Journal of Membrane Science 322 (2008) 416–422
Contents lists available at ScienceDirect
Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci
Effects of the porous structure on conductivity of nanocomposite polymer electrolyte for lithium ion batteries Z.H. Li a,b,c , H.P. Zhang a , P. Zhang a , G.C. Li a , Y.P. Wu a,∗ , X.D. Zhou c a
Department of Chemistry & Shanghai Key Laboratory of Molecular Catalysis and Innovative Materials, Fudan University, Shanghai, 200433 China College of Chemistry, Xiangtan University, Xiangtan, 411105, China c State Key Laboratory of Chemical Engineering, East China University of Science and Technology, Shanghai, 200237, China b
a r t i c l e
i n f o
Article history: Received 28 February 2008 Received in revised form 28 May 2008 Accepted 31 May 2008 Available online 13 June 2008 Keywords: Porous polymer membrane Lithium ion battery Ionic conductivity Polymer electrolyte In situ hydrolysis
a b s t r a c t A kind of porous nanocomposite polymer membranes (NCPMs) based on poly(vinylidene diﬂuoride-cohexaﬂuoropropylene) (P(VdF-HFP)) incorporated with different amounts of TiO2 nanoparticles from in situ hydrolysis of Ti(OC4 H9 )4 was prepared by a non-solvent induced phase separation (NIPS) technology. The SEM micrographs reveal that a porous structure exists in the NCPMs, which changes with the incorporated amount of TiO2 . The NCPMs incorporated with 9.0 wt.% of mass fraction of TiO2 possess the highest porosity, 67.3%, and appear as ﬂexile fracture with an elongation ratio, 74.4%. At this content, the ionic conductivity of the NCPE is up to 0.94 × 10−3 S cm−1 at room temperature and the activation energy for ions transport reaches the lowest, 18.71 kJ mol−1 . It is of great potential application in lithium ion batteries. © 2008 Elsevier B.V. All rights reserved.
1. Introduction The porous polymer membranes based on poly(vinylidene diﬂuoride) (PVdF) have been widely applied in micro-ﬁltration, ultra-ﬁltration, protein adsorption, immobilization and separation, waste water treatment, proton and ionic conductivity , and controlled release of drugs. They can be fabricated by a non-solvent induced phase separation (NIPS) technique or a thermally induced phase separation (TIPS) technique [2–6]. In the NIPS process, the casting polymer solution is ﬁrstly cast on a smooth substrate, and then immersed in a bath of non-solvent with respect to PVdF such as water or alcohol. By exchanging the non-solvent with the solvent, phase separation takes place ﬁnally forming a porous structure in the polymer membranes and it is called “wet NIPS”. But the resulting porous polymer membranes shrink easily after drying at an elevated temperature and exhibit a poor mechanical property. Yet this kind of phase inversion process can be realized by another form, for example, the “dry NIPS” could be obtained by changing the component of the solvent and the non-solvent of polymer solution at an elevated temperature [7–9]. The resulting polymer membranes are dimensionally stable and of good mechanical property.
