High pulse power cell

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention relates generally to electrochemical cells. More particularly, the present invention relates to a nonaqueous high pulse electrochemical cell having an anode comprising an alkali metal and a solid cathode. Preferably, the electrochemical cell of the present invention includes an alkali metal hlloy anode such as lithium-aluminum anode active material, ct.'i anode current collector comprising nickel and a cathode comprising a calendared mixed metal oxide applied to a highly conductive metal screen. A preferred cathode material is silver vanadium oxide (Ag₂V₄O_n) pressed onto a cathode current collector comprising aluminum. This electrochemical system is preferably activated with an electrolytic solution comprising an alkali-metal salt of hexafluorophosphate and more preferably lithium hexafluorophosphate dissolved in a mixture of organic solvents when the anode comprises lithium.

<img file="AU688983B2_D0001.tif" />

WO 96/15562

PCT/US95/12326

- 2 2. Prior Art

The prior art has described various nonaqueous alkali metal/solid cathode electrochemical cells. While many of the prior art cells describe one or more of the components of the present high pulse power cell, none of them describes the combination of an alkali metal anode, an anode current collector comprising nickel, a calendared mixed metal oxide cathode active material such as silver vanadium oxide applied to a cathode current collector comprising aluminum and an electrolytic solution comprising an ion-forming alkali metal of hexafluorophosphate wherein the ion-forming alkali metal is similar to that comprising the anode. It is this cooperative electrochemical system that results in a synergistic improvement in the discharge efficiency of the high pulse power cell of the present invention, which magnitude of discharge efficiency is not provided by any of the prior art cells.

Among the prior art cells describing various components of the present high pulse power cell is U.S. Patent No. 4,830,940 to Keister et al., which is assigned to the assignee of the present invention. Keister et al.

+hcre describes a nonaqueous lithium electrochemical cell_Ahaving a mixed metal oxide cathode associated with a cathode conductor formed from a thin sheet of metal screen, for example titanium or stainless steel screen. The cathode can comprise silver vanadium oxide as the active material combined with a binder such as polytetrafluoroethylene, and additionally may include conductive additives such as graphite powder and acetylene black. The cathode assembly is formed by pouring the cathode active material onto the titanium or stainless steel screen and enclosing the assembly in an envelope of separator material such as polypropylene. The anode comprises an alkali metal,

WO 96/15562

PCT/US95/12326

- 3 preferably lithium or a lithium alloy enclosed within the separator and suitable electrolytes for activating the cell include a lithium salt dissolved in an organic solvent, preferably IM lithium hexafluoroarsenate (LiAsF₆) dissolved in a 50:50 by volume mixture of propylene carbonate and dimethoxyethane (PC:DME).

The electrochemical cell of the present invention comprises a calendared cathode active material pressed onto an aluminum cathode current collector. This combination results in improved cell performance.

U.S. Patent No. 5,154,992 to Berberick et al.

describes a nonaqueous lithium/manganese dioxide electrochemical cell. Other suitable cathode materials include silver vanadium oxide mixed with carbon and polytetrafluoroethylene that has been pressed to form a porous solid structure. The separator is microporous polypropylene bonded to a non-woven polypropylene substrate and the nonaqueous electrolyte includes salts such as LiPF₆and LiAsF₆. This patent does not describe calendaring of the cathode active material mixed with the binder and conductive additives and pressed onto an aluminum current collector to form the cathode component.

U.S. Patent No. 5,114,811 to Ebel et al., which is assigned to the assignee of the present invention, describes an electrochemical cell having a lithium-aluminum alloy anode and a solid cathode. The cathode comprises silver vanadium oxide associated with a current collector that can comprise titanium and stainless steel. Again, this patent does not describe the cathode as being formed in a calendaring step and aluminum is not included as ope of the materials that are useful for the cathode current collector.

WO 96/15562

PCT/US95/12326

- 4 U.S. Patent No. 5,244,757 to Takami et al. describes a lithium secondary electrochemical cell having a carbonaceous negative electrode and a positive electrode comprising a lithium/metal oxide or lithium/mixed metal oxide material, an organic binder material and a conductive additive. Examples of suitable materials for the positive electrode include such compounds as manganese dioxide, a lithium-manganese composite oxide, a lithium-nickel oxide, a lithium-manganese-cobalt oxide, a lithium containing noncrystalline vanadium pentoxide, or chalcogen compounds such as titanium disulfate or molybdenum disulfate. Also,, a part of Co in the lithium-cobalt oxide may be substituted with the metals such as transition metals, Sn, Al, Mg, T, V. The positive electrode is formed by combining these ingredients, kneading the mixture into a sheet and pressing the sheet against a current collector made of a material such as aluminum. This electrochemical system is activated with a lithium ion conductive solution that can include LiBF₄, LiPF₆ or LiCF₃SO₃. However, silver vanadium oxide is not disclosed as a suitable mixed metal oxide, and in that respect, Takami et al. does not disclose the combination of an alkali metal anode, a cathode comprising a calendared silver vanadium oxide associated with an aluminum current collector and an activating electrolytic solution comprising lithium hexafluorophosphate provided as a primary electrochemical cell.

U.S. Patent No. 4,980,250 to Takahishi et al. describes a secondary battery comprising a rechargeable positive electrode, a rechargeable lithium-containing negative electrode, a separator and an organic electrolytic solution. The positive electrode can comprise vanadium oxides combined with a conductive material such as graphite or acetylene black powder, and a binding agent such as

WO 96/15562

PCT/US95/12326

- 5 polytetrafluoroethylene powder pressed into a film-like molded article. Suitable current collectors for both the anode and the cathode comprise metals such as nickel, titanium or stainless steel and the electrolyte is described as including LiAsF₆ dissolved in a non-porotic, high-dielectric organic solvent.

U.S. Patent No. 4,925,571 to Shackle et al. describes a secondary lithium cell having a cathode composition that preferably includes V₆O₁₃ and electrically conductive carbon particles. This cell is activated with an ionicallyconductive electrolyte comprising a single-phase solid solution of an ionizable alkaline metal salt, a solvent for the salt and a polymer which has been polymerized by exposure to actinic radiation, heat or which has been chemically polymerized. Although aluminum is listed along with stainless steel as suitable materials for the cathode current collector, the preferred current collector is described as a nickel foil having particles of nickel electrochemically deposited on the surface of the foil contacting the cathode composition. Further, the cathode active material is not calendared.

The high energy density electrochemical cell of the present invention is an improvement over the prior art alkali metal/mixed metal oxide cells, specifically lithium/silver vanadium oxide cells. In terms of current carrying capacity, energy density and ease of construction, among other advantages, the combination of an alkali metalaluminum alloy anode associated with a nickel current collector, a calendared mixed metal oxide cathode active material pressed onto a cathode current collector comprising aluminum and a nonaqueous electrolytic solution comprising at least one ion-forming alkali metal salt of hexafluorophosphate, exhibits an electrochemically

WO 96/15562

PCT/US95/12326

- 6 cooperative action that results in a synergistic improvement in cell discharge efficiency in comparison to the prior art cells. In particular, the electrochemical cell of the present invention discharges efficiently under pulse currents 1.6 times higher than those described by the prior art. In other words, the electrochemical cell of the present invention can be housed in a casing having a volume 40% to 50% smaller than the prior art cell systems while still being capable of discharging under equivalent current amplitudes. Additionally, the present alkali metal anode and electrolytic solution couple provides improvements in the voltage delay and discharge characteristics of the present high pulse power cell.

These and other improvements and advantages of the present invention will become increasingly more apparent to those skilled in the art by reference to the following descriptions and to the drawings.

SUMMARY OF THE INVENTION

The electrochemical cell of the present invention preferably includes a lithium metal anode and a solid cathode. The solid cathode is preferably formed from an admixture comprising a cathode active material selected from a metal element, a metal oxide, a mixed metal oxide, or carbonaceous compounds and combinations thereof combined with conductive additives and a binder material. These ingredients are thoroughly mixed together and calendared to form a flexible sheet of the cathode active material. The flexible cathode sheet is then cut to size and pressed onto a highly conductive metal screen, preferably comprising· aluminum. A preferred cathode material is silver vanadium oxide (Ag₂V₄O_n) . Both the cathode and anode are preferably encapsulated in a polypropylene separator and wound

WO 96/15562

PC17US95/12326

- 7 together in a flattened jellyroll configuration and housed in a prismatic case. With lithium comprising the ancle, the cell is preferably activated by an electrolytic solution comprising IM LiPF₆ dissolved in a 50:50 by volume mixture of PC/DME.

BRIEF DESCRIPTION OF THE DRAWING

Fig. 1 is a block diagram and flow chart illustrating the steps involved in making a free-standing sheet of cathode active material for use in the high pulse power cell of the present invention.

Fig. 2 is a block diagram and flow chart illustrating the steps involved in manufacturing a cathode component for use in an electrochemical cell from a free-standing sheet of cathode active material.

Figs. 3 and 4 are graphs showing typical discharge curves for representative cells constructed .having a titanium cathode current collector and an aluminum cathode current collector, respectively.

Figs. 5 and 6 are graphs showing discharge curves versus pulse train for cells constructed having titanium and aluminum current collectors, respectively.

Figs. 7 and 8 are graphs showing the voltage under background load and the pl minima and p4 minima pulse train voltages versus delivered capacity of cells having titanium and aluminum cathode current collectors, respectively.

Fig. 9 is a graph showing the thickness of cells activated with either a LiAsF₆ or a LiPF₆ electrolytic solution while being discharged through a 10 ΚΩ resistor.

Figs. 10 to 17 are photographs showing the surface· condition of a variety of aluminum alloys both before and after use in an electrochemical cell.

Figs. 18 to 21 show titanium screens removed from LiPF6 electrolyte cells that were stored at 72°C for various periods of time prior to discharge.

Figs. 22 and 23 are graphs of the pulse discharge voltage versus amp hours for various cells activated with electrolytic solutions having ratios of LiAsFg 5 to LiPF6 ranging from 100:0 to 0:100.

Figs. 24 and 25 are graphs showing discharge curves versus pulse train fro cells activated with different LiPFg/LiAsFe ratios.

Figs. 26 to 28 are graphs of pulse of trains versus time for various cells activated with electrolytic solutions having ratios of LiAsF₆ to LiPF6 ranging from 10 100:0 to 0:100.

Fig. 29 is a graph of the resistance of various cells activated with electrolytic solutions having ratios of LiAsF6 to LiPF6 ranging from 100:0 to 0:100.

Figs. 30 and 31 are graphs of the discharge characteristics of cells 15 activated with electrolytic solutions having LiAsF6 and LiPF6 dissolved therein, respectively.

Fig. 32 is a graph showing the average discharge curves of a plurality of lithium/silver vanadium oxide (Li/SVO) electrochemical cells activated by a solution according to the present invention comprising 1M LiPFe dissolved in a 20 50/50 mixture (by volume) of DME/PC and subjected to the one year accelerated discharge data (ADD) regime. Comparative discharge curves are also shown for similarly built Li/SVO prior art cells activated with an electrolyte comprising a one molar solution of lithium hexafluoroarsenate (1M LiAsFg) dissolved in a 50/50 by volume mixture of DME/PC and subjected to the one 25 year ADD regime.

