- Email: [email protected]
A simple P(VdF-HFP)–LiTf system yielding highly ionic conducting and thermally stable solid polymer electrolytes
Journal of Molecular Liquids 177 (2013) 73–77
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Liquids journal homepage: www.elsevier.com/locate/molliq
A simple P(VdF-HFP)–LiTf system yielding highly ionic conducting and thermally stable solid polymer electrolytes S. Ramesh a,⁎, Soon-Chien Lu b a b
Centre for Ionics University Malaya, Department of Physics, Faculty of Science, University of Malaya, 50603 Kuala Lumpur, Malaysia Centre for Surface Chemistry and Catalysis, Faculty of Bioengineering Science, Katholieke Universiteit Leuven, 3001 Leuven, Belgium
a r t i c l e
i n f o
Article history: Received 21 December 2011 Received in revised form 21 September 2012 Accepted 25 September 2012 Available online 4 October 2012 Keywords: Polymer membrane Material testing Thermal property Poly(vinylidene fluoride-cohexafluoropropylene) (PVdF-HFP) Ionic conductivity
a b s t r a c t In the present work, we report a simple solid polymer electrolyte (SPE) system that is solely based on poly(vinylidene fluoride-co-hexafluoropropylene) [P(VdF-HFP)] and lithium trifluoromethanesulfonate (LiTf). The SPEs produced exhibit high ionic conductivity of ~ 10−4 S cm −1 at ambient temperature when 40 wt.% of LiTf is incorporated. There is an anomaly when moderate amount of LiTf is added into the polymer. This can be related to formation of neutral ion pairs, and substantiated by calculation of activation energy and scanning electron micrographs. TGA thermograms also show that all the samples tested are thermally stable up to 400 °C, even with the incorporation of LiTf. The highest ionic conducting sample achieves highest onset temperature of decomposition. © 2012 Elsevier B.V. All rights reserved.
1. Introduction The development of high energy density, easily portable and safe power sources is one of the key global technological challenges, as driven by the desire of shifting away from the current heavy dependence on fossil fuels. Solid polymer electrolytes (SPEs) fulfill these requirements and overcome the limitations of conventional liquid electrolytes by addressing drawbacks such as electrolyte leakage, flammable organic solvent, and electrolytic degradation of electrolytes [1]. However, SPEs suffer from one of the important industrial aspects, which is low ionic conductivity. There have been many efforts in improving the ionic transport properties of polymer electrolytes to enhance their power application feasibility. Poly(vinylidene fluoride-co-hexafluoropropylene) [P(VdF-HFP)] has received great attention due to its outstanding thermal and electrochemical stability [2]. This copolymer exhibits amorphous PHFP phase that aids ionic conduction, and at the same time crystalline PVdF phase acts as a mechanical support [3]. This satisfies the two contradictory properties of highly ionic conducting and good mechanical strength. Recently, P(VdF-HFP) has shown potential as polymer electrolyte material of rechargeable lithium batteries owing to its high solubility, low crystallinity and glass transition temperature [4]. Many researchers present encouraging results for P(VdF-HFP)-based polymer electrolytes
⁎ Corresponding author. Tel.: +60 3 7967 4391. E-mail address: [email protected] (S. Ramesh). 0167-7322/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.molliq.2012.09.018
with different types of dopant salts, such as lithium fluoroalkyl phosphate [5], lithium perchlorate [6] and lithium bis(trifluoromethanesulfonyl) imide [7]. Lithium trifluoromethanesulfonate salt, LiTf, is one of the most common lithium salt used in polymer electrolyte research. In this study, we utilize only P(VdF-HFP) and LiTf to study about their interaction, electrochemical and thermal properties. Without any other additives or chemicals, we can examine the basic interaction of ionic conduction in polymer and influence on thermal stability. Our studies find that even when P(VdF-HFP) is incorporated with LiTf, which is generally regarded as having lower ionic conductivity as compared to other lithium salts of its family [8], the SPEs produced exhibit exceptional ionic conductivity, thermal stability and mechanically stable with the coexistence of amorphous and crystalline phases. There are great potentials for this system to be further investigated with other additives such as plasticizers and ceramic fillers.
2. Experimental 2.1. Materials Poly(vinylidene fluoride-co-hexafluoropropylene), [P(VdF-HFP)] with average Mw = 455,000, was obtained from Sigma-Aldrich. Lithium trifluoromethanesulfonate salt, LiTf [LiCF3SO3], was obtained from Fluka and dried at 100 °C for 1 hour to eliminate trace amounts of water in the material, prior to the preparation of solid polymer electrolytes (SPEs). Acetone of AR grade was obtained from J.T. Baker.
