European Polymer Journal 47 (2011) 2372–2378
Contents lists available at SciVerse ScienceDirect
European Polymer Journal journal homepage: www.elsevier.com/locate/europolj
Photoinduced free radical polymerization of thermoset lithium battery electrolytes Markus Willgert a, Maria H. Kjell b, Eric Jacques c, Mårten Behm b, Göran Lindbergh b, Mats Johansson a,⇑ a
KTH Royal Institute of Technology, School of Chemical Science and Engineering, Department of Fibre and Polymer Technology, SE-100 44 Stockholm, Sweden KTH Royal Institute of Technology, School of Chemical Science and Engineering, Department of Chemical Engineering and Technology, Applied Electrochemistry, SE-100 44 Stockholm, Sweden c KTH Royal Institute of Technology, School of Engineering Sciences, Department of Aeronautical and Vehicle Engineering, Lightweight Structures, SE-100 44 Stockholm, Sweden b
a r t i c l e
i n f o
Article history: Received 23 May 2011 Received in revised form 9 August 2011 Accepted 21 September 2011 Available online 29 September 2011 Keywords: Solid polymer electrolyte (SPE) Lithium ion Mechanical properties Ionic conductivity Multifunctional batteries Structural batteries
a b s t r a c t Series of solid poly(ethylene oxide)-methacrylate electrolytes have successfully been manufactured with an aim to serve in a multifunctional battery both as mechanical load carrier as well as lithium ion conductor. The electrolytes produced, in a solvent free process with no post cure swelling, hold a broad range of both mechanical as well as ion conducting properties. The monomer and Li-salt mixtures have been irradiated with UV light, initiating free radical polymerization to obtain solid smooth, homogenous specimens to be utilized as ion conducting electrolytes. The storage modulus at 20 °C is ranging from 1 MPa to almost 2 GPa. The conducting ability of the electrolyte ranges from 5.8 1010 up to 1.5 106 S/cm. These large variations in both mechanical properties as well as ionic conductivity are discussed, but also the versatility within the production technique is emphasized. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction The future demand of energy from renewable sources and uses in different applications calls for new solutions to store the energy . As of today, lithium ion batteries are the primary considered storage medium for electrical energy in mobile applications. Batteries however tend to be of considerable weight and also occupy a signiﬁcant space in a product such as a mobile phone or a car. The weight and space issue and ways to get around it becomes essential when constructing for example hybrid electrical vehicles. Much of the design work is in this case a never ending chase to reduce excessive weight in order to be able to get the car as far a distance as possible on as little energy as possible. One way suggested to encounter this problem ⇑ Corresponding author. Tel.: +46 (0) 8 790 92 87; fax: +46 (0) 8 790 82 83. E-mail address: [email protected]
(M. Johansson). 0014-3057/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.eurpolymj.2011.09.018
is to let the battery itself be an integrated part of the construction i.e. it becomes a structural battery [2–4]. This would mean that the battery should be able to withstand a mechanical load while retaining sufﬁcient battery performance. In other words, the battery is given multifunctional properties, thus providing both energy storage and mechanical load carriage at the same time. A key part of a lithium ion cell is the electrolyte where lithium ions are transported at a certain rate resulting in a current. This transport is however restricted if the electrolyte is a solid thermoset polymer. The optimum performance of a structural battery matrices is thus to allow sufﬁcient mobility of the lithium ions whilst retaining a mechanical integrity of the structure with as strong material properties as possible. Solid polymeric electrolytes (SPEs) of today may, at the best, hold conductivities in a region up to 105 S/cm , although excellent liquid electrolytes may possess conductivities as high as 102 S/cm . However, a liquid
M. Willgert et al. / European Polymer Journal 47 (2011) 2372–2378