Fig. 33 is a graph showing trends in the amount of heat dissipated by various lithium/silver vanadium oxide cells activated by electrolytic solutions according to the

WO 96/15562

PCT/US95/12326

- 9 present invention comprising lithium hexafluorophosphate salt (LiPF₆) in comparison to electrolyte salts comprising lithium hexafluoroarsenate (LiAsF₆) and lithium perchlorate (LiClOj .

Fig. 34 is a graph showing the discharge curves of two prior art cells in comparison to the electrochemical cell of the present invention. ,

DETAILED DESCRIPTION OF THE INVENTION

The cell of the present invention has an anode comprising a metal selected from Group IA of the Periodic Table of Elements, including lithium, sodium, potassium, etc., and their alloys and intermetallic compounds including, for example Li-Si, Li-B and Li-Si-B all^s and intermetallic compounds. The preferred anode co.. j.ses lithium, and the more preferred anode comprises a lithium alloy, the preferred lithium alloy being lithium-aluminum with the aluminum comprising from between about 0% to about 50% by weight of the alloy. The greater the amount of

0 aluminum present by weight in the alloy the lower the energy density of the cell.

The form of the anode may vary, but preferably, the anode is a thin metal sheet or foil of the anode metal, pressed or rolled on a metallic anode current collector to 25 form an anode component. A nickel oxide anode current collector is preferred. In the cell of the present invention, the anode component has an extended tab or lead of the same material as the current collector or integrally formed therewith such as by welding and contacted by a weld 30 to a cell case of conductive metal in a case-negative electrical configuration. Alternatively, the anode may be formed in some other geometry, such as a bobbin shape,

WO 96/15562

PCT/US95/12326

- 10 cylinder or pellet to allow an alternate low surface cell design.

The electrochemical cell of the present invention further includes a solid cathode of an electronically conductive composite material which serves as the other electrode of the cell. The electrochemical reaction at the cathode involves conversion Qf ions which migrate from the anode to the cathode into atomic or molecular forms. The solid cathode material may comprise a metal element, a metal oxide, a mixed metal oxide or carbonaceous compounds and combinations thereof. Suitable cathode materials include silver vanadium oxide, copper silver vanadium oxide, manganese dioxide, titanium disulfide, copper oxide, copper sulfide, iron sulfide, iron disulfide, carbon and ' fluorinated carbon. The solid cathode exhibits excellent thermal stability and is generally safer and less reactive than a non-solid cathode.

Preferably, the solid cathode material comprises a mixed metal oxide formed by the chemical addition, reaction or otherwise intimate contact of various metal oxides or metal oxide/elemental metal combinations, preferably during thermal treatment, sol-gel formation, chemical vapor deposition or hydrothermal synthesis in mixed states, as is well known to those skilled in the art. The materials thereby produced contain metals and oxides of Groups IB, IIB, IIIB, IVB, VB, VIB, VIIB and VIIIB of the Periodic table of Elements, which includes the noble metals and/or their oxide compounds .

In one preferred aspect of the mixed metal oxide cathode material of the present invention, readily decomposable compounds consisting of metals from Groups IB, IIB, IIIB, IVB, VB, VIB and VIIB as well as similarly decomposable compounds from Group VIIIB are thermally

WO 96/15562

PCT/US95/12326 treated so as to effect the rapid preparation of the oxides or the respective metal elements themselves to be utilized further in the preparation of the cathode material. Such readily decomposable materials include, but are not limited to, those classes of compounds known as nitrates, nitrites, carbonates and/or ammonium salts. These precursor materials (i.e. nitrates, nitrites, carbonates, ammonium compounds, etc.) may be decomposed in a combined state or individually decomposed and thereafter combined in an oxide-decomposable metal salt compound and subsequently decomposed to form the mixed metal oxide cathode material. For a more detailed description of the preparation of a mixed metal oxide cathode material by the decomposition reaction just described, reference is made to U.S. Patent No. 4,310,609 to Liang et al., which is assigned to the assignee of the present invention and the disclosure of which is hereby incorporated by reference.

For a more detailed description of the preparation of the mixed metal oxide silver vanadium oxide as the cathode active material by a chemical addition reaction of vanadium pentoxide and various silver-containing constituents, reference is made to U.S. patent application Serial Nos. 08/090,574, 08/066,982 and 08/071,518, which are assigned to the assignee of the present invention and the disclosures of which are hereby incorporated by reference. Further, for a more detailed description of the preparation of silver vanadium oxide cathode active material by sol-gel formation, reference is made to U.S. patent application 00/271,154, which is assigned to the assignee of the present invention and the disclosure of which is hereby incorporated by reference.

Such cathode active materials are preferably first formed into a free-standing cathode sheet before being

WO 96/15562

PCT/US95/12326

- 12 pressed onto a suitable current collector for use in the high energy density electrochemical cell of the present invention. As illustrated in the block diagram flow chart in Fig. 1, cathode active material in granular form is first comminuted to reduce its particle size. This is done by comminuting the cathode active material in an attrition or grinding step 10 to reduce its particle size. A ball mill or vertical ball mill is preferred and typical grinding time ranges from between about 10 to 15 minutes. The finely divided cathode active material is preferably mixed with carbon black and/or graphite as conductive diluents and a powder fluoro-resin such as polytetrafluoroethylene powder as a binder material to form a depolarizer cathode admixture in the step designated 12. Other suitable binder materials that are useful with the cathode admixture of the present invention include ethylene propylenedienemonomer (EPDM) and polyvinylidene fluoride (PVDF). Preferably, the admixture comprises about 3 weight percent of the conductive diluents and about 3 weight percent of the binder material with the balance consisting of the cathode active material. This is typically done ir, a solvent of either water or an inert organic medium such as mineral spirits. The mixing process provides for homogeneity in the admixture and for fibrillation of the fluoro-resin to ensure structural integrity in the final cathode sheet product. In some cases, no electronic conductor material is required. In any event, the percent of cathode active material is preferably held between about 80 weight percent to about 99 weight percent.

Following the mixing step, the admixture is removed from the mixer as a paste that can be fed into a series of roll mills which compact or calendar the cathode material into a thin sheet having a tape form, or the cathode

WO 96/15562

PCT/US95/12326

- 13 admixture first can be run through a briquette mill in the step designated 14. In the latter case, the cathode admixture is formed into small pellets which are then fed into the calendar roll mills.

Typically, the compacting step 16 is performed by roll mills comprising two to four calendar mills that serve to press the admixture between rotating rollers to provide the cathode active material as a free-standing sheet in the form of a continuous tape. The cathode tape preferably has a thickness in the range of from between about 0.004 inches to about 0.020 inches. The outer edges of the tape leaving the rollers are then trimmed and the finished tape is wound up on a take-up reel as indicated at 18 to form a roll of the cathode active material. This material is subsequently subjected to a drying step under vacuum conditions to remove any residual solvent and/or water from the cathode active material. Alternatively, the process can include the dropwise addition of a liquid electrolyte into the cathode admixture prior to the admixture being subjected to the compacting step 16. This serves to enhance the performance and rate capacity of an assembled electrochemical cell incorporating the free-standing sheet of cathode active material.

The free-standing sheet of cathode active material can be stored for later use, or unwound and fed on a conveyor belt to a punching machine, as shown at 20. The punching operation 24 forms the continuous tape of cathode active material into any dimension needed for preparation of a cathode component for use in the high pulse power cell of the present invention.

As shown in Fig. 1, the method for manufacturing the free-standing cathode sheet of the present invention contains several feedback loops that serve to recycle the

WO 96/15562

PCT/US95/12326

- 14 cathode active material should the quality control not be up to an acceptable level. This contributes to the process yield as very little cathode material is actually lost to waste. In this manner, after the cathode admixture is pressed during step 16 by the series of calendar mills, if the resulting tape is too thin or otherwise of insufficient quality, the tape is sent to .a recycler, indicated as step 26, that reintroduces the cathode admixture into the feed line entering the calendar mills. If needed, the solvent concentration is adjusted during step 28 as needed to provide a more uniform consistency to the cathode admixture paste for rolling into the free-standing cathode tape.

This first recycle step 26 is also useful for reintroducing trimmings and similar leftover cathode active material back into the feed lin'e entering the calendar mills.

A second recycle loop removes trimmings of the cathode material from the process after the punching operation and feeds this material back into the calendar mills, through the recycler indicated in step 26 and the briquette mill in step 14, if that latter step is included in the process, as previously discussed. Again, the solvent concentration is adjusted during step 28 to produce a paste of the cathode active material that is suitable for rolling into a freestanding cathode tape of uniform cross-sectional thickness.

Upon completion of the drying step 20, the tape of cathode material can be stored for later use, or immediately sent to the punching operation 24. The punching operation serves to cut the sheet material into cathode plates having a variety of shapes including strips, half-round shapes, rectangular shapes, oblong pieces, or others, that are moved during step 30 to a pressing station for fabrication of a cathode component for use in the high pulse power cell of the present invention.

The operation of the pressing station 30 is described in greater detail in Fig. 2, wherein a hydraulic press having platens or fixtures is provided at 32. Receptacle cups for the punched cathode plates are weighted on a tare scale during step 34, and as a quality control check, selected ones of the cathode plates are weighted to ensure cross-sectioned uniformity. To build a cathode component, a first plate is placed in a bottom fixture during step 36 and a cathode current collector, preferably an aluminum metal screen, is placed on top of the plate during step 38. A top press fixture is then positioned on the bottom fixture and over the current collector screen as indicated at 40. This fixture receives a second shaped cathode structure during step 42 followed by a top plug during step 44 which closes the fixture assembly. The fixture assembly is then inserted into the hydraulic press as indicated at 46 which compresses the two cathode structures together 48 with the aluminum cathode current collector sandwiched therebetween to form the cathode component as a laminate. In some situations, it may be desired to press cathode structures on only one side of the current collector.

Following the pressing step, the fixture assembly is removed from the press during step 50 and the cathode component is separated from the press fixture and placed into a collection tray during step 52. Cathode components containing significant amounts of metallic electrode active material, i.e., from between about 80% to 99% by volume are inherently rigid in comparison to carbonaceous materials. As such, metallic materials suitable for use as electrode active materials have sufficient interstitial space between adjacent metallic molecules to provide for intercalation therein, as is well known to those skilled in this art. Being rigid, they can withstand the pressing step without losing their ability to intercalate metal ions produced by the anode, even without the propant function provided by the solvent and water. In that respect, metallic electrode active materials can be dried either before or after being pressed into a cathode component. The cathode

WO 96/15562

PCT/US95/12326

- 16 into—a—eQ-tleet-lon—t-ra-y—du-r-i-ng-step- 58 .—The -cathode component can then be stored for later use, or the cathode component can be incorporated into a high pulse power cell of the present invention.

For a more detailed description of a process for making cathode components for use in the high pulse power cell of the present invention·, reference is made to U.S. patent application Serial No. 08/146,707, which is assigned to the assignee of the present invention and incorporated herein by reference. The prepared cathode components can be used as either a solid cathode prepared by directly pressing the material into a cell can assembly or a spirally wound cathode structure similar to a jellyroll.

