolytes Seok Kim a , Soo-Jin Park b,∗ a b
Advanced Materials Division, Korea Research Institute of Chemical Technology, P.O. Box 107, Yuseong, Daejeon 305-600, Republic of Korea Department of Chemistry, Inha University, 253, Nam-gu, Incheon 402-751, Republic of Korea
a r t i c l e
i n f o
a b s t r a c t
Article history: Received 18 September 2008 Accepted 3 December 2008 Available online 9 December 2008
Polymer composite electrolytes (PCE) based on poly(ethylene oxide) (PEO) and alkylammoniumfunctionalized clay materials were fabricated. Structural modification and the electrochemical properties were investigated in order to understand the effects of organically modified montmorillonite (OMMT) in the polymer matrix on the ion conductivity. Nanosize layer-structured clay fillers were used having an organic modifier with different alkyl lengths and positions. The X-ray diffraction (XRD) results revealed that the interlayer spacing (2.55 nm) for MMT-20A was increased compared with that (1.16 nm) for NaMMT. Both the XRD and the thermal analysis results indicated that the PCE showed reduced crystallinity after the introduction of the OMMT fillers. From the ion-conductivity results, the PCE containing MMT20A showed higher conductivity (6.1 × 10−4 S/cm) than that containing Na-MMT (2.2 × 10−4 S/cm). This indicated that the improved ion conductivity was dependent on the reduced crystallinity that was correlated with the d-spacing of the MMT. Furthermore, the PCE/OMMT showed improved tensile strength. Finally, it was shown that the conducting and mechanical properties were dependent on the interlayer spacing of the modified clays. © 2008 Elsevier Inc. All rights reserved.
Keywords: Composite electrolytes Poly(ethylene oxide) Montmorillonite Conducting behavior Interlayer spacing
1. Introduction Rechargeable lithium batteries with high-specific energies are promising power sources for portable electronic products and electric vehicles [1–5]. However, lithium metal limits the choice of usable electrolytes and impedes the commercialization of secondary Li batteries. Moreover, liquid electrolytes decrease the safety and cycle life of Li batteries. Solid and solid-like polymer electrolytes offer unique advantages such as satisfactory mechanical properties, ease of fabrication as thin films and an application to the formation of proper electrode–electrolyte interfaces. Most of the studies in this field have focused on poly(ethylene oxide) (PEO)-based polymer composite electrolytes [6–9]. However, PEO-based polymer composite electrolytes (PCE) have a conductivity of only 10−4 –10−5 S/cm, which is low for some applications. Electrolytes with high ion conductivities are much sought after for applications to the development of solid-state batteries, electrochromic devices, sensors, and others [10–13]. Organic solid electrolytes, as well as inorganic ones, have been widely studied, and polymeric solid electrolytes with their flexibility and processability, are expected to be extraordinary composites.
*
Corresponding author. Fax: +82 (32) 860 8438. E-mail address: [email protected] (S.-J. Park).
0021-9797/$ – see front matter doi:10.1016/j.jcis.2008.12.006
©
2008 Elsevier Inc. All rights reserved.
Recently, inorganic-filler-containing PCE has been proposed as an alternative to liquid electrolytes in rechargeable lithium batteries. This has spurred development of dry, gelled, and porous types of solid-like polymer electrolytes [14–16]. Recently, a group of materials, through the addition of ceramic powders, organic acid, and inorganic composites [17–19], was examined for the purposes of improving PCE. The addition to polymer films of inert fillers such as β -alumina, γ -LiAlO2 and Al2 O3 [16,20] has also been pursed, to improve the ion conductivity, electrochemical or mechanical properties. Notably, PCE containing nanoparticles exhibit better mechanical strength, higher ion conductivity, and better interfacial stability to lithium anode. The recent interest in PCE containing layer-structured clays such as montmorillonite (MMT) has arisen from the fact that such PCE exhibits dramatic increases in tensile strength, heat resistance, and solvent resistance as well as decreases in gas permeability, all of which properties are desirable in electrolytes for rechargeable batteries. The layered structure of MMT clays consists of two silica tetrahedral sheets and an aluminum octahedral sheet. Stacking of ∼1 nm thick and ∼100 nm wide and long layers by a weak dipolar force yields interlayer galleries. MMT could easily be transformed into organophilic clay as a result of the introduction organic modifiers by cation exchange reaction. Smectite clays offer the properties of cation exchange, intercalation of molecules, and swelling in solvents. The adsorption of organic molecules by natural clays is a
146
S. Kim, S.-J. Park / Journal of Colloid and Interface Science 332 (2009) 145–150
function of the physical and chemical properties of the clay. If the organic molecules react with the hydrophilic clays, the clay layers attract the molecules. Vaia et al. had reported that organic molecules can enter the spaces between the layers of the clay in the following ways [21, 22]: (a) cationic bonding, in which the protonated alkylammonium replaces the sodium ions in the MMT layers, (b) ion–dipole interactions, in which the polar organic molecules are related to the sodium ions in the MMT layers, and (c) dipole–dipole interactions, which include the hydrogen bonding that associates polar organic molecules with hydroxyl groups or oxygen in the clay layers. Vaia et al. also reported that composite electrolytes with a 60 wt% inorganic content fabricated by melt intercalation showed a higher conductivity, 10−6 S/cm, than pristine electrolytes at room temperature [21]. Most studies have established that intercalation of the polymer chains in the silicate galleries appears to suppress their tendency to crystallize, resulting in enhanced ion conductivity. There are reports of annealing treatments, resulting in improved conductivity due to an increase in PEO intercalation into the clay interlayers [23]. PCE can be prepared by a melt process, a solution process, or in-situ intercalative polymerization of host polymers. Organically modified clays can be prepared by treating Na-MMT with various organic modifier agents. To our best knowledge, the influence of the interlayer spacing length of modified fillers on the ion conductivity and mechanical properties of PCE has not been systematically studied. In the present study, therefore, PEO-based PCE were prepared by introducing MMT treated with ammoniumtype modifiers of different alkyl chain lengths and positions. The objective of this investigation was to analyze the influences of the clays-containing PCE interlayer spacing length and microstructure on the electrochemical and mechanical properties. 