electrolyte cannot carry any mechanical load. Hence, in the current application, lower conductivities can be accepted due to the synergetic concept of multifunctionality. One route to create solid electrolytes is to let the structure consist of crosslinked polymer or oligomer segments containing ethylene oxide (EO) groups . Such polymers or oligomers will hereby be named polyethylene oxide (PEO). A lithium ion is able to coordinate to up to four oxygen atoms within those chains, and lithium ion diffusion is possible throughout the electrolyte. The EO chains can be functionalized at the ends and crosslinked in order to form a polymeric network. The crosslinking density in relation to EO groups will strongly affect the conductivity as well as the mechanical properties of the electrolyte. In previous studies [6,8,9], mono methacrylate- and di-methacrylate functionalized EO oligomer have been utilized to build polymeric networks of varied crosslink density. Addition of more di-functionalized monomer will favor the mechanical properties, but the conductivity will be lower due to decreased mobility of ions in the matrix and vice versa. For structural battery applications, it is critical that the modulus reaches an adequate level, at least 100 MPa . Both the conductive- and the mechanical properties will be highly inﬂuenced by the glass transition temperature (Tg) of the material and the crosslink density . Previous studies have demonstrated that the concept of structural batteries holds, and that an optimum balance between electrical and mechanical properties can be obtained . Most studies have involved a thermally induced polymerization of a thermoset polymer with subsequent swelling of the thermoset with the lithium salt together with a co-solvent [11,12]. These procedures induce limitations or drawbacks with respect to design and function of the electrolyte. The thermal cure process takes long time and a signiﬁcant thermal load is imposed on the network. Most thermal free radical initiators also emit gas as they decompose, and this increases the risk of forming voids frozen into the electrolyte which will affect the performance of the electrolyte negatively. Post cure swelling of a thermoset network with an electrolyte is time consuming and can be difﬁcult to control with respect to salt concentration throughout the network. Post cure swelling also reduces the mechanical properties of the ﬁnal system compared to a network polymerized with the lithium salt already present. An alternative to the thermal cure process is photoinduced free radical polymerization techniques. This technique employs UV-light instead of thermal energy to initiate the polymerization using a photoinitiator that decompose to free radicals under UV-irradiation. Several of the aforementioned drawbacks can be reduced or eliminated with this technique compared to thermally induced polymerizations. Photoinduced polymerization is rapid, induce low thermal impact, and well deﬁned networks can be obtained . The technique is well established in several areas such as coating technology [14,15] and dental composites  although less has been presented in the ﬁeld of batteries. Choi et al. has presented work according to this route , although their work did not address the concept of multifunctionality and mechanical load bearing constructions.
UV-induced polymerizations have also been used for several SPE systems with the main focus to retain the geometrical shape of the electrolyte and suppress crystallization of the EO chains [17–19]. These systems have also been studied to correlate the Tg of the system to the electrochemical performance at different temperatures using the William–Landel–Ferry (WLF) equation [20–23]. The mechanical properties of a thermoset may however differ signiﬁcantly above Tg depending on crosslink density and chemical structure between the crosslinks . It is thus important to determine the mechanical properties in relation to conductivity at speciﬁc temperatures to reveal the overall performance. It is furthermore not fully clear how the lithium salt itself contribute to the mechanical properties of a crosslinked SPE. The present work describes a systematic study on photoinduced curing of PEO-dimethacrylate/PEO-monomethacrylate lithium salt mixtures to form thermoset lithium ion battery electrolytes. The efﬁciency of the process as well as the possibility to easily vary both crosslink density and salt concentration over a wide range of compositions is presented. Furthermore, structure – property relationships for the obtained structures is described, as well as the beneﬁts that stem from UV curing in the manufacturing of the electrolytes.
2. Experimental 2.1. Materials All reagents were used as received. Tetraethylene glycol dimethacrylate (SR209), and methoxy polyethylene glycol (350) monomethacrylate (SR550), both depicted A and B respectively in Fig. 1, were kindly supplied by Sartomer Company, Europe. Molecular weights for A and B were 330 g mol1 and 494 g mol1 respectively as reported by the supplier. 2,2-dimethoxy-2-phenylaceto-phenone (Irgacure 651) was obtained by Ciba Specialty Chemicals (Switzerland). Lithium triﬂuoromethanesulfonate (lithium triﬂate) (97%) (Fig. 1), was purchased from Chemtronica AB (Sweden).