74
S. Ramesh, S.-C. Lu / Journal of Molecular Liquids 177 (2013) 73–77
2.2. Preparation of thin films Several polymer electrolyte complexes were prepared as in the compositions according to its weight ratio of polymer to lithium salt, ranging from 95/5 to 60/40. The thin films were prepared by solution casting technique with acetone as solvent. The mixture was cast on a Petri dish and allowed to evaporate slowly inside a fume hood. This procedure yields mechanically stable and free standing films. The thickness of the films was measured using a micrometer screw gauge. 2.3. Instrumentation The samples were cut out resembling shape of stainless steel blocking electrodes used in this study and sandwiched between them. A HIOKI 3532-50 LCR Hi-Tester was used to perform the impedance measurements for each SPE over the frequency range of 50 Hz to 1 MHz. The same measurement was also performed for selected samples from 30 to 80 °C. Those selected SPEs were then subjected to scanning electron microscopy (SEM), with the model Leica S440. Thermogravimetric analysis (TGA) was performed with a Mettler Toledo analyzer that consists of a TGA/SDTA851 e main unit and STARe software with 10 °C min − 1 heat rate between 30 and 500 °C under a nitrogen atmosphere. 3. Results and discussion 3.1. Alternating current (AC) impedance studies Fig. 1 shows two Cole–Cole plots of samples with (a) 5 and (b) 40 wt.% of LiTf. It can be observed in Fig. 1(a) that for low concentration of lithium salt, a semi-circle with a long tail, or more commonly known as electrode spike, is clearly observed. When the salt content increases, the semicircle diminishes completely and leaves only with a sharp tail as in Fig. 1(b). This is typical for a very ionic conductive thin film sample. There are two axes in a Cole–Cole plot, namely real and imaginary part of total impedance Z. The lowest point of imaginary Z axis is extrapolated to determine the bulk resistance, Rb. The ionic conductivity of thin film sample can then be calculated using Eq. (1): ð1Þ
where σ is ionic conductivity, l is thickness of the thin film sample, Rb is bulk resistance determined from impedance spectroscopy and A is surface area of stainless steel blocking electrodes. Fig. 2 illustrates the variation of log ionic conductivity values as a function of lithium triflate (LiTf) weight percentage incorporated into P(VdF-HFP). It can be clearly observed that the ionic conductivity increases exponentially with the addition of LiTf at the beginning, then becomes relatively constant and reaches the highest ionic conductivity of 1.56×10−4 S cm−1 at 40 wt.% of LiTf. The initial increase is simply due to the increased availability of Li+ ions in the system. However, from 15 wt.% onwards, the ionic conductivity decreases gradually until 30 wt.%. This may be due to neutral ion pair formation from Li+ and Tf− ions, or more commonly known as ion-pair effect. High concentration of salt has higher tendency towards this type of formation. This reduces the availability of free mobile ions and possibly obstructs the conducting path, which subsequently contribute to the slight decrease of ionic conductivity. It can also be related to the formation of linkages between the salt itself, causing it to crystallize, thus resulting in the decrease of ionic conductivity [9]. This proposition can be substantiated by activation energy calculation and scanning electron microscopy (SEM) images in latter sections. When the concentration of LiTf exceeds 30 wt.%, the presence of excessive Li+ and Tf− ions starts to disrupt and convert the crystalline phase of P(VdF-HFP) to amorphous, thus changing the morphology of polymer and further enhances the ionic conductivity.
Fig. 1. Cole–Cole plots of samples with (a) 5 and (b) 40 wt.% of LiTf.
The ionic conductivity is also measured at different temperatures, from 30 to 80 °C, in order to evaluate the ionic conduction mechanism in the SPEs. The variation of log ionic conductivity as a function of inverse absolute temperature for selected samples, with 10, 25 and 40 wt.% LiTf respectively, coded as LT10, LT25 and LT40 respectively, -2
Log σ [ lg (S cm -1 ) ]
σ ¼ l=Rb A
-4
-6
-8 0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
Wt% of LiTf Content Fig. 2. Variation of log ionic conductivity as a function of lithium triflate (LiTf) weight percentage incorporated into P(VdF-HFP).