2. Materials and methods 2.1. Preparation of polyethylene oxide/clay PCE In the preparation of the PCE, polyethylene oxide (Mw: 2.0 × 105 ), ethylene carbonate (EC), LiClO4 (Aldrich Co. Ltd.), and fillers were used. The Na-MMT and organically modified MMT (OMMT) (Cloisite 20A, 25A, 30B or MMT-20A, MMT-25A, MMT-30B) supplied by Southern Clay Products (USA) was used as fillers without purification. The detailed specifications are listed in Table 1. PEO and LiClO4 were dried in a vacuum oven for 24 h at 50 ◦ C and 120 ◦ C, respectively. The preparation of the PCE films involved first the dissolution of the PEO in acetonitrile, and then the addition of EC. The solution was agitated for 3 h at 40 ◦ C to completely mix the PEO and EC. The weight ratio of EC to PEO was fixed at 1:4. Then, 1 mol of LiClO4 was added to the solution. After 1 h agitation, 10 wt% Na-MMT or various MMTs was added to the mixture, after which another agitation was sustained for 24 h at 50 ◦ C. The mixtures were then poured into Teflon plates and evaporated slowly at 40 ◦ C in a vacuum oven to yield the PCE film. 2.2. X-ray diffraction (XRD) measurements X-ray diffraction (XRD) patterns of the PCE as a function of Na-MMT or OMMT addition were obtained with a Rigaku Model D/MAX-III B X-ray diffractometer equipped with a rotation anode using CuKα radiation (18 kW rotating anode, λ = 1.5405 Å) at 50 kV and 250 mA with a scanning rate of 2◦ /min. 2.3. DSC analysis The thermal behaviors of the PCE were studied using differential scanning calorimetry (DSC 6 series, Perkin Elmer Co. Ltd.). The
Table 1 Various organically modified MMT used in this study.a Sample
Organic modifier
Modifier concentration
Na+ -MMT MMT-20A
None 2M2HT: dimethyl, dehydrogenated tallow, quaternary ammonium 2MHTL8: dimethyl, dehydrogenated tallow, 2-ethylhexyl quaternary ammonium MT2EtOH: methyl, tallow, bis-2-hydroxyethyl, quaternary ammonium
– 95 meq/100 g clay
MMT-25A
MMT-30B
a
95 meq/100 g clay
90 meq/100 g clay
Data from Southern Clay Products, Inc.
samples were loaded into a sealed aluminum pan, and the measurements were taken in the 30 to 125 ◦ C range with a 5 ◦ C/min heating rate and under N2 atmosphere. 2.4. TEM analysis A Philips CM-20 transmission electron microscope (TEM) was used to examine the morphology of the PCE. The PCE films were sliced to thicknesses of 70 nm with a Boeckeler CRX cryoultramicrotome at the temperature of liquid nitrogen. The accelerating voltage of the TEM was 160 kV. 2.5. Conductivity measurements Impedance spectroscopy was used to determine the ion conductivity of the PCE. The measurements were carried out in the 1 MHz–10 Hz frequency range using a frequency response analyzer (FRA) AUTOLAB with PGSTAT 30 (potentiostat/galvanostat) (Eco Chemie, The Netherlands) electrochemical instrument. The PCE film was sandwiched between stainless steel blocking electrodes. The specimen thickness varied from 0.8 to 1.2 mm. The conductivity values (σ ) were calculated from the bulk resistance ( R b ), which was determined by equivalent circuit analysis software. 2.6. Mechanical strength measurements The tensile strength was measured at room temperature by means of a universal tensile machine (Instron model 5565, Lloyd) at a full-out velocity of 50 mm min−1 . The sample thickness was 100 μm. Measurements were performed five times for each sample, and from which measurements the average value was calculated [24]. 3. Results and discussion 3.1. Crystalline structure and morphology of PCE Our primary objective in the present study was to examine the influences of alkylammonium-modified MMT on the microstructure and conductivity of PCE. We had chosen three kinds of OMMT, each of which varied according to alkyl lengths and positions of the alkylammonium functional group. Cloisite 20A (MMT-20A) was expected to have a large molecular volume size due to the two bulky groups of Hydrogenated Tallow (HT). Cloisite 25A (MMT25A) had only one bulky group of HT and a rather short-length alkyl group. Cloisite 30B (MMT-30B) has one bulky group of Tallow and two short ethyl alcohol groups. XRD is a powerful tool with which to study the structure of clay minerals. The interlayer spacing of the PCE containing the modified clays was determined by the XRD peak, using the Bragg equation
λ = 2d sin θ,
(1)
S. Kim, S.-J. Park / Journal of Colloid and Interface Science 332 (2009) 145–150
Fig. 1. XRD patterns of (a) PEO/LiClO4 and PCE containing 10 wt% (b) Na-MMT, (c) MMT-20A, (d) MMT-25A, and (e) MMT-30B.
147
reflecting the fact that more of the polymer chain could be attracted and penetrated into the interlayer galleries. Fig. 3 shows the TEM results for the PCE containing (a) NaMMT, (b) MMT-20A, (c) MMT-25A and (d) MMT-30B. We could notice the qualitative OMMT d-spacing changes after preparation of the PCE. PCE/Na-MMT (a) showed the stacked platelet morphology of MMT. In the case of MMT-20A (b), the PCE showed rather expanded layered platelets compared with (a). We could conclude that the MMT had not been fully exfoliated, which was confirmed by the existence of the d001 peak in the previous XRD data, shown in Fig. 1. PCE/MMM-25A (c) manifested the stacked platelet morphology of MMT, in which the interlayer spacing length was decreased in comparison with that of MMT-20A (b). PCE/MMT-30B (d) also showed the stacked morphology and the gap spacing was the smaller than that of MMT-20A. In conclusion, our TEM studies confirmed that the d-spacing of MMT in PCE was enhanced and changed by using a different organic modifier. As shown in TEM images, the MMT or O-MMT were dispersed homogeneously in the polymer matrix. Of course, some of MMT clays were existed as delaminated or exfoliated layers. However, most of MMT layered structures were well stabilized and somewhat expanded by organic modifiers or polymer chain intercalation. This expanded OMMT brought the modification of microstructure of composite. Fig. 4 illustrates the wide-range XRD patterns of the various PEO/OMMT PCE. These results showed the PCE crystallinity changes for the different OMMT species. The characteristic diffraction peaks of the PEO crystalline structure were apparent between 2θ = 15 and 30◦ . These diffraction peaks became broader and less prominent with the various MMT additions, resulting in decreased crystallinity. By adding the OMMT to the PEO/LiClO4 system, the intensities of characteristic peaks were decreased the lowest value corresponding to MMT-20A. The crystallinity was investigated in more detail in the following DSC analysis. 3.2. Thermal behaviors
Fig. 2. d-Spacing lengths of (A) filler materials and (B) PCE containing fillers of (a) Na-MMT, (b) MMT-20A, (c) MMT-25A, and (d) MMT-30B.