The high pulse power cell of the present invention includes a suitable separator to provide physical separation between the anode and cathode .active electrodes. The separator is of electrically insulative.serial to prevent an internal electrical short circuit between the electrodes, and the separator material also is chemically unreactive with the anode and cathode active materials and both chemically unreactive with and insoluble in the electrolytic solution. In addition, the separator material has a degree of porosity sufficient to allow flow therethrough of the electrolytic solution during the electrochemical reaction of the cell. Illustrative separator materials include non-woven glass, polypropylene, polyethylene, glass fiber material, ceramics, polytetrafluorethylene membrane commercially available under the designations ZITEa (Chemplast Inc.), polypropylene membrane, commercially available under the designation CELGARD (Celanese Plastic Company Inc.) and DEXIGLAS (C.H. Dexter, Div., Dexter Corp.)

WO 96/15562

PCT/US95/12326

- 17 The form of the separator typically is a sheet which is placed between the anode and cathode electrodes and in a manner preventing physical contact therebetween. Such is the case when the anode is folded in a serpentine-like structure with a plurality of cathode plates disposed intermediate the anode folds and received in a cell casing or when the electrode combination is rolled or otherwise formed into a cylindrical jellyroll configuration.

The electrochemical cell of the present invention further includes a nonaqueous, ionically conductive electrolytic solution operatively associated with the anode and the cathode electrodes. The electrolytic solution serves as a medium for migration of ions between the anode and the cathode during the electrochemical reactions of the cell. The electrolytic solution is comprised of an inorganic salt dissolved in a nonaqueous solvent, and more preferably, an alkali metal salt dissolved in a family of cyclic esters and dithers. The nonaqueous solvents suitable for the present invention are chosen so as to exhibit those physical properties necessary for ionic transport (low viscosity, low surface tension, and wettability). Suitable nonaqueous solvents may be any one or more of the organic solvents which are substantially inert to the anode and cathode electrode materials, such as tetrahydrofuran, propylene carbonate, methyl acetate, diglyme, triglyme, tetraglyme, acetonitrile, dimethyl sulfoxide, dimethyl formamide, dimethyl acetamide, 1,2dimethoxyethane and mixtures thereof. A preferred solvent comprises a 50/50 mixture (by volume) of propylene carbonate (PC) and dimethoxyethane (DME).

The nonaqueous solvent also may be one or a mixture of more than one of the inorganic solvents which can serve as both a solvent and a depolarizer, such as thionyl chloride,

WO 96/15562

PCT/US95/12326

- 18 sulfuryl chloride, selenium oxychloride, chromyl chloride, phosphoryl chloride, phosphorous sulfur trichloride, and others.

The preferred electrolytic solution of the present invention comprises an inorganic salt having the general formula MM'F₆ wherein M is an alkali metal similar to the alkali metal comprising the anode and M' is an element selected from the group consisting of phosphorous, arsenic and antimony. Examples of salts yielding M'F₆ are: hexafluorophosphate (PF₆) and hexafluoroarsenate (AsF₆) and hexafluoroantimonate (SbF₆) . The most preferred electrolytic soli tion comprises at least one ion-forming alkali metal salt of hexafluorophosphate dissolved in a suitable organic solvent wherein the ion-forming alkali metal is similar to the alkali comprising the anode. Thus, in the case of an anode comprising lithium, the alkali metal salt of the electrolytic solution of the present invention comprises lithium hexafluorophosphate dissolved in a 50/50 solvent mixture (by volume) of PC/DNCE. For a more detailed description of a nonaqueous electrolytic solution for use in the high pulse power cell of the present invention, reference is made to U:S. patent application Serial No. 08/169,002, which is assigned to the assignee of the present invention and incorporated herein by reference.

The assembly of the cell described herein is preferably in the form of a wound element cell. That is, the fabricated cathode, anode and separator are wound together in a jellyroll end type configuration or wound element cell stack such that the anode is on the outside of the roll to make electrical contact with the cell case in a case negative configuration. Using suitable top and bottom insulators, the wound cell stack is inserted into a

WO 96/15562

PCT/US95/12326

- 19 metallic case of a suitable size dimension. The metallic case may comprise materials such as stainless steel, mild steel, nickel-plated mild steel, titanium or aluminum, but not limited thereof, so long as the metallic material is compatible for use with components of the cell.

The cell header comprises a metallic disc-shaped body with a first hole to accommodate a glass -to-metal seal/terminal pin feedthrough and a hole for electrolytic solution filling. The glass used is of a corrosion resistant type having from between about 0% to about 50% by weight silicon such as CABAL 12, TA 23 or FUSITE 425 or FUSITE 435. The positive terminal pin feedthrough preferably comprises aluminum although molybdenum, titanium, nickel alloy, or stainless steel can also be used. The cell header comprises elements having compatibility with the other components of the electrochemical cell and is for resistant to corrosion.

The cathode lead is welded to the positive terminal pin in the glass to metal seal and the header is welded to the case containing the electrode stack. The cell is thereafter filled with the electrolytic solution described hereinabove and hermetically sealed such as by close-welding a stainless steel ball over the fill hole, but not limited thereto. This above assembly describes a case negative cell which is the preferred construction of the high pulse power cell of the present invention. As is well known to those skilled in the art, the electrochemical system of the present invention can also be constructed in a case positive configuration.

The electrochemical cell of the present invention operates in the following manner. When the ionically conductive electrolytic solution becomes operatively associated with the anode and the cathode of the cell, an

WO 96/15562

PCT/US95/12326

- 20 electrical potential difference is developed between terminals operatively connected to the anode and the cathode. The electrochemical reaction at the anode includes oxidation to form metal ions during discharge of the cell. The electrochemical reaction at the cathode involves conversion of ions which migrate from the anode to the cathode into atomic or molecular forms. It is observed that the electrochemical systems of this invention have a wide operating temperature range, of between about -55°C to about +225 °C.

The electrochemical cell according to the present invention is illustrated further by the following Examples, which are given to enable those skilled in the art to mor.e clearly understand and practice the present invention. The Examples should not be considered as a limitation of the scope of the invention, but are described as being illustrative and representative thereof.

EXAMPLE I

Comparison Of Titanium And Aluminum Cathode Current Collectors

Seven cells were constructed having an external volume of approximately 5 cc and comprising the silver vanadium oxide cathode active sheet material of the present invention. Four of the cells contained sheeted cathodes applied to the current collector material used in the prior art, i.e., titanium, and a lithium foil anode. These cells were vacuum filled with a IM LiAsF₆ PC/DME (50:50) electrolytic solution. The remaining three cells contained sheeted cathodes applied to the current collector material of the present invention, i.e., aluminum, and a lithium aluminum alloy foil anode. These cells were vacuum filled

WO 96/15562

PCT/US95/12326

- 21 with a IM LiPF₆ PC/DME (50:50) electrolytic solution. The cathode sheet material incorporated into both groups of cells was comprised of a mixed metal oxide, specifically an admixture comprising silver vanadium oxide, conductive carbon additives and a suitable binder. Both the cathode and anode were encapsulated with a 0.001 inch polypropylene separator, wound together in ,a flattened jelly^oll configuration, and housed in a prismatic case. The active surface area of the electrodes was about 114 cm².

A constant resistive load of 7.5 kQ was applied to all seven cells during an initial pre-discharge period. The pre-discharge period is referred to as burn-in and depleted the cells of approximately 1% of their theoretical capacity. Following burn-in, the cells were discharged to 1.4V by the application of a 2.0 Amp pulse train every 30 minutes. The pulse train consisted of four 10 second 2.0 Amp pulses with 15 second rests between each pulse.

- Figs. 3 and 4 are graphs showing typical discharge curves for representative cells constructed according to this example and having a titanium cathode current collector (Fig. 3) and an aluminum cathode current collector (Fig. 4). In Fig. 3, curve 60 was constructed from the pre-pulse open circuit voltage with no load of the representative cell having the titanium cathode current collector, curve 62 was constructed from the minima of the first pulse of each pulse train of this cell and curve 64 was constructed from the minimum of the fourth pulse of each pulse train. Contrastingly, in Fig. 4, curve 66 was constructed from the pre-pulse open circuit voltage with no load of the cell constructed according to the present invention having the aluminum cathode current collector, curve 68 was constructed from the minima of the first pulse of each pulse train of this cell and curve 70 was

WO 96/15562

PCT/US95/12326

- 22 constructed from the minima of the fourth pulse of each pulse train.

Table 1 shows the average (±1 standard deviation) pre burnti-in and post burn-in open circuit (OC) voltage, capacity delivered to 2.0V, 1.7V and 1.5V and cathode efficiency to 2.0V, 1.7V and 1.5V for each experimental cell group having the titanium and aluminum cathode current collectors, respectively. Cathode efficiency is defined as the percentage of theoretical capacity delivered during discharge.

TABLE 1

Titanium Cathode Aluminum Cathode

Current Collector Current Collector Lithium Foil Anode Lithium Aluminum

Foil Anode

Electrical Data

IM LiAsF₆ PC/DME(5 0:50)

IM LiPF₆PC/DME (50:50)

Pre Burn-In

OC Voltage

Post Burn-In

OC Voltage

<td colspan="2"> Capacity</td>

<td> Delivered to..</td><td></td>

<td> 2.0 V</td><td> 0</td>

<td> 1.7 V</td><td> 0</td>

<td> 1.5 V</td><td> 0</td>

Cathode

Efficiency to..

2.0V 56.2+6.8%

1.7V 74.5+3.3%

1.5V 82.2+2.9%

3.290+0.068 V

3.174+0.009 V

491+0.066 Ahrs

651+0.051 Ahrs

719+0.053 Ahrs

3.225+0.007 V

3.169+0.001 V

0.629+0.006Ahrs

0.742+0.007Ahrs

0.786+0.OOSAhrs

71.2+0.5%

84.0+0.6%

88.9+0.5%

The cells built according to the present invention with an aluminum cathode current collector, lithium aluminum alloy foil anode and filled with IM LiPF₆PC/DME(50:50) electrolytic solution were 14.2%, 8.5%, and 6.4% more efficient at voltage cut-offs of 2.0V, 1.7V and 1.5V, respectively, than th? prior art cells built with a

WO 96/15562

PCT/US95/12326

- 23 titanium cathode current collector, lithium foil anode and filled with IM LiAsF₆ PC/DME(50:50) electrolyte. The noted increases in discharge efficiencies are attributable to the cooperative electrochemical action these specific materials have when combined with each other. In other words, the combination of cell materials preferred for the high pulse power cell of the present invention has a synergistic effect on cell performance that is statistically significant at a 99% confidence level at each voltage cut-off.

Further, the electrochemical cells of the present invention activated with the electrolytic solution comprising lithium hexaflourophosphate exhibited reduced reactivity with the lithium aluminum alloy foil, anode in comparison to the prior art cells having the lithium foil anode and activated with a lithium hexaflouroarsenate electrolytic solution. This diminished reactivity reduces the thickness of the passivation film on the anode, which film undesirably causes voltage delay. Additionally, the lower reactivity with the lithium aluminum alloy foil anode results in reduces heat dissipation and in turn further benefits cell efficiency. This is an unexpected benefit of the present cell system since generally improvements in voltage delay are realized at the expense of heat dissipation.

EXAMPLE II

Comparison Of Titanium And Aluminum Cathode Current Collectors

Four hermetically sealed prismatic lithium/silver vanadium oxide cells were constructed in the standard configuration. The cathode admixture consisted of silver vanadium oxide, carbon, graphite, and a suitable binder.