S. Ramesh, S.-C. Lu / Journal of Molecular Liquids 177 (2013) 73–77
is shown in Fig. 3. The regression values of all three chosen samples are close to unity indicating that the temperature-dependent ionic conductivity for this SPE system obeys the Arrhenius rule. As the conductivity–temperature data follow Arrhenius behavior, the nature of cation transport is deduced to be similar to that in ionic crystals, where ions jump into neighboring vacant sites and thus increase the ionic conductivity to a higher value [10]. The motion of ions in solid polymer electrolytes is a liquid-like mechanism, by which the movement of ions through the polymer matrix is assisted by the large amplitude of the polymer segmental motion. Thus, greater segmental motion at higher temperatures either permits the ions to hop from one site to another or provides a pathway for ions to move with faster ionic conduction. Activation energy of several selected samples has been calculated from the Arrhenius equation shown in Eq. (2) below: σ ¼ σ o eð
−RTa Þ E
ð2Þ
where σ is ionic conductivity, σo is the pre-exponential factor, Ea is the apparent activation energy, R is universal gas constant and T is the absolute temperature. The results are tabulated in Table 1. As expected, LT40 has the lowest activation energy, as compared to the other two samples. Interestingly, a slightly steeper slope has been noted for LT25, which means this particular sample needs higher energy for ionic transport. This can be related to the explanation given above. LT25 has much higher activation energy than that of LT10, even though its ionic conductivity is slightly higher compared to that of LT10. This anomaly can be comprehended as the supplied heat provides extra energy for those aggregated neutral ion pairs to redissociate again [11]. As higher amount of free Li + and Tf − ions are released from neutral ion pairs upon heating, the ionic conductivity also raises significantly. Another plausible explanation is by assuming that polymer matrix with high amount of aggregated neutral ion pairs would have smaller pore size. Higher energy is required for Li+ ions to transport through polymer matrix with smaller pore size. Other than salt concentration, the conductivity of thin film also changes with frequency applied to it. The conductance of the samples is calculated with the equation: log σ ¼ log
GL A
ð3Þ
where G denotes the conductance in V, L as the thickness of thin film in cm and A as the surface area of stainless steel blocking electrodes in cm 2. Again LT40 is selected as it shows highest ionic conductivity among all samples. Variation of logarithm conductance as a function of frequency for LT40 at 298, 303, 313, 323, 333, 343 and 353 K has been shown in Fig. 4.
75
Table 1 Activation energies of sample LT10, LT25 and LT40 calculated using Arrhenius's equation. Sample
Calculated activation energy
LT10 LT25 LT40
1.761 kJ mol−1 5.901 kJ mol−1 1.057 kJ mol−1
Generally, the conductance of samples is found to increase with increasing frequency, as well as temperature. At lower frequency region, the charges in the samples are easily accumulated at the electrodes and electrode interfaces, which causes to a decline in free mobile ions and eventually a decrease in conductance at lower frequencies. At higher frequency region, the conductance increases with frequency due to greater mobility of charge carriers and faster hopping of ions. Therefore, the ion exchange process occurs more effectively at higher frequencies. As temperature increases, general trend of conductivity also shifts accordingly. A hump can be clearly observed at all temperatures for LT40. The shifting of hump from higher to lower frequency with increasing temperature can be related to the undissociated LiTf salt present in the sample that is being released as free mobile ions upon heating, resulting further increase in conductance, which has already been discussed in the Ea section above. 3.2. Scanning electron microscopy Scanning electron micrographs of (a) pure P(VdF-HFP), (b) LT10, (c) LT25 and (d) LT40 have been shown in Fig. 5. Very distinguishable changes can be observed from pure P(VdF-HFP), to low and high concentrations of LiTf. Pure P(VdF-HFP) shows normal porous surface with uniform small pore size. The morphology changes, as soon as 10 wt% of LiTf is doped into the polymer, to become separated by layers and greater pore size. However, when it reaches 25 wt%, the morphology has become more similar to that of pure polymer, only with visibly larger pore size. This further confirms the justification regarding neutral ion pairs formation discussed above, apart from the higher activation energy calculated. When more lithium salt is added into the polymer, the morphology changes again to become similar to that of LT10, but with significantly more layered and greater pore size. These features are very similar to those reported by Pandey and Hashmi [12]. According to free-volume model, this microstructure leads to the better conducting pathway for Li+ ions, and consequently leading to higher ionic conductivity [13]. In summary, the observations in these micrographs are congruent with the findings of ionic conductivity in the previous section. 3.3. Thermal studies The samples of the pure P(VdF-HFP), LT10, LT25 and LT40 have been subjected to thermogravimetric analysis (TGA) in nitrogen atmosphere,
Log σ [ lg (S cm -1 ) ]
-3
-4
-5 2.80
2.90
3.00
3.10
3.20
3.30
-1
1000/T (K ) Fig. 3. Variation of log ionic conductivity as a function of inverse absolute temperature for samples LT10, LT25 and LT40.
Fig. 4. Variation of logarithm conductance as a function of frequency for LT40 at 25 to 80 °C.
