CA2146370C - Optimized positive electrode for alkaline cells - Google Patents

Abstract

A positive electrode for use in alkaline rechargeable electrochemical cells comprising: a material comprising a compositionally and structurally disordered multiphase nickel hydroxide host matrix which includes at least one modifier chosen from the group consisting of F, Li, Na, K, Mg, Ln, Se, Nd, Pr, Y, Co, Zn, Al, Cr, Mn, Fr, Cu, Zn, Sc, Sn, Te, Bi, Ru and Pb. A process for forming a high loading uniformly distributed multiphase substantially nitrate free sintered positive electrode for use in an alkaline rechargeable electrochemical cell, the process comprising: (1) fabricating sintered electrode material by forming a slurry of nickel powder, water, carboxy methyl cellulose binder, methyl cellulose binder, and a poly(ethyleneoxide) polymer, spreading the slurry on a preoxidized perforated nickel substrate; drying the slurry; and sintering the slurry; (2) impregnating the sintered electrode material using multiple impregnation cycles to attain high loading.

Description

~O 94/11910 2 14 6 3 7 0 PCI'/US93/10910 OPTII\IIZED POSll'IVE ELECTRODE FOR
ALKALINE CELLS
FIELD OF THE INVENTION
5The present invention relates generally to an optimized nickel hydroxide positive electrode. More specifically, this invention relates to optimized nickel hydroxide positive electrodes for rechargeable ~lk~line cells.
.

BACKGROUND OF THE INVENTION
In rechargeable ~lk~line cells, weight and portability are important considerations.
It is also advantageous for rechargeable ~lk~line cells to have long operating lives without the necessity of periodic mainten~nre. Rechargeable ~1k~line cells may be used as direct replacements for primary AA, C, and D cells in numerous consumer devices such ascalculators, portable radios, and cellular phones. They are often configured into a sealed power pack that is designed as an integral part of a specific device. Rechargeable aLkaline cells can also be configured as larger cells that can be used, for example, in industrial, aerospace, and electric vehicle applications.
The best rechargeable ~1k~1ine cells are ones that can operate as an "install and forget" power source. With the exception of periodic charging, a rechargeable ~lk~linto.
cell should perform without attention and should not become a limiting factor in the life of the device it powers.
There are two basic types of rechargeable ~lk~line cells: nickel c~rlmillm ("NiCd") cells and nickel metal hydride ("Ni-MH") cells.
In a NICd cell, c~(lmium metal is the active material in the negative electrode.NiCd cells use a positive electrode of nickel hydroxide material. The negative and - positive electrodes are spaced apart in the ~lk~lin~ electrolyte.
Upon application of an electrical potential across the m~t~n~ls of a NiCd cell, the negative electrode undergoes to the following reaction:

charge Cd(OH)2 + 2e~ ~ > Cd + 20H-dischuge During discharge, this reaction is reversed, Cd is oxidized to Cd(OH)2 and electrons are re1e~e~1 The reactions that take place at the positive electrode of a Ni-Cd cell are also ~V094/11910 2 ~ ~ 6 ~ 7 Q PCr/US~3/lo91o '_ 2 reversible. For example, the reactions at a nickel hydroxide positive electrode in a nickel cadmium cell are:

chule 2Ni(OH)2 + 20H c > 2NiOOH + 2H20 + 2e ~xh~e In general, Ni-MH cells utilize a negative elecuode that is capable of the reversible 10 electrochemical stora~e of hydrogen. Ni-MH cells usually employ a positive electrode of nickel hydroxide material. The negative and positive electrodes are spaced apart in the alkaline electrolyte.
Upon application of an electrical potential across a Ni-M~-~ cell, the Ni-MH
material of the negative electrode is charged by the electrochemical absorption of 15 hydrogen and the electrochemical generation of hydroxyl ions:
~8e M + H20 + e <~ > M-H + OH-di~ch~e 20 The negative electrode reactions are reversible. Upon discharge, the stored hydrogen is released to form a water molecule and evolve an electron.
The reactions that take place at the nickel llydroxide positive electrode of a Ni-MH cell are:
ch~e Ni(OH)2 + OH ~ > NiOOH + H20 + e ~ e This is the identical reaction that occurs in a NiCd cell.

Ni-MH cells can be funher classifled as V-Ti-Zr-Ni (Ovonic or AB2) based or ABs (mischmetal) alloys depending on the type of hydrogen storage material used as the negative electrode. Both types of material are discussed in detail in applicant's copending Canadian Application No. 2,142,118, filed August 25, 1993.

wo 94/ 1 1 9 1 0 I'cr/ US93/ 1 09 1 ~) $ 3 7 0 3 The first hydrogen stor~ge ~lloys to be investigated as battely electrode materials were TiNi and LaNi5. Many years were spent in studying these simple binary interrnetallics because they were known to have the proper hydrogen bond strength for use in electrochemical applications. Despite extensive efforts, however, researchers found these interrnetallics to be extremely unstable and of marginal electrochemical value due to a variety of deleterious effects such as slow discharge, oxidation, corrosion, poor kinetics, and poor catalysis. The initial use of these simple alloys for battery applications reflect the traditional bias of battery developers toward the use of single element couples of crystalline materials such as NiCd, NaS, LiMS, ZnBr, Ni~e, l~iZn, and Pb-acid. In order to improve the electrochemical properties of the binary interrnetallics while maintaining the hydrogen storage efficiency, early workers began modifying TiNi and LaNi5 systems.
The modification of TiNi and LaNi5 was initiated by Stanford R. Ovshinsky at Energy Conversion Devices (ECD) of Troy, Michigan. Ovshinsky and his team at ECDfound that reliance on simple, relatively pure compounds was a major shortcoming of the prior art. Prior work had determined that catalytic action depends on surface reactions at sites of irregularities in the crystal structure. Relatively pure compounds were found to have a relatively low density of hydrogen storage sites, and the type of sites available occurred accidently and were not designed into the bulk of the material. Thus, the efficiency of the storage of hydrogen and the subsequent release of hydrogen to forrn water was deterrnined to be substantially less than that which would be possible if a greater number and variety of active sites were available.
Ovshinsky had previously found that the number of surface sites could be signif~cantly increased by making an arnorphous film that resembled the surface of the desired relatively pure materials. As Ovshinsky explained in Principles and Applicarions of Amorphicit~ Structural Change arld Oprical Informarion ~r-coding 42 Journal De Physique at C4-1096 (Octobre 19~
Amorphicity is a generic terrn referring to lack of X-ray diffraction evidence of long-range periodicity and is not a sufficient description of a material. To understand amorphous materials, there are several important factors to be .

~0 94/t lglo rCT/US93/10910 7 Q ~

considered~ e type of cllclnical bonding, tl-e number of bonds generated by the local order, that is its coordina~ion, and the innuence of the entire local environment, both chemical and geometrical, upon the resulting varied configurations. Amorphicity is not determined by random packing of atoms viewed as hard spheres nor is the amorphous solid merely a host with atoms imbedded at random. Amorphous materials should bè viewed as being composed of an interactive matrix whose electronic configurations are generated by free energy forces and they can be specifically defined by the chemical nature and coordination of the constituent atoms. Utilizing multi-orbital elements and various preparation techniques, one can outwit the normal relaxations that reflect equilibrium conditions and, due to the three-dimensional freedom of the amorphous state, make entirely new types of amorphous materials chemically modified materials Once arnorphicity was understood as a means of introducing surface sites in a rllm, it was possible to produce "disorder" that takes into account the entire spectrum of local order effects such as porosity, topology, crystallites, characteristics of sites, and distances between sites. Thus, rather than searching for material modifications that would yield ordered materials having a maximum number of accidently occurring surface irregularities, Ovshinky's team at ECD began constructing "disordered" materials where the desired irre.~ularities were tailor made. See, U.S. Patent ~o. 4,623,597, The term "disordered," as used herein corresponds to the meaning of the term as used in the literature, such as the following:
A disordered semiconductor can exist in several structural states. This structural factor constitutes a new variable with which the physical propcrties of the [material~ ... can be controlled. ~urthermore, structural disorder opens up the possibility to prepare in a metastable state new compositions and mixtures that far exceed the limits of thermodynamic equilibrium. Hence, we note the following as a further distinguishing feature. In many disordered [materials1 ... it is possible to control the short-range order parameter and thereby achieve drastic changes in the physical properties of these materials, including forcing new coordination numbers for elements S. R. Ovshinsky, The Shope of Disorder, 32 Journal of Non-Cr~slalline Solids al 22 (1979) (emphasis added).
The "short-range order" of these disordered materials are further explained by Ovshinsky in The Chemical Basis of Amorphicit)u Strucrure and Function, 26:8-9 Re-.
Roum. Phys. at 893-903 (1981):