where λ corresponds to the wavelength of the X-ray radiation (1.5405 Å), d corresponds to the interlayer spacing between the diffractional lattice planes and θ is the measured diffraction angle. Fig. 1 shows the XRD patterns of (a) PEO/LiClO4 , and PCE containing 10 wt% (b) Na-MMT, (c) MMT-20A, (d) MMT-25A and (e) MMT-30B in the 2θ = 2–10◦ region. Pristine PEO/LiClO4 had no characteristic peak in this region. Each PCE pattern had one peak, at 2θ = 6.56, 3.69, 3.82, and 4.07◦ , respectively. These peaks were assigned to the (001) lattice spacing of MMT. As a result of the OMMT introduction, the d-spacing was largely enhanced in compared with Na-MMT. However, each modified-filler-containing PCE showed a d001 peak, indicating that the layered MMT platelets had not been completely exfoliated (or delaminated) and had sustained the stacked structure with an expanded spacing. Fig. 2 illustrates the d-spacing length of the PCE for the different kinds of alkylammonium modified MMT. Graph (A) shows the d-spacing lengths of the Na-MMT, MMT-20A, MMT-25A and MMT-30B, respectively. Among the samples, MMT-20A showed the highest d-spacing value owing to the large molecular volume of the dehydrogenated Tallow functional group. Graph (B) shows the d-spacing lengths of the respective fillers after fabricating the PCE with PEO. The d-spacing lengths were increased, indicating that some parts of the polymer chain had been intercalated into the MMT interlayer. The PCE containing MMT-20A showed the highest d-spacing length, 25.5 Å, whereas the PCE containing Na-MMT showed the lowest, 11.6 Å. Furthermore, in the cases of (c) MMT-25A and (d) MMT-30B, the increased d-spacing length was rather large compared with that of the other samples,
The thermal properties of the PCE were studied by means of the DSC method. The thermal parameters, that is, the melting temperature ( T m ) and the fusion heat are shown in Fig. 5. PEO containing Na-MMT showed decreased T m (52.4 ◦ C) and fusion heat (82.4 J/g) values compared with those (54.6 ◦ C, 99.9 J/g) of pristine PEO, meaning that the Na-MMT were effective in reducing the crystallinity of the PEO. By employing OMMT, the T m and crystallinity of the PCE were effectively reduced further than those of the PCE containing Na-MMT. The smallest T m (47.5 ◦ C) and fusion heat (62.5 J/g) were obtained for the PCE containing MMT-20A. Considered with the previous XRD results, the DSC studies indicated that the PEO crystalline structure had been deteriorated and that the reduced crystallinity was dependent on the d-spacing of the OMMT. As a result, the interlayer spacings of MMT were enhanced, by using organic modifiers that have a large molecular volume. It is thought that the expanded MMT layer silicates may interact effectively with polymer chain. By increasing the filler– matrix interaction, the film morphology or crystalline structure had been changed. The filler effect was enhanced by the expanded MMT by controlling organic modifiers. 3.3. Conductivity behaviors The conductivity was measured by sandwiching the PCE films between the stainless steel electrodes. The conductivity values were calculated according to the intercept of the real part of a complex-plane impedance plot using the equation
σ = t /( R b A ),
(2)
where t is the thickness of the PCE film, R b is the bulk resistance of the film, and A is the area. Fig. 6 plots the conductivity for the
148
S. Kim, S.-J. Park / Journal of Colloid and Interface Science 332 (2009) 145–150
(a)
(b)
(c)
(d)
Fig. 3. TEM images of PCE containing 10 wt% fillers of (a) Na-MMT, (b) MMT-20A, (c) MMT-25A, and (d) MMT-30B (scale bar: 20 nm).
PCE containing various types of MMT at 25 ◦ C. In the result, the highest ion conductivity, 6.1 × 10−4 S/cm, was observed at OMMT 20A. This was about triple the value, 2.2 × 10−4 S/cm, of PEO/NaMMT, indicating that the ion conductivity of the PCE was enhanced by increasing the interlayer spacing of the clays. Fig. 7 shows the impedance plots versus clay contents for the PEO/MMT-20A PCE at 25 ◦ C. An increase in the conductivity, as indicated by the decreased semi-circle in the higher-frequency region, was achieved by adding a small quantity of the OMMT. The maximum ion conductivity (i.e., the smallest semi-circle on the impedance plot) was achieved at 10 wt% OMMT content. When the OMMT was increased beyond 10 wt%, the conductivity decreased slightly from the maximum value. It could be concluded that the addition of the optimum OMMT content (10 wt%) provided the most suitable environment for the ionic transportation and, thereby, achieved the highest conductivity. By increasing the interlayer spacing of the clays, the crystalline structure of PCE had been degraded, meaning the decreased crystallinity. It could lead to the more flexible chain structure of polymer composites. Con-
sequently, lithium ion could move more easily along the flexible polymer main chain, resulting the enhancement of ion conductivity. This increase in ion conductivity by OMMT addition could be related to the functioning of a few clay sheets that were expanded and dispersed in the matrix. Such functions have been explained, by reference to cation interactions, in a previous report [25]. The authors suggested three types of Li cation complexes, those with (a) polyether chains, (b) polyether and silicate layers, and (c) silicate layers. For all types of interaction, the interfacial area among the Li ions, polyether chains, and silicate layers would be critical to the filler effects. In this view, the larger d-spacing of OMMT could probably afford the much higher interaction of fillers with polymer chains or Li ions. 3.4. Mechanical properties The PCE critical strengths are shown in Fig. 8. The tensile strength for PCE/Na-MMT, 0.7 MPa, was higher than that, 0.42 MPa, for pristine PCE containing no filler. Furthermore, the strength
S. Kim, S.-J. Park / Journal of Colloid and Interface Science 332 (2009) 145–150
Fig. 4. XRD patterns of (a) pure PEO, (b) PEO/EC/LiClO4 and PCE containing 10 wt% (c) Na+ -MMT, (d) MMT-20A, (e) MMT-25A, and (f) MMT-30B.
Fig. 5. Thermal parameters of pristine PEO/LiClO4 and PCE containing 10 wt% Na+ MMT, MMT-20A, MMT-25A, and MMT-30B.
Fig. 6. Ion conductivity values of PEO/LiClO4 and PCE containing 10 wt% fillers of (a) Na-MMT, (b) MMT-20A, (c) MMT-25A, and (d) MMT-30B.
149
Fig. 7. Electrochemical impedance spectra of PCE containing MMT-20A with several different weight contents (wt%).
