LBNL-42694

EXPLORATORY TECHNOLOGY

RESEARCH PROGRAM

FOR

ELECTROCHEMICAL ENERGY STORAGE

ANNUAL REPORT

FOR 1998

Environmental Energy Technologies Division

Lawrence Berkeley National Laboratory

University of California

Berkeley, California 94720

Edited by Kim Kinoshita, Technical Manager

June 1999

This work was supportedby the AssistantSecretaryfor EnergyEl%ciencyand RenewableEnergy,OffIceof TransportationTechnologies,OffIceof AdvancedAutomotiveTechnologiesof the U.S. Departmentof

EnergyunderContractNo. DE-AC03-76SFOO098.

—.— .— ———.———-——- . . . -- -.. . . .. . . ... .. ----- . . ..

LI

DISCLAIMER

This repon was prepared as an account of work sponsoredbyanagency of the United States Government. Neitherthe United States Government nor any agency thereof, norany of their employees, make any warranty, express orimplied, or assumes any legal liability or responsibility forthe accuracy, completeness, or usefulness of anyinformation, apparatus, product, or process disclosed, orrepresents that its use would not infringe privately ownedrights. Reference herein to any specific commercialproduct, process, or service by trade name, trademark,manufacturer, or otherwise does not necessarily constituteor imply its endorsement, recommendation, or favoring bythe United States Government or any agency thereof. Theviews and opinions of authors expressed herein do notnecessarily state or reflect those of the United StatesGovernment or any agency thereof.

DISCLAIMER

Portions of this document may be illegiblein electronic image products. Images areproduced from the best available originaldocument.

CONTENTS

INTRODUCTION ....................................................................................................................................... 1

msEmcHPRoJEcTsuMMmEs ........................................................................................................2

ELECTRODE CHARACTERIZATION

Carbon Electrochemistry ................................................................................................................. 2

K. Kinoshita (Lawrence Berkeley National L.uboratory)

Fabrication and Testing of Carbon Electrodes as Lithium-Intercalation Anodes ........................... 5

T.D. Tran (Lawrence Livermore National Laboratory)

Safety and Reactivity of Carbonaceous Anodes Used in for Lithium-Ion Technology ..................6

M.D. Curtis and G.A. Nazri (University of Michigan and GM Research and DevelopmentCenter)

Battery Materials: Structure and Chmactefization ......................................................................... 9

J. iWcBreen (Brookhaven National Laboratoq)

ELECTRODES FOR AQUEOUS ELECTROCHEMICAL CELLS

Preparation of Improved, Low-Cost Metal Hydride Electrodes for Automotive Applications ....11J.J. Reilly (Brookhaven National Laboratory)

Optimization of Metal Hydride Properties in MWNiOOH Cells for Electric Vehicles ................14R.E. White (UniversiQ of South Carolina)

Microstructural Modeling of Highly Porous MWNiOOH Battery Substrates ............................. 16

AM Sastry (University of Michigan)

COMPONENTS FOR NONAQUEOUS CELLS

Novel Lithium/Polymer-Electrolyte Cells .................................................................................... 19

E.J. Cairns and F.R. McLamon (Lawrence Berkeley National Laboratory)

New Cathode Materials ................................................................................................................. 21

A4.S. Whittingham (State University of New York at Binghamton)

...m

..,. ..... - .— -- ,. -—-—.. ..— .—.—. ,,...—— ?-.--” . . . . . . -, . . . . . -,s.---. - .-.-,. .4...- -. . - . . ..-..-— . . . . -,

. , ...A -

Solid Electrolytes .......................................................................................................................... 23

L.C. De Jonghe (Lawrence Berkeley National Laboratory)

Polymer Electrolyte Synthesis for High.Power Batteries .............................................................27

J.B. Kerr (Lawrence Berkeley National Laboratory)

Composite Polymer Electrolytes for Use in Lithium and Lithium-Ion Batteries ......................... 29

S. Khan and P.S. Fedkiw (North Carolina State Universi@)

Polymer Electrolytes for Ambient-Temperature Traction Batteries: Molecular-Level

Modeling for Conductivity Optimization ................................................................................ 31

A4,A. Ratner (Northwestern University)

Corrosion of Current Collectors in Rechargeable Lithium Batteries ............................................ 33

J. W. Evans (Luwrence Berkeley National Laboratory/Universi~ of Cal#omia at Berkeley) i

Development of Novel Chloroaluminate Electrolytes for High-Energy-Density

Rechargeable Lithium Batteries .............................................................................................. 34K.A. Wheeler (Delaware State Universi@)

CROSS-CUTTING RESEARCH I

Analysis and Simulation of Electrochemical Systems .................................................................. 36

J.S. Newman (Luwrence Berkeley National Laboratory/University of California at Berkeley)

Electrode Surface Layers ............................................................................................................... 39

F.R. McLamon (Lawrence Berkeley National Laboratory)

Lithium Electrode Interracial Studies ...........................................................................................42

P.N. Ross, Jr. (Lawrence Berkeley National Laboratory)

ACKNO WLEDGA4ENTS .......................................................................................................................... 44 ~

ACRONYMS .............................................................................................................................................. 44 ~

LIST OF PRIOR ANNUAL REPORTS ..................................................................................................... 4tj ~

FINmcIAL DA TA ....................................................................................................................................47 ~

iv I

INTRODUCTION

The U.S. Department of Energy’s (DOE) Office of Advanced Automotive Technologies conducts

research and development on advanced rechargeable batteries for application in electric vehicles (EVS)

and hybrid systems. Efforts are focused on advanced batteries that offer the potential for high

performance and low life-cycle costs, both of which are necessary to permit significant penetration into

commercial markets.

DOE battery R&D supports two major programs: the United States Advanced

(USABC), which develops advanced batteries for EVS, and the Partnership for a

Battery Consortium

New Generation of

Vehicles (PNGV), which seeks to develop passenger vehicles with a fuel economy equivalent to 80 mpg

of gasoline. This report describes the activities of the Exploratory Technology Research (ETR)

Program, managed by the Lawrence Berkeley National Laboratory* (LBNL). The role of the ETR

Program is to perform supporting research on the advanced battery systems under development by the

USABC and PNGV Programs, and to evaluate new systems with potentially superior performance,

durability and/or cost characteristics. The specific goal of the ETR Program is to identify the most

promising electrochemical technologies and transfer them to the USABC, the battery industry and/or

other Government agencies for further development and scale-up. This report summarizes the research,

financial and management activities relevant to

program, and reports for prior years have been

summary.

the ETR Program in CY 1998. This is a continuing

published; they are listed at the end of this Program

* Participants in the ETR Program include the following LBNL saentists: E. Cairns, J. Evans, J. Kerr, K. Kinoshita,F. McLarnon and J. Newman of the Environmental Energy Technologies Division; and L. De Jonghe and P. Ross of theMaterials Sciences Division.

1

.,-.

RESEARCH PROJECT SUMMARIES

ELECTRODE CHARACTERIZATION

Characterization of electrode morphology and chemical composition are important for the successfuldevelopment of rechargeable electrodes for advanced secondary batteries. Efforts are underway toutilize advanced fabrication techniques and spectroscopy to characterize electrode properties.

Carbon Electrochemistry

Kim Kinoshita90-1142, Lawrence Berkeley National L.uborato~, Berkeley CA 94720(510) 486-7389; fm: (510) 486-4260; e-mail: k_kinoshita@lbl.gov

Objective

● Identify the critical parameters that control the reversible intercalation of Li in carbonaceousmaterials and determine their maximum capacity for Li intercalation.

Approach

“ Couple electrochemical studies with physical measurements to correlate the relationshipbetween the physiochemical properties of carbonaceous materials and their ability tointercalated Li.

● Apply thermal analysis and microscopy techniques to characterize the structure of carbonmaterials for Li-ion batteries.

Accomplishments

● Thermal analysis (TGA and DTA) involving air oxidation of carbons and electrochemicalG

studies suggest that the active surface area, and not the total BET surface area, has aninfluence on the irreversible capacity loss of carbons for Li-ion batteries.

Future Directions

“ Continue collaboration with Superior Graphite and University of Michigan to identify a low-cost carbon with improved rate capability in Li-ion batteries.

● Continue in situ ellipsometry studies of lithiated carbons, and determine the relationshipbetween the type of carbon and the properties of the surface layer formed during the firstcharge cvcle.

The objective of this project is to identifythe critical parameters that control the reversibleintercalation of Li in carbonaceous materials.This project involves investigations of the roleof physiochemical properties of carbonaceousmaterials on their ability to reversibly intercalatedLi. This latter effort is coordinated with theresearch conducted at University of Michigan toevaluate the reactions that occur duringelectrolyte decomposition at carbon electrodes

in Li-ion batteries (see discussion in “Reactivityand Safety Aspects of Carbonaceous AnodesUsed in Lithium-Ion Batteries”).

The collaboration with Superior Graphite(Chicago) continued with a focus to evaluatealternative graphitized carbons for the negativeelectrodes in Li-ion cells. Several meetingswere held with the University of Michigan andSuperior Graphite to identify the most-suitablephysical properties to pursue in the manufacture

2

of improved carbons for higher charge/discharge rate capability. Based on thesediscussions, the desired properties of the carboninclude: (i) graphitized material, (ii) more-spherical particle morphology, (iii) final averageparticle size of 12 microns, and (iv) surface areaof <15 mVg. Several samples were receivedfrom Superior Graphite for physicalcharacterization. XRD analysis indicated thed(O02) spacing is -3.36 ~, typical of a highlygraphitized material. Transmission electronmicroscopy (TEM) examination suggests apreponderance of 2-dimensional structure (i.e.,flake-like) rather than the desired. 3-dimensionalstructure (i.e., spherical). Discussions are inprogress with Superior Graphite to evaluatetheir processing method.