∗ Corresponding author. Tel.: +86 21 55664223. E-mail address: [email protected]
(Y.P. Wu). 0376-7388/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.memsci.2008.05.074
With ﬁrst success in preparation of the micro-porous gel polymer electrolyte (MGPE) based on poly(vinylidene diﬂuorideco-hexaﬂuoropropylene) (P(VdF-HFP)) copolymer in 1996 , the copolymer has been extensively studied for application in polymer lithium ion batteries (PLIBs) because it has relatively low crystallinity compared with PVdF. Moreover, it only swells in organic solvent such as alkyl carbonate, which is usually used as the solvent of non-aqueous electrolyte, resulting in a good mechanical strength. In addition, many kinds of nanoparticles, such as Al2 O3 , LiAlO2 , MgO, SiO2 , TiO2 , ZrO2 , double-layered hydroxide, molecule sieves, and organo-montmorillonite clays, were directly ﬁlled into the polymer membrane to improve the mechanical strength of MGPE as well as ionic conductivity [11–21]. Because of the high speciﬁc surface energy, nanoparticles easily aggregate within polymer matrix. Many kinds of methods, such as in situ polymerization [22–24], in situ hydrolysis [7,25–27], and surface modiﬁcation of the nanoparticles, have been applied to overcome this problem. The “in situ hydrolysis” may be an effective method to prepare the porous nanocomposite polymer membranes because the yielding nanoparticles are incorporated uniformly into the polymer matrix in situ at the time the porous polymer membranes being produced. Although a high porosity leads to a high conductivity, the mechanical properties of the porous polymer membranes will simultaneously trade off. To merge high conductivity and good mechanical properties together, a kind of porous nanocomposite polymer membranes were prepared by in situ hydrolysis of
Z.H. Li et al. / Journal of Membrane Science 322 (2008) 416–422
Ti(OC4 H9 )4 in a solution of P(VdF-HFP) copolymer in this paper. The mechanical properties of polymer membrane were improved by incorporation of TiO2 nanoparticles. The ionic conductivity of the resulting NCPE is in the order of magnitude of 10−3 S cm−1 at room temperature.
2. Experimental 2.1. Preparation of the NCPM and NCPE One gram P(VdF-HFP) (Knyar® 2801) copolymer was ﬁrst dissolved in 9 ml mixture of the solvent (N-methyl pyrrolidone, NMP) and the non-solvent (n-butanol, n-BuOH) (NMP/n-BuOH is 8:1 in volume ratio). A little of concentrated hydrochloric acid solution (36 wt.%) was added into the obtained polymer solution to adjust its pH to about 1.0 before Ti(OC4 H9 )4 was added. After the viscous solution was stirred over 1 h at room temperature, it was cast on a clean glass plate, and then heated at 60 ◦ C for 6 h. A self-supporting polymer membrane was obtained after evaporation of NMP and nBuOH. The polymer membrane was peeled off from the substrate and immersed in water at 80 ◦ C for 5 h. A porous nanocomposite polymer membrane (NCPM) was ﬁnally produced after drying under vacuum at 80 ◦ C for 24 h. The resulting NCPMs were punched into circular pieces with a diameter of 12 mm and stored in a dry glove box (Unilab Mbraun Com.) for 48 h. After the circular pieces were immersed in 1 mol l−1 solution of LiPF6 in the mixture of ethyl carbonate (EC)–dimethyl carbonate (DMC)–diethyl carbonate (DEC) (mass ratio, 1:1:1) overnight, the nanocomposite polymer electrolytes (NCPEs) were achieved.
2.2. Characterization of the NCPMs Scanning electronic micrograph (SEM) was taken by Philips XL30 D6716 digital scanning electron microscope after spraying gold on the surface of the polymer membranes. Their thermal property was measured by PerkinElmer DSCQ10 under nitrogen atmosphere at 10 ◦ C min−1 . The real amount of TiO2 particles in the NCPMs was determined by thermo-gravimetric analysis (TGA) on WRT-3P micro-thermo-balance (Shanghai Balance Instrument Factory) under air atmosphere heated up to 800 ◦ C at a heating rate of 20 ◦ C min−1 . The infrared spectra were recorded on PerkinElmer FTIR 1710 spectrometer, covering a range of 450–4000 cm−1 with a resolution of 1 cm−1 . X-ray diffraction (XRD) patterns were taken by a Bruker D8 diffractometer ranging from 5◦ to 45◦ . All of the tested samples were dried under vacuum at 80 ◦ C overnight with a thickness of 30 m before XRD and FTIR measurements. The mechanical property of the polymer membranes was measured on the Zwick 20TN2S apparatus, at a crosshead speed of 5 mm min−1 , using a mini-tensile bar with length 2.40 cm, thickness 0.010 cm and width 0.45 cm. The tests were carried out at room temperature. The porosity of the polymer membranes, p, was measured as follows. The membranes were immersed in n-butanol for 1 h and their mass was weighed before and after the absorption of n-butanol. The porosity p of the polymer membranes was calculated based on the Eq. (1). p=
(ma − mp )/a (ma − mp )/a + mp /p
In the equation, ma and mp are the mass of the wet and the dry membranes, respectively, while a and p are the density of n-butanol and the polymer matrix, respectively.