~ ~YO 94/11911) rcr/US93/10910 ~ ~ ~6~7~
,~ s IS~hort-range order is not conserved .... Indeed, when crystalline symmetry is destroyed, it becomes impossible to retain the same short-range order.
The reason for this is that the short-range order is controlled by the force fields of the electron orbitals thererore the environment must be fundamentally different in corresponding crystalline and amorphous solids.
In other words, it is the interaction of the local chemical bonds with their surrounding environment WlliCIl deterlnines the electrical, chemical, and physical properties of the material, and these can never be the same in amorphous materials as they are in crystalline materials . . . The orbital relationships that can exist in three-dimensional space in amorphous but not crystalline materials are the basis for new geonletries, many of whicli are inherently anti-crystalline in nature. Distortion of bonds and displacement of atoms can be an adequate reason to cause amorphicily in single component materials. But to sufficiently understand the amorphicity, one must understand the three-dimensional relationships inherent in the amorphous state, for it is they which generate internal topology incompatible with the translational syrnmetry of the crystalline lattice .... What is important in the arnorphous state is the fact that one can make an infinity of materials that do not have any crystalline counterparts, and that even the ones that do are similar primarily in chemical composition. The spatial and energetic relationships of these atoms can be entirely different in the amorphous and crystalline forrns, even though their chemical elements can be tlle same Shol-t-range, or local, order is elaborated on in U S. Patent ~o. 4,520,039 to Ovshil1sky, entilled Cor~lpo~ iol1ally V(7r-ied Malerial.~ a~7d Me~/lod for S~f1~llesizillg t/le Male~ials. Tllis patent discusses how disordered malerials do not require any periodic local order and llow, by using Ovsllinsky's tecllniques, spatial and orientational ~lacement of similal- or dissimilar atoms or groups of atoms is possihle 30 with such increased precisioll and contlol of tlle local conrigulatiolls that it is possit le to produce qualitatively new phellol11el1a In addition, tl1is patel1t discusses that the atoms used need not be restricted to "d band" or "f ban(l" alollls, but can be any atom in whicll the contl-olled aspects of the interaction with tlle local envil-olllllel1t plays a significant role pllysically, electrically, or chelllically so as to affect the 35 pllysical propel-ties and hellce the fullctiol1s of the Inaterials. 'I hese techlliques result in mealls of syntllesizillg new matelials WlliCll are diSOldeled ill SeVel--ll difrerellt senses shllultalleously.
By forming metal hydride alloys from such disordered materials, Ovshinsky and his team were able to greatly increase the reversible hydrogen storage characteristics 40 required for efficient and economical battery applications, and produce batteries having , . .

0 9~ 910 ~ ; 3j 7 ~ PCI /US93/10910 _ 6 1.igh density energy stora~e, efficient reversibility, higll elec~ical efficiency, bulk hydrogen storage witllout structur~l cll~n~e or poisoning, ]ong cycle life, and deer~
discharge capability.
The improved cl~aracteristics of these alloys result from tailoring tlle local chelllic~l S order and hence the local s~uctural order by the incorporation of selected modil icr elements into a host matrix. Disordered metal hyckide alloys have a substantially increased density of catalytically active sites and storage sites compared to conventional ordered matenals. These additional sites are responsible for improved efficiency of electrochemical charging/ciischarging and an inc~ase in electrical energy storage capacity.
10 The nature and number of storage sites can even be designed independently of lhe catalytically active sites. More specifically, these alloys are tailored to allow storage of hydrogen atoms at boncling s~ngths within the range of reversibility suitable for use in secondary battery applications.
Based on the pioneering principles described above, a family of ex~emely efficient 15 elec~ochemical hydrogen storage materials were formulated. These are tlle Ti-V-Zr-Ni type ac~ive materials such as disclosed in U.S. Patent No. 4,551,400 ("the '400 Patent') lo Sapru, llong, E~etcenko, and Venkalesan. These ma~el ials revelsibly fon hydl ides hl order to store hydl-ogell. All the matel ials used in the '400 Patellt utilize a generic Ti-V-Ni compositic)n, whel-e a~ least l i, V, and Ni are r)resent witll al least 20 one or more of Cr, Zr, and /~1. The ma~erials of the '400 Patent are generally Inultiphase materials, which Inay contain, hut are not limi~ed to, one or mol-e phases of Ti-V-Zr-Ni material with C14 and Cl~ type crystal structut-es. Olhel- Ti-V-Zr-Ni alloys may also be used for a recllargeable hydtogell storage negative electrode. One such ramily of materials are tllose described in U.S. Patent No. 4,728,586 ("the '586 Patent") to Venkatesan, Reicllmall, and l;etcenko ror Erll~allced C11arge Retentio Eleclrocl~el1lical l~ydrogell Stol-age Alloys and al1 E/l/7arlced Cllar-ge Re(ertlion Eleclr-ocllernical Cell. Tlle '586 Patent describes a specific sub-class of these Ti-V-Ni-Zr alloys comprising Ti, V, Zr, Ni, and a fiftll compollel]t, Cr. I he '586 patent, 30 mentions tlle possibility of additives and modiriers beyond tlle I i, V, Zr, Ni, and Cr componellts of tlle alloys, and genelally discusses specific additives and modifiers, he amounts and interactiolls Or these nlocliriel-s, and the pal ticular benefits that could ~, be expected flom thelll.
. ~, WO 94/11910 21 ~ 63 70 PCI/US93/10910 The V-Ti-Zr-Ni family of alloys described in the '586 Patent has an inherently higher discharge rate capability than previously described alloys. This is the result of substantially higher surface areas at the metaVelectrolyte interf~ce for electrodes made ~ from the V-Ti-Zr-Ni materials. The surface roughness factor (total surface area divided 5 by geometric surface area) of the V-Ti-Zr-Ni is approxim~tely lO,000. This value inrli~tes a very high surface area. The validity of this value is supported by the inherently high rate capability of these m~tt~ri~1~
The characteristic surface roughness of the metal electrolyte inte-face is a result of the disordered nature of the m~t~ 1 Since all of the con~tit-1ent element~, as well as 10 many alloys and phases of them, are present throughout the metal, they are also represented at the surfaces and at cracks which form in the metaVelectrolyte interface.
Thus, the characteristic surface roughness is descriptive of the interaction of the physical and chemical plupellies of the host metals as well as of the alloys and crystallographic phases of the alloys, in an ~1k~1ine environment These microscopic r,hlomir~l, physical, 15 and crystallographic ~ letcl~ of the individual phases within the hydrogen storage alloy material are believed to be illlp~ in ~çtefmining its macroscopic elec~ochelllical characteri~tics.
In addition to the physical nature of its roughençd surface, it has been observed that V-Ti-Zr-Ni metal hydride alloys tend to reach a steady state surface conditiQn and 20 particle si~. This steady state surface condition is characteri~d by a relatively high cQnrent-~tion of m~t~llic nickel. These obsG.~tions are con~ist~nt with a relatively high rate of removal through precipitation of the oxides of tit~ninm and zirconium from the surface and a much lower rate of nickel solubili7~tion The resultant surface seems to have a higher concentration of nickel than would be expected from the bulk composition 25 of the negative hydrogen storage electrode. Nickel in the met~llic state is electrically conductive and catalytic, ilnp~~ g these p,~.,.lies to the s~1~ce. As a result, the surface of the negative hydrogen storage electrode is more catalytic and conductive than if the surface cont~in~l a higher concentration of insulating oxides.
The surface of the negative electrode, which has a conductive and catalytic compo-30 nent -- the metallic nickel -- appears to interact with chromium alloys in catalyzing various hydride and dehydride reaction steps. To a large extent, many electrode processes, including competing electrode processes, are controlled by the presence of . ' ~V0 94/1 1910 ~ /US93/1091o '_ 8 chromium in the hydrogen storage alloy material, as disclosed in the '586 Patent.
In contrast to the V-Ti-~r-Ni based alloys described above, the early AB5 alloysare ordered materials that have a different chemistry and microstructure, and exhibi~
difrerent electrochemical characteristics compared to the V-Ti-Zr-Ni based alloys.
S However, recent analysis reveals while the early AB5 alloys may have been ordered materials, more recently developed AB5 alloys are not. The performance of the early ordered AB5 materials was poor. However, as the degree of modification increased (that is as the number and amount of elemental modifiers increased) the materials became disordered, and the perforrnance of the AB5 alloys began to improve significantly. This is due to the disorder contributed by the modifiers as well as their electrical and chemical properties. This evolution of AB5 type alloys from a specific class of "ordered" materials to the current multicomponent, multiphase "disordered" alloys is shown in the following patents: (i) U.S. Patent No. 3,874,928; (ii) U.S. Patent No. 4,214,043; (iii) U.S. Patent No. 4,107,395; (iv) U.S. Patent No. 4,107,405; (v) U.S. Patent No. 4,112,199; (vi) U.S.
Patent No. 4,125,688; (vii) U.S. Patent No. 4,214,043; (viii) U.S. Patent No. 4,216,274;
(ix) U.S. Patent No. 4,487,817; (x) U.S. Patent No. 4,605,603; (xii) U.S. Patent No.
4,696,873; and (xiii) U.S. Patent No. 4,699,856. (These references are discussedextensively in U.S. Patent No. 5,096,667.