WO 96/15562 PCT/US95/12326

- 24 The mix was applied to titanium cathode current collectors for the first two cells and to aluminum cathode current collectors for the second two cells. The cathodes of all four cells were encapsulated in a polypropylene film laminate separator. The anode of each cell was a strip of metallic lithium, also encapsulated in a polypropylene film laminate separator. The anodes were wrapped accordion style around the cathode plates. The cell assemblies were then assembled in prismatic housings and vacuum filled with IM LiAsF₆ PC/DME (50:50) electrolyte.

The cells were placed on a 12.1 ΚΩ background load and pulse discharged every 60 days with a 2.0 Amp pulse train. The pulse train consisted of four 10 second pulses with 15 second rest periods between each pulse. Following the application of six such pulse trains, the cells were discharged to 1.0V under the background load.

Table 2 shows selected pulse train voltages, in mV, for each cell tested. The particular pulse train voltages shown are the pre-pulse background voltage (pre-pulse), the minimum voltage record during the first pulse of the pulse train (pl min), the minimum voltage recorded during the fourth pulse of the pulse train (p4 min), the difference between the last voltage recorded during the first pulse of the train and the minimum voltage recorded for that pulse (pl end-pl min) and the difference between the pre-pulse voltage and p4 min (pre pl-p4 min). The average and standard deviation of the selected data points are also indicated in Table 2 for each cathode current collector material.

Pulse Train Data: Ti vs Al Screen Comparison

Pre-Pulse Pl Min

Cell 1 Cell 2 Celli Cell 2

TABLE 2

12.1 kO background load 2.0 Amp pulse train applied every 60 day:

CA

<img file="AU688983B2_D0002.tif" />

P4 Min

Cell 1 Cell 2

Pl End-Pl Min

Cell 1 Cell 2

Pre Pl-Pl Min

Cell 1 Cell 2

Pre Pl

Cell 1 Cell 2

<td> Pulse Train ] Ti Screen Al Screen</td><td> 3333 3329</td><td> 3329 3332</td><td> 2250· 2260</td><td> 2220 2380</td><td> 2350 2380</td><td> 2320 2475</td><td> 120 100</td><td> 110 120</td><td> 1083 1069</td><td> 1109 952</td><td> 983 949</td><td colspan="2"> 1009 857</td>

<td> Pulse Train 2</td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td>

<td> Ti Screen</td><td> 3118</td><td> 3113</td><td> 2300</td><td> 2292</td><td> 2261</td><td> 2253</td><td> 93</td><td> 93</td><td> 818</td><td> 821</td><td> 857</td><td> 860</td><td></td>

<td> Al Screen</td><td> 3113</td><td> 3115</td><td> 2246</td><td> 2356</td><td> 2246</td><td> 2366</td><td> 73</td><td> 129</td><td> 867</td><td> 759</td><td> 867</td><td> 749</td><td></td>

<td> Pulse Train 3</td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td>

<td> Ti Screen</td><td> 2639</td><td> 2629</td><td> 2109</td><td> 2112</td><td> 2063</td><td> 2056</td><td> 0</td><td> 0</td><td> 530</td><td> 517</td><td> 576</td><td> 570</td><td></td>

<td> Al Screen</td><td> 2625</td><td> 2634</td><td> 2163</td><td> 2197</td><td> 2146</td><td> 2158</td><td> 12</td><td> 0</td><td> 462</td><td> 437</td><td> 479</td><td> 476</td><td></td>

<td> Pulse Train 4</td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td>

<td> Ti Screen</td><td> 2537</td><td> 2534</td><td> 1672</td><td> 1653</td><td> 1760</td><td> 1741</td><td> 122</td><td> 117</td><td> 865</td><td> 881</td><td> 777</td><td> 793</td><td></td>

<td> Al Screen</td><td> 2534</td><td> 2537</td><td> 1663</td><td> 1731</td><td> 1848</td><td> 1863</td><td> 192</td><td> 144</td><td> 871</td><td> 806</td><td> 686</td><td> 674</td><td></td>

<td> Pulse Train 5</td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td>

<td> Ti Screen</td><td> 2432</td><td> 2415</td><td> 1616</td><td> 1562</td><td> 1482</td><td> 1406</td><td> 0</td><td> 0</td><td> 816</td><td> 853</td><td> 950</td><td> 1009</td><td></td>

<td> Al Screen</td><td> 2397</td><td> 2422</td><td> 1636</td><td> 1687</td><td> 1533</td><td> 1575</td><td> 0</td><td> 0</td><td> 761</td><td> 735</td><td> 864</td><td> 847</td><td></td>

<td> Pulse Train 6</td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td>

<td> Ti Screen</td><td> 2173</td><td> 2163</td><td> 1067</td><td> 969</td><td> 781</td><td> 623</td><td> 0</td><td> 0</td><td> 1106</td><td> 1194</td><td> 1392</td><td> 1540</td><td></td>

<td> Al Screen</td><td> 2153</td><td> 2170</td><td> 1194</td><td> 1262</td><td> 898</td><td> 952</td><td> 0</td><td> 0</td><td> 959</td><td> 908</td><td> 1255</td><td> 1218</td><td></td>

TABLE 2 (Continued)

Average Pulse Train Data (Note: The average of each data point ± 1 standard deviation is indicated)

Pre-Pulse Pl Min P4 Min Pl End-Pl Min Pre Pl-Pl Min Pre Pl

<td> Pulse Train 1 Ti Screen Al Screen</td><td> 3331 ± 3 3331 ± 2</td><td> 2235 ± 21 2320 ± 85</td><td> 2335 ± 21 2428 ± 67</td><td> 115 + 7 110 ± 14</td><td> 1096 ± 18 1011 ± 83</td><td> 996 ± 18 903 ± 65</td>

<td> Pulse Train 2 Ti Screen Al Screen</td><td> 3116 ± 4 3114 ± 1</td><td> 2296 ± 6 2301 ± 78</td><td> 2257 ± 6 2306 ± 85</td><td> 93 ± 0 101 ± 40</td><td> 820 ± 2 813 ± 76</td><td> 859 ± 2 808 ± 83</td>

<td> Pulse Train 3 Ti Screen Al Screen</td><td> 2634 ± 7 2630 ± 6</td><td> 2111 ± 2 2180 ± 24</td><td> 2060 ± 5 2152 ± 8</td><td> 0 ±0 6 ±8</td><td> 524 ± 9 450 ± 18</td><td> 573 ± 4 478 ± 2</td>

<td> Pulse Train 4 Ti Screen Al Screen</td><td> 2536 ± 2 2536 ± 2</td><td> 1663 ± 13 1697 ± 48</td><td> 1751 ± 13 1856 ± 11</td><td> 120 ± 4 168 ± 34</td><td> 873 ± 11 839 ± 46</td><td> 785 ± 11 680 ± 8</td>

<td> Pulse Train 5 Ti Screen Al Screen</td><td> 2424 ± 12 2410 ± 18</td><td> 1589 ± 38 1662 ± 36</td><td> 1444 ± 54 1554 ± 30</td><td> 0 ± 0 0 ±0</td><td> 835 ± 26 748 ± 18</td><td> 980 ± 42 856 ± 12</td>

<td> Pulse Train 6 Ti Screen Al Screen</td><td> 2168 ± 7 2162 ± 12</td><td> 1018 + 69 1228 ± 48</td><td> 702 ± 112 925 ± 38</td><td> 0 ±0 0 ±0</td><td> 1150 ± 62 934 ± 36 1</td><td> 1466 ± 105 1237 ± 26</td>

WO 96/15562 PCT/US95/1232,

WO 96/15562

PCT/US95/12326

- 27 The pre-pulse voltages of the cells were equivalent, regardless of cathode current collector material, throughout pulse testing. The pl end-pl min voltages were also equivalent for all of the pulse trains except pulse train 4, where the aluminum screen cells had slightly higher voltages. The pl min voltages of the cells with the aluminum cathode current collectors were higher than those with the titanium cathode current collectors for pulse trains 1, 3, 5 and 6. The p4 min voltages of the cells with the aluminum cathode current collectors were higher than those with the titanium cathode current collectors for pulse trains 1, 3, 4, 5 and 6. The pre pl-pl min and pre pl-p4 min voltages were correspondingly lower for the aluminum cathode current collectors for pulse trains 1, 3, 5 and 6 and 1, 3, 4, 5 and 6, respectively. These trends are shown graphically in Figs. 5 and 6. In Fig. 5, curves 72 and 74 were constructed from the average background voltage, i.e., pre-pulse voltage, of the cells construed according to this example having a titanium cathode current collector and an aluminum cathode current collector, respectively, and curves 76 and 78 were constructed from the average pl min (pl) voltage versus the pulse train of these cells, respectively. Fig. 6 again shows the average background voltage of the cells used to construct the discharge curves in Fig. 5, but on an expanded scale, and curves 80 and 82 were constructed from the average p4 min (p4) voltage versus the pulse trains for the cells having the titanium and aluminum cathode current collectors, respectively.

After pulse testing was complete, the cells were discharged to 1.0V under the 12.1 kb background load. Table 3 shows the capacity delivered under the load (excluding the capacity delivered during pulsing) and the

WO 96/15562

PCT/US95/12326

- 28 change in case thickness resulting from discharge under the background load. The cells delivered equivalent amounts of capacity to 2.0V, 1.7V and 1.5V cut-offs and swelled within the same orders of magnitude.

TABLE 3

Capacity Delivered un&lt; er 12.IK Background Load

<td> Screen</td><td> mAhrs to</td><td> mAhrs to</td><td> mAhrs to</td><td> End of</td><td> End of Test</td><td> Change in</td>

<td> Type</td><td> 2.0V</td><td> 1.7V</td><td> 1.5V</td><td> Pulsing Thickness</td><td> Thickness</td><td> Thickness</td>

<td> Ti</td><td> 1765</td><td> 1901</td><td> 1929</td><td> 0.422</td><td> 0.429</td><td> 0.007</td>

<td> 'Ti</td><td> 1734</td><td> 1868</td><td> 1918</td><td> 0.422</td><td> 0.437</td><td> 0.015 </td>

<td> Average</td><td> 1750</td><td> 1885</td><td> 1924</td><td> 0.422</td><td> 0.433</td><td> 0.011</td>

<td> St Dev</td><td> 22</td><td> 23</td><td> 8</td><td> 0.000</td><td> 0.006</td><td> 0.006</td>

<td> Al</td><td> 1719</td><td> 1854</td><td> 1912</td><td> 0.422</td><td> 0.443</td><td> 0.021</td>

<td> Al</td><td> 1750</td><td> 1885</td><td> 1924</td><td> 0.423</td><td> 0.436</td><td> 0.013</td>

<td> Average</td><td> 1735</td><td> 1870</td><td> 1918</td><td> 0.423</td><td> 0.440</td><td> 0.017</td>

<td> St Dev</td><td> 22</td><td> 22</td><td> 8</td><td> 0.001</td><td> 0.005</td><td> 0.006</td>

<td colspan="2"> Figs. 7 and</td><td> 8 show</td><td colspan="2"> the voltage under</td><td colspan="2"> background load</td>

and the pl min and p4 min pulse train voltages versus delivered capacity of the titanium cathode current collector cells and the aluminum cathode current collector cells, respectively. In Fig. 7, curves 84 and 86 were constructed from the background voltage of the two cells in this example having the titanium cathode current collector, curves 88 and 90 were constructed from the pl min voltage versus delivered capacity of these cells, respectively, and curves 92 and 94 were constructed from the p4 min voltage versus delivered capacity of these cells, respectively,

WO 96/15562

PCT/US95/12326

- 29 Contrastingly, in Fig. 8, curves 96 and 98 were constructed from the background voltage of the two cells in this example having the aluminum cathode current collector, curves 100 and 102 were constructed from the pl min voltage versus delivered capacity of these cells, respectively, and curves 104 and 106 were constructed from the p4 min voltage versus delivered capacity of these cells, respectively.