A collaborative study was undertaken withHydroQuebec (HQ) to investigate the propertiesof mesocarbon rnicrobeads (MCMBS) fromOsaka Gas (Japan). The goal of this study wasto investigate the properties of MCMBS afterheat treatment (i.e., 700, 1000, 1200, 1800,2400and 2750”C). Heat treatment was conducted ina graphitizing furnace, and physicalcharacterization involving thermal analysis andTEM were performed at LBNL. The initialMCMBS, which are obtained by processing tarpitch at low temperatures, appeared to consist ofmainly spherical particles (e30-microndiameter). Heat treatment in an inertatmosphere produced a substantial weight loss(-12%) and agglomeration of the particles.These results suggest that the initial MCMBScontained significant fractions of hydrogen andoxygen, indicative on a material that is notcompletely carbonized. In addition, heattreatment produced a significant change inparticle morphology because of agglomeration.The electrochemical evaluation of thesematerials in Li-ion cells is currently underway atHQ.

The collaboration with the University ofMichigan involves analysis of the role of carbonmorphology on the irreversible capacity loss andelectrolyte decomposition in Li-ion cells.Thermal analysis (TGA and DTA) involving airoxidation of carbons suggested that active sitesare present which can be correlated to thecrystallographic parameters, La and LC, and the

3

d(O02) spacing. This finding was extended todetermine the relationship between active siteson carbon and their role in catalyzing electrolytedecomposition leading to irreversible capacityloss (ICL). Electrochemical data from thisstudy with graphitizable carbons and frompublished literature were analyzed to determinethe relationship between the physical propertiesof carbon and the irreversible capacity lossduring the initial charge/discharge cycles.Figure 1 shows the relationship between the ICLand La for carbons examined in this study and

from other research groups. Despite the scatterin the plot, the trend shows that ICL is inverselyproportional to Lw Based on this analysis, we

conclude that the active surface area, and not thetotal BET surface area, has an influence on theirreversible capacity loss of carbons for Li-ionbatteries. This conclusion suggests that thecarbon surface structure has a significant role incatalyzing electrolyte decomposition.

PUBLICATIONS

F.

J.

Kong, J. Kim, X. Song, M. Inaba, K.Kinoshita and F. McLarnon, “ExploratoryStudies of the Carbon/NonaqueousElectrolyte Interface by Electrochemical andIn Situ Ellipsometry Measurements,”Electrochem. and Solid-State Lett., 1, 39(1998).

Kim, F. Kong, X.Y. Song, M. Inaba, K.Kinoshita an~F. McLarno;, “Studies of theCarbon/Nonaqueous Electrolyte Interface byElectrochemistry, In Situ Ellipsometry andImpedance Spectroscopy;’ 193rd Meeting ofthe Electrochemical Society, San Diego, CA,May 3-8, 1998.

T.D. Tran, D.J. Derwin, P. Zaleski, X. Song andK. Kinoshita, “Effects of Heat Treatmentand Morphology of Petroleum Cokes onLithium Intercalation,” 193rd Meeting of theElectrochemical Society, San Diego, CA,May 3-8, 1998.

G. Nazri, B.KinoshitaReactivityLithium

Yebka, M. Nazri, D. Curtis, K.and D. Derwin, “Safety andof Carbonaceous Anode in

Battery Technology;’ 193rd

--,.= -- -. --..,.

Meeting of the Electrochemical Society, San Electrodes,” 194th Meeting of TheDiego, CA, May 3-8, 1998. Electrochemical Society, Boston, MA,

November 1-6, 1998.T.D. Tran, D.J. Derwin, P. Zaleski, X. Song and

K. Kinoshita, “Lithium Intercalation Studies T.D. Tran, X. Song and K. Kinoshita,of Petroleum Cokes of DifferentMorphologies,” 9th International Meetingon Lithium Batteries, Edinburgh, UK, July12-17, 1998.

T.D. Tran and K. Kinoshita, “Carbon for

M,

Supercapacitors and Li-Ion Batteries in G.EVS,” 194th Meeting of TheElectrochemical Society, Boston, MA,November 1-6, 1998.

Inaba, R. Kostecki, K. Kinoshita and F.McLarnon, “Ti-, Zr-, Nb- and Me-dopedLithium Manganese Oxide Positive

“Investigation of Lithiated Carbons byTransmission Electron Microscopy and X-Ray Diffraction Analysis,” MaterialsResearch Society Fall Meeting, Boston, MA,November 30 –December 4, 1998.

Nazri, B. Yebka, M. Nazri, D. Curtis, K.Kinoshita and D. Derwin, “Safety andReactivity of Carbonaceous Anode inLithium-Ion Batteries;’ Materials ResearchSociety Fall Meeting, Boston, MA,November 30 –December 4, 1998.

Soo ~ , , I , , , I 1 , I , , r I ,

Iv

700

600

● This study (coke)o This study (graphite)v Chung (S. Korea)v Zheng (Canada)A Jung (S. Korea)x Xiang (China)A Jung (S. Korea)+ Tatsumi (Japan)o Iijima (Japan)

100 :@o

Q fo I ?+, t I I * -h +“ I , , t I , , t -1

0

-o 200 400 600 800 1000

La (~)

Figure 1. Composite plot of ICL vs. L, for different types of carbonelectrodes. Line indicates trend of the data.

Fabrication and Testing of Carbon Electrodes as Lithium-Intercalation Anodes

Tri D. TranL282, LawrenceLivermoreNationalLaboratory,P.O. Box 808,LivermoreCA 94550(925)422-0915;fm: (925)424-3281;e-mail: tran4@llnl.gov

Objectives

● Evaluate the performance of carbonaceous materials as hosts for Li-intercalation negativeelectrodes.

● Develop reversible Li-intercalation negative electrodes for advanced rechargeable Li-ioncells.

Approach

● Fabricate electrodes from various commercial carbons and graphites and evaluate in small Li-ion cells.

“ Correlate electrode performance (i.e., capacity, irreversible capacity) with carbon structureand properties in collaboration with LBNL.

Accomplishments

● The Li intercalation capacity of Santa Maria coke (x = 0.83) was found to be among thehighest observed for untreated coke materials.

Future Direction

● Project is complete.

Several types of heat-treated cokes fromSuperior Graphite Company were investigatedas electrode materials for Li+ intercalation. Thecokes include fluid coke, pitch coke, SantaMaria coke, electrode coke and nipple coke.These commercial cokes are generally heattreated (calcined) to a maximum temperature ofabout 1400”C to remove volatile components.The electrodes containing these cokes wereprepared according to a standard procedure thatuses pyrolyzed phenolic resin as the binder. Theelectrolyte is 0.5 M LiN(CFJSOz)z, (3M Corp) ina 50:50 mixture of ethylene carbonate anddimethyl carbonate (Grant Chemical).Intercalation experiments were carried out at theC/24 rate using 3-electrode set-ups with Li foilsas both counter and reference electrodes.

The cokes were milled and then sieved to

average particle sizes of 10-25 ~m. The air-milling technique and the particle morphologiesallow for preparation of rather smooth surfacesresulting in low-surface-area powders. Thereversible Li intercalation capacities of these

cokes are in the range of 179 to 310 rnAh/g,corresponding to x values of 0.48 to 0.83 (inLiXC,). The reversible capacity for Santa Mariacokes is considerably larger than that for atypical coke. The voltage profiles of allmaterials show amorphous-like behavior withgradual changes in voltage as the intercalation/deintercalation proceeds. TEMs of severalsamples showed disordered crystallinestructures characteristic of cokes.

The cokes were heat treated at severaltemperatures between 1400 and -2800”C tostudy the effect of heat treatment on themorphology, microstructure and subsequentlythe Li intercalation behavior. The fluid cokeand the Santa Maria coke were selected becauseof their initial amorphous structure and theirparticle morphology, which are more globular innature. In contrast, needle-like cokes such asnipple coke or electrode coke tend to have apreferred structural orientation. It is expectedthat more-spherical particles with graphiticstructures derived from the fluid-type cokes can

lead to higher performances, particularly interms of rate and particle size distribution.Successive heat treatment of the fluid coke at1800 and 2300”C reduced the BET surfaceareas, and correspondingly reduced theirreversible capacity loss. The reversiblecapacity appears to pass through a minimumover this range but the performance for samplesheat treated at 2300”C is only x = 0.55, wellbelow that of graphitized needle-coke that hadbeen treated at a similar temperature. The sametrend was observed for another fluid-type coke(Santa Maria coke). The capacity for SantaMaria coke is among the highest observed foruntreated coke materials. The potential profilesof these materials exhibit gradual changes withLi concentration, consistent with those fromamorphous compounds. TEM observationshows evidence of ordered structure but thed(002)o ,P,Cin,of fluid coke heated to 2300”C(3.39 A) suggests only partial graphitization.

PUBLICATIONS

T.D. Tran and K. Kinoshita, “SurfaceModifications to Improve Performance ofCarbon Lithium Intercalation Anodes,” U.S.

Safety and Reactivity of Carbonaceous

M. David Curtis and Gholarn-Abbas Nazri*

Patent Application submitted August 31,1998

T.D. Tran, D.J. Derwin, P. Zaleski, X. Song andK. Kinoshita “Effects of Heat Treatment andMorphology of Petroleum Cokes on LithiumIntercalation,” 193rd Meeting of theElectrochemical Society, San Diego, CA,May 3-8, 1998.

T.D. Tran, D.J. Derwin, P. Zaleski, X. Song andK. Kinoshita, “Lithium Intercalation Studiesof Petroleum Cokes of DifferentMorphologies,” 9th International Meetingon Lithium Batteries, Edinburgh, UK, July12-17, 1998.

T.D. Tran and K. Kinoshita, “Carbon forSupercapacitors and Li-Ion Batteries inEVS,” 194th Meeting of TheElectrochemical Society, Boston, MA,November 1-6, 1998.