F3C S OLi O
Fig. 1. Chemical structure of monomers A and B as well as the lithium triﬂate salt used.
M. Willgert et al. / European Polymer Journal 47 (2011) 2372–2378
2.2. Techniques 2.2.1. Procedures and test series, photopolymerization The following general procedure was used for the curing of the polymer electrolytes. The sample mixtures were prepared in a glove box under dry conditions (<0.05 ppm H2O), in argon atmosphere. The monomers were mixed in a vial and the lithium triﬂate salt was added. The vials were then sealed and put on a shaking table overnight, allowing the lithium salt to dissolve completely. The photoinitiator (1% w/w, relative to the monomer mixture), was then added to the samples and dissolved. After the photoinitiator was added, 0.3 ml of each sample was transferred into a mold (15 20 1 mm) using a syringe, and cured under UV irradiation at 15 cm distance from the UV light source. Dumbbell specimens for tensile testing were polymerized using the same procedure but using a mold developed for creating dumbbell shaped samples, with an initial length of 10 mm and a cross section area of 3.3 mm2 (1.02 3.28 mm). The light source used for curing was a Blak Ray B-100AP (100 W, 365 nm) Hg UV lamp with an intensity of 5.2 mW cm2 as determined with an Uvicure Plus High Energy UV Intergrating Radiometer (EIT, USA), measuring UVA at 320–390 nm. The temperature of the cured samples did not exceed 42 °C during cure. After curing, the cured solids were taken out of the molds and cut into pieces of appropriate size, using a scalpel, for further characterization. 2.2.2. Electrical Impedance Spectroscopy (EIS) In order to quantify the electrochemical performance, the polymer electrolyte samples, that measures 15 20 1 mm were placed in a four-electrode test cell, consisting of four gold wires (two working electrodes 20 mm apart and two reference electrodes 5 mm apart) and two Plexiglas plates with screws to hold the samples in place. The impedance was measured potentiostatically in the frequency range 1 Hz to 300 kHz, at 10 points per decade using a Gamry Series G 750 Potentiostat/Galvanostat/ZRA interface, in an argon-ﬁlled glove box (<0.05 ppm H2O) at ambient temperature. The conductivity was calculated using the equation
l Rb A
where r, l, Rb and A are the ionic conductivity, the distance between the reference electrodes, the bulk resistance, and the cross-sectional area of the sample, respectively.
10.0 lm. DMA measurements gave values for storage modulus (E0 ), loss modulus (E00 ), loss factor (tan d) and Tg for the electrolytes presented in Table 2. 2.2.4. Tensile tests Uniaxial tensile tests were performed inside a glove box under argon atmosphere using a tensile stage from Deben UK equipped with a 10 mm maximum extensometer and a 300 N load cell. Dumbbell shaped specimens were manufactured with an initial length of 10 mm and an average cross section area of 3.3 mm2 (1.02 3.28 mm). The samples were pulled at a total rate of 0.2 mm/min, providing a strain rate of 2.7 106 s1. This strain rate is <105 s1, which is the rate limit above which the motor speed is assumed to affect the results. The initial cross section area of the sample is considered to be constant, and data were recorded and plotted as stress (r) vs. strain (e) to obtain stress at break (rb), elongation at break (eb) and Young´s modulus. 2.2.5. Fourier-Transform Infrared Spectroscopy (FT-IR) FT-IR analysis was performed using a Perkin-Elmer Spectrum 2000 FT-IR instrument (Norwalk, CT) equipped with a heat-controlled, single reﬂection (ATR: attenuated total reﬂection) accessory unit (golden gate) from Graseby Specac Ltd. (Kent, England). All the IR measurements were performed in reﬂection mode. In order to determine the degree of cure of the products, samples were cut from the produced material and mounted onto the instrument. The disappearance of the vinyl stretch peak at 1637 cm1 relative to the carbonyl peak from the ester group at 1715 cm1 (which was used as a reference peak), was monitored at time = 0 s, 30 s, 60 s, 120 s and 240 s. The conversion is calculated as shown in the following equation:
A0 At A0
where A0 is the area of the vinyl stretch peak at time = 0 and At is the same area at time = t. 3. Results and discussion Two different test series were performed; the ﬁrst series to evaluate curing performance as a function of dose, Table 1, and the second series to evaluate the effect of electrolyte composition, Table 2. 3.1. Curing performance
2.2.3. Dynamical Mechanical Analysis (DMA) Dynamical Mechanical Analysis (DMA) tests were performed on a TA instruments DMA, model Q800 in tensile mode. Samples pieces were cut from the initial sample pieces from the curing so they possessed geometry of 7 5 1 mm. The specimens were tightened in the clamps of the sample holder, and the temperature was then decreased to and held at the starting temperature (50 °C) for 10 min before the measurements were started. The temperature was then increased by 3 °C/min up to a top value of 200 °C as data was recorded. The oscillation frequency was held at 1 Hz at constant amplitude of