The titanium and aluminum cathode current collectors have comparable performance characteristics in terms of deliverable capacity and case swelling behavior. The aluminum cathode current collector cells have better rate capabilities as indicated by the higher pl min and p4 min pulse train voltages.

EXAMPLE III

Swelling Of Cells Containing- LiAsF* Versus Cells Containing LiPF,_;.

A group of cells containing either LiAsF₆ or LiPF₆ at a concentration of 1 mol · L'¹ in a 1:1 volume mixture of propylene carbonate and dimethoxyethane was partially discharged through 2 ΚΩ resistors before being switched to 10 ΚΩ resistors. The thickness of the cells were recorded throughout the 10 ΚΩ portion of the discharges. Plots of cell thickness as a function of decreasing voltage under load are presented in Fig. 9 wherein curve 108 was constructed from the LiAsF₆ electrolyte salt containing cells and curve 110 was constructed from the LiPF₆electrolyte salt containing cells. Average values of cell thickness at 2000 mV and 1000 mV interpolated from the data used in the plots are presented in Table 4. It can be seen in the Fig. 9 and in Table 4 that the LiPF₆ cells were at lower voltages than were the LiAsF₆ cells before the onset of increased swelling. The LiPF₆ cells had greater

WO 96/15562

PCT/US95/12326

<td> thickness at voltages is not typically seen Comparison of</td><td colspan="2"> below about 1400 mV but that voltage in implantable medical devices. TABLE 4 Multiplate SVO cell Swelling</td>

<td></td><td> LiAsF«</td><td> LiPF*</td>

<td> Thickness ® 2V, inches Thickness ® IV, inches</td><td> 0.404±0.010 0.427±0.006</td><td> 0.352±0.003 0.437+0.008</td>

<td></td><td> EXAMPLE IV</td><td></td>

Corrosion Of Aluminum Versus Titanium Current

Collectors Upon Exposure To LiPF₆ In Propylene Carbonate/Dimethoxyethane Electrolytes

Aluminum does not corrode in the LiPF₆ electrolyte. The surface condition of a variety of aluminum alloys are shown in the photographs in Figs. 10 to 17. Figs. 10, 12, 14 and 16 are photographs showing virgin aluminum alloys, i.e., aluminum alloys prior to being used in an electrochemical cell while Figs. 11, 13, 15 and 17 are photographs showing various aluminum alloy cathode current collectors after use in an electrochemical cell. In constructing the various cells, the aluminum alloy materials were first formed into cathode current collector screens by chemical etching or by expanding the metal prior to further treatment. These treating processes are well known to those skilled in the art.

The thusly treated cathode current collectors were then associated with silver vanadium oxide in a calendaring step according to the present invention and incorporated into high pulse power cells that were discharged under a variety of conditions. Table 5 shows the treatments or conditions to which the various aluminum alloys subjected. The various aluminum alloys were designated:

WO 96/15562

PCT/US95/12326

- 31 1100-0: 1100 is wrought aluminum with a minimum puritylevel of 99.00%. 0 (zero) indicates the material has been fully annealed.

1145-H19: 1145 is wrought aluminum with a minimum purity level of 99.45%. H19 indicates the material has been strain hardened to a particular degree without subsequent thermal treatment.

3003-H14: 3003 is wrought aluminum alloy that contains

98.6% aluminum, 1.2% manganese, and 0.12% copper. H14 indicates the material has been strained hardened to a particular degree without subsequent thermal treatment. 5052-H19; 5052 is wrought aluminum alloy that contains

97.2% aluminum, 2.5% magnesium, and 0.25% chromium. H19 indicates the material has been strain hardened to a particular degree without subsequent thermal treatment.

TABLE 5

Aluminum

Figure Alloy

Exposure

Configuration

Condition

<td> Fig. 10</td><td> 1100-0</td><td> none</td><td> expanded metal</td><td> virgin</td>

<td> Fig· 11</td><td> 1100-0</td><td> LiPF₆</td><td> expanded metal</td><td> stored at OC for 3 months @ 37°C</td>

<td> Fig. 12</td><td> 1100-0</td><td> caustic dip</td><td> etched screen</td><td> virgin</td>

<td> Fig. 13</td><td> 1100-0</td><td> caustic dip</td><td> etched screen</td><td> accelerated discharged, then stored at OC for — 3 months © 37°C</td>

<td> Fig. 14</td><td> 1145</td><td> hydrofluoric acid dip</td><td> etched screen</td><td> virgin</td>

<td> Fig. 15</td><td> 1145</td><td> hydrofluoric acid dip</td><td> etched screen</td><td> accelerated discharged, then stored at OC for ~ 3 months @ 37°C</td>

<td> Fig. 16</td><td> 5052-H19</td><td> caustic dip</td><td> etched screen</td><td> virgin</td>

<td> Fig. 17</td><td> 5052-H19</td><td> caustic dip</td><td> etched screen</td><td> accelerated discharged, then stored at OC for — 3 mos. @ 37°C</td>

WO 96/15562

PCT/IJS95/12326

- 32 No corrosion was evident in any of the aluminum ‘ samples.

Accelerated discharge in this example comprised application of a train of pulses to the respective cells each half hour. Each train consisted of four 2 amp pulses of 10 seconds duration with 15 seconds rest between each pulse. Note that each of the photos of an aluminum alloy cathode current collector incorporated into a cell, i.e., Figs. 10, 12, 14 and 16 are spectacularly reminiscent of the virgin material and that no pitting, evident when corrosion takes place, can be seen in these figures regardless of the several treatments investigated.

In great contrast to the photographs showing the . aluminum alloys, Figs. 18 to 21 show titanium screens removed from LiPF₆ electrolyte cells that were stores at 72°C for 39, 61 or 29 days, respectively, prior to discharge as described above. Fig. 18 is a photograph of a virgin titanium screen. Note that the photographs in Figs. 19 to 21 show a line of pits about 1/2 up in each figure whereas no such pits are seen in Fig. 18, the virgin screen. These pits are evidence of corrosion. No such pitting is found in cells using the LiAsF₆ salt electrolyte for either of the above described cathode current collector materials .

EXAMPLE V

Mixed LiPF6/LiAsF6 Salt Cells

A first group of cells containing mixtures of LiAsF₆and LiPF₆ at a combined concentration of 1 mol-L·'¹ in a 1:1 volume mixture of propylene carbonate and dimethoxy ethane were used to investigate the accelerated discharge of mixed salt cells. LiPF₆/LiAsF₆ ratios of 0/100, 25/75, 50/50, 75/25 and 100/0 were used and the cells were discharges

WO 96/15562

PCT/US95/12326

- 33 through application of a train of pulses each half hour.

Each train consisted of four 1.5 amp pulses of 10 second duration with 15 seconds rest between each pulse. The prepulse open circuit voltage and first pulse minima (pl min) for a random selection of cells is presented in Fig. 22, and the pre-pulse open circuit voltage and fourth pulse minima (p4 min) for the same cells is presented in Fig. 23. In Figs. 22 and 23, curve 112 was constructed from the open circuit voltage (OCV) of a cell constructed according to the present invention and activated with an electrolytic solution having dissolved therein LiPF₆ and LiAsF₆ in a ratio of 100:0. Similarly,curves 114, 116, 118 and 120 were constructed from the OCV of cells activated with an electrolyte having the enumerated lithium salts in rations of 75:25, 50:50, 25:75 and 0:100, respectively. In Fig.

22, curve 122 was constructed from the first pulse minima (pl min) of the cell that was previously discharged to construct curve 112. Similarly, curves 124, 126, 128 and 130 were constructed from the first pulse minima of the cells that were previously discharged to construct curves 114, 116, 118 and 120, respectively. In Fig. 23,curve 32 was constructed from the fourth pulse minima (p4 min) of the cell that was discharged to construct curve 112 in Fig. 22. Similarly, curves 134, 136, 138 and 140 were constructed from the fourth pulse minima of the cells that were discharged to construct curves 114, 116, 118 and 120, respectively. Note that although no trend occurs as to decreasing or increasing pulse voltage values within each salt ratio, there are differences with changing ratio.

A second group of cells containing mixtures of LiAsF₆and LiPF₆ were pulse discharged using loads of 17.4 ΚΩ. The pulse trains consisted of four 1.5 amp, 10 second pulses with 15 seconds rest between each pulse. This type

WO 96/15562

PCT/US95/12326

- 34 of discharge is termed in house as one year ADD. Pre-pulse and first pulse voltage minima are plotted versus pulse trains labeled as 0-*l, l-&gt;2, 2-*3,... along the abscissa for each salt ratio group in Fig. 24 while the pre-pulse and fourth pulse voltage are plotted versus these same pulse trains in Fig. 25.

Thus, in Figs. 24 and 25, curve 142 was constructed from the pre-pulse open circuit voltage (OCV) of a cell constructed according to the present invention and activated with an electrolytic solution consisting of LiPF₆and LiAsF₆ in a ratio of 100:0 while curves 144, 146, 148 and 150 were constructed from the OCV of cells activated with an electrolytic solution having the enumerated lithium salts dissolved therein in ratio of 75:25, 50:50, 25:75 and

100:0, respectively. In Fig. 24, curve 152 was constructed from the first pulse minima (pl min) of the cell that was previously discharged to construct curve 142. Similarly, curves 154, 156, 158 and 160 were constructed from the first pulse minima of the cells that were discharged to construct curves 144, 146, 148 and 150, respectively. In Fig. 25, curve 162 was constructed from the fourth pulse minima (p4 min) of the cell that was previously discharged to construct curve 142 in Fig. 24. Similarly, curves 164, 166, 168 and 170 were constructed from the fourth pulse minima of the cells that were discharged to construct curves 144, 146, 148 and 150, respectively.

Note how the pulse minima increase for each train as the amount of LiPF₆ in the electrolyte increases, especially,in the fourth through seventh pulse trains. It 30 is evident that in general, as the LiPF₆/LiAsF₆ ratio increases, the first pulse minima increase.

The difference between first pulse minima and the final first pulse voltage immediately prior to the first

WO 96/15562

PCT/US95/12326

- 35 rest during the pulse train is known as voltage delay. Large voltage delays are undesirable for the intended application of this type of cell, for example in an implantable medical device wherein high currents are needed during device activation. Thus, it should be realized that the use of higher ratios of LiPF₆/LiAsF₆, especially, a ratio of 100/0 offers improved voltage delay characteristics.

