Nonaqueous liquid electrolyte

[0001] This application claims priority to copending provisional application serial No. 60/451,065 entitled Method for Fabricating Composite Electrodes filed Feb. 26, 2003; and to copending provisional application serial No. 60/446,848 entitled Polymer Electrolyte for Electrochemical Cell filed Feb. 11, 2002; and to copending provisional application serial No. 60/443,892 entitled Nonaqueous Liquid Electrolyte filed Jan. 30, 2003; and to PCT/US03/02127 filed Jan. 22, 2003; and to PCT/US0302128 filed Jan. 22, 2003; and is a Continuation-in-Part of co-pending application Ser. No. 10/167,490 filed Jun. 12, 2002, which is a Continuation-in-Part of co-pending application Ser. No. 10/104,352, filed Mar. 22, 2002, the disclosure of each of which is incorporated herein by reference.

[0003] The present invention relates to electrochemical storage devices containing a nonaqueous polymer electrolyte with high ionic conductivity, low impedance, and high thermal stability. More particularly the present invention relates to polysiloxane type electrolytes for electrochemical storage devices.

[0004] The demand for lithium secondary batteries to meet high power and high-energy system applications has resulted in substantial research and development activities to improve their safety, as well as performance. As the world becomes increasingly dependent on portable electronic devices, and looks toward increased use of electrochemical storage devices for vehicles, power distribution load leveling and the like, it is increasingly important that the safety of such devices be paramount, especially as these devices are being used in such environments as airliners and space vehicles. The effort to date has included research in flame-retardants, solid polymer electrolytes, and new electrolyte concepts with improved thermostability. Thus, the development of highly conductive electrolytes, free of any problems associated with volatile and combustible solvents, is of paramount importance. Polymer electrolytes have a number of advantages over the materials currently used in high-performance batteries. Compared to conventional batteries using heavy metals such as lead, nickel, and cadmium, the lithium-based polymer electrolyte batteries would be many times lighter. The proposed designs of these batteries would also allow them to be formed into various shapes besides the traditional cylindrical or rectangular shapes typical in regular batteries. (See, Gray, F. M. Solid Polymer Electrolytes; VCH Publishers: New York, 1991, at 1-4, 5-21). These advantages have fueled commercial interests and research interests into improving the conductive properties of polymer electrolytes. Poly(ethylene oxide) (PEO) has been known to dissolve alkali metal salts and exhibit conductivity since the early 1970's. (See, Fenton, D. E.; Parker, J. M.; Wright, P. V. Polymer 1973, 14, at 589). The properties of PEO have been studied extensively and are outlined in a number of review articles and texts. (See Gray supra; See also, Shriver, D. F.; Bruce, P. G. in Solid State Electrochemistry, P. G. Bruce, ed., Cambridge University Press: New York, 1995, at 95, 97-106). Of more recent interest has been a class of polymers with SiO backbones known as polysiloxanes. (See, Mark, J. E.; Allcock, H. R.; West, R. Inorganic Polymers; Prentice Hall: N.J., 1992, at 3). The SiO backbone of these polymers is highly flexible, yet retains its chemical and thermal integrity. Polysiloxanes also have low glass transition temperatures, T_g, and, when substituted with oligoethylene oxide groups, solvate lithium salts, making the polysiloxanes good candidates for potentially useful polymer electrolytes. (See, Hooper, R.; Lyons, L. J.; Mapes, M. K.; Schumacher, D.; Moline, D. A.; West, R. Macromolecules 2001, 34, 931. R. Hooper, L. J. Lyons, D. A. Moline, R. West, Silicon Chemistry 2002, 1, at 121; Lin, C.; Kao, H.; Wu, R.; Kuo, P. Macromolecules 2002, 35, at 3083; Zhou, G.; Khan, I. M.; Smid, J. Macromolecules 1993, 26, at 2202; Siska, D. P.; Shriver, D. F. Chem. Mater 2001, 13, at 4698; Zhu, Z.; Einset, A. G.; Yang, C.; Chen, W.; Wnek, G. E. Zhu, Z.; Einset, A. G.; Yang, C.; Chen, W.; Wnek, G. E. Macromolecules 1994, 27, at 4076; Zhu, Z.; Einset, A. G.; Yang, C.; Chen, W.; Wnek, G. E.; Fish, D; Khan, I. M.; Wu, E.; Smid, J. British Polymer Journial 1988, 20, at 286; Morales, E.; Acosta, J. L. Electrochimica Acta 1999, 45, at 1049.)