76
S. Ramesh, S.-C. Lu / Journal of Molecular Liquids 177 (2013) 73–77
Fig. 5. Scanning electron micrographs of (a) pure P(VdF-HFP), samples (b) LT10, (c) LT25 and (d) LT40, all with the magnification of 5000×.
to examine their thermal stability. The TGA thermograms in Fig. 6 show that all the samples are thermally stable up to 400 °C, even with the addition of LiTf. This is highly desirable for its potential applications in electrochemical devices. The weight loss percentage increases upon the addition of LiTf. This is mainly due to the presence of Tf− group, which can be burnt off easily as carbon oxides, sulfur oxides and hydrogen fluoride, as compared to the polymer host and Li+ ions (Table 2). However, when the concentration
100 90 80
Wt%
70 60
of LiTf is increased to 25 wt%, the weight loss percentage only decreases marginally. This may be due to the formation of neutral ion pairs, which makes it more difficult to be burnt off, as more Li+ and Tf− ions are bonded ionically. Much energy is needed to break the ionic bond, therefore slightly higher energy is required to decompose LT25, as compared to that of LT10. When the concentration goes up to 40%, the weight loss percentage drastically decreases to 51.22%. This may be due to greater amount of lithium metal ions and higher interaction between lithium salt and polymer host, making the sample more difficult to be decomposed. The decomposition temperature generally increases upon the addition of LiTf, approximately from 400 to 430 °C. Although LT40 suffers from having the slightly higher weight loss percentage than pure P(VdF-HFP), it is compensated with improved decomposition temperature and highest onset temperature of decomposition. This can be explained as the higher
50 40
Table 2 Weight loss percentages, decomposition temperatures and onset temperatures of decomposition of pure P(VdF-HFP), LT10, LT25 and LT40 obtained from TGA thermograms.
30 20 10
Sample
% Weight loss
Decomposition temperature, TD
Onset temperature of decomposition
Pure PVdF-HFP LT10 LT25 LT40
46.21% 64.53% 63.53% 51.22%
405.94 430.38 436.20 432.48
384.74 406.56 405.38 413.61
0 0
50
100
150
200
250
300
350
400
450
500
Temperature (oC) Fig. 6. Normalized TGA thermograms of pure P(VdF-HFP) (−), samples LT10 (●), LT25 (▲) and LT40 (■) in nitrogen atmosphere.
°C °C °C °C
°C °C °C °C
S. Ramesh, S.-C. Lu / Journal of Molecular Liquids 177 (2013) 73–77
level of lithium metal ions present, which requires higher energy to be decomposed. 4. Conclusion A series of solid polymer electrolytes (SPEs) that is solely based on poly(vinylidene fluoride-co-hexafluoropropylene) [P(VdF-HFP)] and lithium trifluoromethanesulfonate (LiTf) has been synthesized by solution casting technique. The SPEs produced reach ionic conductivity as high as 1.56 ×10−4 S cm−1 at 40 wt% of LiTf incorporated. An anomaly has been noted when moderate amount of LiTf is added into the polymer. This can be related to ion-pair effect of lithium and triflate ions. The activation energy calculated using Arrhenius' equation and scanning electron micrographs are congruent with the deduction of neutral ion pair formation. TGA thermograms show that all the selected samples are thermally stable up to 400 °C, even with the addition of LiTf. Although the highest ionic conducting sample suffers from having slightly higher weight loss percentage than pure P(VdF-HFP), it is compensated with improved decomposition temperature and highest onset temperature of decomposition. This system has already shown great potentials even with only a polymer host and a common lithium salt. It is worthy to be further investigated with incorporation of other additives, such as plasticizers, fillers or ionic liquids.
77
Acknowledgment This work was supported by the Exploratory Research Grant Scheme (ERGS: ER017-2011A) and University Malaya Research Grant (UMRG: RG140-11AFR). References [1] S. Ramesh, S.-C. Lu, Journal of Power Sources 185 (2008) 1439–1443. [2] G. Cheruvally, J.-K. Kim, J.-W. Choi, J.-H. Ahn, Y.-J. Shin, J. Manuel, P. Raghavan, K.-W. Kim, H.-J. Ahn, D.S. Choi, C.E. Song, Journal of Power Sources 172 (2007) 863–869. [3] A.M. Stephan, Y. Saito, Solid State Ionics 148 (2002) 475–481. [4] K.M. Kim, J.M. Ko, N.-G. Park, K.S. Ryu, S.H. Chang, Solid State Ionics 161 (2003) 121–131. [5] V. Aravindan, P. Vickraman, European Polymer Journal 43 (2007) 5121–5127. [6] D. Saikia, Y.W. Chen-Yang, Y.T. Chen, Y.K. Li, S.I. Lin, Desalination 234 (2008) 24–32. [7] A. Manuel Stephen, K.S. Nahm, M.A. Kulandainathan, G. Ravi, J. Wilson, European Polymer Journal 42 (2006) 1728–1734. [8] K. Xu, Chemical Reviews 104 (2004) 4303–4418. [9] S. Ramesh, A.K. Arof, Materials Science and Engineering B 85 (2002) 11–15. [10] S. Ramesh, M.F. Chai, Materials Science and Engineering B 139 (2007) 240–245. [11] M.J. Reddy, P.P. Chu, Journal of Power Sources 109 (2002) 340–346. [12] G.P. Pandey, S.A. Hashmi, Journal of Power Sources 187 (2009) 627–634. [13] G.C. Li, P. Zhang, H.P. Zhang, L.C. Yang, Y.P. Wu, Electrochemistry Communications 10 (2008) 1883–1885.