A third group of cells containing mixtures of LiAsF₆and LiPF₆ were partially discharged through loads of 5 ΚΩ before being stored at 50°C for 2 weeks. The cells were equilibrated to 37°C prior to application of multiple pulse trains comprising four 1.5 A, 10 second pulses with 15 seconds rest between each pulse. This test emulates the one year ADD and is used to rapidly discharge cells to a depth of discharge where LiAsF₆ containing cells have historically shown voltage delay. In Fig. 26, curve 172 was constructed from the fourth pulse train of a cell constructed according to the present invention and activated with an electrolytic solution having dissolved therein LiPF₆ and LiAsF₆ in a ratio of 100:0. Likewise, curves 174, 176, 178 and 180 were constructed from the fourth pulse train of cells activated with an electrolytic solution having the enumerated lithium salts dissolved therein in ratios of 75:25, 50:50, 25:75 and 0:100, respectively. Fig. 27 is an expanded graph of the first pulse of the fourth pulse train graphed in Fig. 26. Similarly, in Fig. 28, curves 182, 184, 186. 188 and 190 were constructed from the fifth pulse train of the cells that were previously discharged to construct curves 172, 174, 176, 178 and 180 in Figs. 26 and 27, respectively. It is readily apparent that the higher the ratio of LiPF₆:LiAsF₆, the less effect there is on the voltage delay characteristics of a pulse discharged cell.

WO 96/15562

PCT/US95/12326

- 36 The voltage delay phenomenon is also exemplified in calculations of Rdc, the uncompensated cell resistance, calculated as

Rdc = (prepulse voltacre - pulse minimum) pulse current amplitude

The Rdc for pulses one and four for each electrolyte salt ratio of the cells discharged to construct Figs. 26 to 28 are shown in Fig. 29 and presented in Table 6. It is apparent that mixing LiPF₆ electrolyte salt with LiAsF₆ salt drastically affects the Rdc in these cells.

TABLE 6

Comparison of Pulse Minima Rdc, Ω

LiPF^/LiAsF_ft ratio

100/0

75/25

50/50

25/75

0/100

Pl Min Rdc

0.309±0.013

0.324±0.006

0.302±0.004

0.339±0.027

0.411±0.038

P4 min Rdc

0.261±0.007

0.257±0.009

0.266±0.010

0.262±0.005

0.257±0.009

EXAMPLE VI

One Year ADD Of Cells Containing Single Electrolyte Salt, LiPF&lt; Versus LiAsF&lt;

A group of cells containing either LiAsF_s or LiPF_fl at a concentration of 1 mol’L'¹ in a 1:1 volume mixture of dimethoxyethane/propylene carbonate were discharged using loads of 17.4 ΚΩ. The cells received an application of a

WO 96/15562

PCT/US95/12326

- 37 pulse train comprising four 1.5 A, 10 second pulses with 15 seconds rest between each pulse. One such pulse train was superimposed on the background load each 2 months. The pulse one end pulse four voltage minima and the pre-pulse voltage are plotted in Figs. 30 and 31. Fig. 30 shows discharge curves constructed from a representative cell activated with an electrolytic solution having LiAsF₆dissolved therein, wherein curve 192 was constructed from the pre-pulse open current voltage, curve 194 was constructed from the first pulse minima and curve 196 was constructed from the fourth pulse minima. In contrast, Fig. 31 shows the discharge curves constructed from a representative cell constructed according to the present invention and activated with an electrolytic solution having LiPF₆ dissolved therein, wherein curve 198 was constructed from the pre-pulse OCV, curve 200 was instructed from the pl min and curve 202 was constructed from the p4 min.

Note how the LiPF₆ containing cells do not exhibit voltage delay to the extent that the LiAsF₆ containing cells do, i.e., not as many trains have pulse one minima at voltages lower than pulse four minima. Also note that the voltage minima are not as low in the LiPF₆ containing cells as they in the LiAsF₆ containing cells in the final four pulse trains of the discharge.

EXAMPLE VII

Heat Dissipation Of Li/SVO Cells Containing LiPF₀ Versus

Li As Fa

Fourteen experimental cells were built to test the ability of the electrolytic solution of the present invention comprising a lithium hexafluorophosphate (LiPF₆)

WO 96/15562

PCT/US95/12326

- 38 salt to dissipate heat upon discharge of the cell. In particular, the electrolytic solution comprised IM LiPF₆dissolved in a 50/50 mixture (by volume) of dimethoxyethane and propylene carbonate. The electrolytic solution was used to activate a lithium/silver vanadium oxide cell system similar to that previously described in Example I and having a 1.8 Ah capacity. Two cells were submitted for short circuit testing, three cells were monitored for open circuit voltage (OCV) and heat dissipation, ώϊχ cells were subjected to accelerated pulse discharge testing and three cells were discharged under the one year accelerated discharge data (ADD) regime.

The results from the two cells subjected to short circuit testing compared similarly with results from standard electrolyte cells, i.e., Li/SVO cells activated with an electrolyte solution comprising LiAsF₆.

The heat dissipation from the three cells monitored for open circuit voltage testing ranged from about

7.0 jtzWatts to about 30.0 /zWatts at one month of age and continually declined to between about 4.6 /zWatts and about 10.2 /zWatts at about 14 months of age.

Under accelerated pulse discharge testing, the six cells delivered an average of about 1.34±0.03 Ah to 2 Volts(V), 1.52±0.01 Ah to 1.7V and 1.60+0.01 Ah to 1.5V, respectively. These means are greater than means for Li/SVO cells activated with the LiAsF₆ electrolytic solution, discharged under the same regime.

The average discharge curves for the three cells in this example subjected to the one year accelerated discharge data (ADD) regime are presented in Fig. 32. The discharge curves were recorded by applying pulse trains consisting of four ten second, 1.5 ampere pulses, every 15 seconds, repeated every 30 minutes to the respective

WO 96/15562

PCT/US95/12326

- 39 cells wherein in Fig. 32, curve 204 was recorded at open circuit with no load, curve 206 was constructed from the minima of the first pulse of each train and curve 208 was constructed from the minima of the fourth pulse of each train. These curves are presented in comparison to historic data gathered from a plurality of prior art Li/SVO cells activated with an electrolytic solution comprising LiAsF₆ dissolved in a 50/50 mixture (by volume) of DME/PC. In particular, curve 210 was constructed from the minima of the first pulse of each pulse train applied to the plurality of prior art cells and curve 212 was constructed from the fourth pulse minima of each pulse train applied to the prior art cells.

As is evident in Fig. 32, the cells according to the present invention behaved slightly different than the standard Li/SVO cells containing LiAsF₆ electrolyte. The pulse minima from the present cells were slightly lower at beginning of life (BOL) , however, note the improvement in delivered capacity particularly in the region between about 800 mAh to 1600 mAh for the cells of the present invention over the prior art cells. Voltage delay is present for the standard cells with LiAsF₆ during the fourth pulse train, but there is no evidence of voltage delay during the fourth pulse train in the LiPF₆ cells of the present invention. More importantly, the cells made according to the present invention showed no voltage delay during the middle of life pulse trains.

EXAMPLE VIII

Comparison Of Discharge Of Li/SVO Cells Containing LiPF₆Versus LiAsF&lt;

Fourteen lithium/silver vanadium oxide cells, similar to those cells used in Example VII, were built having a

WO 96/15562

PCT/US95/12326

- 40 1.8 Ah capacity. Four of the cells in this example were control cells and contained a prior art electrolyte solution comprising lithium hexafluoroarsenate (LiAsF₆) salt. In particular, the electrolytic solution for the four control cells comprised IM LiAsF₆ salt dissolved in a 50/50 mixture (by volume) of dimethoxyethane and propylene carbonate. The remaining cells in this example contained an electrolytic solution according to the present invention and in particular IM LiPF₆ dissolved in a 50/50 mixture (by volume) of dimethoxyethane and propylene carbonate. Five of these remaining cells each were subjected to OCV/heat dissipation and accelerated pulse discharge testing.

Five cells in this build having the LiPF_s electrolytic solution of the present invention were predischarged under a 3.57 kOhm load and one pulse train was applied one week after removal of the load. Table 7 compares the performance of the subject cells to data for the control cells of the same configuration having a standard electrolyte salt comprising lithium hexafluoroarsenate.

TABLE 7

Last load volts

Pl min

P4 min

LiPF₆ mean

Control LiAsF₆

3.187±0.004

3.167

3.274±0.003

3.244

2.37±0.075

2.307

2.57910.045

2.576

As is evident in Table 7, the means exceed the minimum criteria set for all categories.

After having been predischarged under a 3.57 ΚΩ load, the five cells in this example were monitored for open circuit voltage. This was done to test for heat dissipation which ranged from about 18.5 ^Watts to about 23.02 gWatts at one month of age compared to about 43.0 μWatts

WO 96/15562

PCT/US95/12326

- 41 dissipated by the one control cell having the LiAsF₆electrolyte solution.

The five cells in this build having the LiPF₆electrolytic solution of the present invention and subjected to the accelerated pulse discharged testing, delivered an average of about 1.34+0.01 Ah to 2V, 1.50+0.01 Ah to 1.7V and 1.56+0.01 Ah to 1.5V. The control cells delivered an average of about 1.35 Ah, 1.50 Ah and 1.55 Ah to the respective cutoffs.

From Examples VII and VIII, it can be concluded that lithium/silver vanadium oxide electrochemical cells using LiPF₆ as an electrolyte salt are comparable in performance to production standards set for similar cell systems using LiAsF₆ as an electrolyte salt. In addition, the use of LiPF₆ salt appears to greatly reduce or even eliminate the phenomenon of voltage delay. The cell heat dissipation is also significantly reduced.

EXAMPLE IX

Reactivity Of LiPF« Versus LiAsF₅ In Li/SVO Cells

Twelve experimental lithium/silver vanadium oxide cells similar to those described in Examples VII and VIII were built to test the lithium/electrolyte salt reactivity of the electrolytic solution of the present invention comprising the lithium hexafluorophosphate salt (LiPF₆) in comparison to electrolyte salts comprising lithium hexafluoroarsenate (LiAsF₆) and lithium perchlorate (LiClO₄) . Six of the twelve cells contained a standard amount of anode active material comprising an average of about 0.542 grams (g.) of lithium while the remaining six cells contained 60% of the standard amount of anode active material comprising an average of about 0.327 g. of lithium. Each set of six cells was divided into thirds and filled with a 50/50 mixture (by volume) of PC/DME

WO 96/15562

PCT/US95/12326

- 42 electrolyte that contained either a IM concentration of LiAsF₆, LiC10₄ or LiPF₆. The IM LiAsF₆ PC/DME electrolyte was prepared in-house, the IM LiC10₄ PC/DME electrolyte was used as received from Tomiyama Pure Chemical Industries, Ltd., and the IM LiPF₆ PC/DME electrolyte was prepared inhouse. The LiC10₄ and LiPF₆ electrolytes contained less than 10 ppm H₂0 by Karl Fischer Titration.

The twelve test ..-ells in this example were then burned-in for 21 hours under a 3.57 ΚΩ load, and acceptance pulse tested one week after burn-in with one 1.5 Amp pulse train con· ’ ting of four 10 second pulses with 15 seconds rest betwe.n each pulse. Heat dissipation was recorded after both burn-in and acceptance pulse testing.

The LiPF₆ cells had the highest burn-in voltages and the LiC10₄ cells had the lowest. The standard lithium and 60% lithium cells had comparable burn-in voltages within each electrolyte group. The LiPF₆, LiAsF₆ and LiClO₄ cells had average preload voltages of about 3731±2mV, 3705±4mV and 3602±15mV, respectively. The average last load voltages of the LiPF₆, LiAsF₆ and LiC10₄ cells were about 3218±lmV, 3210±lmV and 3202±lmV, respectively.