Table 1 Some data of the prepared polymer membranes measured by DSC and their porosity Amount of TiO2 /wt.%
Tc (◦ C)
Hc (J g−1 )
Tm (◦ C)
Hm (J g−1 )
0 3.6 5.1 7.6 9.0 11.4 14.0
101.9 96.2 98.6 99.6 100.1 101.5 97.7
25.27 21.37 22.40 22.17 21.57 22.15 18.34
146.6 143.0 143.3 142.5 142.2 141.3 142.8
19.00 15.16 14.85 14.65 14.27 14.04 13.73
18.2 14.5 14.2 14.0 13.6 13.4 13.1
34.4 43.1 45.8 48.2 67.3 57.1 50.7
2.3. Electrochemical measurement of the NCPEs The NCPEs were sandwiched between two symmetrical stainless steel electrodes and sealed in a model cell, and their electrochemical impedance spectra were measured using an EG&G M273 Potentiostat/Golvannostat in conjunction with M5210 Lockin ampliﬁer electrochemical analysis system in the frequency range 0.1–100 kHz. The ionic conductivity of NCPE was calculated from the Eq. (2). =
d (Rb S)
where is the ionic conductivity, Rb the bulk resistance that was determined by the impedance spectroscopy, d the thickness of the NCPMs, and S the area of the symmetrical electrode. 3. Results and discussion 3.1. Morphology change of the NCPMs Fig. 1 shows the surface and cross-section morphology of the NCPMs incorporated with various amounts of TiO2 nanoparticles. It suggests that the in situ formed nanoparticles have an obvious effect on the porous structure of polymer membranes. Without incorporation with nanoparticles, the polymer membranes had very few pores of irregular shape (Fig. 1a) and a compact polymeric matrix (Fig. 1a1 ). After incorporated with 5.1 wt.% TiO2 nanoparticles, a lot of closed pores like formicary exhibited (Fig. 1b) on the surface of the NCPM. At 9.0 wt.%, many formicary-like pores seem to be open with a diameter of about 2 m. These spherical pores are connected with each other through some smaller pores (Fig. 1c). From its ampliﬁed micrographs (Fig. 1c and c1 ), the formed nanoparticles were dispersed uniformly in polymer matrix with a diameter less than 100 nm and conﬁrmed to be TiO2 by the energy dispersive X-ray spectrum (EDS) (Fig. 1c2 ). Further incorporation of TiO2 up to 14.0 wt.%, the NCPM had some crater-like pores on the surface (Fig. 1d). The hydrolyzed nanoparticles aggregated scarcely from the SEM micrographs when the mass fraction of TiO2 was less than 9.0 wt.%. The data of porosity of the NCPMs were summarized in Table 1. As shown in Table 1, the NCPM containing 9.0 wt.% of nanoparticles has the highest porosity, 67.3%. Without the addition of Ti(OC4 H9 )4 , the non-solvent (n-BuOH) evaporated easily together with NMP from the polymer solution due to its low boiling temperature (118 ◦ C). Before the polymer solution was changed into metastable state, it mostly evaporated away from the casting dope. As a result, polymer chains aggregated slowly in the casting dope forming interlinked spherulites with further evaporation of NMP molecules, as shown in Fig. 1a, since the solid–liquid demixing dominates the phase separation process. The resulting polymer matrix appears compact as shown in Fig. 1a1 . After the addition of Ti(OC4 H9 )4 , the viscosity of the polymer solution increases leading to difﬁcult escaping of solvent and nonsolvent molecules. In this case, the quick liquid–liquid demixing
Z.H. Li et al. / Journal of Membrane Science 322 (2008) 416–422