Simply stated, in the AB5 alloys, like the V-Ti-Zr-Ni metal hydride alloys, as the degree of modification increases, the role of the initially ordered base alloy is of minor imponance compared to the properties and disorder attributable to the panicular modifiers.
In addition, analysis of the current multiple component AB5 alloys indicates that current AB5 a~loy systems are modified following the guidelines established for V-Ti-Zr-Ni based systems. Thus, highly modified AB5 alloys are identical to V-Ti-Zr-Ni based alloys in that both are disordered materials that are characterized by multiple-components and multiple phases and there no longer exists any significant distinction between these two types of multicomponent, multiphase alloys.
Rechargeable alkaline cells can be either vented cells or sealed cells. During normal operation, a vented cell typically perrnits venting of gas to relieve excess pressure as pan of the norrnal operating behavior. In contrast, a sealed cell generally does not permit venting on a regular basis. As a result of this difference, the vent assemblies and ~0 94/1 1910 2 1 4 6 3 7 o PCr/US93,109l0 .

the amounts of electrolyte in the cell container relative to the electrode geometry both differ ~ignific~ntly.
Vented cells operate in a "flooded condition." The term "flooded condition" means ~ that the electrodes are completely immersed in, covered by, and wetted by the electrolyte.
Thus, such cells are somPtimes referred to as "flooded cells." A vented cell is typically designed for very low operating ~lessules of only a few pounds per square inch after which excess pressures are relieved by a vent mech~ni~m In contrast, sealed cells are designed to operate in a "starved" electrolyte configuration, that is with a minimum amount of electrolyte to permit gas recombination.
The enclosure for a sealed cell is normally metallic and the cell may be designed for operation at up to allp~ tPly 100 p.s.i. absolute or higher. Because they are sealed, such cells do not require periodic m~int~n~nce.
Typically, a sealed rechargeable alkaline cell uses a cylindrical nickel-plated steel case as the negative terminal and the cell cover as the positive te~min~l An insulator sep~tes the positive cover from the negative cell can. The electrodes are wound to form a compact "jelly roll" with the electrodes of opposite polarity i~ol~tecl from each other by a porous, woven or non-woven separator of nylon or polypropylene, for example. A tab extend from each electrode to create a single cuIrent path through which current is distributed to the entire electrode area during charging and discharging. The tab on each electrode is elect~ic~lly conn~-cted to its respective terminal.
In sealed cells, the discharge capacity of a nickel based positive electrode is limited by the amount of electrolyte, the amount of active m~tPri~l, and the charging effiri~n~ies. The charge c~pacities of a NiCd negative electrode and a Ni-MH negative electrode are both provided in excess, to maintain the optimum capacity and provide overchalge ~lulec~ion.
The operational lirespan, that is, the available number of charge and discharge cycles of a sealed cell, typically determines the kinds of applirations for which a cell will be useful. Cells that are capable of undergoing more cycles have more potential applirations. Thus, longer lifespan cells are more desirable.
An additional goal in making any type of electrode is to obtain as high an energy density as possible. For small batterito.s, the volume of a nickel hydroxide positive electrode is more ilnpoll~lt than weight and the energy density is usually measured in WO 94/11910 PCI/US93/1091"
21~63~'7 ~ ~

mAh/cc, or an equivalent unit.
At present, sintered, foamed, or pasted nickel hydroxide positive electrodes areused in NiCd and Ni-MH cells. The process of making sintered electrodes is well known in the art. Convention~1 sintered electrodes normally have an energy density of around 480-500 mAh/cc. In order to achieve signifi~nt1y higher loading, the current trend has been away from sintered positive electrodes and toward foamed and pasted electrodes that can be manufactured with an energy density of greater than 550 mAh/cc.
In general, sintered positive electrodes are constructed by applying a nickel powder slurry to a nickel-plated steel base followed by sintering at high t~.nl~e.~ture. This process causes the individual particles of nickel to weld at their points of contact resulting in a porous material that is approxim~t~ly 80% open volume and 20% solid metal. This sintered material is then impregnated with active m~teri~1 by so~king it in an acidic solution of nickel nitrate, followed by con~ ion to nickel hydroxide by reaction with sodium hydroxide. After impregn~tion, the m~teri~1 is subjected to electrochemical lS formation in ~lk~lin~ so1~1tion to convert the nickel hydroxide to nickel oxyhydroxide.
In all rechargeable cells using a nickel hydroxide posilive electrode, the nickel hydroxide changes back and forth be~. ~n Ni(OH)2 and NiOOH as the cell is charged and discharged. These reartions involve a ~ignifir~nt density change during the charge/discharge reaction~ This eYp~n~ion and contraction causes a "swelling" of the electrode. This swelling is a cQmmon cause of failure in cells using a nickel hydroxide positive electrode. Failure occurs bec~se as the posilive electrode swells, it absorbs free electrolyte from the sepa~alol until the sepalalor dries out.
U.S. Patent No. 5,077,149, describes a cell system to avoid swelling of the positive electrode. The described cell uses a Ni-MH negative electrode, a nickel hydroxide positive electrode, and a sulfonated, non-woven polypropylene separator all of which contain a zinc compound. The zinc compound prevents electrolyte migration to thepositive electrode by facilitating electrolyte retention in the negative electrode and the sep~ator. This reduces the expansion of the positive electrode. This patent states that expansion of the positive electrode causes a change in the electrolyte distribution and an increase in intern~1 resi~t~nre, and that the use of zinc oxide in the cell, rather than the fabrication of the electrode is the soll1tion to this problem.
Various "poisons", introduced during the production of the positive electrode or ~vO 94/11910 PCI/US93/10910 ~ 214637~

generated during the operation of the cell, can also cause cell failure. For example, residual nitrates and Fe are both known poisons.
u~1 nitrates occur during impregnation processes that use nickel nitrate.
~ Unfol~unal~ly, even parts per million levels of nitrate can result in undesirable self-5 discharge mech~ni~m~ through the formation of the nitrate shuttle reaction.
In both NiCd and Ni-MH cells, free Fe can be leached from insufficiently plated can or tab connPction~. In addition, some Ni-MH alloys contain Fe, and these materials oxidize and corrode. Once Fe gets into the aqueous electrolyte so1utioll, it is deposited on the nickel hydroxide and reduces the oxygen overvoltage, errecLively, poisoning the l0 positive electrode. A reduction in the oxygen overvoltage means that oxygen evolution will occur at the posiLive electrode before the positive electrode is fully charged, resulting in a reduction in capacity.
It has become standard practice in the Ni-Cd industry to avoid even the smallestFe i~ y in the cell by substituting pure Ni for Fe and by the extensive use of heavy 15 nickel plating. In a~lrlition, previously unknown poisoning mech~ni~m~ such as deposition or dissolution of met~llic species such as oxides of Ti, Zr, or V have been shown to affect the nickel hydroxide electrode in adverse ways such as reduction in capacity, lowered cycle life, and in~cased self discharge.
In summary, prior art nickel hydroxide positive electrodes have a number of 20 d~firienries that prevent the re~1i7~tion of the full potential of ill.p~ ,d Ni-MH negative electrodes. For example, sintered positive electrodes have energy density limitations. In ad~lition, while the use of foamed and pasted electrodes avoid these energy density problems, presently available nickel hydroxide positive electrodes undergo swelling that ultim~s~ly results in separator dryout, are susceptible to poisoning, may have poor rate 25 capability, and are susceptible to poisoning.

Summary of the Invention One object of the present invention is a sintered nickel hydroxide positive 30 electrode having an energy density of > 560- mAh/cc.
Another object of the present invention is a nickel hydroxide positive electrode that is resist~nt to swelling.