Serial number, electrolyte salt type, anode blank weight (g.), pre-load, first load, last load and post load burn-in voltages, and pre-pulse, pulse 1 minimum, pulse 1 end, pulse 4 minimum and pulse 1 end minus pulse 1 minimum acceptance pulse voltages are presented in Table 8.

TABLE 8

Burn-In Acceptance Pulse

3.57K for 21 hrs 1.5 Amp Puke Train

<td> Cell Serial Number</td><td> Salt Type</td><td> Anode Blank Weight(g)</td><td> Pre-Load Voltage</td><td> First Load Voltage</td><td> Last Load Voltage</td><td> Post Load Voltage</td><td> PrePulse Voltage</td><td> Pl Min Voltage</td><td> Pl End Voltage</td><td> P4Min Voltage</td><td> Pl EndPl Min</td>

<td> 76956</td><td> LiAsF_fi</td><td> 0.542g</td><td> 3700</td><td> 3698</td><td> 3207</td><td> 3209</td><td> 3276</td><td> 2422</td><td> 2646</td><td> 2629</td><td> 224</td>

<td> 76957</td><td> LiAsF₆</td><td> 0.542g</td><td> 3707</td><td> 3705</td><td> 3209</td><td> 3210</td><td> 3274</td><td> 2434</td><td> 2661</td><td> 2646</td><td> 227</td>

<td> 76966</td><td> LiAsF₆</td><td> 0.328g</td><td> 3705</td><td> 3702</td><td> 3209</td><td> 3210</td><td> 3274</td><td> 2483</td><td> 2700</td><td> 2671</td><td> 217</td>

<td> 76967</td><td> LiAsF₆</td><td> 0.327g</td><td> 3708</td><td> 3705</td><td> 3209</td><td> 3210</td><td> 3274</td><td> 2444</td><td> 2683</td><td> 2664</td><td> 239</td>

<td> 76960</td><td> LiC10₄</td><td> 0.542g</td><td> 3611</td><td> 3609</td><td> 3200</td><td> 3202</td><td> 3267</td><td> 2534</td><td> 2703</td><td> 2673</td><td> 169</td>

<td> 76961</td><td> LiC10₄</td><td> 0.542g</td><td> 3618</td><td> 3615</td><td> 3202</td><td> 3203</td><td> 3269</td><td> 2573</td><td> 2695</td><td> 2668</td><td> 122</td>

<td> 76968</td><td> LiC10₄</td><td> 0.329g</td><td> 3589</td><td> 3586</td><td> 3201</td><td> 3202</td><td> 3264</td><td> 2605</td><td> 2690</td><td> 2666</td><td> 85</td>

<td> 76969</td><td> LiC10₄</td><td> 0.330g</td><td> 3590</td><td> 3588</td><td> 3201</td><td> 3202</td><td> 3264</td><td> 2607</td><td> 2693</td><td> 2668</td><td> 86</td>

<td> 76964</td><td> LiPF₆</td><td> 0.541g</td><td> 3729</td><td> 3726</td><td> 3216</td><td> 3218</td><td> 3284</td><td> 2305</td><td> 2651</td><td> 2637</td><td> 346</td>

<td> 76965</td><td> LiPF₆</td><td> 0.541g</td><td> 3728</td><td> 3724</td><td> 3215</td><td> 3217</td><td> 3284</td><td> 2317</td><td> 2666</td><td> 2649</td><td> 349</td>

<td> 76970</td><td> LiPF₆</td><td> 0.324g</td><td> 3732</td><td> 3727</td><td> 3216</td><td> 3219</td><td> 3286</td><td> 2173</td><td> 2629</td><td> 2603</td><td> 456</td>

<td> 76971</td><td> LiPF₆</td><td> 0.324g</td><td> 3733</td><td> 3728</td><td> 3215</td><td> 3218</td><td> 3286</td><td> 2180</td><td> 2642</td><td> 2620</td><td> 462</td>

ELECTROLYTE: All salts were dissolved in 50/50 v/o PC/DME in IM concentration.

WO 96/15562 PCT/US95/12326

WO 96/15562

PCT/US95/12326

-44EXAMPLE X

Comparison Of Accelerated Pulse Testing- Of LiPF&lt; Verses LiAsFx In Li/SVO Cells

After acceptance pulse testing, the six standard lithium cells used in Example IX were subjected to accelerated pulse testing. Accelerated pulse testing comprised a plurality of 1.5 Amp pulse trains, in particular 68 to 72 such pulse trains, each train consisting of four 10 second duration pulses with 15 seconds rest between each pulse. This corresponds to between about 1133 mAhrs to 1200 mAhrs of delivered capacity. Heat dissipation was recorded before and after accelerated pulse testing. The six 60% lithium cells used in Example IX were also subjected to accelerated pulse testing comprising between 68 to 70 pulse trains similar to those to which the standard lithium cells were subjected, i.e., 1.5 Amp pulse trains, consisting of four 10 second pulses with 15 seconds rest between each pulse. These cells delivered between about 766 mAhrs and 814 mAhrs of capacity to a 1.5V cut-off.

Table 9 shows the serial number, salt type, anode blank weight (g.), heat dissipation (/zWatts) after burn-in (B-I) and acceptance pulse testing, the number of pulse trains applied and capacity (mAhrs) delivered to 2.0 Volts(V), 1.7V and 1.5V during accelerated pulse testing, heat dissipation (^Watts) after accelerated pulse testing, the capacity (mAhrs) removed during 1 ΚΩ discharge, heat dissipation (/zWatts) after 1 ΚΩ discharge and end of life (EOL) thickness in inches (in.) for each cell tested.

Both the standard and 60% lithium LiAsF₆, LiClO₄ and LiPF₆ cells dissipated an average of about 35.04 /zWatts, 52.30 /zWatts and 17.87 /zWatts respectively, of heat after burn-in and acceptance pulse testing. This results in a statistically significant difference in the heat dissipated between the electrolyte salt groups at a confidence of &gt;99.9%. Additionally, both the standard and 60% lithium LiAsF₆, LiClO₄ and LiPF₆ cells dissipated about

WO 96/15562

PCT/US95/12326

- 45 27.63 ^Watts, 35.21 gWatts and 19.37 ^Watts of heat respectively, after accelerated pulse testing. There thus exists a statistically significant difference in the heat dissipated after accelerated pulse testing between the electrolyte salt groups at a confidence &gt;97.5%. Typically, statistically significant differences are not found to exist between lithium groups within each electrolyte salt set. Thus, the lithium hexafluorophosphate salt clearly exhibits superior heat dissipation characteristics with respect to the other tested electrolyte salts (i.e., low reactivity with the lithium anode) after burn-in and acceptance pulse testing and accelerated pulse testing.

TABLE 9

<td rowspan="2"> Cell Serial Number</td><td rowspan="2"> Salt Type</td><td rowspan="2"> Anode Blank Weight (g)</td><td rowspan="2"> gWatts After BI &amp; Ace. Pulse</td><td rowspan="2"> Number of Pulse Trains</td><td colspan="3"> Accelerated Pulse Testing Capacity mAhrs to</td><td rowspan="2"> gWatts After Accel. Pulse</td><td rowspan="2"> IK Load Discharg e Capacity (mAhrs)</td><td rowspan="2"> gWatts After IK Load Discharge</td><td rowspan="2"> EOL Thickness</td>

<td> 2.0V L5V</td><td colspan="2"> 1.7V</td>

<td> 76956</td><td> LiAsF₆</td><td> 0.542 g</td><td> 34.81</td><td> 70</td><td> 1167*</td><td> 1167*</td><td> 1167*</td><td> 31.30</td><td> 180</td><td> 41.76</td><td> 0.4235 in</td>

<td> 76957</td><td> LiAsF₆</td><td> 0.542 g</td><td> 34.18</td><td> 70</td><td> 1167*</td><td> 1167*</td><td> 1167*</td><td> 29.74</td><td> 181</td><td> 40.80</td><td> 0.4230 in</td>

<td> 76966</td><td> LiAsF₆</td><td> 0.328 g</td><td> 38.72</td><td> 70</td><td> 774</td><td> 797</td><td> 807</td><td> 22.85</td><td> 126</td><td> 23.63</td><td> 0.4210 in</td>

<td> 76967</td><td> LiAsF₆</td><td> 0.327 g</td><td> 32.46</td><td> 70</td><td> 771</td><td> 795</td><td> 805</td><td> 26.61</td><td> 117</td><td> 19.56</td><td> 0.4210 in</td>

<td> 76960</td><td> LiC10₄</td><td> 0.542 g</td><td> 51.24</td><td> 72</td><td> 1200*</td><td> 1200*</td><td> 1200*</td><td> 39.75</td><td> 178</td><td> 36.39</td><td> 0.4235 in</td>

<td> 76961</td><td> LiC10₄</td><td> 0.542 g</td><td> 52.96</td><td> 72</td><td> 1200*</td><td> 1200*</td><td> 1200*</td><td> 32.24</td><td> 179</td><td> 35.74</td><td> 0.4230 in</td>

<td> 76968</td><td> LiC10₄</td><td> 0.329 g</td><td> 52.96</td><td> 70</td><td> 768</td><td> 788</td><td> 799</td><td> 35.06</td><td> 107</td><td> 22.69</td><td> 0.4200 in</td>

<td> 76969</td><td> LiC10₄</td><td> 0.330 g</td><td> 52.02</td><td> 69</td><td> 766</td><td> 780</td><td> 786</td><td> 33.80</td><td> 121</td><td> 23.95</td><td> 0.4210 in</td>

<td> 76964</td><td> LiPF₆</td><td> 0.541 g</td><td> 18.22</td><td> 68</td><td> 1133*</td><td> 1133*</td><td> 1133*</td><td> 19.88</td><td> 182</td><td> 10.63</td><td> 0.4210 in</td>

<td> 76965</td><td> LiPF₆</td><td> 0.541 g</td><td> 17.12</td><td> 68</td><td> 1133*</td><td> 1133*</td><td> 1133*</td><td> 18.78</td><td> 182</td><td> 6.44</td><td> 0.4200 in</td>

<td> 76970</td><td> LiPF₆</td><td> 0.324 g</td><td> 16.03</td><td> 68</td><td> 769</td><td> 799</td><td> 814</td><td> 21.91</td><td> 107</td><td> 10.49</td><td> 0.4205 in</td>

<td> 76971</td><td> LiPF₆</td><td> 0.324 g</td><td> 20.10</td><td> 68</td><td> 772</td><td> 793</td><td> 807</td><td> 16.90</td><td> 112</td><td> 14.55</td><td> 0.4205 in</td>

The pulse 4 minimum voltages of these cells was above the indicated cut-offs. The indicated capacity is the amount of capacity delivered by the cells during accelerated pulse testing.

ELECTROLYTE: All salts are dissolved in 50/50 v/o PC/DME in IM concentration.

PCT/US95/12326

WO 96/15562

PCT/US95/12326

-47 EXAMPLE XI

Comparison Of Heat Dissipation Of LiPF₆ Versus LiAsF* And LiCIO, In Li/SVO Cells

After undergoing accelerated pulse testing in Example X, the energy potential of the 60% lithium cells returned to about 2.6V. This high open circuit voltage indicated that not all the lithium was utilized during discharge. In order to determine how much of the heat dissipation was due to lithium/electrolyte reactivity, the 60% lithium cells were then discharged under a 1 ΚΩ load at 37°C to approximately 0 mV, which they reached after about 65 hours on this load. The 60% lithium cells delivered between about 107 mAhs and 126 mAhs of capacity.