The conversion as function of dose was measured on samples 0:0–1:8 using FT-IR analysis, Fig. 2. The conversion was determined by measuring the decrease of the peak at 1637 cm1 using the carbonyl peak at 1715 cm1 as an internal reference. The use of a photo fragmenting initiator implies that the main crosslinking mechanism is conventional free radical addition polymerization. The presence of the PEO-segments however introduces abstractable hydrogens why some amount of crosslinking may occur via this route, although probably to a low extent. The results show that the monomer system
M. Willgert et al. / European Polymer Journal 47 (2011) 2372–2378 Table 1 Samples for curing performance evaluation. All samples had the monomer composition 50/50 5 by weight (A/B) and lithium salt content and applied UV dose as depicted in the table.a
Li-salt (% w/w)
Dose (J cm2)
0:0 1:1 1:2 1:3 1:4 0:1 1:5 1:6 1:7 1:8
12 12 12 12 12 0 0 0 0 0
0 0.16 0.31 0.63 1.25 0 0.16 0.31 0.63 1.25
on different compositions was subsequently chosen to be 1.25 J cm2 in order to obtain a fully cured material. The route in which the materials were made showed to be a fast and easy way to manufacture electrolytes, and the created materials became smooth, transparent ﬁlms, free from voids or non-dissolved salt particles. They were easily taken out of the mold for further processing and/or analysis. 3.2. Properties of the polymer electrolyte
1 s irradiation correspond to a dose of 5.2 mJ cm2.
polymerizes readily under the set conditions, and a high conversion is rapidly obtained, Fig. 3. However, the use of the methacrylate carbonyl peak at 1715 cm1 as an internal reference underestimates the conversion since the absorbtivity of the carbonyl is reduced when the alkene conjugated to it is reacted i.e. the reference peak to some extent change. The conversion data in Fig. 3 should thus be considered as comparative trends rather than absolute values, and the conversion curve should in reality be shifted slightly upwards to higher conversions. The almost complete conversion can also be seen on the disappearance of the acrylate peak at 815 cm1 in Fig. 2. It is not possible to ﬁnd a reference peak that is totally unaffected why the carbonyl peak was chosen as the best compromise. Temperature measurements on the sample also revealed that the maximum temperature at the sample surface was 42 °C indicating that the thermal load on the system was low. A comparison between polymerizations performed with or without lithium salt present was also performed. The presence of the lithium salt reduced the polymerization rate slightly but a similar ﬁnal conversion could be obtained as without salt present. The small rate difference is proposed to be due to diluting effects of the lithium salt and possibly different thermal history i.e. less reaction heat per unit volume. The dose for the test series
The majority of the electrolytes presented were below their Tg:s at room temperature, and this is probably also the desired case when designing multifunctional batteries. Depending on if the electrolyte is above or below its Tg at the predominate temperature, the ionic conductivity as well as the mechanical properties of the electrolyte will be affected extensively. 3.2.1. Effect of crosslink density, mechanical properties The effect of the crosslink density on the mechanical properties can be seen comparing samples 2:1–2:9 in Table 2 and Figs. 4–6. The storage modulus at 20 °C is ranging from 1 MPa to almost 2 GPa. The results clearly show that there is a distinct increase of E0 , rb and Young’s modulus with an increased amount of crosslinking resin A, Table 2 and Fig. 6. The glass transition temperature, Tg, also follow this pattern. The loss modulus decreases accordingly as the viscous contribution becomes less expressed. It can also be seen in Figs. 4 and 5, that the Tg-region becomes wider with increased crosslink density. This is indicative of a more dense and heterogeneous network . The curves for samples 2:1 and 2:3 are affected by experimental difﬁculties in stabilizing the samples below 40 °C why the drop in modulus below 40 °C should be considered as an experimental artifact. The change in crosslink density is also clearly seen on the modulus curves where the level of the modulus above Tg is directly related to the effective chain length between crosslinks, Mc, of the polymers. The tensile tests were performed at ambient temperature and the results reveal that the polymer goes from a
Table 2 Test series results for different monomer compositions as well as salt contents.a Sample no.