Fig. 8. Tensile strengths of PEO/LiClO4 and PCE containing 10 wt% fillers of (a) NaMMT, (b) MMT-20A, (c) MMT-25A, and (d) MMT-30B.
value was further upgraded with the introduction of OMMT. Notably, PEO/MMT-20A showed the highest strength value, 1.25 MPa. It is thought that the d-spacing of MMT also affected the PCE mechanical properties. Probably, the wider gap spacing of MMT could effect a higher interfacial area between the filler and polymer chains, which could help to improve the mechanical properties of PCE, by means of fabricated nanocomposites. For these samples, we performed the mechanical property measurements as a function of the filler contents. The results are shown in Fig. 9. By increasing the filler content incrementally to 15 wt%, the tensile strength was increased in proportion to the content. However, the value became saturated leveling off over 15 wt%, indicating that a higher filler content over the optimum point would not effectively improve the mechanical strength. This phenomenon differed in the case of the conductivity changes as a function of the filler contents. In the conductivity results, the maximum value was obtained for 10 wt%, the value being degraded for filler contents over 10 wt%. However, the mechanical strength was not decreased over 10 wt%. Considering the all of the results together, it was concluded that the filler effect on the conductivity and mechanical strength was somewhat variant, because the MMT filler effect on the conductivity consisted of multiple complex factors including filler–Li ion interaction, polymer–filler interaction,
150
S. Kim, S.-J. Park / Journal of Colloid and Interface Science 332 (2009) 145–150
tion of the clay platelets with the Li ions or polymer chains. This enhancement in the ion conductivity of the PCE system was attributed to a reduction in the degree of crystallinity. The maximum conductivity, 6.1 × 10−4 S/cm, had been obtained for PEO/MMT20A. The MMT-20A had a larger interlayer spacing than any other MMT sample. Consequently, the ion conductivity of the PCE could be increased using the organically modified clays, which effected the decrease of the crystallinity that was correlated with the dspacing. Moreover, the mechanical strength was upgraded by introducing OMMT, in place of Na-MMT. Significantly, PEO/MMT-20A showed the highest strength value, 1.25 MPa. It could be concluded that the mechanical properties of PCE were also dependent on the d-spacing of MMT. By increasing the filler content by increments to 15 wt%, the tensile strength was increased and then saturated.
Fig. 9. Tensile strengths of PCE containing MMT-20A with several different weight contents (wt%).
and polymer–Li ion interaction [25]. Relevant studies are underway and are being conducted separately. 4. Conclusion Various MMT of different alkyl chain lengths and positions were introduced into the PEO PCE. The d-spacing and stacked platelet morphology of the PCE were confirmed by XRD and TEM results, respectively. The XRD and thermal analysis results indicated that the PCE showed the reduced crystallinity by the introduction of organically modified fillers. The XRD results revealed that the interlayer spacings of the OMMT were much larger than those of the Na-MMT. Notably, the PCE containing Cloisite-20A (MMT-20A) showed the largest interlayer spacing (25.5 Å) of MMT, owing to the large volume size of the modifier and the partial insertion of some parts of the polymer chain. Accordingly, PEO/MMT-20A showed the lowest melting temperature, 47.5 ◦ C and the smallest fusion heat value. The reduced crystallinity was dependent on the changed d-spacing of the MMT in the PCE. We confirmed the dependency of the interlayer spacing on the electrochemical or physical properties of PCE. This study demonstrated that the addition of OMMT enhanced the ion conductivity of the PCE. Also, the OMMT could influence the ion-conduction mechanism of the PEO/LiClO4 system by means of some interac-
References [1] J.M. Tarascon, M.B. Armand, Nature 414 (2001) 359. [2] F.M. Gray, Solid Polymer Electrolytes: Fundamental and Technological Applications, VCH Publishers, New York, 1991. [3] J.R.M. Callum, Polymer Electrolyte Reviews, Elsevier, Amsterdam, 1987. [4] M. Wakihara, Mater. Sci. Eng. R 33 (2001) 109. [5] P.V. Wright, Br. Polym. J. 7 (1975) 319. [6] A.L. Oleksiak, H.D. Inerowicz, J. Power Sources 81–82 (1999) 813. [7] M. Mclin, C.A. Angell, Solid State Ionics 56 (1992) 1027. [8] S.A. Hashmi, A. Chandra, S. Chandra, in: B.V.R. Chowdari, et al. (Eds.), Solid State Ionics: Materials and Applications, World Scientific, Singapore, 1992, p. 567. [9] M.K. Park, H.S. Kim, J.H. An, J. Kim, J. Ind. Eng. Chem. 11 (2005) 222. [10] F.B. Dias, L. Plomp, B.J. Veldhuis Jakobert, J. Power Sources 88 (2000) 169. [11] S.J. Park, B.R. Jun, J. Colloid Interface Sci. 284 (2005) 204. [12] S. Kim, Y.J. Jung, Electrochem. Commun. 9 (2007) 249. [13] S. Kim, S.J. Park, Electrochim. Acta 52 (2007) 3477. [14] H.J. Rhoo, H.T. Kim, J.K. Park, T.S. Hwang, Electrochim. Acta 42 (1997) 1571. [15] S.R. Starkey, R. Frech, Electrochim. Acta 42 (1997) 471. [16] M. Nookala, B. Kumar, S. Rodrigues, J. Power Sources 111 (2002) 165. [17] J. Ennari, L. Pietila, V. Virkkunen, F. Sundholm, Polymer 43 (2002) 5427. [18] S. Kim, J.Y. Kang, S.G. Lee, J.R. Lee, S.J. Park, Polymer (Korea) 29 (2005) 403. [19] X.A. Fu, S. Qutubuddin, J. Colloid Interface Sci. 283 (2005) 373. [20] X. Qian, N. Gu, Z. Cheng, X. Yang, E. Wang, S. Dong, Electrochim. Acta 46 (2001) 1829. [21] R.A. Vaia, S. Vasudevan, W. Krawiec, L.G. Scanlon, E.P. Giannelis, Adv. Mater. 7 (1995) 154. [22] R.A. Vaia, R.K. Teukolsky, E.P. Giannelis, Chem. Mater. 6 (1994) 1017. [23] B. Liao, M. Song, H.J. Liang, Y.X. Pang, Polymer 42 (2001) 10007. [24] H.V. Olphen, An Introduction to Clay Colloid Chemistry, Interscience Publishers, New York, 1963. [25] H.W. Chen, F.C. Chang, Polymer 42 (2001) 9763.