[0005] The present inventors synthesized the polysiloxane polymers of the present invention with the goal of creating low viscosity, low molecular weight samples to enhance ionic conductivity.

[0006] A primary objective of the present invention is to provide a high conductivity, safe liquid electrolyte for use in electrochemical storage devices, especially lithium secondary batteries, as well as super capacitors and hybrid devices.

[0007] Another objective is to provide a safe electrolyte that is easy and inexpensive to manufacture.

[0008] Yet another objective is to provide an electrolyte for medical and high energy applications.

[0009] Accordingly, the present invention relates to electrochemical storage devices, especially lithium-based secondary batteries, with a nonaqueous and nonvolatile polymeric electrolyte with high room temperature ionic conductivity and enhanced safety.

[0010] To fulfill the above objectives, the nonaqueous and nonvolatile polymeric electrolyte solutions in the present invention are prepared by using the compositions which comprise poly(siloxane-g-ethylene oxide) in a liquid state with proper viscosity and a dissolved lithium salt.

[0011] FIG. 1 is a visual summary of the synthesis of poly(siloxane-g-3 ethylene oxide) (n8), (Sample W100).

[0012] FIG. 2 is a plot of conductivity derived from impedance measurements for several side chain oxygen to lithium (EO:Li) ratios.

[0013] FIG. 3 is a plot of potential vs. current density for Sample W100 Doped with LiTFSI (EO:Li32:1).

[0014] FIGS. 4A and 4B represent the structures of the three linear polysiloxane polymers of Example 2.

[0015] FIG. 5 is a plot of conductivity derived from impedance measurements of the samples of Example 2.

[0016] Poly(siloxane-g-ethylene oxide) (see General Formula I below) electrolyte shows outstanding flame resistance. Its propagation rate is significantly lower than conventional liquid electrolyte for lithium ion batteries and low molecular weight polyethylene oxide electrolyte. This is because the inorganic siloxane backbone naturally acts as a combustion inhibitor, resulting in a two stage combustion process. The initial stage involves decomposition of the stable polymer into flammable volatile by-products. For polysiloxanes such as our electrolyte, this requires a lot of energy due to the strength of the SiO backbone. This means that in order for combustion to occur, the heat added must be sufficient to decompose the polymer, ignite the by-products, and transfer enough heat back to the polymer to the sustain the reaction. Additionally, our polymer is more thermally stable than the carbonates presently used in the industry as a standard. This is stability arises from the polymer's large molecular size, nonvolatile nature and the higher temperatures required to vaporize. The SiO backbone also gives this polymer the added benefit of being nontoxic. The thermal stability and nontoxicity of this polymer electrolyte make it particularly well suited for medical device applications, especially implanted batteries for such devices as cardiac assist pumps, insulin pumps, neuromuscular stimulators, cardiac pacemakers, automatic defibrillators, cochlear implants, and other bioelectronic devices. The usage of this polymer in place of the traditional carbonates in medical device batteries would substantially improve safety. The polymer electrolyte of the present invention is also well suited for high energy applications such as electric and hybrid vehicles, submarines, satellites, and load-leveling installations.

[0017] Referring to FIG. 1, a visual summary of the synthesis, the liquid polymer electrolyte of the form visually depicted by General Formula I was synthesized.

[0018] In particular, a species of this type with (n8 and m0), labeled W100 poly(siloxane-g-3 ethylene oxide) (n8) (General Formula II) was synthesized using commercially available precursors involved in two major steps, as described below. Note R4 is O-(alkylene oxide)k-R11 wherein R11 is alkyl group. Note further that viscosity increases with n, becoming a solid at values exceeding about 20. Preferably, n should range from 4 to 20, more preferably from 4 to 12, and most preferably approximately 8. R1, R2, R3, R8, R9 and R10 are preferably chosen from the group consisting of: methyl, ethyl, propyl, and butyl. R5, R6 and R7 are preferably chosen from the group consisting of: methyl, ethyl, propyl, and butyl.

[0019] Step 1: Ring Opening Polymerization Synthesizing a MD_n^HM (n8) Intermediary

[0020] The commercially available compounds 1,3,5,7-tetramethylcyclotetrasiloxane (D₄^H, Gelest Inc., 68.59 g or 0.286 mol) (General Formula III) and hexamethyldisiloxane (HMDS, Aldrich, 23.15 g or 0.143 mol) (a disiloxane described by General Formula IV for the case wherein, R1R2R3R8R9R10CH3) were used as precursor materials.