dominates the phase separation process so the polymer solution develops into a polymer-rich and a polymer-poor phases, which would form the polymer matrix and the macropores, respectively. In addition, as Ti(OC4 H9 )4 is gradually hydrolyzed during drying at elevated temperature, the formed TiO2 nanoparticles from hydrolysis become the nuclei of the polymer chains to accelerate their crystallization resulting in trapping of solvent and non-solvent in polymer gel. Furthermore, the incompletely hydrolyzed TiO2 contains some hydrophilic groups (–OH) and hydrophobic groups (–OC4 H9 ) on the surface, which are compatible with the polymerpoor and polymer-rich phases. It increased the interfacial stability
between the polymer-rich phase and the polymer-poor phase. Besides that, owing to the surface tension most n-BuOH was dispersed in the polymer-poor phase like many spherical micelles and left a larger number of macropores within polymer membrane after exchanging with H2 O during the further immersion in water. When a little amount of Ti(OC4 H9 )4 , for example 5.1 wt.%, was added, the in situ hydrolyzed Ti(OH)4 nanoparticles accelerated the re-crystallization of the polymer chains resulting in immediate phase separation and thus a little number of macropores formed in the polymer-poor phase as shown in Fig. 1b1 . At 9.0 wt.%, the incompletely hydrolyzed Ti(OH)x (OC4 H9 )4−x (0 < x < 4) stabilized
Fig. 1. Micrographs of the partial prepared polymer membranes containing different amounts of TiO2 . Surface: (a) 0 wt.%, (b) 5.1 wt.%, (c) 9.0 wt.%, (d) 14.0 wt.%; cross-section: (a1 ) 0 wt.%, (b1 ) 5.1 wt.%, (c1 ) 9.0 wt.%, (d1 ) 14.0 wt.%; and the ampliﬁed micrographs of sample c: (c ), (c1 ) and the energy dispersive X-ray spectrum (EDS) of sample c (c2 ).
Z.H. Li et al. / Journal of Membrane Science 322 (2008) 416–422
Fig. 1. (Continued ).
the interface between the polymer-rich and -poor phases, a larger number of macropores produced in the polymer-poor phase as shown in Fig. 1c1 . The viscosity of the polymer solution increased further with the addition of Ti(OC4 H9 )4 (14.0 wt.%) resulting in an increase of surface tension of the casting dope. It indicates that solvent and non-solvent molecules escape difﬁcultly from the surface of casting dope forming the cater-like pores on the surface of polymer membrane as shown in Fig. 1d1 . Moreover, a few of micelles stayed in the polymer membrane even after immersion in water, leading to a decrease of the porosity. 3.2. DSC measurement of the NCPMs The DSC curves of the NCPMs incorporated with various amounts of TiO2 particles are depicted in Fig. 2. It suggests that the melting temperature (Tm ) of the NCPMs is about 2 ◦ C lower than that of the virginal polymer membranes because of the interaction between TiO2 nanoparticles and polymer chains. However, the amount of TiO2 has little effect on the Tm of the NCPMs. The
crystallinity (Xc ) of NCPMs can be calculated from Eq. (3).
Hm ∅ Hm
∅ is the melting heat of the virginal ␣-PVdF crystalline, where Hm 104.7 J g−1  and Hm the heat of fusion of NCPMs. Hm can be calculated from the integral area of the DSC curves. The values of Xc are summarized in Table 1. The crystallinity of NCPMs decreases slightly with the increase of TiO2 amount. During the phase separation the rearrangement of polymer chains may be disordered due to the nucleation effect of nanoparticles, which results from the interaction between the Lewis acid groups (–OH) on the surface of nanoparticles and the polar CF2 groups of the polymer chains. As a result, the crystallinity of polymeric matrix decreases slightly and the melting temperature lowers. This is consistent with the addition of nanoﬁllers in PEObased polymer electrolytes .