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Liquids journal homepage: www.elsevier.com/locate/molliq
A simple P(VdF-HFP)–LiTf system yielding highly ionic conducting and thermally stable solid polymer electrolytes S. Ramesh a,⁎, Soon-Chien Lu b a b
Centre for Ionics University Malaya, Department of Physics, Faculty of Science, University of Malaya, 50603 Kuala Lumpur, Malaysia Centre for Surface Chemistry and Catalysis, Faculty of Bioengineering Science, Katholieke Universiteit Leuven, 3001 Leuven, Belgium
a r t i c l e
i n f o
Article history: Received 21 December 2011 Received in revised form 21 September 2012 Accepted 25 September 2012 Available online 4 October 2012 Keywords: Polymer membrane Material testing Thermal property Poly(vinylidene fluoride-cohexafluoropropylene) (PVdF-HFP) Ionic conductivity
a b s t r a c t In the present work, we report a simple solid polymer electrolyte (SPE) system that is solely based on poly(vinylidene fluoride-co-hexafluoropropylene) [P(VdF-HFP)] and lithium trifluoromethanesulfonate (LiTf). The SPEs produced exhibit high ionic conductivity of ~ 10−4 S cm −1 at ambient temperature when 40 wt.% of LiTf is incorporated. There is an anomaly when moderate amount of LiTf is added into the polymer. This can be related to formation of neutral ion pairs, and substantiated by calculation of activation energy and scanning electron micrographs. TGA thermograms also show that all the samples tested are thermally stable up to 400 °C, even with the incorporation of LiTf. The highest ionic conducting sample achieves highest onset temperature of decomposition. © 2012 Elsevier B.V. All rights reserved.
1. Introduction The development of high energy density, easily portable and safe power sources is one of the key global technological challenges, as driven by the desire of shifting away from the current heavy dependence on fossil fuels. Solid polymer electrolytes (SPEs) fulfill these requirements and overcome the limitations of conventional liquid electrolytes by addressing drawbacks such as electrolyte leakage, flammable organic solvent, and electrolytic degradation of electrolytes [1]. However, SPEs suffer from one of the important industrial aspects, which is low ionic conductivity. There have been many efforts in improving the ionic transport properties of polymer electrolytes to enhance their power application feasibility. Poly(vinylidene fluoride-co-hexafluoropropylene) [P(VdF-HFP)] has received great attention due to its outstanding thermal and electrochemical stability [2]. This copolymer exhibits amorphous PHFP phase that aids ionic conduction, and at the same time crystalline PVdF phase acts as a mechanical support [3]. This satisfies the two contradictory properties of highly ionic conducting and good mechanical strength. Recently, P(VdF-HFP) has shown potential as polymer electrolyte material of rechargeable lithium batteries owing to its high solubility, low crystallinity and glass transition temperature [4]. Many researchers present encouraging results for P(VdF-HFP)-based polymer electrolytes
⁎ Corresponding author. Tel.: +60 3 7967 4391. E-mail address: [email protected] (S. Ramesh). 0167-7322/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.molliq.2012.09.018
with different types of dopant salts, such as lithium fluoroalkyl phosphate [5], lithium perchlorate [6] and lithium bis(trifluoromethanesulfonyl) imide [7]. Lithium trifluoromethanesulfonate salt, LiTf, is one of the most common lithium salt used in polymer electrolyte research. In this study, we utilize only P(VdF-HFP) and LiTf to study about their interaction, electrochemical and thermal properties. Without any other additives or chemicals, we can examine the basic interaction of ionic conduction in polymer and influence on thermal stability. Our studies find that even when P(VdF-HFP) is incorporated with LiTf, which is generally regarded as having lower ionic conductivity as compared to other lithium salts of its family [8], the SPEs produced exhibit exceptional ionic conductivity, thermal stability and mechanically stable with the coexistence of amorphous and crystalline phases. There are great potentials for this system to be further investigated with other additives such as plasticizers and ceramic fillers.
2. Experimental 2.1. Materials Poly(vinylidene fluoride-co-hexafluoropropylene), [P(VdF-HFP)] with average Mw = 455,000, was obtained from Sigma-Aldrich. Lithium trifluoromethanesulfonate salt, LiTf [LiCF3SO3], was obtained from Fluka and dried at 100 °C for 1 hour to eliminate trace amounts of water in the material, prior to the preparation of solid polymer electrolytes (SPEs). Acetone of AR grade was obtained from J.T. Baker.