[0021] Note that alternate precursor materials may be used such as cyclical polysiloxane with three to ten silicon-oxygen repeating units, for example:

[0022] D₃^H: 1,3,5-trimethylcyclotrisiloxane,

[0023] D₄^H: 1,3,5,7-tetramethylcyclotetrasiloxane (as used in Example 1),

[0024] D₅^H: 1,3,5,7,9-pentamethylcyclopentasiloxane,

[0025] D₆^H: 1,3,5,7,9,1-hexamethylcyclohexasiloxane,

[0026] D₇^H: 1,3,5,7,9,11,1 3-heptamethylcycloheptasiloxane,

[0027] D₈^H: 1,3,5,7,9,11,13,15-octamethylcyclooctasiloxane,

[0028] D₉^H: 1,3,5,7,9,11,13,15,17-nonamethylcyclononasiloxane, or

[0029] D₁₀^H: 1,3,5,7,9,11,13,15,17,19-decamethylcylcodecasiloxane.

[0030] Preferred is from about three to about five silicon-oxygen repeating units.

[0031] A ring opening polymerization of the cyclic compound was performed through the addition of a chain-stopping compound, HMDS (in a 1:2 molar ratio), in the presence of concentrated sulfuric acid (2.6% by wt, Fisher Scientific, 1.85 g) and stirred at 60 C. for 24 hours. The 1:2 ratio was used in this preferred example; however, the ratio is not limited to 1:2. Preferably, the chain-stopping compound is added in an amount sufficient to limit final chain lengths to n4 to 20. For example, if the ratio is 2 D₄^H: 1 disiloxane, n will be 8; for 3 D₄^H: 1 disiloxane, n will be 12; etc.

[0032] The resulting mixture was then allowed to return to room temperature and washed with 10% NaHCO₃ (315 ml) and deionized water (610 ml). This liquid was dissolved in diethyl ether (200 ml) and then dried over Na₂SO₄. The diethyl ether solvent was removed on a rotary evaporator. The sample was then again dried at 70 C. under vacuum (0.05 torr) overnight (about 16 hours). At this step Nuclear Magnetic Resonance (NMR) characterization was performed on the current product, MD_n^HM (n8) (General Formula V, with a more generalized chemical structure of the Intermediary shown in General Formula VI), yielding a spectrum consistent with the proposed structure (Table 2). R1, R2, R3, R8, R9 and R10 are preferably chosen from the group consisting of: methyl, ethyl, propyl, and butyl. R5, R6 and R7 are preferably chosen from the group consisting of: methyl, ethyl, propyl and butyl.

[0033] wherein at least R4 or R5 is H.

TABLE 2
NMR Spectra of MDnHM (n8)
NMR Spectra (in CDCl3)	Line 1 (ppm)	Line 2 (ppm)
1H	4.70	0.210.12
	(broad, 1H, SiH)	(m, 5.3H, SiCH3)
29Si	9.55	35.17
	(m, OSi(CH3)3),	(SiH)

[0034] Step 2: Addition of a PEO Side-chain The above sample MD_n^HM (n8) (48.54 g, 0.617 mol SiH) was then mixed in a 500 ml flame dried flask containing tri(ethyleneglycol) monomethyl ether (PEOMe) (k3) (Aldrich, 101.22 g or 0.617 mol) (described by the General Formula VII for the case wherein k3). Also added to the mixture was tris(pentafluorophenyl)boron, B(C₆F₅)₃, (Aldrich, 0.16 g or 0.31 mmol) (General Formula VIII), which served as a catalyst for the dehydrocoupling reaction which occurs between the SiH groups of the MD_n^HM (n8) and the OH groups of the PEOMe (m3). Note, the dehydrocoupling catalyst should be loaded as a percentage of the moles of SiH groups present in the substrates, preferably between 0.01% and 10%, more preferably between 0.01% and 2.00%, and most preferably about 0.05% on a per mole basis. This mixture was vacuum pumped down (0.05 torr) and then filled with argon, four times in succession. Next, a solvent (150 mL toluene, dried over sodium) was added and the entire mixture was heated to 70-75 C., which caused bubbling to occur. Other solvents such as benzene may be used in place of toluene, and the mixture may be heated to between 40 C. and 200 C. The mix was then stirred until this bubbling ceased, approximately 17 hours. The sample was then further dried at 125 C. under vacuum (0.02 torr) to generate a colorless oil (137.6 g). Infrared spectroscopy (IR) showed the absence of HO or SiH groups, an indication that our reaction proceeded to completion. Table 3 shows the spectra resulting from the NMR, similarly indicative of the absence of HO or SiH groups, and that the reaction proceeded to completion (Table 3). Detecting traces of boron via mass spectroscopy, FTIM, X-ray diffraction, and/or neutron diffraction in a polysiloxane, and in particular in poly(siloxane-g-3 ethylene oxide) (n8), could be used to reveal that our inventive employment of a boron catalyst was utilized. It is noted that, although boron is considered the best catalyst, alternate catalysts including alkali metal or alkaline earth hydroxides, alkali metal or alkaline earth carbonates, triethylamine, and pyridine may be used. See, S. Kohama, Y Umeki, J., Applied Polymer Sci., 21, 863 (1977). For example, and without limitation, zinc octanoate, triethylamine, pyridine, potassium hydroxide, magnesium hydroxide, potassium carbonate may be used. Experiments have further demonstrated that a rhodium complex, RhCl(PPh₃)₃, is an efficient catalyst but produces a dark color in the polymer which is difficult to remove.