T.D. Tran, X. Song and K. Kinoshita,“Investigation of Lithiated Carbons byTransmission Electron Microscopy and X-Ray Diffraction Analysis,” MaterialsResearch Society Fall Meeting, Boston, MA,November 30 –December 4, 1998.

Anodes Used in Lithium-Ion Technology

Department of Chemistry, University of Michigan, Ann Arbor MI 48109-1055*Physics and Physical Chemistq Department, GM Research and Development Center, Warren, MI 48090 !.

(734) 763-2132; fu: (734) 763-2307; e-mail: mdcurtis@umich. edu /gnazri@cmsa.gmr. com

Objective

● Evaluatein Li-ion

Approach

safety, reactivity, and performances of various carbonaceous anodes for applicationbattery.

● Design and fabricate electrodes and cells from various carbonaceous materials, and correlateelectrochemical performances with structural parameters of carbon, in order to develop anaccelerated screening strategy for anode materials.

● Evaluate reactivity of various carbonaceous electrodes toward electrolyte decomposition, gasgeneration, and surface film formation.

● Evaluate thermal stability and the nature of surface films formed on carbonaceous electrodesduring operation of Li-ion batteries.

6

Accomplishments

● Novel electrode fabrication methods mdcelldesigns have been used, wtichdlow red-timeanalysis of gaseous species generated in Li-ion cells during charge-discharge cycles.

● Carbonaceous electrodes with fewer edge sites and extended basal plane are less reactive andsafer to use in Li-ion technology.

“ Carbonaceous electrodes with fewer surface functional groups, trapped and adsorbedimpurities, and defect sites are less reactive and safer to be used in Li-ion systems.Functional groups and impurities can be removed at 1000°C. However, annealing of defectsites requires heat treatment up to 2800”C in the presence of soft carbon.

● A more-compact and more-stable surface film formed on a carbonaceous electrode when highinitial charge-discharge rates were applied. The thickness of the surface film grows to a limitbeyond which electron-tunneling process between the anode and electrolyte solventmolecules is no longer possible.

Future Directions

● Improve the safety of carbonaceous anode by surface modification and surface coating toavoid electrolyte decomposition and gas generation.

● Optimize the performance of carbonaceous anodes for high-power and high-energyapplications.

The objective of this project is to evaluatethe influence of carbon structural parameterstoward electrolyte decomposition, gasgeneration, and surface film formation duringinitial charge-discharge cycling, in order todesign a low-cost high-performance electrodefor application in Li-ion batteries. Althoughcarbonaceous anodes are very desirable for Li-ion technology, they have two major problemsthat require special research attention:1) excessive electrolyte decomposition duringthe first lithiation process, and 2) thegeneration of gaseous species. The gases aredetrimental to the performance of Li cells andthey are considered a safety hazard, as themajority of the gaseous species formed areflammable; also some of the gases may alsodissolve in organic electrolyte and participatein parasitic reactions.

A unique electrode fabrication method andcell have been designed to examine thereactivity of various carbonaceous electrodestoward electrolyte decomposition. Figure 2shows the cell designed for in situ gas analysis.

“)Qukkmwd(lOxlOcm-It

ToMSRI/

Figure 2. Schematic of cell used for gasanalysis during charge-discharge cycling.

As shown in this work, the degree ofgraphitization of the carbonaceous electrodeand structural parameters of graphite crystalliteplay critical roles in the safety of Li-ionbatteries. A major effort was directed tounderstand the relationship betweenphysiochemical properties of carbonaceousmaterials and their electrochemicalperformances. We found that the degree ofcrystallinity of the carbonaceous materials andthe nature of electrolyte used are importantfactors for the electrochemical stability of Li-ion cells. The nature of gases generated duringthe initial Li insertiotiextraction process

7

. . ..—— . . ..— ——— —.-.. >. . . . ..- .-.--F.,...... ... ., -. - ..- ---

depends on electrolyte compositions and thenature of the carbon electrodes. We reportresults for a matrix of experiments, designed tostudy the electrochemical behavior of severalcombinations of carbonaceous electrodes invarious carbonate-based electrolytes.Electrolytes and solvents used are given inTable 1.

Table 1. Solvents and electrolytes usedEC DEC DMC

EC- LiPF6 DEC-EC DMC-ECEC-DEC- DEC-LiPF, DMC-DECLiPF,EC-DMC- DEC-DMC- DMC-LiPF,LiPF LiPF,EC-;C-LiPF, DEC-PC-LiPF, DMC-PC-LiPF,

We have also designed a unique diagnostic toolconsisting of mass spectrometry and gaschromatography, to measure and identifygaseous species generated in Li-ion cells duringcharge-discharge cycling. The volume andcomposition of gaseous species generated on 18different carbonaceous anodes in differentelectrolytes were recorded. We found twoclasses of gaseous species forming oncarbonaceous electrodes; 1) the first class isusually generated before the electrolytedecomposition potential is reached, and isrelated to adsorbed and trapped impurities incarbon electrode, and 2) the second classincludes gaseous species generated at about 0.8volt vs. Li due to electrolyte decomposition; gascompositions are very sensitive to theelectrolyte composition. The composition ofmajor gaseous species detected during the initiallithiation of carbonaceous anodes in rnixed-carbonate solvents are CHd, CzHd, CO, andminimal amount of CzHb, H2, and COZ.However, when PC-containing electrolytes wereused, the major gas components were C~Hc andCO. A significant amount of hydrogen wasevolved when a trapped impurity such asmoisture was present. We reported a completetable of gas compositions vs. the nature ofgraphites and electrolytes in our previous report.In order to cross check our results, we have nowmeasured the Hz, CO and C02 amounts withspecially designed gas chromatographyequipped with sensitive columns for thedetection of H,, CO, CO,.

In order to correlate the structuralparameters of carbonaceous anodes to theirperformances we studied the morphology andcrystal structure of variety of carbonaceousmaterials using SEM. The crystallite parametersand the degree of graphitization were measuredby X-ray diffraction and Raman spectroscopyrespectively. Surface area of carbonaceouselectrode was measured using the BETtechnique. A correlation between the crystalliteparameters in terms of degree of disorder andsize of basal plane, La, vs. irreversible capacityloss and total volume of gas generated duringthe initial lithiation process is shown in Fig 3.

70

G

~ ‘0

d~ 50

40 : ...

c

? +— ICL (mAh/g)

“,,‘.

\. . .

“W-””Gas volu~ 430

25

30~ ,0.- .. .,

900 .’. ”

-:800 : ‘.

.,::.

700 : ‘,.,,,.,

~ 600 :‘.,

f

.!,4“ “..

500 :

400 -

300 -

1 I I I I200 u m 4 ~ w 0.2

Ac) 3? i g

m u5

Figure 3. Correlation between structuralparameters of carbonaceous anodes in terms ofdegree of disorder, R, and the size of basalplane, La, vs. irreversible capacity loss, and totalvolume of gas generated during initial Iithiationprocess for several carbonaceous electrodes.

(.

8 I

In order to develop a practical low-cost andhigh-performance anode, graphitic electrodeswith low-edge surface area and a high degree ofcrystallinity should be considered. Thegraphitic materials provide significantly longercycle life, higher volumetric energy density anda high-quality flat electrode potential profile(300-50 mV). The reactivity of the graphiticelectrode can be significantly improved byadjusting the crystallite parameters, mainly byreducing the edge and defect sites containingsp3 carbons and removing adsorbed and trappedimpurities by heat treatment. Furthermodifications of graphitic electrodes require

surface modification and surface coating whichare the subject of our future investigations.

REFERENCES

G.A. Nazri, B. Yebka, M.D. Curtis, K.Kinoshita, D. Derwine, ECS, ProceedingsVolume, Boston, Spring 1997. 194thMeeting of the Electrochemical Society,Boston, MA, November 1-6, 1998.

G.A. Naqi, B. Yebka, M.D. Curtis, K.Kinoshita, D. Derwine, MRS ProceedingsVolume 465, 1999.

Battery Materials: Structure and CharacterizationJames A4cBreenBrookhavenNationalLaboratory,DAS-Bldg.480, P.O. Box 5000, UptonNY 11973-5000(516)344-4513,fm: (516)344-4071; e-mail: jmcbreen@bnl.~ov

Objective

s Elucidate the molecular aspects of battery materials and processes by in situ high-resolutionX-ray diffraction (XRD) and X-ray absorption (xAS).

Approach

● Apply in situ XRD and XAS to study low-cost LiXNi02 and LiXM~O, based cathodes, andtin-based composite oxide anodes.

Accomplishments

● In situ XRD studies were completed on LiXNi02, LixCo02 and LiXNi,.YCoY02cathodes inrechargeable Li cells.

● In situ XAS was used to study the charge/discharge processes in a tin-based composite oxideo ).anode (snB0.56po.4M0.4 3.47

Future Direction

● Utilize in situ XRD and XAS to study cycling behavior and failure mechanisms of Liintercalation electrodes.

● Develop diagnostic methods to study Li-ion battery materials that operate at high-rate chargeand discharge.

The objective of this project is to elucidate in situ high-resolution XRD and EXAFSthe molecular aspects of materials and electrode studies on both anodes and cathodes for Li-ionprocesses in Li-ion batteries, and to use fhis batteries.information to develop electrode and electrolyte X-ray Absorption Studies of Tin-Basedstructures with low cost, good performance, and Composite Oxides (TCO): A tin-basedlong life. Work during the year included both composite oxide (TCO) material with a nominal

9

composition of SnBo.5Gp0.4Alo.403.47was studiedby in situ XAS. During the first charge cycle(Li insertion) the electrode accepted a charge of708 mAh/g which corresponds to a LdSn ratioof 5.4. On the subsequent discharge thecapacity was 424 mAh/g. Results obtained withthe as-prepared material confirmed theamorphous nature of the TCO material andshow that Sn in TCO is coordinated with threeoxygen atoms at a distance of 2.12 A. Uponcharging, Li initially interacts with theelectrochernically active Sri-O center to formmetallic Sn. The removal of oxygen from theSn was quantitative. Upon further chmge, Lireacts with Sn forming initially LHn and thenLi7Sn3 at a Li/Sn ratio of 5.4. Figure 4 showsthe Fourier transform of the Sn K edge EXAFSduring first charge (i.e., Li insertion into theoxide).