74
S. Ramesh, S.-C. Lu / Journal of Molecular Liquids 177 (2013) 73–77
2.2. Preparation of thin films Several polymer electrolyte complexes were prepared as in the compositions according to its weight ratio of polymer to lithium salt, ranging from 95/5 to 60/40. The thin films were prepared by solution casting technique with acetone as solvent. The mixture was cast on a Petri dish and allowed to evaporate slowly inside a fume hood. This procedure yields mechanically stable and free standing films. The thickness of the films was measured using a micrometer screw gauge. 2.3. Instrumentation The samples were cut out resembling shape of stainless steel blocking electrodes used in this study and sandwiched between them. A HIOKI 3532-50 LCR Hi-Tester was used to perform the impedance measurements for each SPE over the frequency range of 50 Hz to 1 MHz. The same measurement was also performed for selected samples from 30 to 80 °C. Those selected SPEs were then subjected to scanning electron microscopy (SEM), with the model Leica S440. Thermogravimetric analysis (TGA) was performed with a Mettler Toledo analyzer that consists of a TGA/SDTA851 e main unit and STARe software with 10 °C min − 1 heat rate between 30 and 500 °C under a nitrogen atmosphere. 3. Results and discussion 3.1. Alternating current (AC) impedance studies Fig. 1 shows two Cole–Cole plots of samples with (a) 5 and (b) 40 wt.% of LiTf. It can be observed in Fig. 1(a) that for low concentration of lithium salt, a semi-circle with a long tail, or more commonly known as electrode spike, is clearly observed. When the salt content increases, the semicircle diminishes completely and leaves only with a sharp tail as in Fig. 1(b). This is typical for a very ionic conductive thin film sample. There are two axes in a Cole–Cole plot, namely real and imaginary part of total impedance Z. The lowest point of imaginary Z axis is extrapolated to determine the bulk resistance, Rb. The ionic conductivity of thin film sample can then be calculated using Eq. (1): ð1Þ
where σ is ionic conductivity, l is thickness of the thin film sample, Rb is bulk resistance determined from impedance spectroscopy and A is surface area of stainless steel blocking electrodes. Fig. 2 illustrates the variation of log ionic conductivity values as a function of lithium triflate (LiTf) weight percentage incorporated into P(VdF-HFP). It can be clearly observed that the ionic conductivity increases exponentially with the addition of LiTf at the beginning, then becomes relatively constant and reaches the highest ionic conductivity of 1.56×10−4 S cm−1 at 40 wt.% of LiTf. The initial increase is simply due to the increased availability of Li+ ions in the system. However, from 15 wt.% onwards, the ionic conductivity decreases gradually until 30 wt.%. This may be due to neutral ion pair formation from Li+ and Tf− ions, or more commonly known as ion-pair effect. High concentration of salt has higher tendency towards this type of formation. This reduces the availability of free mobile ions and possibly obstructs the conducting path, which subsequently contribute to the slight decrease of ionic conductivity. It can also be related to the formation of linkages between the salt itself, causing it to crystallize, thus resulting in the decrease of ionic conductivity [9]. This proposition can be substantiated by activation energy calculation and scanning electron microscopy (SEM) images in latter sections. When the concentration of LiTf exceeds 30 wt.%, the presence of excessive Li+ and Tf− ions starts to disrupt and convert the crystalline phase of P(VdF-HFP) to amorphous, thus changing the morphology of polymer and further enhances the ionic conductivity.
Fig. 1. Cole–Cole plots of samples with (a) 5 and (b) 40 wt.% of LiTf.
The ionic conductivity is also measured at different temperatures, from 30 to 80 °C, in order to evaluate the ionic conduction mechanism in the SPEs. The variation of log ionic conductivity as a function of inverse absolute temperature for selected samples, with 10, 25 and 40 wt.% LiTf respectively, coded as LT10, LT25 and LT40 respectively, -2
Log σ [ lg (S cm -1 ) ]
σ ¼ l=Rb A
-4
-6
-8 0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
Wt% of LiTf Content Fig. 2. Variation of log ionic conductivity as a function of lithium triflate (LiTf) weight percentage incorporated into P(VdF-HFP).