TABLE 3
NMR Spectra of Sample W100
NMR Spectra
(in CDCl3)	Line 1 (ppm)	Line 2 (ppm)	Line 3 (ppm)
1H	3.703.30	3.15	0.050.10
	(m, 12H, CH2)	(s, 3H, OCH3)	(m, 5.1H, SiCH3)

[0035] Ionic Conductivity: When doped, Sample W100, by virtue of its low molecular weight and viscosity, displays high levels of ionic conductivity. The sample was initially dried on a high vacuum line (pressure reached 9.510⁵ torr) following synthesis. Before testing the samples were doped with the lithium salt, LiN(SO₂CF₃)₂ (LiTFSI) at various concentrations. These concentrations were calculated based on the molar ratios between the amount of side-chain oxygen molecules in the sample and lithium cations present in the salt (Table 4). Alternate candidate alkali metal salts include the lithium salts: LiClO₄, LiBF₄, LiAsF₆, LiPF₆, LiCF₃SO₃, Li(CF₃SO₂)₂N, LiC(CF₃SO₂)₃, LiN(SO₂C₂F₅)₂, lithium bis(oxalato) borate (LiBOB), lithium alkyl fluorophosphates, and mixtures thereof. Other alkali metal salts may be used, particularly those comprising at least one quaternary ammonium salt having an anion selected form the following groups: ClO₄, BF₄, AsF₆, PF₆, CF₃SO₃; (CF₃SO₂)₂N, C(CF₃SO₂)₃C, (C₂F₅SO₂)₂N, PF₃(C₂F₅)₃, PF₃(CF₃)₃ and B(C₂O₄)₂.

TABLE 4
LiTFSI Doping of Sample W100
EO:Li	Sample W100 (g)	LiTFSI (g)
24.1:1	1.58	0.24
15.1:1	1.49	0.37

[0036] This doping was achieved through one of two methods. The first involves direct doping of the salt and polymer by placing them in a Nalgene cup in an argon atmosphere dry box. Once inside the box the mixture was placed in a drying tube containing a Teflon stir bar. Next the tube was removed from the dry box and placed on a Schlenk line to enable the sample and salt to mix under an argon flow until homogenous mixing occurred. The criterion used to evaluate this level of mixing was the absence of salt crystals based on unaided visual inspections. This was achieved following several hours (overnight) of constant stirring. The second, or solution, method involves placing the desired amount of sample in a Nalgene cup that is then transferred to a sealed drying tube with a Teflon stir bar within the dry box. The salt is transferred into the tube through a syringe containing 0.052 M LiTFSI in a tetrahydrofuran (THF) solution under an argon flow. The mix was then allowed to stir to achieve homogeneity on a Schlenk line. The THF solvent is removed on the Schlenk line and the tube is then placed on a high vacuum line until a pressure below 310⁵ torr is reached. The direct doping method was used in the preparation of doped Sample W100.

[0037] The solution doping method was used when less than 15 mg of salt will be used since that small amount of salt cannot be measured precisely in a dry box. When using 50 to 100 mg of salt, precise amounts can be measured in a dry box allowing the use of the direct method, which does not expose the sample to additional solvent (which then needs to be removed). Sample W100 was mixed with the direct method since about 100 mg of salt was added to the pure polymers. It is quicker and ensures that no additional solvent needs to be used.