2:1 2:2 2:3 2:4 2:5 2:6 2:7 2:8 2:9 2:10 2:11 2:(6) 2:12 2:13 a
Monomer Composition (% w/w) A
5 10 20 30 40 50 60 70 80 50 50 50 50 50
95 90 80 70 60 50 40 30 20 50 50 50 50 50
Li -salt content (% w/w)
E020 C (MPa)
E00peak min (MPa)
92 97 >98 >98 96 98 97 93 98 98 >98 98 >98 >98
12 12 12 12 12 12 12 12 12 6 9 12 15 18
1.5 106 1.4 106 1.2 106 5.2 107 1.5 107 2.9 108 5.3 109 5.8 1010 1 4.0 108 4.1 108 2.9 108 3.4 108 6.1 108
26 22 9 23 38 59 75 92 110 56 57 59 49 65
1 2 7 90 320 780 1200 1940 1240 440 740 780 500 800
80 140 720 790 700 770 560 390 160 200 600 770 850 270
1.39 1.13 0.76 0.52 0.50 0.41 0.35 0.34 0.33 0.35 0.39 0.41 0.36 0.40
Tg is deﬁned as the temperature at tan dpeak
M. Willgert et al. / European Polymer Journal 47 (2011) 2372–2378
A sorb Abs bance (a.u ( u.)
Fig. 5. E0 vs. temperature for samples 2:1, 2:3, 2:5, 2:7 and 2:9, to show the effect of varied crosslink density on the mechanical properties. 3200.0 2800 2400 2000 1800 1600 1400 1200 1000
cm -1 Fig. 2. IR Spectrum for 0:1(upper spectrum) and 1:8 (lower spectrum) mixtures. Note especially the dissapearance of the acrylate group stretch peak at 1637 cm1 and at 815 cm1.
Fig. 6. Stress (r) vs. strain (e) for samples 2:4, 2:6, 2:7, 2:8, 2:10 and 2:13 to emphasize the differences in mechanical properties due to varied crosslink density, as well as the small differences in mechanical properties with varied salt content.
Fig. 3. Conversion vs. dose for samples 0:0–1:8 in Table 1, one series without lithium triﬂate salt and one with the salt present. The lines should only be regarded as trend lines.
Fig. 4. Tan d (E00 /E0 ) vs. temperature for samples 2:1, 2:3, 2:5, 2:7 and 2:9, to show the effect of varied crosslink density on the mechanical properties.