WO 94/11910 PCI/US93/1091~' , Yet another object of the present invention is a nickel hydroxide positive electrode that is resistant to poisoning.
These and other objects of the present invention are s~ti~fied by a positive electrode for use in ~lk~line rechargeable electrochçmicPl cells comprising: a m?tçri~l comprising a compositionally and structurally disordered multiphase nickel hydroxide host matrix which includes at least one modifier, preferable three modifiers, chosen from the group consisting of F, Li, Na, K, Mg, Ba, Ln, Se, Nd, Pr, Y, Co, Zn, Al, Cr, Mn, Fe, Cu, Zn, Sc, Sn, Sb, Te, Bi, Ru, and Pb.
Other objects of the present invention are sPti~fied by a positive electrode for use in rechargeable elecL,uchel,lical cells comprising a sintered nickel hydroxide electrode lacking c~-lmillm and having an energy density of > 560 mAh/cc, and a cycle life of >
500 cycles.
Still other objects of the invention are ~ti~fi~-l by a positive electrode for use in ~lk~linr rechargeable elec~ochf "ir~l cells, said positive electrode lacking c~(lmillm, having an energy density of > 560 mAh/cc, having a self discharge in a sealed Ni-MH
cell of < 30% in 30 days at 20~C, and having residual nitrates present in an amount less than 200 ppm.
Additional objects of the invention are s~ti~fiçd by a posilive electrode for use with V-Ti-Zr-Ni metal hydride alloy rechargeable ele~ ocl-~.,.ir~l cells comprising a sintered nickel hydroxide electrode lacking ca~lminm and having an energy density of > 560 mAh/cc, a cycle life of > 500 cycles, and a self dischalge in a sealed Ni-MH cell of <
30% in 30 days at 20~C.
Objects of the invention are also s~ti~fi~rl by a sintered positive electrode for use in ~lk7~1ine rechargeable electroch~mir~l cells comprising: a nickel substrate that is preoxidi7ed and ~or~ted; a nickel sinter having pores and an outer surface on said nickel substrate; and nickel hydroxide and cobalt hydroxide precipitate in said pores and on said outer surface; where said sintered positive electrode contains < 200 ppm residual nitrates, has an energy density of > 560 Ah/cc, has a self discharge in a sealed Ni-MH
cell of < 30% in 30 days at 20~C, and lacks Cd.
Objects of the present invention are also s~ti~fiçcl by a process for f~bric?ting sintered electrode m~t~ri?l from which a sintered positive electrode for use in an alk~line rechargeable electrochemical cell can be produced, said process comprising: forming a WO 94/11910 21~ 63 70 PCI/US93/10910 slurry of nickel powder, water, carboxy methyl cellulose binder, methyl cellulose binder, and a poly(ethylene oxide) polymer; spreading said slurry on a preoxidized perforated nickel substrate; dIying said slurry; and sintering said slurry.
Yet other objects of the invention are s~tis~e~l by a process for impregnating sintered 5 electrode material from which a high loading unirolll-ly distributed multiphase substantially nitrate free sintered positive electrode for use in an ~lk~linP rechargeable electrochemical cell can be produced, said process comprising: impregnating said sintered electrode m~teri~l using from multiple impregnation cycles to attain high lo~ling, where each impregnation cycle comprises the steps of: placing said sintered electrode material 10 on a rack; dipping said rack into nickel nitrate; allowing said rack to drip dry; dipping said dried rack into NaOH solution; spraying said rack in a first tank with deionized water overflowing from a second tank; dipping said rack in said second tank filled with deionized water overflowing from a third tank; dipping said rack in said third tank filling with deionized water at a rate of 8-10 gpm; drying said rack; and flipping said rack to 15 attain uniform deposition of material; where in the median dip cycle and in the final dip cycle of said multiple impregnation cycles, said step of dipping said rack into nickel nitrate is replaced by a step of dipping said rack into cobalt nitrate.
The objects of the present invention are also s~ticfi~fl by a process for forming a high loading uni~ol~nly distributed mllltiph~ce subst~nti~lly nitrate free sintered positive 20 electrode for use in an ?. lk~linç rechargeable electr~h~-mic~l cell, said process comprising:
(1) fab~c?ting sintered electrode m~teri~l by forming a slurry of nickel powder, water, carboxy methyl cellulose binder, methyl cellulose binder, and a poly(ethylene oxide) polymer; spreading said slurry on a preoxidized perforated nickel substrate; drying said slurry; and sintering said slurry; (2) impregnating said sintered electrode m~t~ri~l using 25 multiple impregnation cycles to attain high lo~ling, where each impregnation cycle comprises the steps of: placing said sintered electrode m~t~ri?l on a rack; dipping said rack into nickel nitrate; allowing said rack to drip dry; dipping said dried rack into NaOH
solution; spraying said rack in a first tank with deionized water overflowing from a second tank; dipping said rack in said second tank filled with deionized water overflowing 30 from a third tank; dipping said rack in said third tank filling with deionized water at a rate of 8-10 gpm; drying said rack; and flipping said rack to attain uniform deposition of material; where in the median dip cycle and in the final dip cycle of said multiple WO 94/11910 PCI/US93/1091t' - 21~637'~

impregnation cycles, said step of dipping said rack into nickel nitrate is replaced by a step of dipping said rack into cobalt nitrate to produce an enriched cobalt surface; and (3) forming said impregnated sinter into positive electrode material by presoaking said impregnated sinter in NaOH presoak tanks to substantially elimin~te nitrates; brushing the 5 presoaked impregnated sinter in a surface brushing station; charging the brushed impregnated sinter; discha~ g the charged impregnated sinter; rinsing the discharged impregnated sinter; and drying said rinsed impregnated sinter to complete the formation of positive electrode material.

Brief Desc,il)~ion of the Drawin~s Figure l is a cross sectional ,c~lesenlaLion of the sintering line.
Figure 2 is a schem~tir of the impregnation process.
Figure 3 is a cross sectional l~lese~ tion of the deionized water rinse system. Figure 4 is a cross sectional le~resçnt~ticn of the formation line.
Detailed Des~ ,tion of the Invention A sinle.~,d posiLive electrode of the present invention embodies a new kind of active m~t.ori~1, that has a energy capacity equivalent to foamed or pasted electrodes that have a porosity of 90% or greater, in ~ ition the active material of the present invention has 20 a lower concentration of residual nitrates, and a greater resi~t~nce to poisoning than prior art m~tt~ri~
It is well known that Ni-MH negative electrode material is a much more efflcientstorage mylillm than nickel catlmimn negative electrode m~teri~1 This makes it possible to decrease the thi~n~ss of the negative electrode by 50% in a Ni-MH cell compared to 25 a Ni-Cd cell and increase the thi-~ness of the positive electrode in a Ni-MH cell by 50%.
So, instead of two electrodes that are approximately 0.025" thick in a conventional C size Ni-Cd cell, Ni-MH cells have a 0.038" thick positive electrode and a 0.012" thick negative electrode.
In order to make a thicker nickel hydroxide positive electrode, the present inventors 30 discovered that the viscosity of the slurry mixture used to prepare the nickel sinter structure must be much higher than convention~11y used in the Ni-Cd industry. A
standard slurry for p~e~;ng a nickel sinter structure for use in Ni-Cd cells has a viscosity ~VO 94/11910 21~ 63 7U PCI/US93/10910 of approximately 40,000 centipoise ("cp"). For a final electrode thickness of 0.025", the slurry is initially applied at a thirkness of approximately 0.05" followed by drying and sintering where the final thirkness is reduced by a~yuu~i,nately 50% or 0.025". After the slurry is applied to the substrate via a doctor blade appaldtus, the coated substrate is S passed through a vertical drying tower, which is usually followed by a hori7Ont~l sintering operation.
To make a 50% thicker final electrode, the coating of the slurry onto the substrate must be almost 0.090". This thirkness is beyond the capability of convention~1 slurries in that the normal viscosity is far too low to avoid sagging, running, and other10 imperfections as the coated substrate enters a vertical drying tower. To correct this problem, the present invention involves a subst~nti~1 increase in slurry viscosity from 40,000 cp to ayl"u~ llately 72-75,000 cp. This higher viscosity allows a much thicker coating of the slurry without the problems discussed above.
In addition to making a s~bst~nti~lly thicker elect~ode than conventional nickel15 hydroxide electrodes, the present in~,ell~ol~ discovered that it is also desirable to make a nickel hydroxide electrode that has higher capacily than previously known. Conventional sintered nickel hydroxide electrodes have a base sinter structure that is ayplo~ te1y 80%
porous. Conventional thinking was that to make a higher CdpaCily electrode, it was n~o~ess~ry to make the nickel sinter even more porous, up to as high as 90%. Because 20 90% porous m~t~n~l that is sintered is much more fragile and prone to swelling, the industry trend has been to develop foamed and pasted structures of 90% porosity.In fact, many con,panies have developed pasted nickel hydroxide electrodes utilizing foam metal or fiber metal substrates having 90% pOlu~ily. Active material is usually applied by pasting techniques using special active materials. The desired objecive in 25 using these pasted materials was pnm~rily an increase in energy density from ayplo~inlately 500 mAh/cc for standard sintered structures to over 550 mAh/cc.
The trend toward foamed metal substrate electrodes away from sintered electrodes has advantages as well as disadvantages. The primary motivating factors for foam metal substrate electrodes was a desire to have a higher energy density than standard sintered 30 electrodes, to lower the cost in high production vo1-1mes, and to simplify processing. On the other hand, a great number of problems exist which have still not been fullyaddressed. The major problem with pasted electrodes that has not been resolved is that WO 94/11910 PCI/US93/1091~