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Interlayer spacing effect of alkylammonium-modified montmorillonite on conducting and mechanical behaviors of polymer composite electrolytes Seok Kim a , Soo-Jin Park b,∗ a b
Advanced Materials Division, Korea Research Institute of Chemical Technology, P.O. Box 107, Yuseong, Daejeon 305-600, Republic of Korea Department of Chemistry, Inha University, 253, Nam-gu, Incheon 402-751, Republic of Korea
a r t i c l e
i n f o
a b s t r a c t
Article history: Received 18 September 2008 Accepted 3 December 2008 Available online 9 December 2008
Polymer composite electrolytes (PCE) based on poly(ethylene oxide) (PEO) and alkylammoniumfunctionalized clay materials were fabricated. Structural modification and the electrochemical properties were investigated in order to understand the effects of organically modified montmorillonite (OMMT) in the polymer matrix on the ion conductivity. Nanosize layer-structured clay fillers were used having an organic modifier with different alkyl lengths and positions. The X-ray diffraction (XRD) results revealed that the interlayer spacing (2.55 nm) for MMT-20A was increased compared with that (1.16 nm) for NaMMT. Both the XRD and the thermal analysis results indicated that the PCE showed reduced crystallinity after the introduction of the OMMT fillers. From the ion-conductivity results, the PCE containing MMT20A showed higher conductivity (6.1 × 10−4 S/cm) than that containing Na-MMT (2.2 × 10−4 S/cm). This indicated that the improved ion conductivity was dependent on the reduced crystallinity that was correlated with the d-spacing of the MMT. Furthermore, the PCE/OMMT showed improved tensile strength. Finally, it was shown that the conducting and mechanical properties were dependent on the interlayer spacing of the modified clays. © 2008 Elsevier Inc. All rights reserved.
Keywords: Composite electrolytes Poly(ethylene oxide) Montmorillonite Conducting behavior Interlayer spacing
1. Introduction Rechargeable lithium batteries with high-specific energies are promising power sources for portable electronic products and electric vehicles [1–5]. However, lithium metal limits the choice of usable electrolytes and impedes the commercialization of secondary Li batteries. Moreover, liquid electrolytes decrease the safety and cycle life of Li batteries. Solid and solid-like polymer electrolytes offer unique advantages such as satisfactory mechanical properties, ease of fabrication as thin films and an application to the formation of proper electrode–electrolyte interfaces. Most of the studies in this field have focused on poly(ethylene oxide) (PEO)-based polymer composite electrolytes [6–9]. However, PEO-based polymer composite electrolytes (PCE) have a conductivity of only 10−4 –10−5 S/cm, which is low for some applications. Electrolytes with high ion conductivities are much sought after for applications to the development of solid-state batteries, electrochromic devices, sensors, and others [10–13]. Organic solid electrolytes, as well as inorganic ones, have been widely studied, and polymeric solid electrolytes with their flexibility and processability, are expected to be extraordinary composites.
*
Corresponding author. Fax: +82 (32) 860 8438. E-mail address: [email protected] (S.-J. Park).
0021-9797/$ – see front matter doi:10.1016/j.jcis.2008.12.006
©
2008 Elsevier Inc. All rights reserved.
Recently, inorganic-filler-containing PCE has been proposed as an alternative to liquid electrolytes in rechargeable lithium batteries. This has spurred development of dry, gelled, and porous types of solid-like polymer electrolytes [14–16]. Recently, a group of materials, through the addition of ceramic powders, organic acid, and inorganic composites [17–19], was examined for the purposes of improving PCE. The addition to polymer films of inert fillers such as β -alumina, γ -LiAlO2 and Al2 O3 [16,20] has also been pursed, to improve the ion conductivity, electrochemical or mechanical properties. Notably, PCE containing nanoparticles exhibit better mechanical strength, higher ion conductivity, and better interfacial stability to lithium anode. The recent interest in PCE containing layer-structured clays such as montmorillonite (MMT) has arisen from the fact that such PCE exhibits dramatic increases in tensile strength, heat resistance, and solvent resistance as well as decreases in gas permeability, all of which properties are desirable in electrolytes for rechargeable batteries. The layered structure of MMT clays consists of two silica tetrahedral sheets and an aluminum octahedral sheet. Stacking of ∼1 nm thick and ∼100 nm wide and long layers by a weak dipolar force yields interlayer galleries. MMT could easily be transformed into organophilic clay as a result of the introduction organic modifiers by cation exchange reaction. Smectite clays offer the properties of cation exchange, intercalation of molecules, and swelling in solvents. The adsorption of organic molecules by natural clays is a
146
S. Kim, S.-J. Park / Journal of Colloid and Interface Science 332 (2009) 145–150
function of the physical and chemical properties of the clay. If the organic molecules react with the hydrophilic clays, the clay layers attract the molecules. Vaia et al. had reported that organic molecules can enter the spaces between the layers of the clay in the following ways [21, 22]: (a) cationic bonding, in which the protonated alkylammonium replaces the sodium ions in the MMT layers, (b) ion–dipole interactions, in which the polar organic molecules are related to the sodium ions in the MMT layers, and (c) dipole–dipole interactions, which include the hydrogen bonding that associates polar organic molecules with hydroxyl groups or oxygen in the clay layers. Vaia et al. also reported that composite electrolytes with a 60 wt% inorganic content fabricated by melt intercalation showed a higher conductivity, 10−6 S/cm, than pristine electrolytes at room temperature [21]. Most studies have established that intercalation of the polymer chains in the silicate galleries appears to suppress their tendency to crystallize, resulting in enhanced ion conductivity. There are reports of annealing treatments, resulting in improved conductivity due to an increase in PEO intercalation into the clay interlayers [23]. PCE can be prepared by a melt process, a solution process, or in-situ intercalative polymerization of host polymers. Organically modified clays can be prepared by treating Na-MMT with various organic modifier agents. To our best knowledge, the influence of the interlayer spacing length of modified fillers on the ion conductivity and mechanical properties of PCE has not been systematically studied. In the present study, therefore, PEO-based PCE were prepared by introducing MMT treated with ammoniumtype modifiers of different alkyl chain lengths and positions. The objective of this investigation was to analyze the influences of the clays-containing PCE interlayer spacing length and microstructure on the electrochemical and mechanical properties. 