[0038] Sandwich conductivity cells sealed with o-rings were used to measure conductivity. The cells were placed in the dry box and had their dimensions measured to enable the calculation of conductivity according to the equation, $=\frac{L}{\mathrm{RA}},$

[0039] where is conductivity (S/cm), L is the length of the containment ring (cm), R is the resistance () and A is area (cm²). Three (3) different containment rings were used on our cells, giving geometric factors $\frac{L}{\left(A\right)}$

[0040] between 0.208 and 0.293 cm¹. Resistance was derived from impedance measurements according to the following equation, RZcos(), where Z is impedance () and corresponds to the phase angle. These values were measured by a Princeton Applied Research Potentiostat/Galvanostat Model 273A with a Model 1025 Frequency Response Analyzer using PAR PowerSine software. The parameters for these tests were a frequency range of between 75.0 Hz and 100 kHz and a default AC amplitude of 10 mV. Variations in temperature were achieved using a condenser connected to a Brinkman MGW Lauda RM 6 circulating bath.

[0041] FIG. 2 shows the results of impedance measurements yielding conductivity calculations, which were recorded at various temperatures and plotted. It is apparent that an EO:Li ratio of 15:1 yields the best conductivity for our Sample W100. This data was then fit to the Vogel-Tamman-Fulcher (VTF) equation, $={\mathrm{AT}}^{-1/2}\exp \left[\frac{-B}{T-T_0}\right],\mathrm{where}A\frac{{\mathrm{SK}}^{1/2}}{\left(\mathrm{cm}\right)}$

[0042] and B (K1) are constants and To (K) is the ideal glass transition temperature. The equation parameter B, is related to activation energy (E_a) by a constant such that, $E_a=B\left(8.31\frac{J}{\mathrm{mol}K}\right)$

[0043] (Table 5).

TABLE 5
Conductivity Data for Sample W100
	25 C.	37 C.	Ea	To
EO:Li	(S/cm)	(S/cm)	(kJ/mol)	(K)
24.8:1	2.01E4	3.08E4	3.07	202
15.1:1	2.66E4	4.58E4	4.37	196

[0044] Table 6 summarizes the conductivity measurements and VTF derived data for both the 24.8:1 and 15:1 EO:Li doping ratios. The values of temperature and conductivity are presented in the forms of 1/T1000 and log , respectively, so that the data can be easily plotted (as in FIGS. 2 and 6) and related to the VTF equation. The corresponding calculated log values derived from the VTF fit are also presented for data point for comparison.

TABLE 6
Conductivity Measurements and VTF Derived Data
24.8:1	15:1
	Calculated		Calculated
Experimental Data	Data	Experimental Data	Data
1/T 1000	log	VTF fit	1/T 1000	log	VTF fit
3.5430	4.0117	4.0231	3.5581	4.0268	4.0289
3.4680	3.8781	3.8821	3.4620	3.7992	3.8035
3.3495	3.6963	3.6861	3.3518	3.5746	3.5754
3.2263	3.5113	3.5108	3.2191	3.3382	3.3361
3.1158	3.3744	3.3735	3.1099	3.1659	3.1633
3.0225	3.2651	3.2700	2.9936	2.9947	2.9986
2.9382	3.1868	3.1846	2.9235	2.9097	2.9081

[0045] Electrochemical Stability. For the measurement of the electrochemical stability window of the polymer electrolyte, stainless steel type 2032 button cell assemblies were used, with a stainless steel disc as a working electrode and a lithium metal disc as a counter electrode. The measurement cell was assembled in an argon-filled dry box.

[0046] FIG. 3 shows the electrochemical stability of the polymer electrolyte measured by using ZAHNER electrochemical workstation IM6 with scan rate of 5 mV/sec from 2.8 V to 6.0 V measured at 25 C. The first scan cycle displays small current increase from 4.0 V onward. Successive scans show that Sample W100 doped with LiTFSI can be cycled up to 4.5 V without an additional decomposition.

[0047] Viscosity (): The viscosity of Sample W100 (see Table 7) was measured by using a Brookfield type viscometer (DV-II) with a spindle speed of 50 RPM measured at 25 C.

