Solid State Ionics 136–137 (2000) 1175–1179 www.elsevier.com / locate / ssi
Composite electrolytes based on low molecular weight polyglycols ~ M. Marcinek, A. Zalewska, G. Zukowska, W. Wieczorek* Faculty of Chemistry, Warsaw University of Technology, ul. Noakowskiego 3, 00 -664 Warsaw, Poland
Abstract In the present work ionic conduction is studied in poly(ethylene glycol) (PEG)–LiClO 4 –a-Al 2 O 3 electrolytes over wide salt concentration and temperature ranges. Composite electrolytes with a-Al 2 O 3 containing inert surface groups are examined. Conductivity enhancement was observed compared to PEG–LiClO 4 electrolytes in the 0.5–5 mol kg 21 salt concentration range. Conductivity studies are coupled to viscosity and DSC experiments which helps us to establish the relation between salt concentration and viscosity (flexibility) of the polymer host. The ion–polymer interactions are studied on the basis of FT-IR spectroscopy and, particularly, changes in the intensity of the IR modes characteristic of C–O–C vibrations. Finally, the ionic associations are discussed on the basis of the Fuoss–Krauss formalism. These studies show the reduction in the ion-pair formation for composite polymeric electrolytes in the entire concentration range studied. 2000 Elsevier Science B.V. All rights reserved. Keywords: Poly(ethylene glycol); Lithium perchlorate; Composite electrolytes; Ionic associations
1. Introduction Polyether-based electrolytes doped with lithium salts have attracted considerable attention due to the possibility of application in ambient and moderate temperature lithium or lithium ion batteries [1,2]. It has been widely recognized that the properties of semicrystalline poly(ethylene oxide)-based systems can be improved by the addition of inorganic fillers, e.g. aluminas, silica, etc. [3–5]. The addition of these fillers results in an enhancement of ionic conductivity, due to the reduction of the polymer host crystallinity, an improvement of the mechanical and thermal stability of the electrolytes and the reduction in the formation of resistive passive layers on the *Corresponding author. Fax: 1 48-22-628-2741. E-mail address:
[email protected] (W. Wieczorek).
lithium electrode–polymer electrolyte interface [6]. Recently it has been shown that the conductivity enhancement can be achieved by the addition of aluminum oxide bearing the surface Lewis acid type groups to the amorphous poly(ethylene glycol) (PEG)–LiClO 4 electrolytes [7]. However the conductivity enhancement was limited to a narrow LiClO 4 concentration range (0.5–3 mol kg 21 PEG). In the present paper the effect of the addition of a-Al 2 O 3 with inert surface groups on the conductivity, ion–ion and ion–polymer interactions is analyzed using impedance spectroscopy, DSC, rheological and FT-IR experiments performed in wide salt (10 26 –5 mol kg 21 ) and temperature (20–958C) ranges. The observed conductivity enhancement is related to the decrease of the electrolyte viscosity as well as the increase in the fraction of charge carriers (free ions and triplets) compared to the pure PEG–
0167-2738 / 00 / $ – see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0167-2738( 00 )00580-4
1176
M. Marcinek et al. / Solid State Ionics 136 – 137 (2000) 1175 – 1179
LiClO 4 systems. The ion associations are analyzed using the Fuoss–Kraus formalism [8].