The TCO material is amorphous anddisplays only a Sri-O interaction. Upondischarging (Li removal), metallic Sn isproduced with a Sn-Sn distance intermediate tothose of gray and white Sn. The coordinationnumber is also low (<3). This indicates that Snexists in a highly dispersed state in a matrix ofLi20. After complete discharge there is noevidence of Sri-O interactions.

I dTcO

0.94 [Lusn)

2

W.L

----

G ,,,... too (Ulsn)3 ... ... - -

.=c ,, \t#

;;-’$, Z.si {1.usn), ‘, ;, ~ . . . . . . . .,,

z, ,,,, :! .c.)i 4.f4 (Lus”)

k J , ,.. !1 \ —.—-... ,’ ,>:..j .,. .) :, , ~Y~!

~

5.35 (U/sn)-. . . . .

. ‘,i, \? ,“.

o-=+%S. +}. : , ‘ ‘~,: + ,-?,

o 1 2 3

Radial Distance (~)

Figure 4. Fourier transform of the Sn K edgeEXAFS during first charge of the TCOelectrode.

X-ray Diffraction Studies (XRD) ofLiNiOz, LiCoOz and LiNil.XCoXOz: In situXRD was carried out on cells with LiNiOz,LiCoOz and LiNil.XCoXOz (x = 0.1, 0.2, 0.3)cathodes that were charged at the C/10 rate.

Cathodes were fabricated by casting a mixtureof the cathode material (LiNil-XCoX02, LiCo02,or LiNi02), acetylene black and a binder (PVdF)on an aluminum foil sheet. Disk electrodes(2.82 cm’) were punched from the sheet andassembled in the cell with a Celgard separatorand a Li foil anode. The electrolyte was 1 MLiPF6 in solvent mixture of 1EC:3EMC. ]?3situXRD on Lil-XNiOz, during charge, in the 3.8 to5.1 V region, revealed the formation of threehexagonal phases. In a similar voltage rangeLil-XCo02 showed three CdC12-type hexagonalphases and one Cd12-type phase. Unlike reportsof several other investigators, no monoclinicphases were observed. In the case of Lil-XNi02no phase change was detected until x = 0.4;whereas a phase change is detected in Li 1-XC002as early as x = 0.1. Substitution of even smallamounts (x = 0.1 ) of Co in Li 1-XNi02completelychanges the phase behavior. The most obviousdifference is that first phase change appears at x~ 0.1. Figure 5 shows a set of XRD spectraduring charge of LiNi02.

-----ill!’! } 1 4

kit’WI-+’-13,.,

1314151617283032343638

20 20

XRD spectra during charge ofFigure 5.LiNi02. Scan 1 was recorded at the beginningof charge and scan 14 at the end of charge.

10

PUBLICATIONS

Y. Ein-Eli, S.H. Lu, W.F. Howard, S. Mukerjee,J. McBreen, J.T. Vaughey, and M.M. Y.Thackeray, “LiMnz.XCuXOA(O.1< x <.5): ANew Class of 5 V Cathode Materials for LiBatteries. I. Electrochemical, Structural andSpectroscopic Studies:’ J. Electrochem. Sot.145, 1238 (1998).

J. McBreen, S. Mukerjee and X.Q. Yang, “InSitu X-ray Studies of Battery and Fuel Cell

Materials”, Synchrotrons Radiation News,11(3), 18 (1998).

Ein-Eli, S.H. Lu, M.A. Rzeznik, S.Mukerjee, X.Q. Yang, and J. McBreen,“LiMnz.XCuXOq (O.1 < x < 0.5): A NewClass of 5 V Cathode Materials for LiBatteries. H. In Situ Measurements,” J.Electrochem. Soc 145,3383 (1998).

ELECTRODES FOR AQUEOUS ELECTROCHEMICAL CELLS

Electrodes are being developed to identify low-cost metal hydrides for MH/NiOOH cells andimproved metal oxides for electrochemical capacitors.

Preparation of Improved, Low-Cost Metal Hydride Electrodes for Automotive Applications

James J. ReillyBrookhavenNationalLaboratory,DAS-Bldg.815, P.O.Box 5000, UptonNY 11973-5000(516)344-4502,fu: (516)344-4071;e-mail: reilly@bnl.gov

Objectives -

“ Determine the properties and behavior of metal hydride electrodes with a view towardsimproving their lifetime and storage capacity, and decreasing their costs.

● Determine structure of complex AB, electrode alloys.

● Create a realistic mathematical model of the behavior of the MHXelectrode.

Approach

“ Determine thermodynamic properties of alloy hydrides used in electrodes.

● Fabricate and test hydride electrodes.

● Employ and develop X-ray absorption spectroscopy (xAS) methods to determine theelectronic structure, the local atomic environment and corrosion of alloy hydride electrodematerials.

Q Employ neutron diffraction to determine crystal structure.

Accomplishments

● Determined the crystal structure of L~i,,,CoO,,,MnO,M.,DX and La(Ni,Sn),+x alloys.

● Demonstrated that cobalt-free La(Ni,Sn),+X electrodes have storage capacity and cycle lifeequal to or better than the commercial alloy MmNi~.=CoO#U~MnO,q

Future Direction

● Project was completed.

11

.. . . . .-——— —..———- ——— - .--—— .--. .-—.. . .. . -. . --—.. — ... .. .“—. ...

Crystal Structure of La(’8Ni0.~7cc2Ni0.cx)~3~Co0.7~~0.~~AIO~Dd.7J.The deuteridealloy, La(’*Ni0,,,’2Ni0,62~,,5,Co0,,,Mn0,,M0,,0D,,,,was prepared and the hexagonal space groupP6/mmm (#191) was used as a model todetermine the crystal structure. The sitepreference of cobalt to occupy the 3g over the2C sites was independently confirmed withintwo standard deviations from the value obtainedfor the uncharged alloy. Again, Al and Mn weredetermined to exclusively occupy the 3g sites.Four of the possible five deuterium sites werefound to be occupied in the following order: the12n site with n D atoms = 2.28(3) (D(3)), the6m site with n = 1.70(3) (D(2)), the 120 sitewith n = 0.55(2) (D(1)) and the 4h site with n =O.17(1) (D(4)). This leads to a stoichiometry ofAB,D,,O and corresponds to a lattice expansionof 17.6%. After the experiment the D contentwas determined via thermal decomposition to be4.73 D atoms per formula unit, in excellentagreement with that obtained from thecrystallographic refinement.

Crystal Structure of Non- StoichiometricLa(Ni,Sn),+x Alloys. The La(Ni,Sn),+X alloyswere prepared in an arc furnace under inert gasand then annealed for three days at 1223 K.One alloy, E564, was obtained from acommercial vendor, Tribacher AuermetProductions, GmbH. X-ray diffraction analysisindicated that the alloys were single phase. Atotal of six alloys (Table 1) were examined, fourof which were La(Ni,Sn)~+X alloys. Thecomposition is indicated by weight, the AB~stoichiometry normalized with respect to therare earth component(s) and the structuralformula as determined from the crystalstructure. High-resolution X-ray powderdiffraction patterns were collected at theNational Synchrotrons Light Source at BNL.

The X–ray Diffraction Pattern ofLaNi,USno,,. E594 indicates that the sample ishighly crystalline. The P6/mmm space group ofstoichiometric LaNi~ was preserved, as in thecase of La(Ni,Cu)~X. The B(3,3) atomicdisplacement parameter indicates a largedisplacement of La along the c axis compared tothe in-plane displacement. Aligning Ni

dumbbells along the c axis in a random fashionwill lead to this type of displacement parameterfor the disordered dumbbell A-sites. Therefinement also indicates that the electrondensity of the B-site in the basal plane (2c) canbe fully accounted for by Ni occupation.Placing Sn on this site would lead to vacancies,which in view of the stoichiometry is not verylikely. There is a substantial increase in the caxis and a decrease of the a axis of LaNid ~Sno~7(E594) relative to the stoichiometric LaNi,,SnO~(E492). The contraction of the a axis indicatesthat the bigger Sn atom (atomic radius of 163pm) does not occupy the basal plane 2C sites.The mid-plane 3g sites are occupied by Ni andall of the Sn. No indications of long-rangesuperstructures could be found. The crystalstructures of the other La-deficient alloys (E600,E601) were also determined and were found tosimilarly substitute Ni dumbbells on vacant Asites. The structure of the stoichiometric alloy,LaNi,,SnO, (E492) was also determined and Snwas found to exclusively occupy the 3g site.

Cycle Life of La(Ni,Sn),+X Alloys. TheLa(Ni,Sn),+x alloys were fabricated intoelectrodes containing 0.075 g of active materialand cycled in an open cell containing 6 M KOH.The counter electrode was Pt and the referenceelectrode was Hg/HgO. The electrode wasactivated in situ via successive electrochemicalcharge and discharge cycles.

The results of the cycle tests are summarizedin Table 2. The corrosion rate is based on theslope of the linear portion of the capacity vs.cycle life plots divided by Q~aX.The number ofhydrogen atoms/unit cell, n, is calculated fromthe maximum uptake, Q~U. Note that as thenumber of Ni dumbbell sites increase on the Asublattice, as indicated by the structuralformulas, the corrosion rate decreases with onlya small penalty in storage capacity. Finallywhen 4.4590 of the A sites are occupied by Nidumbbells the storage capacity exceeds that ofthe commercial alloy (E564) while the corrosionrate is less.