2:14~3:7~01 _ pasted electrodes generally have a low cycle life and poor power capability. The more porous structure is weak and very prone to swelling resulting in electrolyte redistribution problems in a sealed cell which shortens the cell's life. In ~ ition, the higher porosity structure is also less cond11rtive and the distance between active m~teri~1 and current 5collection is usually increased. Both problems, swelling and reduced conductivity, have led to the development of very special active m~t~ri~1~ such as high density spherical nickel hydroxide. Intensive efforts are made to achieve high density in order to achieve a high final electrode capacity. In a~l~lition, special form~ tinn~ of active material are utili~d to reduce swelling. However, these electrodes are very costly due to the fact that l0the foam metal substrate and the high density nickel hydroxide are expensive to produce.
Consequently, a sintered positive electrode having comparable energy storage capacity to pasted foam electrodes without reduced power and cycle life is very desirable.
Quite une~pectedly, the present inventors found that there was an abundant amount of active m~teri~l present in a conventinn~1 nickel hydroxide positive electrode, but that 15it was not fully uti1i7~ In forming a sintered electrode, nickel powder is sintered to form an çle~tric~l and physical ~leton Nickel hydroxide is deposited on this ~rleton in an impregnation process to form an electrically conductive matrix. The present ~s found that in the nickel powder sk~1etQn, pores greater than appro~ te1y 30 microns in ~ meter could not fully utilize the nickel hydroxide active m~teri~1 and that 20in convention~1 nickel hydroxide posili~c electrodes, many of the pores were from 40-80 microns in ~i~meter. With this discovery, the present inventors reali~d that impregnation of more active material was not as important as effectively utilizing the active m~teri~1 that was already present. The sintered electrodes of the present invention predomin~nt1y limit pore size to a~ imately 30 microns in size, therefore, they have 25greater capacity even though their overall porosity, like that of convention~1 sintered electrodes, is around 80%.
One step in the construction of conventit~n~1 sintered nickel hydroxide electrodes involves forming a slurry of nickel and methyl cellulose binder. The present inventors discovered that this simple, single binder slurry is not effective in providing good mixing 30and subsequent uni~olln pore si~ distribution. Methyl cellulose is a relatively ineffirient binder and, as a result, the nickel forms clumps and pockets. The resulting sinter, if viewed under a mic~o~col e, shows the nickel ske1eton to be uneven with many pores of Uv094/11910 1 ~ 63 70 PCI/US93/10910 greater than 30 microns in size. The inefficiency of the nickel/methyl cellulose sluIry is exacerbated when trying to make a 0.038" thick positive electrode as desired in a Ni-MH
cell. In order to get a final thickn~ss of 0.038", it is n~cessary to doctor blade the nickel slurry onto the substrate in a layer that is ~p~o~ ately 0.090" thick. This is 5 ap~lo~il--ately double the thic~n~ss of nickel slurry used to produce a converltion~1 nickel hydroxide positive electrode. As previously stated, in order to achieve such a thick electrode, it is necessary to substantially increase the slurry viscosity. The higher viscosity makes the mixing pl'~ C~ ~ies of the methyl cellulose even worse. Applying such a thick slurry is extremely difficult because conventinn~11y forrm-1~ted slurries lack l0 s1)fflci~nt viscosity- to be self-s~1staining at this thickn~ss The problems associated with prior art slurnes are o~lco..-e by a unique binder system we have developed. This binder system is formulated by combining a poly(ethylene oxide) polymer, carboxymethyl cellulose, and methyl cellulose with water.
This slurry system is preferably formulated using 43-53 wt% nickel, 45-55 wt% water, 15 0.3-1.3 wt% POLYOX (a trade mark of Union Carbide for poly(ethylene oxide) polymer), 0.1-0.9 wt% c~l,o~y.l-ethyl cçllnlose~ and 0.1-0.7 wt% methyl cellulose. Without wishing to be bound by theory, it is believed that this binder system makes it possible to produce higher capacity positive electrodes, bccause it mer1i~tes the even mixing and formation of a homogen~ou~ so111tion of nickel particles in a slurry that is extremely viscous yet can 20 still be doctor bladed into a unifol~l layer.
The slurry of the present invention is fc.rm~ ted by first mixing the dry ingre~i~nt~, the poly(ethylene oxide) polymer, carboxymethyl ce1h11Ose, and methyl cellulose with INCO 255 ~) nickel powder (m~nnfactmed by the Intern~tiona1 Nickel Company) for 15 to 60, preferably 30 minlltes by tumbling at from l to 3, preferably 2 RPM, in a smooth 25 wall cont~iner. Water is added tO the t.1mb1e-1 mixture and the tumbling resllmç~l at from 1 to 3, preferably 2 RPM until the slurry has a viscosity of from 68,000 Centipoise (cp) to 76,000 cp; and a density of from 1.65 to 1.71 g/cc.
The prepared slurry is doctor bladed onto a preoxidi7~d perforated nickel substrate, followed by drying and sintering. Preoxidized p~lruldled nickel substrate is pler~lcd for 30 twû reasons. It has been found that a~lhesion of the nickel powder is enh~nred during sintering by preoxi~tioll This is important in that del~min~tion of the nickel powder matrix is a common difficulty in the preparation of these m~teri~1~. It has also been found WO 94/11910 PCI/US93/1091~
2I4~370 that the preoxi-li7ç~ m~tPri~l is more immune to corrosion or "acid attack" during impregnation. It is also envisioned that a full range of porous, solid nickel substrates may be used in addition to the perforated nickel substrate specifically described herein.
Figure l illustrates a roll-to-roll m~chinç for doctor blading the slurry of the invention S onto a substrate, drying the slurry, and sintering the slurry as a continuous operation. In Figure l, preoxifli7Pcl perforated solid nickel substrate is fed from a payout roll 1 through a slurry box 2. Previously l,lep~d slurry is fed from the slurry barrel 3 into the slurry box 2. The thi~knçss of the slurry on the substrate is determined as the screen moves by the doctor blade 4. The slurry is dried in the vertical drying oven 5 and si~d using a l0 c~1end~r roll 6. The si~d dry slurry is sintered in a continuous sintering furnace 8, which can have multiple heating zones and cooling zones. Samples for quality control inspection are taken from the resulting sinter formed on the solid nickel substrate in an on line punch press device 9. Tension on the continuous ribbon of sinter as it moves through the machine is m~int~inecl through the use of proximity switches such as proximity switch 7 lS which can be connçcted through a control device such as a micro computer (not shown) to the payout roll 1 and the take up roll l0.
Other unique aspects of the cint~ring process are the use of a special doctor blade ~p~aluS which substantially ~olimin~tes the form~tion of "dogbone" shaped edges. It is common for the slurry to be wiped away in strips which can later be utilized as integral 20 tabs for current col1~-ction. During the wiping, tremendous hydrostatic prcssul~s are built up c~lcing a spring back effect of the slurry at the outer edges near the wiped areas. The doctor blade &lJpdlaluS reduces this problem through the use of a tapered or stepped d~ctrring blade. The problem of "dogboning" is undesirable in that uneven con~lc,ssion can result in the cell unless expensive milling is done, essenti~11y grinding away of 25 m~tPri~1 Another unique so1-ltion to this problem is the use of a calendering mill positioned after drying. but before sintering. This allows the "dogbone" to be gently squeezed whereas post sinter c~1endering can cause a high density of nickel at the surface, inhihiting the pelleL,alion of nickel nitrate during impregn~tion Finally, it should be noted that an atmosphere of 7% H2 with the balance of N2 at a len.p~latu c of 30 ~plo~--ately 900~C is used during sintering. This atmosphere is most effective in providing a reducing atmosphere and high strength con,pa-~;d to other gas compositions.