3.3. XRD and FTIR characterization of the NCPMs
Fig. 2. DSC curves of the prepared polymer membranes.
Fig. 3 shows that the characteristic diffraction peaks of the P(VdF-HFP) crystals located at the angles of 20.47◦ and 18.09◦ can be assigned as the planes of (110, ␣) and (020, ␣), respectively . After incorporating with TiO2 nanoparticles, these two peaks hardly shift indicating that there are few changes of the crystal structure of the polymer matrix. However, the half widths of the crystal peaks decrease with the increase of the nanoparticle amount, indicating an increase in the amount of the amorphous phase, which is consistent with the above DSC measurement. Fig. 4 shows the FTIR spectra of the virginal P(VdF-HFP) copolymer and the NCPMs. In the case of the virginal P(VdF-HFP), its characteristic peaks are clearly present, for example 3027 (CH2 asymmetric stretching), 2988 (CH2 symmetric stretching), 1688 (–CH = CF– skeletal bending), 1408 (–C–F stretching), 1267–1164 (–C–F and –CF2 – stretching), 1074 (C–C skeletal vibration), 880 (vinylidene group), 839 (CH2 rocking), 772 (CF3 stretching), 684 (CH2 bending), 511 (CF2 bending) and 484 cm−1 (CF2 wagging) [31–35].
Z.H. Li et al. / Journal of Membrane Science 322 (2008) 416–422
Fig. 5. Stress–strain curves for some polymer membranes.
Fig. 3. XRD patterns of the prepared polymer membranes.
In the case of the nanocomposite polymer membranes, the skeletal bending vibration peak (1688 cm−1 ) disappears while some new vibration peaks appear at 3119, 1636, and 1507 cm−1 . The new wide absorption peaks at 3119–3278 and at 1640 cm−1 are assigned to the stretching and bending vibration of –OH groups on the surface of nanoparticles, respectively. The vibration peak at 1507 cm−1 perhaps belongs to the TiO2 . The integral areas of the absorption peaks at 1408, 511, and 484 cm−1 of P(VdF-HFP) decrease due to the interactions between the nanoparticles and the polymer chains suggesting that the polarity of the CF2 groups decreases. Because the position of the characteristic vibration peaks of CH2 groups are nearly unchanged, there are no hydrogen bonds between these groups and the added nanoparticles. Nevertheless, the Lewis acid effect of nanoparticles decreased the polarity of CF2 groups, facilitating to weaken the interaction between Li+ ions and polymer chains.
brane is up to 19.40 MPa, and its fracture elongation ratio is 16.3%. After adding nanoparticles, the maximum stress of the NCPMs decreases while the fracture elongation ratio increases obviously. The enhanced fracture elongation ratio is owed to the deformation of the pores within the polymer matrix and the interfacial layers between nanoparticles and polymer matrix. The fracture of the virginal P(VdF-HFP) membrane appears fragile whereas the NCPMs show a ﬂexile fracture. At 9.0 wt.%, the maximal fracture elongation ratio of the NCPM increases to 74.4% while its maximum stress was down to 9.20 MPa. It suggests that the NCPMs have a good reﬂexibility beneﬁting the fabrication of polymer lithium-ion batteries (PLIBs). 3.5. Ionic conductivity of the NCPEs
Fig. 5 shows the stress–strain curves of some prepared polymer membranes. The maximum stress of the virginal P(VdF-HFP) mem-
Fig. 6 shows the dependence of ionic conductivity on temperature for the gelled polymer electrolytes. The ionic conductivity of the NCPEs increases with the increase of temperature for all of the polymer electrolytes. The ionic conductivity at room temperature, as shown in Fig. 7, increases ﬁrstly with the amount of the nanoparticles, and then decreases. At 9.0 wt.%, the NCPE arrives at the highest ionic conductivity, 0.94 × 10−3 S cm−1 . The imitated curves in Fig. 6 are linear, and the conductive behavior of the NCPEs
Fig. 4. FTIR spectra of the prepared polymer membranes.