S. Ramesh, S.-C. Lu / Journal of Molecular Liquids 177 (2013) 73–77
is shown in Fig. 3. The regression values of all three chosen samples are close to unity indicating that the temperature-dependent ionic conductivity for this SPE system obeys the Arrhenius rule. As the conductivity–temperature data follow Arrhenius behavior, the nature of cation transport is deduced to be similar to that in ionic crystals, where ions jump into neighboring vacant sites and thus increase the ionic conductivity to a higher value [10]. The motion of ions in solid polymer electrolytes is a liquid-like mechanism, by which the movement of ions through the polymer matrix is assisted by the large amplitude of the polymer segmental motion. Thus, greater segmental motion at higher temperatures either permits the ions to hop from one site to another or provides a pathway for ions to move with faster ionic conduction. Activation energy of several selected samples has been calculated from the Arrhenius equation shown in Eq. (2) below: σ ¼ σ o eð
−RTa Þ E
ð2Þ
where σ is ionic conductivity, σo is the pre-exponential factor, Ea is the apparent activation energy, R is universal gas constant and T is the absolute temperature. The results are tabulated in Table 1. As expected, LT40 has the lowest activation energy, as compared to the other two samples. Interestingly, a slightly steeper slope has been noted for LT25, which means this particular sample needs higher energy for ionic transport. This can be related to the explanation given above. LT25 has much higher activation energy than that of LT10, even though its ionic conductivity is slightly higher compared to that of LT10. This anomaly can be comprehended as the supplied heat provides extra energy for those aggregated neutral ion pairs to redissociate again [11]. As higher amount of free Li + and Tf − ions are released from neutral ion pairs upon heating, the ionic conductivity also raises significantly. Another plausible explanation is by assuming that polymer matrix with high amount of aggregated neutral ion pairs would have smaller pore size. Higher energy is required for Li+ ions to transport through polymer matrix with smaller pore size. Other than salt concentration, the conductivity of thin film also changes with frequency applied to it. The conductance of the samples is calculated with the equation: log σ ¼ log
GL A
ð3Þ
where G denotes the conductance in V, L as the thickness of thin film in cm and A as the surface area of stainless steel blocking electrodes in cm 2. Again LT40 is selected as it shows highest ionic conductivity among all samples. Variation of logarithm conductance as a function of frequency for LT40 at 298, 303, 313, 323, 333, 343 and 353 K has been shown in Fig. 4.
75
Table 1 Activation energies of sample LT10, LT25 and LT40 calculated using Arrhenius's equation. Sample
Calculated activation energy
LT10 LT25 LT40
1.761 kJ mol−1 5.901 kJ mol−1 1.057 kJ mol−1
Generally, the conductance of samples is found to increase with increasing frequency, as well as temperature. At lower frequency region, the charges in the samples are easily accumulated at the electrodes and electrode interfaces, which causes to a decline in free mobile ions and eventually a decrease in conductance at lower frequencies. At higher frequency region, the conductance increases with frequency due to greater mobility of charge carriers and faster hopping of ions. Therefore, the ion exchange process occurs more effectively at higher frequencies. As temperature increases, general trend of conductivity also shifts accordingly. A hump can be clearly observed at all temperatures for LT40. The shifting of hump from higher to lower frequency with increasing temperature can be related to the undissociated LiTf salt present in the sample that is being released as free mobile ions upon heating, resulting further increase in conductance, which has already been discussed in the Ea section above. 3.2. Scanning electron microscopy Scanning electron micrographs of (a) pure P(VdF-HFP), (b) LT10, (c) LT25 and (d) LT40 have been shown in Fig. 5. Very distinguishable changes can be observed from pure P(VdF-HFP), to low and high concentrations of LiTf. Pure P(VdF-HFP) shows normal porous surface with uniform small pore size. The morphology changes, as soon as 10 wt% of LiTf is doped into the polymer, to become separated by layers and greater pore size. However, when it reaches 25 wt%, the morphology has become more similar to that of pure polymer, only with visibly larger pore size. This further confirms the justification regarding neutral ion pairs formation discussed above, apart from the higher activation energy calculated. When more lithium salt is added into the polymer, the morphology changes again to become similar to that of LT10, but with significantly more layered and greater pore size. These features are very similar to those reported by Pandey and Hashmi [12]. According to free-volume model, this microstructure leads to the better conducting pathway for Li+ ions, and consequently leading to higher ionic conductivity [13]. In summary, the observations in these micrographs are congruent with the findings of ionic conductivity in the previous section. 3.3. Thermal studies The samples of the pure P(VdF-HFP), LT10, LT25 and LT40 have been subjected to thermogravimetric analysis (TGA) in nitrogen atmosphere,
Log σ [ lg (S cm -1 ) ]
-3
-4
-5 2.80
2.90
3.00
3.10
3.20
3.30
-1
1000/T (K ) Fig. 3. Variation of log ionic conductivity as a function of inverse absolute temperature for samples LT10, LT25 and LT40.
Fig. 4. Variation of logarithm conductance as a function of frequency for LT40 at 25 to 80 °C.
76
S. Ramesh, S.-C. Lu / Journal of Molecular Liquids 177 (2013) 73–77
Fig. 5. Scanning electron micrographs of (a) pure P(VdF-HFP), samples (b) LT10, (c) LT25 and (d) LT40, all with the magnification of 5000×.