2. Materials and methods 2.1. Preparation of polyethylene oxide/clay PCE In the preparation of the PCE, polyethylene oxide (Mw: 2.0 × 105 ), ethylene carbonate (EC), LiClO4 (Aldrich Co. Ltd.), and fillers were used. The Na-MMT and organically modified MMT (OMMT) (Cloisite 20A, 25A, 30B or MMT-20A, MMT-25A, MMT-30B) supplied by Southern Clay Products (USA) was used as fillers without purification. The detailed specifications are listed in Table 1. PEO and LiClO4 were dried in a vacuum oven for 24 h at 50 ◦ C and 120 ◦ C, respectively. The preparation of the PCE films involved first the dissolution of the PEO in acetonitrile, and then the addition of EC. The solution was agitated for 3 h at 40 ◦ C to completely mix the PEO and EC. The weight ratio of EC to PEO was fixed at 1:4. Then, 1 mol of LiClO4 was added to the solution. After 1 h agitation, 10 wt% Na-MMT or various MMTs was added to the mixture, after which another agitation was sustained for 24 h at 50 ◦ C. The mixtures were then poured into Teflon plates and evaporated slowly at 40 ◦ C in a vacuum oven to yield the PCE film. 2.2. X-ray diffraction (XRD) measurements X-ray diffraction (XRD) patterns of the PCE as a function of Na-MMT or OMMT addition were obtained with a Rigaku Model D/MAX-III B X-ray diffractometer equipped with a rotation anode using CuKα radiation (18 kW rotating anode, λ = 1.5405 Å) at 50 kV and 250 mA with a scanning rate of 2◦ /min. 2.3. DSC analysis The thermal behaviors of the PCE were studied using differential scanning calorimetry (DSC 6 series, Perkin Elmer Co. Ltd.). The
Table 1 Various organically modified MMT used in this study.a Sample
Organic modifier
Modifier concentration
Na+ -MMT MMT-20A
None 2M2HT: dimethyl, dehydrogenated tallow, quaternary ammonium 2MHTL8: dimethyl, dehydrogenated tallow, 2-ethylhexyl quaternary ammonium MT2EtOH: methyl, tallow, bis-2-hydroxyethyl, quaternary ammonium
– 95 meq/100 g clay
MMT-25A
MMT-30B
a
95 meq/100 g clay
90 meq/100 g clay
Data from Southern Clay Products, Inc.
samples were loaded into a sealed aluminum pan, and the measurements were taken in the 30 to 125 ◦ C range with a 5 ◦ C/min heating rate and under N2 atmosphere. 2.4. TEM analysis A Philips CM-20 transmission electron microscope (TEM) was used to examine the morphology of the PCE. The PCE films were sliced to thicknesses of 70 nm with a Boeckeler CRX cryoultramicrotome at the temperature of liquid nitrogen. The accelerating voltage of the TEM was 160 kV. 2.5. Conductivity measurements Impedance spectroscopy was used to determine the ion conductivity of the PCE. The measurements were carried out in the 1 MHz–10 Hz frequency range using a frequency response analyzer (FRA) AUTOLAB with PGSTAT 30 (potentiostat/galvanostat) (Eco Chemie, The Netherlands) electrochemical instrument. The PCE film was sandwiched between stainless steel blocking electrodes. The specimen thickness varied from 0.8 to 1.2 mm. The conductivity values (σ ) were calculated from the bulk resistance ( R b ), which was determined by equivalent circuit analysis software. 2.6. Mechanical strength measurements The tensile strength was measured at room temperature by means of a universal tensile machine (Instron model 5565, Lloyd) at a full-out velocity of 50 mm min−1 . The sample thickness was 100 μm. Measurements were performed five times for each sample, and from which measurements the average value was calculated [24]. 3. Results and discussion 3.1. Crystalline structure and morphology of PCE Our primary objective in the present study was to examine the influences of alkylammonium-modified MMT on the microstructure and conductivity of PCE. We had chosen three kinds of OMMT, each of which varied according to alkyl lengths and positions of the alkylammonium functional group. Cloisite 20A (MMT-20A) was expected to have a large molecular volume size due to the two bulky groups of Hydrogenated Tallow (HT). Cloisite 25A (MMT25A) had only one bulky group of HT and a rather short-length alkyl group. Cloisite 30B (MMT-30B) has one bulky group of Tallow and two short ethyl alcohol groups. XRD is a powerful tool with which to study the structure of clay minerals. The interlayer spacing of the PCE containing the modified clays was determined by the XRD peak, using the Bragg equation
λ = 2d sin θ,
(1)
S. Kim, S.-J. Park / Journal of Colloid and Interface Science 332 (2009) 145–150
Fig. 1. XRD patterns of (a) PEO/LiClO4 and PCE containing 10 wt% (b) Na-MMT, (c) MMT-20A, (d) MMT-25A, and (e) MMT-30B.
147
reflecting the fact that more of the polymer chain could be attracted and penetrated into the interlayer galleries. Fig. 3 shows the TEM results for the PCE containing (a) NaMMT, (b) MMT-20A, (c) MMT-25A and (d) MMT-30B. We could notice the qualitative OMMT d-spacing changes after preparation of the PCE. PCE/Na-MMT (a) showed the stacked platelet morphology of MMT. In the case of MMT-20A (b), the PCE showed rather expanded layered platelets compared with (a). We could conclude that the MMT had not been fully exfoliated, which was confirmed by the existence of the d001 peak in the previous XRD data, shown in Fig. 1. PCE/MMM-25A (c) manifested the stacked platelet morphology of MMT, in which the interlayer spacing length was decreased in comparison with that of MMT-20A (b). PCE/MMT-30B (d) also showed the stacked morphology and the gap spacing was the smaller than that of MMT-20A. In conclusion, our TEM studies confirmed that the d-spacing of MMT in PCE was enhanced and changed by using a different organic modifier. As shown in TEM images, the MMT or O-MMT were dispersed homogeneously in the polymer matrix. Of course, some of MMT clays were existed as delaminated or exfoliated layers. However, most of MMT layered structures were well stabilized and somewhat expanded by organic modifiers or polymer chain intercalation. This expanded OMMT brought the modification of microstructure of composite. Fig. 4 illustrates the wide-range XRD patterns of the various PEO/OMMT PCE. These results showed the PCE crystallinity changes for the different OMMT species. The characteristic diffraction peaks of the PEO crystalline structure were apparent between 2θ = 15 and 30◦ . These diffraction peaks became broader and less prominent with the various MMT additions, resulting in decreased crystallinity. By adding the OMMT to the PEO/LiClO4 system, the intensities of characteristic peaks were decreased the lowest value corresponding to MMT-20A. The crystallinity was investigated in more detail in the following DSC analysis. 3.2. Thermal behaviors
Fig. 2. d-Spacing lengths of (A) filler materials and (B) PCE containing fillers of (a) Na-MMT, (b) MMT-20A, (c) MMT-25A, and (d) MMT-30B.