TABLE 7
Temp (C.)	EO:Li	(cPs)
25.0	32:1	26.50

[0048] To further explore the properties of related polymer electrolytes, the three linear polysiloxane polymers shown in FIGS. 4A and 4B were synthesized. The liquid samples were doped with LiTFSI at various concentrations before measuring conductivity. A previous study found the 32:1 EO:Li ratio to be the optimum ratio for maximum conductivity, but the polymers examined in that study were double-comb polysiloxanes. (See, Hooper, supra.)

[0049] Past studies have measured conductivities that could be considered commercially viable (Ibid.), but the polysiloxanes that were studied had to be meticulously synthesized at the laboratory level. Trying to convert these syntheses to commercial volumes would likely prove to be too difficult and costly. Thus, the methyl polysiloxane samples have the advantage of having been synthesized by much simpler schemes from the readily available starting materials poly(methylhydrosiloxane), PMHS, and poly(ethylene glycol), PEO, as a step toward making the transition from laboratory to market.

[0050] For comparison purposes, the liquid polymers W22P, W76, W100, and W102 were synthesized at the Organosilicon Research Center at the University of Wisconsin-Madison. The liquid samples were dried on a high vacuum line until ultimate pressure was reached (W22P, 1.510⁵ torr; W76, 2.210⁵ torr; W100, 9.510⁵ torr; W102, 9.510⁵ torr).

[0051] Specifically, and by way of example, Sample W76 (FIG. 4A) was synthesized in the same way as W100 except omitting Step 1 (Ring Opening Polymerization Synthesizing a MD_n^HM (n8) Intermediary) above. The materials used were:

[0052] 55%PMHS-co-PDMS (M_w9001200 from Gelest Inc., 25.0 g, 0.167 mol SiH)

[0053] Tri(ethyleneglycol) monomethyl ether (Aldrich, 27.38 g, 0.167 mol);

[0054] B(C₆F₅)₃ (Aldrich, 0.77 g, 1.5 mmol);

[0055] Toluene (150 ml).

[0056] Sample W76 (27 g) was thus obtained and tested. Spectra data (IR, NMR) for W76 were similar to that for W100 except for the presence of an additional absorption band at 19-22 ppm on the ²⁹Si NMR spectrum. Sample W76, where the PEO side chains are attached to the siloxane backbone through SiO bonds, is a novel material for use as an electrolyte. (Note that in terms of General Formula I, the precursor for Sample W76, n7 and n6.3. Generally, in the present invention, m may range from 0 to about 20, or more preferably from 0 to about 8, and most preferably should be about 0).

[0057] Similarly, Sample W102 (FIG. 4B) was synthesized according to Step 2 in the Sample W100 preparation above (i.e., omitting step 1). The materials used were:

[0058] PMHS (n34) (Gelest Inc., 41.59 g, 0.638 mol SiH);

[0059] Tri(ethyleneglycol) monomethyl ether (Aldrich, 104.59 g, 0.638 mol);

[0060] B(C₆(F₅)₃ (Aldrich, 0.17 g, 0.33 mmol);

[0061] Toluene (220 ml)

[0062] Sample W102 (139.3 g) was thus obtained and tested. Spectra (IR, NMR) data for W102 were similar to W100.

[0063] Referring to FIG. 4B, it may be seen that Sample 22P is almost identical to Sample W102. The difference in the length of the silicon oxide backbone results from the use of potassium carbonate (K₂CO₃) as a catalyst in formulating Sample W22P versus the use of tris(pentafluorophenyl)boron (B(C₆F_S)₃) as a catalyst in formulating Sample W102. Importantly, the boron-containing catalyst results in more precise control of the length of the silicon oxide backbone, and a significant improvement in performance. The use of a boron-containing catalyst is therefore much preferred to other catalysts.

[0064] Viscosity (): The viscosity of Samples W22P, W76, and W102 (see Table 8) was measured using the same method as Sample W100 (using a Brookfield type viscometer (DV-II) with a spindle speed of 50 RPM measured at 25 C.).

	TABLE 8
	Sample	Temp (C.)	EO:Li	(cPs)
	W22P	25 C.	32:1	96.00
	W76	25 C.	32:1	113.20
	W102	25 C.	32:1	64.13

[0065] Polymer Electrolyte Preparation.

[0066] Doped polymer samples were prepared by direct mixing of the salt with the polymer. Both polymer and the calculated amount of salt (side-chain oxygen to lithium ion ratio) were placed in a Nalgene cup in the dry box and sealed in a custom drying tube with a stir bar. Once out of the dry box, the tube was placed on a Schlenk line to allow the polymer and salt to stir under an argon flow until a homogeneous mixture was achieved.