Fig. 6. log – 1/T curves for the prepared polymer electrolytes.
3.4. Mechanical properties of the NCPMs
Z.H. Li et al. / Journal of Membrane Science 322 (2008) 416–422
9.0 wt.% TiO2 nanoparticles. At this content, the highest ionic conductivity of the obtained NCPE is up to 0.94 × 10−3 S cm−1 at room temperature and its apparent activation energy for ions transport is minimal, 18.71 kJ mol−1 . Acknowledgements Financial support from the National Basic Research Program of China (973 Program No: 2007CB209700), Chinese Postdoctoral Science Fund (20060400618), Natural Science Foundation of Hunan Province (06JJ4093), the Open Fund from State Key Laboratory of Chemical Engineering in East China University of Science and Technology (HF06008), and the Doctoral Starting Fund from Xiangtan University (05QDZ18) is greatly appreciated. References Fig. 7. The relationship between ionic conductivity at room temperature and apparent activation energy for ions transport of the polymer electrolytes with the added amount of TiO2 nanoparticles.
follows the Arrhenius Eq. (4) in the measured temperature range. = 0 exp
where is the ionic conductivity, 0 , Ea , R, and T are the pre-exponential factor, the apparent activation energy for ions transport, the gas constant, and the absolute temperature, respectively. Therefore, the apparent activation energy for ions transport can be calculated from the Arrhenius equation and is also presented in Fig. 7. The apparent activation energy for ion transport of the NCPE was down to the minimum value, 18.71 kJ mol−1 when the NCPM is incorporated with 9.0 wt.% nanoparticles. It is well known that there are three phases in the porous gel polymer electrolyte: solid-state polymer matrix, the gelled polymeric matrix coming from the swelling of organic components in non-aqueous electrolyte, and absorbed liquid electrolyte. The solidstate polymer matrix ensures the mechanical strength of the gel polymer electrolyte, and the others provide a high ionic conductivity. In our case, there is a new phase, TiO2 nanoparticles, in the polymer electrolytes. The nanoparticles inﬂuence, on one hand, the porous structure of the polymer membrane. On the other hand, the Lewis acid–base effect, which comes from the interaction between –OH groups on the surface of nanoparticles and polar CF2 groups of polymer chains, can weaken the interaction between Li+ ions and polymer chains facilitating their migration. However, there is a trading-off between the Lewis acid–base effect to decrease the amount of the –OH groups on the surface of nanoparticles and the aggregation of nanoparticles. At 9.0 wt.%, the NCPM reaches to the highest porosity, so its apparent activation energy for ions transport is down to the minimum value and the ionic conductivity reaches to the highest value. 4. Conclusion By in situ hydrolysis of Ti(OC4 H9 )4 , the yielding TiO2 nanoparticles were incorporated in situ into the porous P(VdF-HFP) copolymer membranes that were prepared by a dry non-solvent induced phase separation technique. The crystallinity of the NCPMs decreases with the increase of the incorporated amount of TiO2 nanoparticles, and the porous structure of the NCPMs changes with the amount of TiO2 nanoparticles. Due to the deformation of the pores within the polymer matrix and the interfacial layers between the nanoparticles and the polymer matrix, the NCPM exhibits a ﬂexile fracture with an elongation ratio of 74.4% when it contains