12

TABLE 1. ALLOY COMPOSITIONS

Element Ratio Normalized Exp. # Structural FormulaStoichiometry

‘%92Ni4611sn0,Wl LaNi,MSno31 E600 L%,,,,(Ni.~,),Ni. ~~,,sn.,,,,

LA,,Ni.wSnO,0 LaNi484Sn0,2 E594 L%mg,(Ni.mJNi,sm s%,,

LA 9Ni4 msnn m LaNi4 ,,Snn,, E601 L~gR,@Jioo1w)2Ni,,,2,Snn,~,

LaNi4,0Sn,)3,) LaNi,,nSn,,,o E492 LaNi.,nSnn,()

MmNil ,,Co0,,Mn04Alnq, MmNil ,,Con,,MnndAln,. E564 MmNi, ,,Con,,MnOdAln,.

LaNi4 ,Mnn @10, LaNi4,Mn0,Aln, E132 LaNi.lMnn,@n,

*Mm refers to mischrnetal with a composition 30 wt~o La, 48$Z0Ce, 1590 Nd and 6~o Pr.

TABLE 2. ELECTRODE PERFORMANCE

Structural FormulaExp.# Mol. Q.= VH, ~

3 H atoms %“VIV %0 Decay Stoich.wt. InAMg unit cell cycle x inABx

L~qf,5 (NiOW,)zNid,,,,Sn~?,l~ 600 450 283 4.75 0.064 5.326

MmNi, ,,Co0,,Mnn4Alnq 564 423 272 3.13 4.29 15.5 0.067 5.00

L~mRi(NiO~,, 2) WJh,”g 594 450 292 2.95* 4.89 16.1 0.081 5.157

‘%9R9@in mm 2) NL9An~,~~ 601 451 301 5.06 0.154 5.077

LaNi,,SnO, - 492 450 301 3.26 5.05 18.4 0.239 5.00

LaNi4,Mn04Aln, 132 421 324 3.26 5.09 18.5 0.485 5.00

. Atomic volume of D in deuteride phase

PUBLICATIONS

T. Vogt, J.J. Reilly, J.R. Johnson, G.D. Adzic,E.~.” Ticianeili, S. Mukerjee and J.McBreen, “High Cycle Life, Cobalt Free,AB, Metal Hydride Electrodes,” in SelectedBattery Topics, G. Halpert, M.L. Gopikanth,K.M. Abraham, W.R. Cieslak, W.A. Adams,eds., PV 98-15, Electrochemical Society,Pennington NJ, (1998).

E.A. Ticianelli, S. Mukerjee, J. McBreen, J.J.Reilly, J.R. Johnson and G.D. Adzic, “X-Ray Absorption Spectroscopy and ReactionKinetic Studies of Yttrium Containing Metal

T.

13

Hydride Electrodes;’ inTopics, G. Halpert, M.L.

Selected BatteryGopikanth, K.M.

Abraham, W.R~ Cieslak, W.A. Adarns, eds.,PV 98-15, Electrochemical Society,Pennington NJ, (1998). Boston,Massachusetts, Fall 1998.

Vogt, J.J. Reilly, J.R. Johnson, G.D. Adzicand J. McBreen, “Structure ofStoichiometric and Non-Stoichiometric AB5Type Alloys and Their Properties as MetalHydride Electrodes,” Proc. of the RareEarths(1998).

Conference, Fremantle, Australia,

Optirnization of Metal Hydride Properties in MH/NiOOH Cells for

Electric Vehicle Applications

Ralph E. WhiteDepartment of Chemical Engineering, University of South Carolina, Columbia, SC 29208(803) 777-7314; fa: (803) 777-8265; e-mail: rew@sun.che.sc.edu

Objectives

● Optimize the alloy composition of metal hydride electrodes by rnicroencapsulation ofhydrogen-storage alloys.

● Develop a theoretical model to evaluate the equilibrium potential and the exchange currentdensity as a function of state of charge of the electrode.

● Develop improved metal hydride electrodes for MH/NiOOH batteries.

Approach

● Prepare bare and Co-coated LaNi,,,SnO,,, electrodes for determination of transport andelectrochemical kinetic parameters.

● Determine the cycle lives of bare and electroless-coated alloys. .

Accomplishments

● A ten-fold increase in cycle life and a 100% charge-discharge efficiency wasCo-P coating for LaNiA,,,Sn02,alloys (250 cycles compared to 25 cycles at 0.3

● No change in particle size for the Co-coated alloy is seen with cycling,impedance and laser measurements.

Future Direction

“ Proiect was corndeted.

achieved by theC rate).

consistent with

A theoretical model for the impedance ofmetal hydrides was developed that simulated theimpedance response of a hydride electrodeunder diffusion and kinetic limitations. Theeffect of diffusion coefficient and particle sizeon the Nyquist plots was also studied, and it wasfound that decreasing the diffusion coefficientor increasing the particle size shifts the onset ofthe transition region to lower frequency values.Impedance data from the Co-plated LaNi,,,Sno2,electrode were obtained in the frequency rangeof 0.001 Hz to 10+5Hz.

Nyquist plots of the Co-plated electroderevealed three regions at low frequencies,namely semi-infinite diffusion, transition anddiffusion-limited regimes. The diffusioncoefficient was calculated from the slope of theNyquist plots in the transition region. It wasfound that the diffusion coefficient increases

with an increase in the concentration ofhydrogen in the particles. The decrease indiffusion coefficient with state-of-charge (SOC)was postulated to be due to the transformationfrom the ~-phase to the u-phase at low SOC.

Electrochemical impedance spectroscopy(EIS) was used to study the effect of cycling onthe alloy particle size. The impedance responseof LaNidzTSnO,zdelectrodes was obtained in thefrequency range of 1 mHz to 100 kHz. Theslopes of the transition regions in the impedanceprofile were used to calculate the particle size ofthe alloy. Pulverization behavior for theLaNid2TSn0.2d electrode was obtained bydetermining the particle size at the end of eachdischarge cycle.

More pulverization was observed withbigger particles and it gradually slowed down inan exponential manner. EIS was used to

14

determinate the particle size during cycling ofNi-MH batteries. Cycle-life studies on aLaNiA,27Sn0,malloy showed capacity decay evenafter the particle size stabilization, which showsthat pulverization is not the only reason forcapacity decay. X-ray studies showed thepresence of a large amount of lanthanum oxidesin the cycled alloy, indicating that alloyoxidation is an important factor in the capacityfade of metal hydrides.

The multi-step process for electroless platingof Co, Co-P, Ni, Ni-P and Co-Ni-P alloy wassimplified to a one-step technique. ALaNi,,,,SnOx alloy was microencapsulated withCo and Co-Ni using this one-step depositionprocess, and characterization studies showedthat Co has several advantages over otherelectroless coatings. The capacity fade in thebare alloy was absent and a constant capacitywith cycling was obtained. The effect ofdeposition temperature and coating thickness onthe alloy electrode performance was studied.Figure 6 shows the cycling behavior of the bareand Co-coated alloy. The performance ofLaNi,,2,Sn02, alloy mixed with Co is alsopresented for comparison. A ten-fold increasein cycle life is seen due to the Comicroencapsulation. The capacity fade in thebare alloy is absent and a constant capacity withcycling is obtained. Further, thin Co filmsfowed on the surface of the alloy enhance theoverall capacity of the alloy significantly.

100

t

Bare

o’I

o 20 40 60 80 100 120 140 160 18C

Number of cycles

Figure 6. Cycle life studies on bare and Co-coated alloys.

PUBLICATIONS

B.S. Harm, B.N. Popov and R.E. White,“Theoretical Analysis of Metal HydrideElectrodes: Studies on Equilibrium Potentialand Exchange Current Density,” J.Electrochem. Sot. 145,4082 (1998).

B.S. Haran, B.N. Popov and R.E. White,“Determination of Hydrogen DiffusionCoefficient in Metal Hydrides by ImpedanceSpectroscopy,” J. of Power Sources, 75, 56(1998).

B.S. Haran, B.N. Popov and R.E. White,“Studies on Electroless Cobalt Coatings forMicroencapsulation of Hydrogen StorageAlloys:’ J. Electrochem. Sot. 145, 3000(1998).

B.N. Popov, G. Zheng, B.S. Haran and R.E.White, “Studies on Metal HydrideElectrodes with Different Weights andBinder Contents,” accepted for publicationin J. Appl. Electrochemistry, May (1998).

B.N. popov, B.S. Haran and R.E. White,“Development of High Performance MetalHydride Alloys by a Novel ElectrolessPlating Process,” 193th Meeting of theElectrochemical Society, San Diego, CA,May 3-8, 1998.

B.S. Haran, B.N. POpOV, T. Cupolo and R.E.

A.

A.

White, “Capacity Deterioration of MetalHydride Alloys — Theoretical andExperimental Study:’ 193rd Meeting of theElectrochemical Society, San Diego, CA,May 3-8, 1998.

Durairajan, B.S. Haran, B.N. Popov and R.E.White, “Studies on Hydrogen Diffusion inMetal Hydrides Using ImpedanceSpectroscopy;’ 194th Meeting of theElectrochemical Society, Boston, MA,November 1-6, 1998.

Durairajan, B.S. Haran, B.N. Popov and R.E.R.E. White, “Development of HighPerformance Metal Hydride Alloys byCobalt Microencapsulation,” Proceedings ofthe Symposium on Batteries for the 21stCentury, 194th Meeting of theElectrochemical Society, Boston, MA,November 1-6, 1998.

15

-. ..- .,--- ...-..- ------------ --- ---, - z-... . ...>------- m,-. .-. .-.-..<’-. . ..”.,

t . . . .. .... ., -..

Microstructural Modeling of Highly Porous MH/NiOOH Battery Substrates

Ann Marie SastryDepartment of Mechanical Engineering and Applied Mechanics, The University of Michigan, Ann Arbor,MI 48109-2I25(313) 764-3061; ftzx: (313) 747-3170; e-mail: amsastiy@engin. umich.edu

Objectives

● Develop stochastic geometry models for key morphologies of Li-ion electrode materials, tomap connectivity of these microstructure over useful ranges of manufactured materials.