vO 94/11910 21 ~ 63 7~ PCI'/US93/10910 ~ hemi~1 conversion as a method of impregnating nickel hydroxide into a sinter is well known. However, the ch~mir~1 conversion of the present invention is unique and not suggested by the prior art. At least one prior art method of impregnating the sinter also had other disadvantages. This method involves dipping the sintered nickel skeleton 5 into the acidic nickel nitrate solution which resulted in tremendous of corrosion of the nickel sinter.
Prior art methods of impregnation were also inappl~liate for the purpose of providing high loading of active m~teri~l in order to achieve the desired high energy density.
Frequently, problems of surface loading prevented interior loading of active material, 10 thereby inhibiting the goal of thoroughly loading tne available porosity from sintering with active material. Other problems with prior art çh~mir~1 impregnation involve insufficient drying after nitrate loading and rinsing, which p,e~ents high density nickel hydroxide loading. Still other problems involved too high a level of residual nitrates and carbonate h~ ies due to ins11ffiri~nt rinsing and purity of the NaOH conversion bath. Other 15 problems with prior art impregnation involved the use of excessive temperatures during drying, c~ ing the creation of elec~ e...ir~11y inactive nickel oxides. Further problems resulted from a failure to colll~cllsate for excessive loading variances due to gravity.
Finally, the active material itself is in~1ffici~nt for use in Ni-MH cells. This problem is in ad-lition to difficulties in producing a thick, highly loaded, high ~1ti1i7~tion electrode 20 free from excessive impurities, ~ cussed above. Prior art active m~t~ri~l~ did not have to address the problems of el~lochc...ir~1 formation difflclllties due to high 1c-a~ing, poisoning resist~nce to new unique potentiaI poisons from a V-Ti-Zr-Ni-Fe based alloys, nor the higher sensitivity of Ni-MH cells to self-discharge col,lpar~,d to NiCd cells.
Further, electrode charge effici~ncy, esperi~11y under quick charge conditions had to improve. This is in part due to the fact that practical current d~nsitit:s increase in high energy Ni-MH cells col~ ed to NiCd cells because end users still require ten hours of slow charging and 15 to 60 minutes of fast charging, even though the absolute current required to one-hour charge a 5.0 Ah Ni-MH C size cell is 5.0 Amps collll,~cd to a Ni-Cd C si~ cell where the one hour charge current is only 2-2.4 Amps.
In addition to the difficulties ~ cllcse~ above, problems also arise in the electrochemir~l formation process. We have observed that conventional prior art formation is totally inadequate in the areas of residual nitrate reduction and activation of WO 94/119lO PCI/US93/1091~
21~6370 the nickel hydroxide to çlimin~te discharge reserve. Further, we have also learned that the composition of the alk~lin~ electrolyte used in formation, its ten~ ture, the charge and discharge current densities, and the manner of rinsing and drying of the electrode all play important roles in the proper functioning of positive electrodes used in a high perform~nce Ni-MH battery.
Together, these factors illustrate that the sintered nickel hydroxide positive electrodes of the present invention are superior for use in NiCd cells and particularly superior for use in state of the art Ni-MH cells. The inadequacy of prior art sintered nickel hydroxide posili-re electrodes is underscored by the industry wide movement to foam based pasted posilive electrodes in Ni-MH cells as ~ cncsed above.
The present invention avoids the problems of the prior art. Rather than using nickel plated steel as the substrate, the present invention uses a preoxidi~d perforated nickel substrate, as tli~cllcsed above. This promotes better adhesion of the nickel powder particle to the substrate metal as well as signific~ntly limiting the amount of corrosion that occurs during the il.,p,egnation steps. In ~ tion~ the sintpring process of the invention provides an ~d~lition~l degree of pre-oYi~l~ti-n to the nickel powder particles making them more resistant to corrosion during impregn~tion The impregnation process of the present invention is accomplished using multiple, .
success.ve, lmpregnatlon cycles.
A further aspect of the unique impregnation process of the present invention is that the posilive electrode of the present invention cont~in~ a higher ~ e.~age of co-~.~c;~ ed cobalt than do the prior art m~t~ri~l~. While the use of co-precipitated cobalt is known, particularly for nickel~lminm cells, the Co content of these cells is only a~p~ tely 1-3%. In contrast, a positive electrode of the present invention con~ins greater than 6 wt%, preferably 9-10 wt% co-p~eci~ ted cobalt.
In the impregnation process of the present invention, the high cobalt content of the posilive electrode is further acce-~lu~ed by the use of a cobalt nitrate dip in the median dip cycle and then again at the final dip cycle. Herein, the phrase "median dip cycle"
is used to refer to the dip cycle halfway through the series of succes~ive impregnation cycles and the phrase "final dip cycle" to refer to the dip cycle at the end of the series of successive impregnation cycles. Using the cobalt nitrate median and final dip cycles produces an enriched cobalt surface which provides better conductivity, poisoning WO 94/11910 ~ PCr/US93~10910 63 76~

resistance, and suppresses ~2 evolution.
Without wishing to be bound by theory, it is believed that the final cobalt nitrate dip in the impregnation process also result in surface enriched cobalt. The median dip cycle is believed to improve utilization and accelerate activation by increasing the overall 5 conductivity of the aggregate active m~teri~l. The outer surface is critical for poisoning resistance. In reality this means that while the composition of cobalt in the active m~ttori~1 is greater than 6 wt%, preferably 9-10 wt%, the concentration of cobalt hydroxide on the surface of the electrode is much higher.
Cobalt hydroxide is more resistant to poisoning than pure nickel hydroxide. The use l0 of the described concentration of pure cobalt hydroxide and a higher concentration of the co-precipitated cobalt is crucial for use in V-Ti-Zr-Ni metal hydride alloy based Ni-MH
batteries.
Another unique aspect of the materials of the present invention believed to result from this unique impregnation and form~tion process is that the fini~he~l electrode has a very 15 low concentration of residual nitrates. Resi~lu~1 nitrates result primarily because the impregnation cycles of the impregnation process use nickel nitrate or cobalt nitrate that are converted into their r~s~,ecliv~ hydroxides as described in detail below. In the prior art, the co~ ion of nitrates into hydroxides was much less than 100% çfflcient and, as a result, some nickel nitrate was locked into the matrix of the fini~hç~ positive electrode 20 materials. In the present invention, the impregnation process is ~ignifi(~ntly more effic-ent which results in a re-lnctiQn of the residual nitrate to no more than 200 ppm, maxlmum.
The impregnation process of the present invention is schem~tic~11y illustrated in Figure 2. The sinter is coiled onto a rack and dipped into nickel nitrate or cobalt nitrate 25 depending on the impregnation cycle. During the first impregnation cycle, a nickel nitrate dip, the sinter is dipped into 0.02 N HO3 in nickel nitrate (2.5 M M(NO3)2) for lS
min11tes. During the rem~in~ler of the impregnation cycles that use 0.02 N HO3 in nickel nitrate the sinter is dipped in 0.04 N HO3 in nickel nitrate (2.~ M Ni(NO3)2) for l~
minutes. The nickel nitrate solutions are m~int~ined at appluAi-nately 45~C. The reduced 30 acid concentration on the first cycle acts to inhibit corrosion especially at the substrate/sinter in1~r~ce On subsequent dips, corrosion is minimi7Ç~ by the nickel hydroxide reaction product itself. Each of the two cobalt nitrate impregnation cycles are WO 94/11910 214 6 3 i O PCI'/US93/1091~' for lS minntes The cobalt nitrate solutions are m~int~ined at approximately 20~C.
The dipped sinter is then dried until "bone dry". As used herein, "bone dry" means drying until no further weight loss occurs. The dipped sinter is usually bone dry after approximately 60 minutes at a It.npel~ture not higher than 80~F. It has been learned that S higher drying Lenl~e~a~ures were detrimental due to the formation of electrochemically inactive nickel oxide. Drying time is m~int~inPcl at commercially practical level by an emphasis on the circulation of large air volume as opposed to higher lem~cldture.
The bone dry sinter is then dipped in a 70~C, 30 wt%, solution of NaOH for al)plu~inlately 15 min~1tes, preferable 19-21 mim1tes, most preferably 20 mimltes Following the NaOH dip, the sinter is rinsed using a three step deionized water rinse system as schem~ti~11y dia~lanllllcd in Figure 3, employing a spray tank l, and two dip tanks 2,3. The water in the spray tank 1 is the cascade overflow from dip tank 2, and the water in dip tank 2 is the c~c~de overflow from dip tank 3. The water in the spray tank l is drained off at a rate of 8-l0 gpm, the same rate at which the deionized water is lS rep1eni~hed to dip tank 3. The water in the spray tank l is sprayed from the top and bottom at the rate of ap~ro~cilllately l00 gpm. The term "deionized water" as used herein refers to water having a maximum conductivity of 0.83 micro ohms, a minimum resistivity of 12,000 megaohms-cm, and a pH of from 4.5-7.5. The teml)el~ule of the water as it is added to dip tank 3 is ~pl~ te1y 70-80~C. A sinter is rinsed in each 20 rinsing station for applo~ lately 60 minlltes. We have discovered that a combination of sprayfi"",l~ion is superior to either a total spray or total immersion system. The system of the present invention exploits the better quality rinse resulting from immersion and the efflrien~y of a spray to remove the initial NaOH quickly. After the completion of rinsing the sinter is again dried until bone dry. Impregnation cycles are repeated until the loading 25 process is complete.
As previously menti-~ned, it is common in industry for the bottom material to be more heavily loaded than the top due to gravity. This is undesirable since loading is directly proportional to cell capacity and a low as possible capacity distribution is desirable. Some companies combat this problem by rest~r~ing racks at certain points in the dip process.
30 An aspect of the present invention involves a far superior solution. In the present invention, a rack flip device is used to flip each impregnation rack so that the top of the rack becollles the bottom and the bottom becomes the top. This rotation takes only ~O 94/11910 21 1 ~ 3 7 ~ PC~/US93/10910 seconds, and the rack/flip device is easily incorporated into the processing line. Using the rack flip device after each impregnation cycle insures uniform loading which is vital for achieving a uniformly high capacity electrode.
Nitrate ions, as mentioned above, are generated in the electrode fabrication process of 5 the present invention because of the use of nickel and cobalt nitrates. When a sinter is immersed in a nickel or cobalt nitrate solution it soaks up the solution into its pores. The subsequent drying step drives off the water and leaves behind nickel nitrate or cobalt nitrate salt in the pores. When the dry sinter is immersed in sodium hydroxide solution precipitation occurs. For nickel nitrate, for example, this precipitation can be expressed l0 as follows:

Ni(NO3)2 + 2NaOH ~ Ni(OH)2 + 2NaNO3 The nickel or cobalt hydroxide precipitate is held in the pores and on the surface of the lS nickel sinter and the sodium nitrate dissolves in the hydroxide. Some of the nitrate can remain occh1de~ in the nickel hydroxide or cobalt hydroxide precipitate.
In prior art meths)-l~, the impregnated sinter inevitably retains small q~1~ntities of nitrate in spite of any subsequent rinsing operation. We have dcte~ ed that the actual amount of nitrate held depends on the number of dips (the loading) and the amount of 20 residual nitrates in the dip tanks and rinse water. In the present invention, the three part rinse following the NaOH dip removes nitrate residue and gives an active m~teri~l filled only with nickeVcobalt hydroxides. However, it is now nPcess~ry to follow the rise of sodium nitrate in the alkali dip solution and also in the rinse water. Prior art methods that did not use as many impregnation cycles andJor the three part dip described above perrnit 25 the concentration of nitrates to build up to levels s11ffl~ient to effect self discharge.
Consequently, an aspect of the present invention monitors the nitrate ion level in the NaOH bath and when the level reaches a maximum of 30,000 ppm, the solution is replaced with fresh NaOH.
In ~r~ition~ the form~tiQn process of the present invention removes nitrates that remain 30 in the impregnated sinter. In addition, the formation process increases the surface area of the electrode and increases the electrolyte uptake for quicker activation. The formation process occurs after the nickel hydroxide has been deposited inside the pores in the WO 94/l 1910 PCI/US93/109l1' sintering step and involves an electroch~mic~l formation cycle that is a one cycle charge/discharge prior to the positive electrode material going into the battery. After the completion of the formation process, the amount of residual nitrates present is small, despite the fact the electrode is a~plo~i",ately 50% thicker and 10-15% heavier due to 5 loaded active m~teri~l, both factors that would contribute to higher levels of residual nitrate rather than lower levels. The effectiveness of the formation process is a result of high efficiency through 200% overcharge followed by complete discharge.
Figure 4 schem~ti~lly illustrates the formation process of the present invention in which the irnpregnated sinter is formed into sintered positive electrode material. The 10 formation process begins by winding impregnated sinter from the irnpregnation racks onto formation spools.
A formation spool 1 is fed into a presoak tank 2 containing NaOH electrolyte at 40-50~C, preferably 45~C. We have found that the electrolyte absorbed by the electrode during the presoak greatly f~cilit~tçs charge effiriçncy during the initial charging step.
15 The electrode material is unwound from the form~sion spool 1 and fed past opposing brushes 3 to remove surface loading and loose partirul~tes The brushes have variable speed and pr~s~ to allow adjuctm~nt for specific incoming m~teri~l and conditionc The brushing step further f~cilit~t~s electrolyte uptake by removing surface loaAing- In lition, brushing i,npro~,s electric~l contact b~ ,en the m~t~ri~l and contact rollers 20 because it removes surface nickel hydroxide which has low conductivity.
Using a series of wetted contact rollers, the m~t~ri~l is moved through a chargesection 3. The charge section 3 concictc of a series of tanks, preferably four tanks, cont~ining electrolyte m~int~inPd at the same t.,l"pe~ture as the presoak tank 2. A
counter electrode is present in each tank of the charge section 3. The counter electrode 25 is col-nPcl~cl to the negative terminal of a power supply and the wetted contact rollers are conn~cted to the positive terminal of the power supply.
The charge section is designed to provide 200% of the theoretical capacity of the m~teri~l However, no amount of charge input will be effective if it is not accepted by the active m~t~ri~l We have invented several innovative approaches to assure charge 30 acceptance. As mçnti~ n~cl above, the use of a presoak tank and brushes improves electrolyte penetration. However, the sohltion itself is also important. It is common in industry to form in KOH rather than NaOH since KOH is used in the final cell for reasons ~'V0 94/11910 2 1 4 6 3 7 0 PCr/US93/10910 of charge efficiency, le,llye~lule characteristics, cycle life, etc. Consequently, our discovery that NaOH is a more effective form~tit~n electrolyte is surprising. A related aspect of the present invention is the discovery that heating the NaOH to 45~C rather than conventional formation at room temperature further ~ccent-l~tes the benefit. We have 5 observed substantially higher capacity on the first cycle in the sealed cell when using NaOH at 45~C to the degree that almost 100% of expected capacity is provided on the first cycle.
Still another aspect of the invention is a means to provide greater charge acceptance by an innovative counter electrode design which co"lpensates for voltage drop across the 10 electrode m~teri~l Normal formation uses a single flat plate counter electrode with terminal connections at the top of the bath. This is a problem in that the terminal connection of the material itself is the contact roller, which is also at the top to the bath.
The resistance of the nickel electrode is substantial and over the entire length of the counter electrode, the voltage drop is signific~nt A co"venlional single plate counter 15 electrode causes ~ignifi~nt v~rian~es in the current density from the top to the bottom of the mat~rial- This sitU~tion results in most of the applied current being wasted on gas evolution instead of being used to charge the material. We conr~ eA this conclllsjon using static tests on the m~rhin~ that showed the characteri~tic color change from green (nickel hydroxide) to black (nickel o~yhy~u~ide) oc-;u~l~d only in the upper 10% of the 20 available charge section Our innovative sol~ltion to this problem was to break the single plate counter electrode into five segrnents, where each segment is separated by a resistor designed to match the voltage drop of the positive electrode m~tPri~l This configuration provides a very uniform current density to the materi~l, facilit~ting charge. This same approach f~ t~tes 25 discharge as well, and is repeated throughout the form~tion process.
The collective form~tion process provides electrodes having virtually a 100% real depth of discharge, greatly reduced levels of residual nitrates, increased surface area (which allows easy electrolyte uptake during cell fabrication as mentioned above), and yields cells that exhibit virtually 100% of their expected capacity even on their first cycle.
30 As a further demonstration of the effectiveness of the described formation process, convention~l Ni-Cd positive electrodes not sintered or impregnated in the mannerdescribed above yield cells having very high self-discharge. By making no change in the WO 94/11910 PCr/US93/1091~' 214637~ _ fabrication of such electrodes except to use the improved formation described above, self-discharge rates were reduced by 50%. Thus, the cells of the present invention have a self discharge in a sealed Ni-MH cell of < 30% in 30 days at 20~C.
The discharge section 4 of Figure 4 is similar to the charge section 3. The discharge 5 section 4, also consists of a series of tanks co.-l~it~ g electrolyte at the same temp~ tulG
as the presoak tank 2, counter electrodes, and wetted contact rollers. However, in the discharge section 4, it is necess~ry to have only two tanks because discharge is accepted at higher rates than charge, and only 100% of capacity is required to be removed, (unlike charge where 200% input is required and charging alone has "activated" the material).
10 The counter electrodes are connecterl to the positive temlin~l of a power supply, and the wetted contact rollers are connected to the negative terminal of the power supply. The object of the discharge section 4 is to remove all of the charge provided in the charge section 3. Since not all electrode material has identir~l capacity, the discharge section is ~esign~d to provide approximately 6% overdischarge (on average) in order to ensure that 15 all m~tPri~l is fully discharged. Generally, nomin~lly o~r~ha,g~,d m~teriRl has a characteristic "grayish" color.
The rinse S of the form~tion process preferably uses three tanks of deioni7Pd water having a counlc~ t flow rate of 3-5 gpm and an initial ltll-pc~lu,e of 75-85~C.
Brushes 6 may optionally be present in at least one of the rinse tanks to remove surface 20 loading and particulate matter.
The dryer 7 must be capable of drying the m~teri~l until it is bone dry. Any kind of app,upliate dryer may be used such as an infrared dryer.
Finished positive electrode m~teri~l is taken up on the take up spool 8.
The multiple impregnation cycles of the present invention result in spatial and 25 orient~tiQn~l pl~re~llen~ of similar or fli~simil~r atûms or groups of atoms that produce qualitatively new ~clrollll~nt~e levels for sintered pûsitive electrodes. The multiple impregnation cycles of the present invention result in a disordered multicomponent m~t~ri~l comprising a nickel hydroxide hûst matrix into which cobalt is incorporated as a modifier in a manner similar to the negative alloy disordered matèrials described above.
30 The disordered positive alloy m~teri~l~ of the present invention do not have periodic local order.
By forming nickel hydroxide posilive electrodes that are disordered materials, we have ~vo 94/11910 PCI/US93/10910 '- 2t~6370 greatly increased the porosity and performance of these electrodes. Generally, the improved characteristics of these alloys result from tailoring the local chemical order and hence the local structural order by the incorporation of at least one modifier, most preferably three modifiers, chosen from the group consisting of F, Li, Na, K, Mg, Ba, Ln, 5 Se, Nd, Pr, Y, Co, Zn, Al, Cr, Mn, Fe, Cu, Zn, Sc, Sn, Sb, Te, Bi, Ru, and Pb. Like the metal hydride negative alloys ~liccucsed above, disordered positive electrode m~ter~
have a substantially increased density of catalytically active sites and storage sites compared to the prior art single or multi-phase crystalline m~t~ri~lc These adflition~l sites are responsible for i~ oved effiriency of electrochemic~l charging/discharging and 10 an increase in el~ctri~l energy storage capacity.
The choice of disordered materials has filnd~mental scientific advantages: as seen, a substantial number of elementc can be includ~d in the list of c~n~ tes for electrodes.
These elemçntc offer a variety of bonding possibilities due to the multi-direction~lity of d-orbitals, and less so due to f-orbitals which, although extending in still more directions 15 than d-orbitals, are closer to the nucleus of the metal atom and, hence, less ~cescible.
Where prior art sintered electrode m~t~ri~lc had an energy density of only around 500 mAh/cc, the m~t~ri~lc of the present invention have an energy density of > 560 mAh/cc, preferably 600 mAh/cc.
The present invention is e~pl~in~d further in the following non-limiting Examples.