The standard lithium cells used in Example X were then discharged for a similar length of time, i.e., 65 hours, under a 1 KO load with between about 178 mAhrs to 182 mAhrs of capacity being delivered. Heat dissipation was recorded after 1 ΚΩ discharge.

As is evident from Table 9, statistically significant separation between standard lithium and 60% lithium cells is evident within the LiAsF₆ and LiClO₄ salt groups. The standard and 60% lithium LiAsF₆ cells dissipated an average of 41.28 gWatts and 21.60 gWatts of heat respectively, while the standard and 60% lithium LiC10₄ cells dissipated an average of 36.07 gWatts and 23.32 gWatts of heat, respectively. Statistically significant differences do not exist between lithium groups within the LiPF₆ electrolyte salt set as the standard and 60% lithium LiPF₆ cells dissipated an average of 8.54 gWatts and 12.52 gWatts of heat, respectively. Collectively, the LiPF₆ cells dissipated an average of about 10.53 gWatts of heat. Statistically significant differences appear to exist in the heat dissipated after 1 ΚΩ discharge between the standard lithium LiAsF₆ and LiClO₄ cells, the 60% lithium LiAsF₆ and LiC10₄ cells, and both the standard and 60% LiPF₆ cells.

WO 96/15562

PCT/US95/12326

-48 Also noteworthy from Table 9 is the end-of-life (EOL) thicknesses of the various cells. The standard LiPF₆ cells had EOL thicknesses that were less than the EOL thicknesses of the standard lithium LiAsF₆ and LiC10₄ cells but comparable to the 60% lithium LiAsF₆ and LiC10₄ cells. The standard lithium LiAsF₆ and LiC10₄ cells had an average EOL thickness of about 0.4233±0.0003 in. and the 60% lithium LiAsF₆ and LiClO₄ cells had an average EOL thickness of about 0.4205±0.0006 in. The average EOL thickness of the standard and 60% lithium LiPF₆ cells was about

0.4208±0.0003 in.

Fig. 33 illustrates the trends in the amount of heat dissipated by the various experimental cells used in Examples IX to XI, after burn-in and acceptance pulse testing (Example IX), accelerated pulse testing (Example X) and 1 ΚΩ discharge testing (Example XI), as indicated in Table 9. In particular, curve 214 was recorded from the average discharge of the two lithium/silver vanadium oxide cells (Li/SVO) indicated as Serial Nos. 76964 and 76965 and having a standard amount (0.542 grams) of active anode material and an electrolyte comprising LiPF₆ according to the present invention; curve 216 was recorded from the average discharge of the two Li/SVO cells indicated as Serial Nos. 76970 and 76971 and having 60% of the standard amount (0.327 grams) of the active anode material and an electrolyte comprising LiPF₆; curve 218 was recorded from the average discharge of two standard Li/SVO cells indicated as Serial Nos. 76956 and 76957 and having an electrolyte comprising LiAsF₆ according to the prior art;

curve 220 was recorded from the average discharge of the two 60% of standard Li/SVO cells indicated as Serial Nos. 76966 and 76967 and having the LiAsF₆ electrolyte;

curve 222 was recorded from the average discharge of the two standard Li/SVO cells indicated as Serial Nos. 76960 and 76961 and having an electrolyte comprising LiClO₄according to the prior art; and curve 224 was recorded from the average discharge of the two 60% of standard Li/SVO

WO 96/15562

PCT/US95/12326

-49 cells indicated as Serial Nos. 76968 and 76969 and having the LiC10₄ electrolyte.

From Fig. 9 and Examples IX to XI it is evident that the LiC10₄ cells dissipated the greatest amount of heat after both burn-in and accelerated pulse testing. Differences were not apparent in the heat dissipation response for the standard lithium and 60% lithium cells filled with LiAsF₆ or the LiC10₄ electrolytic solutions after burn-in and accelerated pulse testing. However, after 1 ΚΩ discharge testing, differences were apparent. The standard lithium cells dissipated more heat than the 60% lithium cells. This indicates that a portion of the heat dissipated by the lithium anode cells containing LiAsF₆ or LiClO₄ can be attributed to lithium/electrolyte reactivity.

The LiPF₆ cells dissipated less heat after each test in Examples IX to XI than the LiAsF₆ and LiClO₄ cells, regardless of the amount of lithium originally contained in the cells. Additionally, separation between the standard and 60% lithium LiPF₆ cells was not evident after burn-in, accelerated pulse testing or 1 ΚΩ discharge testing. This leads to the conclusion that LiPF₆/lithium reactivity appears to be minimal when compared to other processes within the cell contributing to heat dissipation. Thus, based on the heat dissipation characteristics of the cells studied, the LiAsF₆ and LiClO₄ electrolytes appear to be more reactive toward lithium than LiPF₆.

Thus, it is shown by the results from Examples III to XI that the use of an electrolytic solution comprising an ion-forming alkali metal salt of hexafluorophosphate dissolved in an organic solvent and coupled with an alkali metal anode greatly reduces the voltage delay of the high pulse power cell of the present invention, as well as, benefiting the heat dissipation characteristics of such a cell. This is particv’larly the case when the anode comprises lithium and the electrolytic solution comprises lithium hexafluorophosphate as the ion-forming salt.

WO 96/15562

PCT/US95/12326

-50Further, when the electrolytic solution of the present invention is used in conjunction with an aluminum cathode current collector, there is markedly reduced evidence of reactivity between these two cell components in comparison with other conductor materials such as titanium used in an electrochemical cell activated with lithium hexafluorophosphate.

EXAMPLE XII

Comparison Of Cell Performance Of High Pulse Power Cell Versus Li/SVO Cells Activated With LiAsF_ft

Two prior art cells and a high pulse power cell system according to the present invention were constructed in a similar manner as those described in Example I. These cells were electrically tested to further demonstrate the improvements in cell performance attributable to the preferred electrochemical system of the present inventior Both prior art cells utilized pressed, multiplate cathodes, titanium cathode current collectors and ribbon lithium anodes, and they are referred to as prior art cell type A and prior art cell type B in this example.

In comparison, the high pulse power cell according to the present invention was built having a lithium aluminum alloy anode, a nickel oxide anode current collector, a calendared silver vanadium oxide cathode sheet pressed onto an aluminum cathode current collector. The cell components were then configured in a jellyroll electrode assembly.

The prior art cells were vacuum filled with a IM LiAsF₆ PC/DME(50:50) electrolytic solution, while the high pulse power cell of the present invention was vacuum filled with IM LiPF₆ PC/DME (50:50) electrolytic solution. These cells were then burned-in and accelerated discharge tested. A constant current pulse was applied to the cells every thirty minutes during the accelerated discharge testing. The pulse train consisted of four 10 second pulses with 15 second rests between each pulse. The pulse amplitude was 2.00 amperes, 1.23 amper'es and 2.00 amperes for the prior

WO 96/15562

PCT/US95/12326

-51 art cell type A, prior art cell type B, and high pulse power cell, respectively.

Fig. 34 is a graph showing typical discharge curves for each o.Z the prior art cell., and for the high pulse power cell of the present invention. In that respect, curve 226 was constructed from the prepulse open circuit volta„. ’’ί th no load of the prior art cell type A, curve 228 was constructed from the minima of the first pulse of each pulse train of this cell and curve 230 was constructed from the minimum of the fourth pulst of each pulse train of the prior art type A cell. Similarly, curve 232 was constructed from the prepulse open circuit voltage with no load of the prior art cell type B, curve 234 was constructed from the minima of the first pulse of each pulse train of the cell and curve 236 was constructed from the minima of the fourth pulse of each pulse train of the prior art type B cell.

In comparison, curve 238 was constructed from the prepulse open circuit voltage with no load of the high pulse power cell of the present invention, curve 240 was constructed from the minima of the first pulse of each pulse train of this cell and curve 242 was constructed from the minima of the pulse of each pulse train of the present high pulse power cell.

Table 10 shows the external cell volume, electrode surface area, current amplitude, capacity delivered to 2.0V, 1.7V, and 1.5V, cathode efficiency to 2.0V, 1.7V and 1.5V, and the energy density at 2.0V, 1.7V and 1.5V for the prior art type A and type B cells and the high pulse power cell of the present invention.

WO 96/15562

FCT/US95/12326

-52 TABLE 10

Cell Prior Art

Characteristic Type A

Prior Art

Type.. B

High Pulse

Power Cell

External Cell Volume

10.3 cc

6.4 cc

5.0 cc

Electrode Surface

Area 86 cm² cm² 114 cm²

Current Amplitude

2000 mAmp

1230 mAmp

2000 mAmp

Delivered

<td> Capacity to..</td><td></td><td></td><td></td><td></td><td></td><td></td>

<td> 2.0V</td><td> 1.465</td><td> Ahrs</td><td> 0.827</td><td> Ahrs</td><td> 0.576</td><td> Ahrs</td>

<td> 1.7V</td><td> 1.686</td><td> Ahrs</td><td> 0.975</td><td> Ahrs</td><td> 0.678</td><td> Ahrs</td>

<td> 1.5V</td><td> 1.758</td><td> Ahrs</td><td> 1.033</td><td> Ahrs</td><td> 0.720</td><td> Ahrs</td>

<td> Cathode Efficiency</td><td></td><td></td><td></td><td></td><td></td><td></td>

<td> to. .</td><td></td><td></td><td></td><td></td><td></td><td></td>

<td> 2.0V</td><td> 62.9%</td><td></td><td> 72.4%</td><td></td><td> 71.9%</td><td></td>

<td> 1.7V</td><td> 72.4%</td><td></td><td> 85.4%</td><td></td><td> 84.7%</td><td></td>

<td> 1.5V</td><td> 75.5%</td><td></td><td> 90.5%</td><td></td><td> 89.9%</td><td></td>

Energy Density

2.0V

1.7V

1.5 V

0.34 Whrs/cc

0.39 Whrs/cc

0.40 Whrs/cc

0.31 Whrs/cc

0.36 Whrs/cc

0.37 Whrs/cc

0.31 Whrs/cc

0.36 Whrs/cc

0.38 Whrs/cc

The synergistic advantages gained by the combination of materials, namely the lithium aluminum alloy anode, nickel anode current collector, silver vanadium oxide cathode active material, aluminum cathode current collector and lithium hexafluarophosphate, dissolved in a mixture of organic solvents, used to fabricate the high pulse power cell are quite evident in Table 10. The high pulse power cell is approximately 20% small than the prior art type B cell, has two times more electrode surface area and discharges as efficiently with current amplitudes that are 1.6 times higher than those used to discharge the prior art type B cell. Also, the high pulse power cell discharges more efficiently than the prior art type A cell with an equivalent current amplitude even though it is approximately 50% smaller. Further, the energy density of

WO 96/15562

PCT7US95/12326

- 53 the high pulse power cell of the present invention is not compromised by the reduced volume of the cell container and, in fact, the energy density is comparable to the energy densities of both the prior art cells.

It is intended that the foregoing description and examples be only representative of the present invention and that the present invention be only limited by the hereinafter appended claims.

WO 96/15562 PCT/US95/12326