● Identify failure modes in cycled Li-ion and metal hydride/NiOOH cells and to understand theunique physical load mechanisms in negative-electrode substrate materials.

● Validate modeling efforts with electrochemical experiments, employing post-morteminvestigations of material properties.

Approach

● Perform transport and mechanics simulations on stochastic geometry models for keymorphologies of Li-ion electrode materials, including whiskers (-1 pm), fibers (-10 pm) andspheroidal particles (c1O pm) based on carbonaceous materials.

● Evaluate physical degradation of electrode materials with simulations using the stochastic IDand 2D (finite element) models.

● Characterize darnage progression through continued simulations, testing the validity ofnumerical approaches.

Accomplishments

● Identified ranges of applicability of a novel stochastic network generation techniquecompared with full-field finite-element solutions.

● Identified corrosion effects on material morphology by image analysis, and by mechanicaland transport (electrical resistivity) testing of substrates.

● Refined darnage simulations and validated experimentally the degradation hypothesis ofmass transfer from particle to fiber morphology.

Future Directions

● Continue investigation of scale effects in simulation of mechanical damage and transportbehavior in random networks.

● Conduct full-scale cell testing of Li-ion cells to assess to benchmark theoretical results fromdamage analysis.

● Develop new standards for cell testing to better assess electrode performance.

The objective is to develop predictive MH/NiOOH substrate materials have beencapability for the electrochemical performance completely characterized experimentally.of Li-ion and MH/NiOOH cells. The Simulations based on the evolution hypothesis

16

depicted schematically in Fig. 7 have now beenvalidated with experimental data. Goodcorrelations were found from the simulations ofelectrical resistivity (Figs. 8 and 9). Thesesimulation results demonstrated the utility of thestochastic network approach in bothcharacterizing key morphological changes insubstrates, and in suggesting design guidelinesin the manufacture of superior materials. Thehypothesis that corrosion and electrodepositioncause changes in substrate architecture wasdemonstrated to be valid qualitatively andquantitatively, and the models showed that thesechanges could be tractably represented by themethods developed.

(a) @)

Figure 7. Evolution hypothesis for morphologyof the positive plate, Ni/MH cells. Small Niparticles in the as-received material (a), whichare thought to form an interconnected networkwith the fibers, are consumed in theelectrochemical reaction and redeposited on thesubstrate (b).

Simulations also showed that rigorousbounds on behavior of fibrous media do notprovide narrow-enough predictions of propertiesto allow design of substrates. Indeed, thestochastically arranged networks often fail to“percolate” for the technologically importantcase of very low volume fractions. Particularlyin this regime, variance in both mechanical andtransport properties is large, and the average iswell below the upper bound. This has importantimplications where particles represent asignificant portion of the material. These are thesubjects of ongoing study.

Implications for Design of Substrates.Although compression of cells could eventuallybe eliminated with improved materials design, itis a commonly performed step to assure good

transport over cycle life. Thus, materialcompression must be taken into account indesign for conductivity. Also, the consumptionand redeposition of particle mass in the cellcauses very significant changes in the networksmodeled. Doubling of aspect ratios of fibers hasimportant consequences for transport properties;indeed, these morphological changes aloneexplain much of the reduction in conductivitymeasured experimentally. In essence, one mustdesign for minimum conductivity in the post-cycled condition if substrates comprised ofmixtures of fibers and particles are to be used.Further corrosion of the substrates is possible inmore-aggressive cycling conditions and afterprolonged cycling. The best strategy inmodeling, however, is probably to separateelectrodeposition modeling from the networkbehavior, thus carrying out design in two steps:first, asse~sing the likely changes in architectureduring cycling, and second, modeling theconductivity of the final microstructure todetermine minimum requirements on mass(volume). Simulations here show that thesecond step can be performed with excellentaccuracy, and which are guided by experimentsthat deliver estimates for the determination ofthe microstructure.

I— parallel model

0.25 — experiments

+ simul~”on / /

0.05 H i

,10 25 20 25 30

volume fraction

Figure 8. Simulations vs. experimental data.Resistivities of compressed materials vs. upperbound predictions and simulation results, foreach material type (10, 14 and 29910 volumefractions).

17

0.3 -— parallel model

— experiments

0.25 ~ simulation

j 0.2x’-=0 15 -~.

0.1

0.05 .

n-o 5 25 30

Jume fgction ~Yo)

Figure 9. Simulations vs. experimental data.Resistivities of post-cycled materials vs. upper-bound predictions and simulation results, foreach material type (10, 14 and 299t0 volumefractions).

The mechanical properties and models serveas “proof tests” of material evolution; they alsoindirectly address the matter of swelling. The 2-D simulations showed that bond stiffness mustbe quite low to explain the observed behavior, inthe absence of plasticity in the model.Experimental evidence suggests very littleoverall permanent deformation of materials aftercycling, motivating the elastic models, butdisplacements and rotations can be locally highin stochastic systems. The 3-D problem ofmaterial resistance to failure in transverse planes(i.e., loss of conductivity of the positive platedue to transverse failures) must incorporateobservations and simulations of the bondbehavior in the plane. It is probably reasonable

to assume that bonds behave much the sameway, regardless of location, but the long beamsections which span the thickness of thematerials would inherently offer much differentbalances of loads in bending, torsion and tensionthan the shorter beams in the plane of thenetworks studied, necessitating fully 3-Dsimulations for those materials.

The comparisons of rnicromechanicalmodels and experimental data have producedgood agreement for the materials studied.Results point to use of higher-aspect-ratio fibersfor use in substrates, with modeling alsonecessary to account for structural changesduring cycling.

PUBLICATIONS

A.M. Sastry, S.B. Choi and X. Cheng, “Damagein Composite NiMH Positive Electrodes,”ASME J. Engin Mats. and Tech. 120, 280-283, (1998).

A.M. Sastry, X. Cheng and C.W. Wang,“Mechanics of Stochastic FibrousNetworks,” J. of Thermoplastic CompositeMats. 30,288-296, (1998).

C.W. Wang, X. Cheng and A.M. Sastry,“Failure of Stochastic Fibrous Networks,”Proceeding of the 13th Annual TechnicalConference of the American Society forComposites, Baltimore, MD, September 21-23, 1998.

/. I

18

COMPONENTS FOR AMBIENT-TEMPERATURE NONAQUEOUS CELLS

Metal/electrolyte combinations that improve the rechargeability of ambient-temperature, nonaqueouscells are under investigation.

Novel Lithium/Polymer-Electrolyte Cells

Elton J. Cairns and Frank R. A4cLumon90-1142, Lawrence Berkeley National Laboratory, Berkeley CA 94720(510) 486-4636, fm:(510) 486-4260

Objectives

● Investigate the behavior of advanced electrode materials in high-performance rechargeablebatteries, and develop means for improving their lifetime and performance.

● Improve the cycle life and utilization of the sulfur electrode in Li/polymer/S cells.

● Identify new electrode structures and compositions which will eliminate or minimize thisfundamental mechanism of capacity loss of Li/Mn02 cells

Approach

“ Fabricate and test Li/polymer electrolyte/sulfur cells.

● Use a completely new approach to synthesize MnOX and related materials for Li cells.

Accomplishments

● Greatly improved sulfur utilization at 60 C and room temperature were achieved, both in therange of 400 mAh/g S

● Novel aerogel LiXMnYO,powders with very high surface area have demonstrated specificcapacities up to 175 rnAh/g, but show significant capacity fading with cycling.

Future Directions

● Use in situ UV/VIS spectroscopy to determine the mechanism by which the sulfir electrodeoperates.

● Improve the performance of MnOx, making use of post-test characterization to identifi thestructure of the active component.

Li/S Cells. Effort continues to develop aspectroelectrochemical cell to identify theproducts of Li/S cell reactions duringgalvanostatic cycling. The goal is to identifythe cause(s) of the rapid capacity fade which ischaracteristic of this high-energy cell. Attemptsto prepare LiAl negative electrodes by electro-alloying Li into an expanded Al mesh resultedin a brittle structure that either disintegrated intothe electrolyte or punctured the polymerelectrolyte and then shorted the cell. Negative

electrodes prepared by electrodepositing Li ontoa Ni grid were also unsuccessful because the Lieither dissolved into the electrolyte or formed agrainy, highly reactive film that corroded beforeuse. A cell sandwich with annular sulfur and Lifoil electrodes is now used which allows thelight beam to pass only through the polymerelectrolyte. This cell design avoids thedifficulty of preparing semi-transparentelectrodes, and has allowed us to collectpreliminary spectra. Galvanostatic studies on

19

.,,.. ... ... e.+ — —., . —. .. —-, .x..,.-. . ,.-.=..,..,-..,,.,.,7....... ....—.-a=,=, .- ., -w, ,

Li/S cells indicated that the first (higher-voltage) plateau is retained longer and capacityfade is reduced by limiting discharges to -25%of the theoretical cell capacity. These resultssuggest that during this stage of discharge,polysulfides are formed. The effects of othervariables such as operating temperature anddischarge/charge currents on cell lifetime werealso studied, however no significant trends wereidentified for the limited ranges of thesevariables that were studied.

Li/S cells prepared with polyethylene-methylene oxide (PEMO) electrolyte werecycled at room temperature and at 60”C.Positive electrodes with 50% sulfur, 35%electrolyte and 15% carbon achieved am,aximum utilization of 45?i0, or about 375mAh/g-cathode at 60”C. The capacity fadedfairly rapidly upon continued cycling to thelower voltage cutoff of 1.7 V vs. Li/Li+. Thecapacity of 375 mAh/g-cathode is -6590 of thegoal of 70% utilization at 60°C. Experimentsare in progress to achieve a capacity goal of-400 mAh/g-cathode at room temperature. Thisgoal was recently met using a gel electrolyteconsisting of thermally cross-linked fumedsilica and butyl methacrylate as the mechanicalbackbone, with PEGDME as the plasticizer.The conductivity of this electrolyte at roomtemperature obtained by ac impedancemeasurements is >10-3 S/cm*, which is amongthe highest conductivities for a gel electrolytethat contains no carbonate-based plasticizer.The positive electrode of this Li/S cell contains50% sulfur, 30% PEGDME + salt, 5% PEO and15% carbon.