ductile to a brittle fracture behavior with increased crosslink density, Fig. 6. It should however be noted that even the more stiff samples exhibit a rather ductile behavior with an eb of more than 10%. This is important in the present application since the lithiation/delithiation process in
the electrodes normally is associated with a change in size of the same. A too brittle electrolyte would thus lead to fractures with subsequent loss of electrochemical performance. In tensile testing eb and rb is widely affected by the composition of monomers in the thermoset. 3.2.2. Effect of crosslink density, conducting properties The conduction ability of the electrolyte is promoted by a more ﬂexible structure containing a greater part of monofunctionalized PEO. This has, as shown in Table 2 and Fig. 7, a pronounced inﬂuence on the conductivity. The diffusion of the lithium ions becomes easier as the network becomes more open. Conductivities of up to 1.5 106 S/cm can be obtained for the present systems. The results clearly show that there is an expected increase of E0 with increased amount of crosslink resin A which is also accompanied by a corresponding decrease of the conducting abilities, Fig. 7. A storage modulus larger than 100 MPa is obtained for systems having 40% or more of monomer A. The results also show that an optimal composition can be chosen depending on the speciﬁc demands on the mechanical properties for this system. It should however be noted that this is true for the present monomer composition in combination with the speciﬁc salt used and that other monomers and salts may exhibit a different behavior. The conductivity data are
M. Willgert et al. / European Polymer Journal 47 (2011) 2372–2378
Fig. 7. The connected behavior of conductivity and modulus. The lines should only be regarded as trend lines.
Fig. 8. tan d (E00 /E0 ) vs. temperature for samples 2:10, 2:6, and 2:13, to show the effect of varied lithium triﬂate salt content on the mechanical properties.
Tensile test furthermore shows only small changes in modulus with varied lithium triﬂate salt content, Fig. 6. There is also no signiﬁcant trend in the values for conductivity with differences in the salt concentration, Table 2. This indicates that the ion transport mainly is restricted by the polymer network rather than the ion concentration in the electrolyte i.e. the number of transported ions are controlled either by the number of coordination sites in the polymer or to the free volume available in the network. A rather strong interaction between the polymer matrix and the salt could explain the rather small differences in mechanical properties between different salt concentrations. A more loose interaction between the salt and the polymer would probably give a softening effect with subsequent decrease in modulus if more salt is added. 4. Conclusions
Fig. 9. E0 vs. temperature for samples 2:10, 2:6, and 2:13, to show the effect of varied lithium triﬂate salt content on the mechanical properties.
within the same range as previously reported for corresponding thermally cured systems . 3.2.3. Effect of lithium salt concentration The behaviors presented above in the previous paragraph, with increasing storage modulus, are valid with increasing crosslink density. However, when varying the lithium salt content, the effect is much smaller and the Tg remains in the same region for all different concentrations, Table 2, Fig. 8 and 9.
Series of solid PEO methacrylate electrolytes have successfully been manufactured using photoinduced polymerization. The electrolytes hold a broad range of both mechanical as well as ion conducting properties. The polymer electrolytes are created in a rapid, solvent free process, where the lithium salt and a photoinitiator has been dissolved in the monomer mixtures before curing. The mixtures have then been irradiated with UV light, initiating radical polymerization to obtain solid smooth, homogenous ion conducting electrolytes. The storage modulus at 20 °C is ranging from 0.8 MPa to well above 1.5 GPa depending on the monomer composition. However, the conductive performance starts off at fairly good conductivities, such as 1.5 106 S/cm and drops dramatically as the material gets stiffer. These large variations in both mechanical properties as well as electrical conductivity emphasize versatility of the production technique, but also in the properties and performance of the materials produced. It also accentuates the importance of ﬁnding an acceptable balance between mechanical and conductive performance. Acknowledgements Funding by the Swedish Foundation for Strategic Research (SSF), framework Grant RMA08-0002 is gratefully
M. Willgert et al. / European Polymer Journal 47 (2011) 2372–2378
acknowledged. The monomers were supplied by Sartomer Company, which is thankfully acknowledged. Finally, the entire KOMBATT group is acknowledged for valuable and giving discussions.