 M. Ulbricht, Advanced functional polymer membranes, Polymer 47 (2006) 2217.  Z.H. Xu, L. Li, F.W. Wu, S.J. Tan, Z.B. Zhang, The application of the modiﬁed PVdF ultraﬁltration membranes in further puriﬁcation of Ginkgo biloba extraction, J. Membr. Sci. 255 (2005) 125.  Y.H. Zhao, B.K. Zhu, X.T. Ma, Y.Y. Xu, Porous membranes modiﬁed by hyperbranched polymers: I. Preparation and characterization of PVdF membrane using hyperbranched polyglycerol as additive, J. Membr. Sci. 290 (2007) 222.  G.L. Ji, C.H. Du, B.K. Zhu, Y.Y. Xu, Preparation of porous PVdF membrane via thermally induced phase separation with diluent mixture of DBP and DEHP, J. Appl. Polym. Sci. 105 (2007) 1496.  X.F. Li, X.L. Lu, Morphology of polyvinylidene ﬂuoride and its blend in thermally induced phase separation process, J. Appl. Polym. Sci. 101 (2006) 2944.  C. Lee, J.H. Kim, J.Y. Bae, Polymer gel electrolytes prepared by thermal curing of poly(vinylidene ﬂuoride)-hexaﬂuoropropene/poly(ethylene glycol)/propylene carbonate/lithium perchlorate blends, Polymer 44 (2003) 7143.  X.M. He, Q. Shi, X. Zhou, C.R. Wan, C.Y. Jiang, In situ composite of nano SiO2 P(VdF-HFP) porous polymer electrolytes for Li-ion batteries, Electrochim. Acta 51 (2005) 1069.  S.R. Huang, G.Z. Wu, S.M. Chen, Preparation of microporous poly(vinylidene ﬂuoride) membranes via phase inversion in supercritical CO2 , J. Membr. Sci. 293 (2007) 100.  W.H. Pu, X.M. He, L. Wang, C.Y. Jiang, C.R. Wan, Preparation of PVdF-HFP microporous membrane for Li-ion batteries by phase inversion, J. Membr. Sci. 272 (2006) 11.  A.S. Gozdz, C.N. Schmutz, J. M. Tarascon, Lithium secondary battery extraction method, US Pat. 554074l (1996).  N.T. Kalyana Sundaram, A. Subramania, Nano-size LiAlO2 ceramic ﬁller incorporated porous PVdF-co-HFP electrolyte for lithium-ion battery applications, Electrochim. Acta 52 (2007) 4987.  A.M. Stephan, K.S. Nahm, T.P. Kumar, M.A. Kulandainathan, G. Ravi, J. Wilson, Nanoﬁller incorporated poly(vinylidene ﬂuoride–hexaﬂuoropropylene) (PVdFHFP) composite electrolytes for lithium batteries, J. Power Sources 159 (2006) 1316.  K.M. Kim, N.G. Park, K.S. Ryu, S.H. Chang, Characteristics of P(VdF-HFP)/TiO2 composite membrane electrolytes prepared by phase inversion and conventional casting methods, Electrochim. Acta 51 (2006) 5636.  J.Y. Xi, X.P. Qiu, L.Q. Chen, PVdF-PEO/ZSM-5 based composite microporous polymer electrolyte with novel pore conﬁguration and ionic conductivity, Solid State Ionics 177 (2006) 709.  M. Wachtler, D. Ostrovskii, P. Jacobsson, B. Scrosati, A study on PVdF-based SiO2 -containing composite gel-type polymer electrolytes for lithium batteries, Electrochim. Acta 50 (2004) 357.  H.P. Zhang, P. Zhang, Z.H. Li, M. Sun, Y.P. Wu, H.Q. Wu, A novel sandwiched membrane as polymer electrolytes for lithium ion battery, Electrochem. Commun. 9 (2007) 1700.  A. Subramania, N.T.K. Sundaram, A.R.S. Priya, G.V. Kumar, Preparation of a novel composite micro-porous polymer electrolyte membrane for high performance Li-ion battery, J. Membr. Sci. 294 (2007) 8.  