Novel Cathode Materials. AerogelLiXMnYO,powders with very high surface areawere prepared. Initial tests showed goodperformance with a first discharge capacity of175 mAh/g, however severe capacity fadingaccompanied cycling. Some of the capacityfading was attributed to the presence of water inthe aerogel structure. Using a uniquely

designed drying tube and a new electrodefabrication method, water has been removedfrom the aerogel powder and reabsorption ofwater was prevented. The new electrodesexhibit a lower first discharge capacity of 150mAh/g, but the rate of capacity fade duringcycling decreased. Several electrode fabricationtechniques using different orders of mixing theelectrode components, current collectors, anddrying temperature were evaluated to maximizethe utilization of the active material. No majorimprovements were observed using the varioustechniques, and the cause of the severe capacityfade remains under investigation. Elementalanalysis will be used to determine the exactcomposition of the aerogel, and the manganeseoxidation state of the aerogel will be determinedusing an iron sulfate chemical titrationtechnique.

Other compositions of lithium metal oxideaerogel materials will be synthesized using sol-gel methods followed by CO, supercriticaldrying. The aerogel powders will becharacterized by XRD to verify an amorphousstructure, thermogravimetric analysis todetermine dryness and thermal stability, andSEM to measure the micropore size. The newmaterial will be used to produce test cells forperformance and cycle-life evaluation.

PUBLICATION AND PRESENTATION

T.C. Adler, F.R. McLarnon and E.J. Cairns,“Investigations of a New Family ofAlkaline-Fluoride-Carbonate Electrolytesfor Zinc/Nickel Oxide Cells,” Ind. Eng.Chem. Research, 37,3237-41 (1998).

D. Marmorstein, H.-J. Ahn, K. Striebel, F.McLarnon and E.J Cairns,“Lithium/Polymer/Sulfur Cells,” 194thMeeting of the Electrochemical Society,Boston, MA, November 1-6, 1998.

L

20

New Cathode Materials

ill Stanley WhittinghamChemistry and Materials Research Center, State University of New York at Binghamton, Binghamton,NY 13902-6000(607) 777-4623, fa:(607) 777-4623; e-mail: stanwhit@binghamton. edu

Objectives

● Synthesize and evaluate oxides of manganese oxides for alkali-metal intercalation electrodes.Complete work on vanadium oxides.

● Identify new intercalation compounds for positive electrodes in advanced nonaqueoussecondary batteries.

Approach

● Synthesize metal oxidesintercalation of Li ions.

● Characterize the metalelectrochemical cells.

Accomplishments

that have an appropriate crystallographic structure to permit facile

oxide structures by XRD analysis and evaluate materials in

“ Synthesized a range of transition metal substituted layered manganese oxides: AxMYMnO, (A= Li, Na, K and M = Fe, Co, Ni) and showed that the cobalt compounds gave the best cyclingresults.

“ Identified severaI new phases of vanadium-stabilized manganese oxides.

“ Measured electrochemical cycling of the new phase ~V~08, and found good cyclability.

~ufire Directio~

● Continue synthesis and electrochemical studies of manganese oxides.

“ Use more-stable crystalline lattices.

The objective of this project is to synthesizeand evaluate first-row transition metal oxidesfor Li intercalation electrodes in advancednonaqueous rechargeable batteries. Mildhydrothermal techniques are being used for thesynthesis, and in cases where the highestoxidation states are not obtained, other methodswill be explored to drive the transition metal toits highest oxidation state. Manganese oxidesare particularly attractive as the cathode, andmeans of stabilizing the layered manganesedioxide structure are being explored. ~

An attempt to incorporate nickel, in place ofan alkali ion, into layered manganese dioxide bythe hydrothermal decomposition of TMAOMnOd

in the presence of Ni salts led to the formationof a totally new structure with the formulaHNiMnO~ rather than a simple layeredmanganate Ni~n02. Unfortunately, reaction inan electrochemical cell is very slow. Thus, wesynthesized a range of layered manganates,AXM~nO,, (A= Li, Na, K and M = Fe, Co, Ni)in the temperature range 700-950”C. InitiallyNi was chosen to “alloy” with Mn, as the totalelectron count would be more like that ofLiCo02. The reactants, N~CO,, Mn2C0, andNiO were reacted to give the expectedhexagonal structure with a=2.876(6) A andc=ll.16(2) ~. The compound was ion-exchanged to the Li form as follows:

21

LiBr

NaMnmNil~O, > LiM~NilnOzhexanol, 160”C

The electrochemical cycling of NaMnmNi,,,02 isshown in Fig. 10 (left). The compound must becharged first as little Li is incorporated during afirst discharge. Essentially all of the Na wasremoved below 5 V, and 0.8 Li wasincorporated on the subsequent discharge.However, only 0.6 of that Li could be removedand capacity fading was rapid. Thus, a secondsample was first ion-exchanged with Li to give

(Li,.YNaY)MnvsNi,,@2 and then cycledelectrochemically. The results are shown in Fig.10 (right); 0.55 Li was removed below 4.6 V,and the subsequent discharge and charge cycleare shown.

X-ray analysis indicates that NiO is presentas an impurity. In contrast, the Co- and Fe-substituted compounds appear to be COO- andFeO-free for the composition KYCo0,,Mn090zOur present work therefore targets thosematerials, with particular emphasis on the K

compound because K+ ions are preferable to Na+or Li+ ions in the lattice. Electrochemical datacollection is now underway on these materials.

The pure trivalent manganese compoundswere found to have a high resistivity, typically10’ to 107 ohm-cm at room temperature, similarto MnOOH in dry cells. In contrast, Co dopingof the Mn sites such as in KYCoO,MnOgOzdropsthe resistivity by two orders of magnitude.

Summary of Vanadium OxideCompounds. Studies of vanadium oxidecompounds were completed. The tunnel/layeredcompound H2V,0, was obtained bysolvothermal synthesis in acetic acid. Itsstructure and electrochemical behavior is shownin Fig. 11. The repeatability of the dischargecurves suggests that the structure is maintainedduring cycling. The synthesis of a number ofvanadium-substituted manganese oxides isunderway and an interesting new materialappears to have been formed that has ahexagonal structure and has an open pipemorphology.

Figure 10.

black andelectrolyte.

The

10%

1 -NaMnmNi,n02: 1.5M LiPFf in DMO.EC=2

I I I I I i L&J_L_d

LI M~Nim02- lMLIPF in DMC EC=l.2

o-1 -0.8 -0.6 -0.4 -0.2 0 0.2x in (LINa),Mn,J Ni,n 02 -0.4 -0.3 -0.2 -0.1 0 0.1

x in Li ,+XMnmNi O1/3 2

electrochemical cycling of the cell made by using 80~0 (Na,Li)MnZ~Nil,qOz, 1091iocarbonTeflon powder as cathode, Li metal ribbon as anode and LiAsF, in PC:DMC as

,.:~0.50 1.0 15 2.0 2.5

x in LixH2V30%

Figure 11. Structure of H,V,O, (left) and electrochemical behavior (right).

22

PUBLICATIONS

T.A. Chirayil, P. Y. Zavalij, M. S. Whittingham,“Hydrothermal Synthesis of VanadiumOxides,” Chem Mater. 10, 2629-2640(1998).

F,R. Chen, P.Y. Zavalij, M.S. Whittingham, “The

Hydrothermal Synthesis andCharacterization of New OrganicallyTemplated Layered Vanadium Oxides byMethylamine,” Mat. Res. Sot. Proc., 497,173-180 (1998).

M.S. Whittingham, “The Relationship BetweenStructure ‘md Cell Properties of the Cathodefor Lithium Batteries, “ in Lithium Batteries,O. Yarnamoto and M. Wakihara, eds.,Kodansha, Tokyo, Japan (1998).

Zhang, P.Y. Zavalij and M.S. Whittingham,“Hydrothermal Synthesis andCharacterization of a Series of Novel ZincVanadium Oxides as Cathode Materials,”Mat. Res. Sot. Proc., 496,367-372 (1998).

Solid Electrolytes

Lutgard C. De Jonghe62-203, Lawrence Berkeley National Laboratory, Berkeley CA 94720(510) 486-6138, fa: (510) 486-4881; e-mail: lcdejonghe@lbl.gov

Objectives

● Develop highly conductive novel composite electrolytes that combine the best features ofsingle-ion conductors with those of polymer or gelled electrolytes, and to understand theirtransport properties.

● Develop high-capacity, robust Mn02 cathode materials for rechargeable Li and Li-ion cellsbased on the N~,tiMn02 tunnel structure.

Approach

“ Develop methods for understanding transport properties of complex electrolyte systems.

● Fabricate composite electrolytes containing a single-ion conducting component with aconventional polymer or gelled component

● Synthesize and characterize Mn02 based on the N~,aMn02 structure to achieve improvedreversible capacity, cycle life and resistance to abuse (e.g., overcharge and overdischarge) inLi-polymer and Li-ion cells.

Accomplishments

● Characterized transport properties of PPO-LiCF$O~ and PPO-LiN(CF$O,), polymerelectrolyte systems.

● Designed cell to determine effect of electro-osmotic solvent drag in gelled, liquid orcomposite electrolytes.

● Demonstrated reversible capacity of 90-100 mAh/g (approximately half the theoretical capacity)for LiXMn02 based on the N~WMn02 structure, at voltages compatible in metallic Li/polymerconfigurations. .

● Demonstrated stability of LiXMnO, based on the N~.MMn02 structure between 1-5 V vs. Li.