[0067] FIG. 5 is a plot of conductivity derived from impedance measurements of the samples of Example 2. O-ring sealed sandwich conductivity cells were loaded in the dry box of each sample using one of three containment rings. The geometric factor, 1/A, for the rings ranged from 0.208 to 0.293 cm¹. Ionic conductivities were calculated with the equation s1/R*1/A, where s is conductivity and R is resistance. Resistance was calculated from the impedance (Z) using the equation, RZ*cos(q) where q is the phase angle and Z is the impedance. The impedance was measured on a Princeton Applied Research Potentiostat/Galvanostat Model 273A with a Model 1025 Frequency Response Analyzer operated under computer control using PAR PowerSine software. The frequency ranged from 75.0 Hz to 100 kHz and the default AC amplitude of 10 mV was used for each measurement. Measurements were taken at temperatures ranging from 0 to 70 C. by placing the conductivity cell inside a condenser attached to a Brinkman MGW Lauda RM 6 variable temperature, circulating bath. Conductivity was calculated from the impedance measurements and plotted with respect to temperature. The plot of FIG. 5 was fit to the VTF equation, (See, Fulcher, G. S. J. Am. Ceram. Soc. 1925, 8, 339) in order to calculate the activation energy from the equation parameters. The samples W100 (tested at EO:Li doping ratios of 15:1 and 24.8:1) and W102 (tested at a EO:Li doping ratio of 24.8:1) outperformed the other samples tested. It is anticipated that W102 would also perform well at the 15:1 level. From FIG. 5, there appears to be a trend over all samples that there is an optimum EO:Li doping ratio range of about 5:1 to 50:1, more preferably about 12:1 to 28:1, even more preferably about 15:1 to 25:1, and most preferably about 15:1.

[0068] FIG. 6 is a plot of VTF derived conductivity for various concentrations of LiTFSI. It is evident that the sample W100 doped at 15:1 outperforms all other samples tested throughout the test range. It should be apparent that the present invention solves the long-felt need to create safe, high energy, lightweight electrochemical storage devices having liquid electrolytes. Batteries containing the present electrolyte would be inherently safer than those with more volatile, flammable, and unstable electrolytes, and have significantly better performance due to the lower impedance and increased conductivity. Additionally, the cost of manufacturing the electrolyte of the present invention is anticipated to be lower than other alternate electrolytes.

[0069] Following the same synthetic procedures as in Example 1, but varying the amount of starting materials, Sample W119 was synthesized.

[0070] Step 1: Ring Opening Polymerization Synthesizing a MD_n^HM (n4)

[0071] Commercially available Hexamethyldisiloxane HMDS (Aldrich, 41.44 g or 0.255 mol), 1,3,5,7-tetramethylcyclotetrasiloxane D₄^H (Gelest Inc., 61.24 g or 0.255 mol) (General Formula III), and concentrated sulfuric acid H₂SO₄ (Fisher Scientific, 2.46 g) were used as precursors for the product MD_n^HM(n4) (78 g).

[0072] NMR analysis showed the following data: ¹H NMR (in CDCl₃): 4.70 ppm (broad, 1H, SiH), 0.21 ppm0.12 ppm (m, 7.4H, SiCH₃). ²⁹Si NMR (in CDCl₃): 8.78 ppm (m, OSi(CH₃)₃), 36.54 ppm (m, SiH).

[0073] Step 2: Addition of PEO side chains

[0074] Tri(ethyleneglycol) monomethyl ether (Aldrich, 65.85 g, 0.402 mol), MD_n^HM (n4) (40.39 g, 0.402 mol SiH), Tris(pentafluorophenyl)boron B(C₆F₅)₃ (Aldrich, 0.104 g, 0.202 mmol) were used to afford the colorless liquid product designated Sample W119 (96.7 g). Referring to General Formula II (as well as the final product in FIG. 1), n4.

[0075] Spectroscopic Data: IR showed no HO groups (at 33003500 cm¹) and no SiH (at 2160 cm¹) present. ¹H NMR (in CDCl₃): 3.703.30 ppm (m, 12H, CH₂), 3.15 ppm (s, 3H, OCH₃), 0.050.10 ppm (m, 7.4 H, SiCH₃).

[0076] Viscosity was measured at 30.00 cPs.