where λ corresponds to the wavelength of the X-ray radiation (1.5405 Å), d corresponds to the interlayer spacing between the diffractional lattice planes and θ is the measured diffraction angle. Fig. 1 shows the XRD patterns of (a) PEO/LiClO4 , and PCE containing 10 wt% (b) Na-MMT, (c) MMT-20A, (d) MMT-25A and (e) MMT-30B in the 2θ = 2–10◦ region. Pristine PEO/LiClO4 had no characteristic peak in this region. Each PCE pattern had one peak, at 2θ = 6.56, 3.69, 3.82, and 4.07◦ , respectively. These peaks were assigned to the (001) lattice spacing of MMT. As a result of the OMMT introduction, the d-spacing was largely enhanced in compared with Na-MMT. However, each modified-filler-containing PCE showed a d001 peak, indicating that the layered MMT platelets had not been completely exfoliated (or delaminated) and had sustained the stacked structure with an expanded spacing. Fig. 2 illustrates the d-spacing length of the PCE for the different kinds of alkylammonium modified MMT. Graph (A) shows the d-spacing lengths of the Na-MMT, MMT-20A, MMT-25A and MMT-30B, respectively. Among the samples, MMT-20A showed the highest d-spacing value owing to the large molecular volume of the dehydrogenated Tallow functional group. Graph (B) shows the d-spacing lengths of the respective fillers after fabricating the PCE with PEO. The d-spacing lengths were increased, indicating that some parts of the polymer chain had been intercalated into the MMT interlayer. The PCE containing MMT-20A showed the highest d-spacing length, 25.5 Å, whereas the PCE containing Na-MMT showed the lowest, 11.6 Å. Furthermore, in the cases of (c) MMT-25A and (d) MMT-30B, the increased d-spacing length was rather large compared with that of the other samples,
The thermal properties of the PCE were studied by means of the DSC method. The thermal parameters, that is, the melting temperature ( T m ) and the fusion heat are shown in Fig. 5. PEO containing Na-MMT showed decreased T m (52.4 ◦ C) and fusion heat (82.4 J/g) values compared with those (54.6 ◦ C, 99.9 J/g) of pristine PEO, meaning that the Na-MMT were effective in reducing the crystallinity of the PEO. By employing OMMT, the T m and crystallinity of the PCE were effectively reduced further than those of the PCE containing Na-MMT. The smallest T m (47.5 ◦ C) and fusion heat (62.5 J/g) were obtained for the PCE containing MMT-20A. Considered with the previous XRD results, the DSC studies indicated that the PEO crystalline structure had been deteriorated and that the reduced crystallinity was dependent on the d-spacing of the OMMT. As a result, the interlayer spacings of MMT were enhanced, by using organic modifiers that have a large molecular volume. It is thought that the expanded MMT layer silicates may interact effectively with polymer chain. By increasing the filler– matrix interaction, the film morphology or crystalline structure had been changed. The filler effect was enhanced by the expanded MMT by controlling organic modifiers. 3.3. Conductivity behaviors The conductivity was measured by sandwiching the PCE films between the stainless steel electrodes. The conductivity values were calculated according to the intercept of the real part of a complex-plane impedance plot using the equation
σ = t /( R b A ),
(2)
where t is the thickness of the PCE film, R b is the bulk resistance of the film, and A is the area. Fig. 6 plots the conductivity for the
148
S. Kim, S.-J. Park / Journal of Colloid and Interface Science 332 (2009) 145–150
(a)
(b)
(c)
(d)
Fig. 3. TEM images of PCE containing 10 wt% fillers of (a) Na-MMT, (b) MMT-20A, (c) MMT-25A, and (d) MMT-30B (scale bar: 20 nm).
PCE containing various types of MMT at 25 ◦ C. In the result, the highest ion conductivity, 6.1 × 10−4 S/cm, was observed at OMMT 20A. This was about triple the value, 2.2 × 10−4 S/cm, of PEO/NaMMT, indicating that the ion conductivity of the PCE was enhanced by increasing the interlayer spacing of the clays. Fig. 7 shows the impedance plots versus clay contents for the PEO/MMT-20A PCE at 25 ◦ C. An increase in the conductivity, as indicated by the decreased semi-circle in the higher-frequency region, was achieved by adding a small quantity of the OMMT. The maximum ion conductivity (i.e., the smallest semi-circle on the impedance plot) was achieved at 10 wt% OMMT content. When the OMMT was increased beyond 10 wt%, the conductivity decreased slightly from the maximum value. It could be concluded that the addition of the optimum OMMT content (10 wt%) provided the most suitable environment for the ionic transportation and, thereby, achieved the highest conductivity. By increasing the interlayer spacing of the clays, the crystalline structure of PCE had been degraded, meaning the decreased crystallinity. It could lead to the more flexible chain structure of polymer composites. Con-
sequently, lithium ion could move more easily along the flexible polymer main chain, resulting the enhancement of ion conductivity. This increase in ion conductivity by OMMT addition could be related to the functioning of a few clay sheets that were expanded and dispersed in the matrix. Such functions have been explained, by reference to cation interactions, in a previous report [25]. The authors suggested three types of Li cation complexes, those with (a) polyether chains, (b) polyether and silicate layers, and (c) silicate layers. For all types of interaction, the interfacial area among the Li ions, polyether chains, and silicate layers would be critical to the filler effects. In this view, the larger d-spacing of OMMT could probably afford the much higher interaction of fillers with polymer chains or Li ions. 3.4. Mechanical properties The PCE critical strengths are shown in Fig. 8. The tensile strength for PCE/Na-MMT, 0.7 MPa, was higher than that, 0.42 MPa, for pristine PCE containing no filler. Furthermore, the strength
S. Kim, S.-J. Park / Journal of Colloid and Interface Science 332 (2009) 145–150
Fig. 4. XRD patterns of (a) pure PEO, (b) PEO/EC/LiClO4 and PCE containing 10 wt% (c) Na+ -MMT, (d) MMT-20A, (e) MMT-25A, and (f) MMT-30B.
Fig. 5. Thermal parameters of pristine PEO/LiClO4 and PCE containing 10 wt% Na+ MMT, MMT-20A, MMT-25A, and MMT-30B.
Fig. 6. Ion conductivity values of PEO/LiClO4 and PCE containing 10 wt% fillers of (a) Na-MMT, (b) MMT-20A, (c) MMT-25A, and (d) MMT-30B.
149
Fig. 7. Electrochemical impedance spectra of PCE containing MMT-20A with several different weight contents (wt%).