2. Experimental
2.1. Sample preparation PEG (Mw 5 350, Aldrich, monomethyl capped) was filtered and then dried on a vacuum line first at | 608C for 72 h and then, under a vacuum of 10 25 Torr, stringently freeze dried using freeze–pump– thaw cycles. While still under vacuum, the polymer was transferred to an argon-filled dry-box (moisture content lower than 2 ppm) where the salt was dissolved into the polymer using a magnetic stirrer. Salt concentration varied from 10 26 to 5 mol kg 21 of polymer. Samples of the salt concentration from 5 mol kg 21 down to 0.5 mol kg 21 were prepared by the direct dissolution of salt in a polymer. Samples of the highest salt concentration were heated up to 508C to facilitate the dissolution process. Samples of lower salt concentration were prepared by the successive dilution of a batch containing electrolyte with 0.5 mol kg 21 LiClO 4 . LiClO 4 (Aldrich, reagent grade) was dried under vacuum at 1208C prior to the dissolution. The composite electrolytes were obtained by the dispersion of a-Al 2 O 3 in a PEG–LiClO 4 solution. The concentration of a-Al 2 O 3 in the composite electrolytes was equal to 10 mass%. a-Al 2 O 3 (Aldrich, reagent grade, fraction of the grain size lower than 5 mm with the neutral surface groups) was dried under a vacuum of 10 25 Torr at 1508C for over 72 h prior to the addition to the polymer–salt mixture. All samples were equilibrated at ambient temperature for at least 1 month before undertaking any experiments.
2.2. DSC studies DSC data were obtained between 2 110 and 1508C using a UNIPAN 605M scanning calorimeter with a low-temperature measuring head and a liquidnitrogen-cooled heating element. Samples in aluminum pans were stabilized by slow cooling to 2 1108C and then heated at 108C min 21 to 1508C. An empty aluminum pan was used as a reference.
2.3. Conductivity measurements Ionic conductivity was determined using the complex impedance method in the temperature range 20–908C. The samples were sandwiched between stainless-steel blocking electrodes and placed in a temperature-controlled oven. The experiments were performed in a constant-volume cylindrical cell with a 7.8-mm-diameter electrode and fixed electrolytes of 1.6-mm thickness. The impedance measurements were carried out on a computer-interfaced Solartron– Schlumberger 1255 impedance analyzer over the frequency range 1 Hz–1 MHz.
2.4. FT-IR Infrared absorption spectra were recorded on a computer-interfaced Perkin-Elmer 2000 FT-IR system with a wavenumber resolution of 2 cm 21 . FT-IR studies were performed at 258C. Electrolytes were sandwiched between two NaCl plates and placed in the FT-IR temperature-controlled cell; the accuracy of the temperature was estimated to be 618C.
2.5. Rheological experiments Rheological experiments were conducted at 258C using a Bohlin Visco 88BV viscometer in two coaxial cylinders geometry. The measurements were performed within a shear rate range of 24–1200 cm 21 .
3. Results Fig. 1 presents the conductivity isotherms depicted as the function of salt concentration at 258C. Up to a LiClO 4 concentration equal to 0.25 mol kg 21 , the conductivities of both systems (PEG–LiClO 4 and PEG–LiClO 4 –a-Al 2 O 3 ) do not differ from each other by more than | 30–40%. For higher salt concentrations conductivities measured for electrolytes containing a-Al 2 O 3 are much higher. This difference rises to up to 30–40 times for systems with the highest salt concentration. Fig. 2 shows the temperature dependence of conductivity for electrolytes containing 0.001 and 2 mol kg 21 LiClO 4 (for comparison, data for electrolytes with and without a-Al 2 O 3 are shown). For
M. Marcinek et al. / Solid State Ionics 136 – 137 (2000) 1175 – 1179
Fig. 1. Changes in ionic conductivity as a function of salt concentration measured at 258C for: d, PEG–a-Al 2 O 3 –LiClO 4 electrolytes; and s, PEG–LiClO 4 electrolytes.
Fig. 2. Changes in ionic conductivity of PEG–LiClO 4 electrolytes (open symbols) and PEG–a-Al 2 O 3 –LiClO 4 electrolytes (filled symbols) as a function of inverse temperature. Samples of various molal salt concentrations: d, s, 0.001 mol kg 21 ; m, n, 2 mol kg 21 .