WO 94/11910 PCr/US93/109'~

ExamPles Table 1 SLURRY FORMULATION
quantity in kg Nickel Powder 54.9 Water 58.5 Carboxy methyl cellulose .626 Methyl cellulose .478 Polyox(g 956 ---Total--- 11 5.45 A slurry was l,lep~d using nickel powder, water, carboxy methyl cellulose binder, methyl cellulose binder, and POLYOX~ poly(ethylene oxide) polymer in the q~l~ntities in~lir~tçd in Table 1.
All the materials except water were added to a mixing drum which was rotated forthirty minutes at 2 rpm. Water was then added and mixed in with a stirring rod to remove air and reduce lumps to less than 0.5 inches in diameter. The drum was again sealed and mixed at 2 rpm. After 48 hours, the density of the resulting slurry was chçr~ed and any visible lumps broken. Viscosity was adjusted to 72,000 Centipoise (cp) (+/- 4,000 cp) and density to 1.68 g/cc (+/- 0.03 g/cc).
Sintering took place in a five zone furnace with each zone set at applo~i.l.ately 910~C.
Prior to sintering, the slurry was doctor bladed onto a preoxidized perforated solid nickel substrate and dried in a two zone drying tower at le-l-~el~lules of 107~C and 88~C. The air flow was m~int~ined at 5 SCF~II. Drying took place under a 7% hydrogen, 93%
nitrogen atmosphere. The physical parameters of these materials at this point are shown in Table 2.

~0 94/11910 PCI/US93/10910 :~ 2Ig6370 Table 2 after drying tower after sintering after doctor blade thi~kness (in) 0.094 0.045 0.036 area weight (g/in2) 2.21 1.241 1.18 Density (g/in3) 23.5 27.6 32.6 Sintered m~t~ri~l was then coiled on impregnation racks. The impregnation process involved 14 individual impregnation cycles. Each impregnation cycle involved a nitrate dip in nickel or cobalt nitrate, dlying, a ~lk~lin~- dip in NaOH, rin~ing, and drying.
The nitrate dip of impregnation cycle 1 was in 0.02 N HO3 in 2.5 M Ni(NO3)2 hexahydrate for 20 minutes The nitrate dips for impregnation cycles 6 and 14 were in 15 cobalt l~itrate hexahydrate for 20 ~inl~les The nitrate dips for all other impregnation cycles used 0.4 N HO3 in 2.5 M Ni(NO3)2 hexahydrate for 15 minutes.
For each impregnation cycle, following the nitrate dip in the a~lopliate nitratesolution, the impregnation rack was lifted out of the solution and allowed to drip dry for 30 mimltes The rack was then placed in a forced air recirculation dryer at 80~C for 60 minlltes where the dryer had a flow rate of 2,000 ft2/minute.
Rinsing following the nitrate dip, was done in a three tank system having a counter current flow from tank 3 to 2 to 1. Tank 1 was a spray rinse and tanks 2 and 3 were immersion rinses. Deionized water was used throughout.
Following the final irnpregnation cycle, impregnated positive material was uncoiled from the impregnation racks and coiled onto formation spools. The m~teri~l was then fed continuously at a rate of 8"/minute into the form~tion machine. In the formation machine the formation spools were placed in a presoak tank cont~ i. g 30% NaOH electrolyte at 45~C. The m~teri~l was then fed from the formation spools through opposing nylonbrushes (to remove surface loading and loose particulates), and into a charge section.
The charge section consisted of four tanks cont~ining counter electrodes and contact WO 94/11910 PCI'/US93/1091"
2l~6~37a.' ~

rollers. In the charge section, the contact rollers were connected to the positive terminal of a power supply while counter electrodes were connected to the negative terminal. This section provided the electrode material with at least 90% of its theoretical state of charge in order to encourage electrolyte absorption and cell capacity, as well as to remove 5 electrochemically inactive charge reserve. Electrolyte in the charge section was maintainPd at a te,l,peldture of 45~C to assist the charging reaction efficiency. Also, the electrolyte was recirculated and sprayed onto the contact rollers and tension on the belt of the material was kept high in order to provide optimal conductivity between the rollers and the material.
The matçrial was then passed into a discharge section consisting of two tanks were it received approximately a 6% overdischarge, on an average, in order to insure that all material was fully discharged.
The material was then rinsed using deionized water and nylon brushes.
Finally, the m~ten~l was dried using an infrared heater. The resulting positive electrode belt was slit, punched, cut to length, and fabricated into standard positive electrodes.
Ni-MH negative electrode material having the following composition V,8Til5Zr,8Ni29Cr5C~7Mn8 was fabricated into negative electrodes as flescribed in copending U.S. Application No.
07/879,823 the contents of which are incorporated by reference.
Standard nickel ca~lmillm negative electrode mat~rial~ were fabricated as described in Falk and SaLkind, Alkaline ~torage Batteries (1969).
Prepared negative electrodes, s~ dLor, nickel hydroxide positive electrodes of the present invention, and 30 % KOH electrolyte were assembled into "C" cells as described in detail in U.S. Patent Application No. 07/879,823. The specific negative electrode, separator, and positive electrode used in each cell is indir~tçd in Table 3, below. The fini~hPcl cells were subjected to charging and discharging con~lition~ and their charge retention determined as in~lic~ted in Table 3.

'~0 94/11910 PCI/US93/10910 ~ 21 ~63 70 Table 3 positive alloy negative alloy separator charge retention (~)C rate to l.OV

3 days 14 days I-pos Ni-MH treated pp 92 79 I-pos Ni-MH pp 83 62 I-pos Ni-MH nylon 8 l 52 S-pos Ni-MH treated pp 88 67 S-pos Ni-MH pp 78 42 S-pos Ni-MH nylon 76 39 'Ni-MH- stands for a Ni-MH alloy having the CV "r~s ~ ~ Vl"TilsZrl,,Ni2,,CrsCo,Mn";
I-pos eletl,udes are posibve elett,od~s fabficated acccrding to the present invention;
~S-pos- e,ec~udes are standard, prior art positive elett~udes ~nylon- separators are standard nylon separators;
15 ~pp- separators are standard untreated pu!y,,,ù~,ylene separators; and ~treated pp separators are radiation grafted poly,~,v~ ne separators as described in detail in U S
Patent Application No 07/879,823 It is obvious to those skilled in the art that the positive electrode materials of the 20 present invention may be pl~a~d by additional methods without departing from spirit and scope of the present invention.
The drawings, discussion, descriptions, and examples of this specification are merely illustrative of particular embo liment~ of the invention and are not meant as limitations upon its practice. It is the following claims, including all equivalents, that define the 25 scope of the invention.

Claims (11)

What is claimed is:

1. A positive electrode for use in alkaline rechargeable electrochemical cells comprising: a material comprising a compositionally and structurally disordered multiphase nickel hydroxide host matrix which includes at least one modifier chosen from the group consisting of F, Li, Na, K, Mg, Ba, Ln, Se, Nd, Pr, Y, Co, Zn, Al, Cr, Mn, Fe, Cu, Zn, Sc, Sn, Sb, Te, Bi, Ru, and Pb.

2. The positive electrode as claimed in claim 1, comprising at least three modifiers chosen from the group consisting of F, Li, Na, K, Mg, Ba, Ln, Se, Nd, Pr, Y, Co, Zn, Al, Cr, Mn, Fe, Cu, Zn, Sc, Sn, Sb, Te, Bi, Ru, and Pb.

3. The positive electrode as claimed in claim 1 for use with V-Ti-Zr-Ni metal hydride alloy rechargeable electrochemical cells.

4. The positive electrode as claimed in claim 1 for use with V-Ti-Zr-Ni metal hydride alloy rechargeable electrochemical cells wherein said positive electrode further comprises a sintered nickel hydroxide electrode lacking cadmium and having a cycle life of ~ 500 cycles.

5. The positive electrode as claimed in claim 1 for use with V-Ti-Zr-Ni metal hydride alloy rechargeable electrochemical cells wherein said positive electrode further comprises a sintered nickel hydroxide electrode lacking cadmium and having an energy density of ~ 560 mAh/cc, a cycle life of ~ 500 cycles, and a self discharge in a sealed Ni-MH cell of ~ 30% in 30 days at 20°C.

6. The positive electrode claimed in claim 5, wherein residual nitrates are present in an amount less than 200 ppm.

7. The positive electrode claimed in claim 5, wherein said energy density is greater than 600 mAh/cc.

8. The positive electrode claimed in claim 5, comprising a sintered electrode material having pores approximately 30 microns in size.

9. The positive electrode claimed in claim 5, wherein said positive electrode contains greater than 6 wt% coprecipitated cobalt.

10. The positive electrode claimed in claim 5, wherein said positive electrode contains 9-10 wt% coprecipitated cobalt.

11. The positive electrode claimed in claim 8, wherein the overall porosity of said sintered electrode material is around 80%.