Y.X. Jiang, Z.F. Chen, Q.C. Zhuang, J.M. Xu, Q.F. Dong, L. Huang, S.G. Sun, A novel composite microporous polymer electrolyte prepared with molecule sieves for Li-ion batteries, J. Power Sources 160 (2006) 1320.  Z.H. Li, G.Y. Su, X.Y. Wang, D.S. Gao, Micro-porous P(VdF-HFP)-based polymer electrolyte ﬁlled with Al2 O3 nanoparticles, Solid State Ionics 176 (2005) 1903.  X.J. Wang, J.J. Kang, Y.P. Wu, S.B. Fang, Novel composite polymer electrolytes based on poly(ether-urethane) network polymer and modiﬁed montmorillonite, Electrochem. Commun. 5 (2003) 1025.  X.J. Wang, H.P. Zhang, J.J. Kang, Y.P. Wu, S.B. Fang, Novel composite polymer electrolytes based on poly(ether-urethane) network polymer and fumed silicas, J. Solid State Electrochem. 11 (2006) 21.  B. Lepoittevin, N. Pantoustier, M. Devalckenaere, M. Alexandre, D. Kubies, C. Calberg, R. Jerome, P. Dubois, Poly(-caprolactone)/clay nanocomposites
Z.H. Li et al. / Journal of Membrane Science 322 (2008) 416–422 by in-situ intercalative polymerization catalyzed by dibutyltin dimethoxide, Macromolecules 35 (2002) 8385. X. Fan, C. Xia, R.C. Advincula, Grafting of polymers from clay nanoparticles via in situ free radical surface-initiated polymerization: monocationic versus bicationic initiators, Langmuir 19 (2003) 4381. M. Avella, M.E. Errico, E. Martuscelli, Novel PMMA/CaCO3 nanocomposites abrasion resistant prepared by an in situ polymerization process, Nano Lett. 1 (2001) 213. Y.J. Wang, D. Kim, Crystallinity, morphology, mechanical properties and conductivity study of in situ formed PVdF/LiClO4 /TiO2 nanocomposite polymer electrolytes, Electrochim. Acta 52 (2007) 3181. Z.H. Li, H.P. Zhang, P. Zhang, Y.P. Wu, X.D. Zhou, Macroporous nanocomposite polymer electrolyte for lithium ion batteries, J. Power Sources, doi:10.1016/j.jpowsour.2008.02.068. Z.H. Li, P. Zhang, H.P. Zhang, Y.P. Wu, X.D. Zhou, A lotus root-like porous nanocomposite polymer electrolyte, Electrochem. Commun. 10 (2008) 791. Y. Rosenberg, A. Sigmann, M. Narkis, S. Shkolnik, The sol/gel contribution to the behavior of ␥-irradiated poly(vinylidene ﬂuoride), J. Appl. Polym. Sci. 43 (1991) 535.
 F. Croce, G.B. Appetecchi, L. Persi, B. Scrosati, Nanocomposite polymer electrolytes for lithium batteries, Nature 394 (1998) 456.  K.M. Kim, J.C. Kim, K.S. Ryu, Characteristics of P(VdF-HFP)/TiO2 composite electrolytes prepared by a phase inversion technique using dimethyl acetamide solvent and water non-solvent, Macromol. Mater. Eng. 291 (2006) 1495.  G.Q. Zhai, E.T. Kang, K.G. Neoh, Inimer graft-copolymerized poly(vinylidene ﬂuoride) for the preparation of arborenscent copolymers and “surface-active” copolymer membranes, Macromolecules 37 (2004) 7240.  F.A. Landis, R.B. Moore, Blends of a perﬂuorosulfonate ionomer with poly(vinylidene ﬂuoride): effect of counterion type on phase separation and crystal morphology, Macromolecules 33 (2000) 6031.  Y.T. Shieh, T.T. Hsiao, S.K. Chang, CO2 pressure effects on melting, crystallization, and morphology of poly(vinylidene ﬂuoride), Polymer 47 (2006) 5929.  T. Boccaccio, A. Bottino, G. Capannelli, P. Piaggio, Characterization of PVdF membranes by vibrational spectroscopy, J. Membr. Sci. 210 (2002) 315.  X.Z. Tian, X. Jiang, B.K. Zhu, Y.Y. Xu, Effect of the casting solvent on the crystal characteristics and pervaporative separation performances of P(VDF-co-HFP) membranes, J. Membr. Sci. 279 (2006) 479.