1177
both sets of electrolytes the conductivities of the systems with a-Al 2 O 3 fillers are higher. The difference in conductivities decreases with an increase in temperature and is higher for the system with the higher salt concentration. Fig. 3 presents the changes in the glass-transition temperature as a function of salt concentration for the PEG–LiClO 4 and PEG–LiClO 4 –a-Al 2 O 3 electrolytes. T g is almost invariant up to the salt concentration 0.25 mol kg 21 . For higher salt concentrations T g rises for both sets of electrolytes. However, T g values measured for PEG–LiClO 4 –a-Al 2 O 3 electrolytes in this salt concentration range are lower. Fig. 4 shows the variation of the electrolyte viscosities as a function of LiClO 4 concentration. The trends shown are similar to the DSC results. The viscosities of both sets of electrolytes varied in the narrow range up to the salt concentration 0.1 mol kg 21 PEG. Similar viscosities for both sets of electrolytes are measured. For higher salt concentrations the viscosity increases and the increase is faster in the case of PEG–LiClO 4 electrolytes than for systems with a-Al 2 O 3 additives. For electrolytes with the highest salt concentration the viscosities of
Fig. 3. Changes in the T g as a function of salt concentration. Data for: d, PEG–a-Al 2 O 3 –LiClO 4 electrolytes; and s, PEG–LiClO 4 electrolytes.
1178
M. Marcinek et al. / Solid State Ionics 136 – 137 (2000) 1175 – 1179
Fig. 4. Viscosity as a function of salt concentration. Data obtained at 258C for: d, PEG–a-Al 2 O 3 –LiClO 4 electrolytes; and s, PEG– LiClO 4 electrolytes.
PEG–LiClO 4 electrolytes are roughly 20 times higher than for PEG–a-Al 2 O 3 –LiClO 4 systems. Fig. 5 shows the variation in the position of the maximum of C–O–C stretch as a function of salt
Fig. 5. Changes in the maximum of the C–O–C stretching mode as a function of salt concentration. Data for: d, PEG–a-Al 2 O 3 – LiClO 4 electrolytes; and s, PEG–LiClO 4 electrolytes.
concentration for the PEG–LiClO 4 and PEG– LiClO 4 –a-Al 2 O 3 electrolytes. The position of the peak maximum is an indication of the intensity of the polyether–salt interaction [2]. The downshift of this maximum may result from the formation of inter- or intratransient crosslinks by Li 1 cations or Li 1 ClO 2 4 Li 1 charged triplets [9,10]. As shown in Fig. 5 the position of the maximum decreases with an increase in the salt concentration for LiClO 4 concentrations higher than 0.01 mol kg 21 . For a salt concentration higher than 1 mol kg 21 the position of the peak maximum is observed at a higher wavenumber for electrolytes with a-Al 2 O 3 fillers. Fig. 6 presents changes in the fraction of ‘free’ ions, ion pairs and triplets calculated for both sets of electrolytes on the basis of the Fuoss–Kraus formalism [8]. The detailed description of the applied procedure has been published elsewhere [7]. The data used for calculation are included in Table 1. As can be seen from Fig. 6 the fraction of neutral ion pairs which do not participate in the ionic transport is much lower for PEG–LiClO 4 –a-Al 2 O 3 electrolytes. For the samples with the highest salt concentration the fraction of charge carries is approximately twice as high for systems containing a-Al 2 O 3 .
Fig. 6. Changes in the fraction of: d, s, ‘free’ ions; j, h, ion pairs; and m, n, triplets as a function of LiClO 4 concentration calculated at 258C on the basis of the Fuoss–Kraus formalism for PEG–a-Al 2 O 3 –LiClO 4 electrolytes (filled symbols) and PEG– LiClO 4 electrolytes (open symbols).
M. Marcinek et al. / Solid State Ionics 136 – 137 (2000) 1175 – 1179 Table 1 Physicochemical data used for the calculation of ionic association on the basis of the Fuoss–Kraus formalism (all calculations for 258C)
L I0 (S cm 21 kg mol 21 ) l T0 (S cm 21 kg mol 21 ) d (g cm 23 ) h (Pa s) KI (mol 21 kg) KT (mol 21 kg)
PEG–LiClO 4
PEG–a-Al 2 O 3 –LiClO 4
0.00336 0.00224 1.094 a 0.020 b 9.4 3 10 4 48.8
0.00291 0.00194 1.382 a 0.029 c 611.3 21.2
a Density of PEG and PEG–a-Al 2 O 3 found from picnometric determinations. b Data for PEG. c Data for PEG–a-Al 2 O 3 .