References  Wetzel ED. Reducing weight: Multifunctional composites integrates power, communications, and structure. AMPTIAC Quart 2004;8(4):91–5.  Liu P, Sherman E, Jacobsen A. Design and fabrication of multifunctional structural batteries. J Power Sour 2009;189(1):646–50.  Snyder JF, Carter RH, Wetzel ED. electrochemical and mechanical behavior in mechanically robust solid polymer electrolytes for use in multifunctional structural batteries. Chem Mater 2007;19(15):3793–801.  Thomas JP, Qidwai MA. The design and application of multifunctional structure-battery materials systems. JOM 2005;57(3):18–24.  Zhang H, Kulkarni S, Wunder SL. Blends of POSS-PEO(n = 4)8 and high molecular weight poly(ethylene oxide) as solid polymer electrolytes for lithium batteries. J Phys Chem B 2007;111(14):3583–90.  Snyder JF, Wetzel ED, Watson CM. Improving multifunctional behavior in structural electrolytes through copolymerization of structure- and conductivity-promoting monomers. Polymer 2009;50(20):4906–16.  Fenton DE, Parker JM, Wright PV. Complexes of alkali metal ions with poly(ethylene oxide). Polymer 1973;14(11):589.  Gerbaldi C et al. UV-cured polymer electrolytes encompassing hydrophobic room temperature ionic liquid for lithium batteries. J Power Sour 2010;195(6):1706–13.  Cokbaglan L et al. 2-Mercaptothioxanthone as a novel photoinitiator for free radical polymerization. Macromolecules 2003;36(8):2649–53.  Gerbaldi C et al. UV-curable siloxane-acrylate gel-copolymer electrolytes for lithium-based battery applications. Electrochim Acta 2009;55(4):1460–7.  Choi Y et al. Ultraviolet radiation curing of acrylates for lithium polymer electrolytes. J Appl Electrochem 1997;27(9):1118–21.
 Li Z et al. Novel network polymer electrolytes containing ﬂuorine and sulfonic acid lithium prepared by ultraviolet polymerization. J Appl Polym Sci 2008;108(4):2509–14.  Decker C. Photoinitiated crosslinking polymerization. Prog Polym Sci 1996;21(4):593–650.  Decker C. New developments in UV radiation curing of protective coatings. Surf Coat Int, Part B: Coat Trans 2005;88(B1):9–17.  Watts MPC, et al. Advances in the fabrication of surface modiﬁed microﬂuidic devices in nonﬂuorescing UV cured materials. In: Proceedings of SPIE, 2008; 6882(Micromachining and Microfabrication Process Technology XIII). p. 688203/1-688203/5.  Murray GA, Yates JL, Newman SM. Ultraviolet light and ultraviolet light-activated composite resins. J Prosthet Dentist 1981;46(2):167–70.  MacCallum JR, Smith MJ, Vincent CA. Effects of radiation-induced crosslinking on the conductance of lithium perchlorate/polyethylene oxide electrolytes. Solid State Ionics 1984;11(4):307–12.  Allcock HR et al. Polyphosphazenes with etheric side groups: prospective biomedical and solid electrolyte polymers. Macromolecules 1986;19(6):1508–12.  Bennett JL et al. Radiation crosslinking of poly[bis(2-(2methoxyethoxy)ethoxy)phosphazene]: effect on solid-state ionic conductivity. Chem Mater 1989;1(1):14–6.  Cheradame H, Souquet JL, Latour JM. Ionic conductivity of macromolecular network. I. Polyether ﬁlled with sodium tetraphenylboride s. Mater Res Bull 1980;15(8):1173–7.  Killis A et al. Ionic conductivity of polyether-polyurethane networks containing sodium tetraphenylborate: a free volume analysis. Makromol Chem 1982;183(11):2835–45.  Killis A et al. Ionic conductivity of polyether-polyurethane networks containing alkali metal salts. An analysis of the concentration effect. Macromolecules 1984;17(1):63–6.  Watanabe M et al. Temperature dependence of ionic conductivity of crosslinked poly(propylene oxide) ﬁlms dissolving lithium salts and their interfacial charge transfer resistance in contact with lithium electrodes. Polym J (Tokyo, Jpn) 1984;16(9):711–6.  Ozawa K. Lithium ion rechargeable batteries. Weinheim: WileyVCH; 2009. 1st reprint [Chapter 9].  Ward IM. Mechanical properties of polymers. 2nd ed. New York: Wiley-Interscience; 1985 [Chapter 8].