to examine their thermal stability. The TGA thermograms in Fig. 6 show that all the samples are thermally stable up to 400 °C, even with the addition of LiTf. This is highly desirable for its potential applications in electrochemical devices. The weight loss percentage increases upon the addition of LiTf. This is mainly due to the presence of Tf− group, which can be burnt off easily as carbon oxides, sulfur oxides and hydrogen fluoride, as compared to the polymer host and Li+ ions (Table 2). However, when the concentration
100 90 80
Wt%
70 60
of LiTf is increased to 25 wt%, the weight loss percentage only decreases marginally. This may be due to the formation of neutral ion pairs, which makes it more difficult to be burnt off, as more Li+ and Tf− ions are bonded ionically. Much energy is needed to break the ionic bond, therefore slightly higher energy is required to decompose LT25, as compared to that of LT10. When the concentration goes up to 40%, the weight loss percentage drastically decreases to 51.22%. This may be due to greater amount of lithium metal ions and higher interaction between lithium salt and polymer host, making the sample more difficult to be decomposed. The decomposition temperature generally increases upon the addition of LiTf, approximately from 400 to 430 °C. Although LT40 suffers from having the slightly higher weight loss percentage than pure P(VdF-HFP), it is compensated with improved decomposition temperature and highest onset temperature of decomposition. This can be explained as the higher
50 40
Table 2 Weight loss percentages, decomposition temperatures and onset temperatures of decomposition of pure P(VdF-HFP), LT10, LT25 and LT40 obtained from TGA thermograms.
30 20 10
Sample
% Weight loss
Decomposition temperature, TD
Onset temperature of decomposition
Pure PVdF-HFP LT10 LT25 LT40
46.21% 64.53% 63.53% 51.22%
405.94 430.38 436.20 432.48
384.74 406.56 405.38 413.61
0 0
50
100
150
200
250
300
350
400
450
500
Temperature (oC) Fig. 6. Normalized TGA thermograms of pure P(VdF-HFP) (−), samples LT10 (●), LT25 (▲) and LT40 (■) in nitrogen atmosphere.
°C °C °C °C
°C °C °C °C
S. Ramesh, S.-C. Lu / Journal of Molecular Liquids 177 (2013) 73–77
level of lithium metal ions present, which requires higher energy to be decomposed. 4. Conclusion A series of solid polymer electrolytes (SPEs) that is solely based on poly(vinylidene fluoride-co-hexafluoropropylene) [P(VdF-HFP)] and lithium trifluoromethanesulfonate (LiTf) has been synthesized by solution casting technique. The SPEs produced reach ionic conductivity as high as 1.56 ×10−4 S cm−1 at 40 wt% of LiTf incorporated. An anomaly has been noted when moderate amount of LiTf is added into the polymer. This can be related to ion-pair effect of lithium and triflate ions. The activation energy calculated using Arrhenius' equation and scanning electron micrographs are congruent with the deduction of neutral ion pair formation. TGA thermograms show that all the selected samples are thermally stable up to 400 °C, even with the addition of LiTf. Although the highest ionic conducting sample suffers from having slightly higher weight loss percentage than pure P(VdF-HFP), it is compensated with improved decomposition temperature and highest onset temperature of decomposition. This system has already shown great potentials even with only a polymer host and a common lithium salt. It is worthy to be further investigated with incorporation of other additives, such as plasticizers, fillers or ionic liquids.
77
Acknowledgment This work was supported by the Exploratory Research Grant Scheme (ERGS: ER017-2011A) and University Malaya Research Grant (UMRG: RG140-11AFR). References [1] S. Ramesh, S.-C. Lu, Journal of Power Sources 185 (2008) 1439–1443. [2] G. Cheruvally, J.-K. Kim, J.-W. Choi, J.-H. Ahn, Y.-J. Shin, J. Manuel, P. Raghavan, K.-W. Kim, H.-J. Ahn, D.S. Choi, C.E. Song, Journal of Power Sources 172 (2007) 863–869. [3] A.M. Stephan, Y. Saito, Solid State Ionics 148 (2002) 475–481. [4] K.M. Kim, J.M. Ko, N.-G. Park, K.S. Ryu, S.H. Chang, Solid State Ionics 161 (2003) 121–131. [5] V. Aravindan, P. Vickraman, European Polymer Journal 43 (2007) 5121–5127. [6] D. Saikia, Y.W. Chen-Yang, Y.T. Chen, Y.K. Li, S.I. Lin, Desalination 234 (2008) 24–32. [7] A. Manuel Stephen, K.S. Nahm, M.A. Kulandainathan, G. Ravi, J. Wilson, European Polymer Journal 42 (2006) 1728–1734. [8] K. Xu, Chemical Reviews 104 (2004) 4303–4418. [9] S. Ramesh, A.K. Arof, Materials Science and Engineering B 85 (2002) 11–15. [10] S. Ramesh, M.F. Chai, Materials Science and Engineering B 139 (2007) 240–245. [11] M.J. Reddy, P.P. Chu, Journal of Power Sources 109 (2002) 340–346. [12] G.P. Pandey, S.A. Hashmi, Journal of Power Sources 187 (2009) 627–634. [13] G.C. Li, P. Zhang, H.P. Zhang, L.C. Yang, Y.P. Wu, Electrochemistry Communications 10 (2008) 1883–1885.