[0077] Conductivity was measured at 2.1010⁴ Scm¹ at 25.1 C. (at doping level of O/Li24, using LiTFSI).

[0078] Following the same synthetic procedures as in Example 1, but varying the amount of starting materials, Sample W168 (n6) was synthesized.

[0079] Step 1: Ring Opening Polymerization Synthesizing a MD_n^HM (n6)

[0080] Commercially available Hexamethyldisiloxane HMDS (Aldrich, 19.59 g, 0.121 mol), 1,3,5,7-tetramethylcyclotetrasiloxane D₄^H (Gelest Inc., 43.42 g, 0.181 mol), and concentrated sulfuric acid H₂SO₄ (Fisher Scientific, 1.59 g) were used as precursors for the product MD_N^HM (n6) (46.5 g).

[0081] NMR analysis yielded the following data: ¹H NMR (in CDCl₃): 4.70 ppm (broad, 1H, SiH), 0.21 ppm0.12 ppm (m, 6.1H, SiCH₃). ²⁹Si NMR (in CDCl₃): 8.92 ppm (m, OSi(CH₃)₃), 36.72 ppm (m, SiH).

[0082] Step 2: Addition of PEO side chains

[0083] Tri(ethyleneglycol) monomethyl ether (Aldrich, 39.72 g, 0.242 mol), MD_n^HM (n6) (21.56 g, 0.243 mol SiH), and Tris(pentafluorophenyl)boron B(C₆F₅)₃ (Aldrich, 0.150 g, 0.293 mmol) were used to afford the colorless liquid product designated Sample W168 (59.0 g). Referring to General formula II (as well as the final product in FIG. 1), n6.

[0084] Spectroscopic Data: IR showed no HO groups (at 33003500 cm¹) and no SiH (at 2160 cm¹) present. ¹H NMR (in CDCl₃): 3.703.30 ppm (m, 12H, CH₂), 3.15 ppm (s, 3H, OCH₃), 0.05-0.10 ppm (m, 5.9H, SiCH₃).

[0085] Following the same synthetic procedures as in Example 1, but varying the amount of starting materials, Sample W169 (n11) was synthesized.

[0086] Step 1: Ring Opening Polymerization Synthesizing a MD_n^HM (n11)

[0087] Commercially available Hexamethyldisiloxane HMDS (Aldrich, 8.28 g, 0.051 mol), 1,3,5,7-tetramethylcyclotetrasiloxane D₄^H Gelest Inc., 42.84 g, 0.179 mol), and concentrated sulfuric acid H₂SO₄ (Fisher Scientific, 1.58 g) were reacted at 75 C. for 72 hours to afford the product MD_n^HM (n1) (46.5 g).

[0088] NMR analysis yielded the following data: ¹H NMR (in CDCl₃): 4.70 ppm (broad, 1H, SiH), 0.21 ppm0.12 ppm (m, 4.7 H, SiCH₃). ²⁹Si NMR (in CDCl₃): 8.87 ppm (m, OSi(CH₃)₃), 36.31 ppm (m, SiH).

[0089] Step 2: Incorporation of PEO side chains:

[0090] Tri(ethyleneglycol) monomethyl ether (Aldrich, 45.77 g, 0.279 mol), MD_n^HM (n11) (21.93 g, 0.291 mol SiH), and Tris(pentafluorophenyl)boron B(C₆F₅)₃ (Aldrich, 0.101 g, 0.197 mmol) were used to afford the colorless liquid product designated Sample W169 (62.8 g). Referring to General formula II (as well as the final product in FIG. 1), n11.

[0091] Spectroscopic Data: IR showed no HO groups (at 33003500 cm¹) and no SiH (at 2160 cm¹) present. ¹H NMR (in CDCl₃): 3.703.30 ppm (m, 12H, CH₂), 3.15 ppm (s, 3H OCH₃), 0.050.10 ppm (m, 4.7H, SiCH₃).

[0092] The specific implementations disclosed above are by way of example and for enabling persons skilled in the art to implement the invention only. We have made every effort to describe all the embodiments we have foreseen. There may be embodiments that are unforeseeable or which are insubstantially different. We have further made every effort to describe the invention, including the best mode of practicing it. Any omission of any variation of the invention disclosed is not intended to dedicate such variation to the public, and all unforeseen or insubstantial variations are intended to be covered by the claims appended hereto. Accordingly, the invention is not to be limited except by the appended claims and legal equivalents.