4. Discussion It is shown that the addition of a-Al 2 O 3 fillers containing neutral surface groups results in an increase in the conductivity of PEG–LiClO 4 electrolytes in the 0.5–5 mol kg 21 salt concentration range. These results are qualitatively similar to the previous data on the effect of the a-Al 2 O 3 filler with Lewis acid surface groups and can be explained by the previously proposed mechanism highlighting the importance of the Lewis acid–base type interactions between polymer, salt and filler particles [7]. There are, however, a few important differences. The conductivity enhancement over the pure PEG– LiClO 4 electrolytes observed in the recent studies is higher, especially for samples with the highest salt concentration. This results from the lower electrolyte viscosity and higher fraction of charge carriers in PEG–a-Al 2 O 3 –LiClO 4 systems as shown in Figs. 4 and 6. Moreover, the addition of inert a-Al 2 O 3 results in a lower increase in the viscosity than has been previously observed for systems with a-Al 2 O 3 bearing Lewis acid surface groups [7]. This is probably due to the weaker interactions between the polymer basic groups and the surface acid centers of the filler. The weaker polymer–filler interactions are also confirmed by the FT-IR studies (see data for PEG–a-Al 2 O 3 –LiClO 4 electrolytes in Fig. 5). It should be stressed that the observed enhancement in the ambient temperature conductivity agrees quite well with the combined decrease in the viscosity (factor of | 20) and the increase in the charge
1179
carriers concentration (factor of | 2) for PEG–aAl 2 O 3 –LiClO 4 electrolytes as compared to the pure PEG–LiClO 4 system. The conductivity enhancement corresponds to the region in which the formation of 1 transient crosslinks via the Li 1 ClO 2 triplets is 4 Li observed. The observed reduction in the polymer– salt interactions as well as the increase in the concentration of the charge triplets combined with the viscosity reduction suggests that the a-Al 2 O 3 fillers are very efficient in breaking inter- or intratransient crosslinks, thus creating extra charge carriers. This is in agreement with the recent studies by Scrosati et al. [11], who reported an increase in cation transport numbers for PEO–LiClO 4 (TiO 2 , Al 2 O 3 ) composite polymeric electrolytes.
Acknowledgements This work was financially supported by the President of the Warsaw University of Technology via Research Grant No. 504 / 164 / 853 / 8.
References [1] B. Scrosati (Ed.), Applications of Electroactive Polymers, Chapman and Hall, London, 1993. [2] P.G. Bruce (Ed.), Solid State Electrochemistry, Cambridge University Press, Cambridge, 1995. ´ [3] W. Wieczorek, Z. Florjanczyk, J.R. Stevens, Electrochim. Acta 40 (1995) 2251. [4] F.M. Gray, Solid Polymer Electrolytes — Fundamentals and Technological Applications, VCH, Weinheim, Germany, 1991. [5] B. Scrosati, R. Neat, in: B. Scrosati (Ed.), Applications of Electroactive Polymers, Chapman and Hall, London, 1993, Chapter 6. [6] P.G. Bruce, D.F. Shriver, F.M. Gray, in: P.G. Bruce (Ed.), Solid State Electrochemistry, Cambridge University Press, Cambridge, 1995, Chapters 5 and 6. ~ ´ [7] W. Wieczorek, P. Lipka, G. Zukowska, H. Wycislik, J. Phys. Chem. B 102 (1998) 6968. [8] R.M. Fuoss, F. Accascina, Electrolytic Conductance, Interscience, NY, 1959. [9] R. Oleander, A. Nitzan, J. Chem. Phys. 102 (1995) 7180. [10] J. Cruickshank, H.V. Hubbard, N. Boden, I.M. Ward, Polymer 36 (1995) 3779. [11] F. Croce, G.B. Appetechi, L. Perci, B. Scrosati, Nature 394 (1998) 456.