Fig. 8. Tensile strengths of PEO/LiClO4 and PCE containing 10 wt% fillers of (a) NaMMT, (b) MMT-20A, (c) MMT-25A, and (d) MMT-30B.
value was further upgraded with the introduction of OMMT. Notably, PEO/MMT-20A showed the highest strength value, 1.25 MPa. It is thought that the d-spacing of MMT also affected the PCE mechanical properties. Probably, the wider gap spacing of MMT could effect a higher interfacial area between the filler and polymer chains, which could help to improve the mechanical properties of PCE, by means of fabricated nanocomposites. For these samples, we performed the mechanical property measurements as a function of the filler contents. The results are shown in Fig. 9. By increasing the filler content incrementally to 15 wt%, the tensile strength was increased in proportion to the content. However, the value became saturated leveling off over 15 wt%, indicating that a higher filler content over the optimum point would not effectively improve the mechanical strength. This phenomenon differed in the case of the conductivity changes as a function of the filler contents. In the conductivity results, the maximum value was obtained for 10 wt%, the value being degraded for filler contents over 10 wt%. However, the mechanical strength was not decreased over 10 wt%. Considering the all of the results together, it was concluded that the filler effect on the conductivity and mechanical strength was somewhat variant, because the MMT filler effect on the conductivity consisted of multiple complex factors including filler–Li ion interaction, polymer–filler interaction,
150
S. Kim, S.-J. Park / Journal of Colloid and Interface Science 332 (2009) 145–150
tion of the clay platelets with the Li ions or polymer chains. This enhancement in the ion conductivity of the PCE system was attributed to a reduction in the degree of crystallinity. The maximum conductivity, 6.1 × 10−4 S/cm, had been obtained for PEO/MMT20A. The MMT-20A had a larger interlayer spacing than any other MMT sample. Consequently, the ion conductivity of the PCE could be increased using the organically modified clays, which effected the decrease of the crystallinity that was correlated with the dspacing. Moreover, the mechanical strength was upgraded by introducing OMMT, in place of Na-MMT. Significantly, PEO/MMT-20A showed the highest strength value, 1.25 MPa. It could be concluded that the mechanical properties of PCE were also dependent on the d-spacing of MMT. By increasing the filler content by increments to 15 wt%, the tensile strength was increased and then saturated.
Fig. 9. Tensile strengths of PCE containing MMT-20A with several different weight contents (wt%).
and polymer–Li ion interaction [25]. Relevant studies are underway and are being conducted separately. 4. Conclusion Various MMT of different alkyl chain lengths and positions were introduced into the PEO PCE. The d-spacing and stacked platelet morphology of the PCE were confirmed by XRD and TEM results, respectively. The XRD and thermal analysis results indicated that the PCE showed the reduced crystallinity by the introduction of organically modified fillers. The XRD results revealed that the interlayer spacings of the OMMT were much larger than those of the Na-MMT. Notably, the PCE containing Cloisite-20A (MMT-20A) showed the largest interlayer spacing (25.5 Å) of MMT, owing to the large volume size of the modifier and the partial insertion of some parts of the polymer chain. Accordingly, PEO/MMT-20A showed the lowest melting temperature, 47.5 ◦ C and the smallest fusion heat value. The reduced crystallinity was dependent on the changed d-spacing of the MMT in the PCE. We confirmed the dependency of the interlayer spacing on the electrochemical or physical properties of PCE. This study demonstrated that the addition of OMMT enhanced the ion conductivity of the PCE. Also, the OMMT could influence the ion-conduction mechanism of the PEO/LiClO4 system by means of some interac-
References [1] J.M. Tarascon, M.B. Armand, Nature 414 (2001) 359. [2] F.M. Gray, Solid Polymer Electrolytes: Fundamental and Technological Applications, VCH Publishers, New York, 1991. [3] J.R.M. Callum, Polymer Electrolyte Reviews, Elsevier, Amsterdam, 1987. [4] M. Wakihara, Mater. Sci. Eng. R 33 (2001) 109. [5] P.V. Wright, Br. Polym. J. 7 (1975) 319. [6] A.L. Oleksiak, H.D. Inerowicz, J. Power Sources 81–82 (1999) 813. [7] M. Mclin, C.A. Angell, Solid State Ionics 56 (1992) 1027. [8] S.A. Hashmi, A. Chandra, S. Chandra, in: B.V.R. Chowdari, et al. (Eds.), Solid State Ionics: Materials and Applications, World Scientific, Singapore, 1992, p. 567. [9] M.K. Park, H.S. Kim, J.H. An, J. Kim, J. Ind. Eng. Chem. 11 (2005) 222. [10] F.B. Dias, L. Plomp, B.J. Veldhuis Jakobert, J. Power Sources 88 (2000) 169. [11] S.J. Park, B.R. Jun, J. Colloid Interface Sci. 284 (2005) 204. [12] S. Kim, Y.J. Jung, Electrochem. Commun. 9 (2007) 249. [13] S. Kim, S.J. Park, Electrochim. Acta 52 (2007) 3477. [14] H.J. Rhoo, H.T. Kim, J.K. Park, T.S. Hwang, Electrochim. Acta 42 (1997) 1571. [15] S.R. Starkey, R. Frech, Electrochim. Acta 42 (1997) 471. [16] M. Nookala, B. Kumar, S. Rodrigues, J. Power Sources 111 (2002) 165. [17] J. Ennari, L. Pietila, V. Virkkunen, F. Sundholm, Polymer 43 (2002) 5427. [18] S. Kim, J.Y. Kang, S.G. Lee, J.R. Lee, S.J. Park, Polymer (Korea) 29 (2005) 403. [19] X.A. Fu, S. Qutubuddin, J. Colloid Interface Sci. 283 (2005) 373. [20] X. Qian, N. Gu, Z. Cheng, X. Yang, E. Wang, S. Dong, Electrochim. Acta 46 (2001) 1829. [21] R.A. Vaia, S. Vasudevan, W. Krawiec, L.G. Scanlon, E.P. Giannelis, Adv. Mater. 7 (1995) 154. [22] R.A. Vaia, R.K. Teukolsky, E.P. Giannelis, Chem. Mater. 6 (1994) 1017. [23] B. Liao, M. Song, H.J. Liang, Y.X. Pang, Polymer 42 (2001) 10007. [24] H.V. Olphen, An Introduction to Clay Colloid Chemistry, Interscience Publishers, New York, 1963. [25] H.W. Chen, F.C. Chang, Polymer 42 (2001) 9763.