23

-........ -~.w-.— ..= —..-.= ---- -- -V.--—- -

Future Directions

● Project was redirected to develop robust tunnel-containing manganese oxide structures forcathode materials.

● Improve cycle life and rechargeable capacity (>200 rnAh/g theoretical capacity) of MnO, byelectrochemical and structural characterization.

● Demonstrate an increased capacity at -3 V to at least 150 mAh/g by doping manganese oxidewith titanium.

The initial goal for the project was tofabricate novel composite electrolytes thatcombine the advantages of single-ionconductors with that of gelled or polymerelectrolytes. Another goal was to develop newmethods to determine transport properties ofgelled and composite electrolytes. This programwas redirected in the fall of 1998, and the newemphasis is to develop novel, high-capacityelectrodes based on tunnel-containingmanganese oxides for rechargeable Li and Li-ion batteries.

Two studies of transport property of binarysaltipolymer systems were completed incollaboration with J. Kerr’s group (PolymerElectrolyte Synthesis for High Power Batteries).Results on polypropylene oxide (PPO)-LiCF~SOq and Li(NCF~SOz)z show that higherLi+ transference numbers and conductivities areobtained when the latter salt is used. It has beenrecognized that, in addition to salt concentrationgradients in polymer or gelled electrolytes thatarise during passage of current, electro-osmoticsolvent drag can also occur. In extreme cases,drying or flooding of cell components canoccur, causing premature cell failure. Figure 12shows a schematic representation of anexperimental cell to quantify this phenomenon.

The device consists of two cell halves thatare sealed with an O-ring seal. An HDPE-encapsulated plunger holds a stainless steelplatform onto which the Li electrode is pressed,and electrical contact is made via tungsten wiresembedded in the plungers. The electrodes canbe placed up to l-cm apart. The test membrane,which can be a gelled or composite polymerelectrolyte or a separator material such asCelgard, is clamped into a threaded holder

placed between the two electrodes. To preventbubbles and to ensure complete wetting of themembrane, the device is vacuum-filled througheither of the two fill valves. The device holds-30 mL of electrolyte solution when filled.Most of the internal volume is occupied by theHDPE plungers, and the solution must flowaround them. Two graduated capillary tubes areattached via Swagelok connectors, and a 1-mmdistance on the capillary tube corresponds to1.87 x 10-7ml, so the device is very sensitive tovolume changes. No metal except Li contactsthe electrolyte solution and there are no bubblesor leaks observed when the cell is properlyassembled.

The work in this laboratory was recentlyredirected towards a study of novel tunnel-containing Mn02 based on the N~J,aMn02structure. Lithium ion-exchanged materialsmade from N~aMn02 have a high degree ofreversibility toward intercalation processes andthey are resistant to degradation when over-charged or over-discharged. Furthermore,conversion to the thermodynamically favoredspinel structure does not occur below about400”C, nor does degradation occur upon cyclingat elevated temperatures. N%.MMn02 has anunusual double tunnel structure, which canaccommodate intercalation and de-intercalationof either Na or Li ions without undue stresses,rendering it extremely robust (Fig. 13) toovercharge and overdischarging. The presentresearch is directed towards improving thevoltage profile by doping to increase thecapacity in the 3 V region (for Li batteries) or inthe 4 V region (for Li-ion batteries).

24

Figure 12. Cell set-up for electro-osmotic drag measurements.

Figure 13. Structure of N%.@4nOz, lookingdown the c-axis. The spheres represent sodiumatoms or sites for Na atoms. In the as-madematerial, the sites in the small tunnels are fillyoccupied while the sites in the large S-shapetunnels are approximately half occupied.

.-5 , , 1

2 2.5 3 3.5 4 4.5

Cell Potential, V

Figure 14. Differential capacity as a function ofpotential between 3.9 and 2.5 V for aLi/LiXMn02 cell stepped 10 mV every fourhours. The electrolyte is 1 M LiPFG in 1:1EC/PC. The cell was charged prior to theexperiment. The lower trace represents celldischarge and the upper trace cell charge.

No significant degradation of Li-exchangedN~~Mn02 cathodes occurs between 1 and 5 Vvs. Li during slow cyclic voltammetry scans incells with liquid electrolytes. The remarkablestability and reversibility of this material can beattributed to its unusual double tunnel structure(Fig. 13), which is flexible enough to undergointercalation processes without stress, andrequires substantial rearrangement to form aclose-packed structure such as the spinel.Figure 14 shows the results of a steppedpotential experiment on a LiXMn02/Li cell witha liquid electrolyte, which illustrates theextraordinary reversibility of this material.

Evidence obtained in this laboratorysuggests that it may be possible to insert morethan 0.66 Li/Mn in the NaXMn02 structure dueto the smaller size of the Li ion; this implies thatcapacities above 200 rnAh/g may be achievable.In practice, the capacity in Li-cell configurationsvaries with the degree of exchange, temperature,and voltage limits used. About 140 mAh/g isutilized in partially exchanged materialsdischarged between 3.7 and 2.5 V at 85°C inLi/polymer electrolyte cells. Completelyexchanged materials exhibit less capacity and ahigher average voltage between these limits, aswell as having a smaller unit cell size thananalogs containing residual Na.

25

-. .. ,,X------ -$. -. :-m. -“ - :, :, J..- -,.r.-.--l@lTn:-7------ -v : .—.->,= ~t,. .. .,: .: $~–—- ‘-- -.-7- --,. .“ ----

The ECPS experiments show that Liintercalation and extraction processes between3.9 and 2.5 V vs. Li are extremely reversible.Most of the capacity is located between about2.9-3.4 V vs. Li and the voltage profile isgradually sloping in this region with severalsmall plateaus. When charged past acomposition of about Lio.3Mn02, the voltagerises steeply. At 3.9 V vs. Li, the composition isapproximately Lio.28Mn02, corresponding to acapacity of -85 mAh/g left in the structure.

Leaving residual Na in the structure, whichincreases the unit cell size, allows more capacityto be utilized below 3.7 V vs. Li in Li cells.Under normal cycling conditions below thisvoltage, further exchange is unlikely to occurfor compositions containing -0.14 Na/Mn,except at elevated temperatures. It is possible,however, that Na ions would be de-intercalatedif the cell is charged above 4.2 V vs. Li, causinga decrease in unit cell size and lowered capacityin the normal operating range. Because cells inEV battery packs are susceptible to over-charging, this is undesirable. An alternativeapproach to increasing the capacity is to replacesome of the Mn in the structure with a transitionmetal that is easier to oxidize, such as Ti.

So far, we have succeeded in synthesizingcompounds in which 11, 22 and 55% of the Mnhas been replaced with Ti. Both N~.aMn02

and N%.44Ti0.ssMnO.@z belong to the samespace group (Pbmn), but the latter has asomewhat larger cell volume than the former(712 ~ compared to 680 ~’). XRD studiesindicate that N~.uTiol ,Mno8@2 andN~~Tioz2MnOy802 are isostructural, and have

slightly larger unit cells than N~.MMn02.Provided that this size differential is maintainedduring Li-ion exchange, it is likely that therewill be a beneficial effect on utilization atpotentials practical for Li battery operation.

PUBLICATIONS

M.M. Doeff, P. Georen, J. Qiao, J. Kerr andL.C. De Jonghe, “Transport Properties of aHigh Molecular Weight Polypropyleneoxide)-LiCF~SO~ System,” 193rd Meeting ofthe Electrochemical Society, San Diego, CA,May 3-8, 1998. LBNL-42081

M.M. Doeff, J. Hou, J.B. Kerr, J. Qiao and M.Tian, “Synthesis and Testing of Comb-branch Polymers for Lithium Batteries andElectrochromic Windows,” 193rd Meetingof the Electrochemical Society, San Diego,CA, May 3-8, 1998.

M.M. Doeff and J. Reed, “Li Ion Conductors

A.

A.

Based on Laponite/Poly(Ethylene Oxide)Composites,” Solid State Ionics, 115, 109(1998).

Ferry, M.M. Doeff and L.C. De Jonghe,“Transport Property and RamanSpectroscopic Studies of the Polymer 1.Electrolyte System P(EO)n-NaTFSI,” J.Electrochem. Sot. 145, 1586 (1998).

Ferry, M.M. Doeff and L.C. De Jonghe,“Transport Property Measurements - ofPolymer Electrolytes,” Electrochim. Acts.43, 1387 (1998).

T.J. Richardson, P.N. Ross, Jr. and M.M. Doeff,“XRD Studies of LithiumInsertion/Extraction in Cathodes Derivedfrom N~,aMn02,” 194th Meeting of TheElectrochemical SocieQ, Boston, MA,November 1-6, 1998.

M.M. Doeff, P. Georen, J.B. Kerr, S. Sloop andL.C. De Jonghe, “Effect of MicrophageSeparation on the Measurement of TransportProperties of Polymer Electrolytes,” 194thMeeting of The Electrochemical Society,Boston, MA, November 1-6, 1998.

26

Polymer Electrolyte Synthesis for High-Power Batteries

John B. Kerr62-203,LawrenceBerkeleyNationalLaboratoq, Berkeley, CA 94720(510) 486-6279; fax: (510) 486-4995; e-mail: jbkerr@lbl.gov

Objectives

● Develop methods to prepare, purify and characterize polymer electrolyte materials.

“ Measure Li-ion transference numbers in polymer electrolytes.

c Develop structure-function relationships involving polymer structure and (i) Li+-ion transportthrough the bulk polymer and electrode interfaces, (ii) stability towards electrode materials indynamic cycling and (iii) mechanical and thermal properties.

Approach

● Prepare comb-branch polymers (polyvinylether, polyacrylate ethers and polyepoxide ether)containing 2 to 8 ethylene oxide groups in the side chain.

c Prepare single-ion conductor materials by fixing anions to the side chains of the comb-branchpolymers.

● Measure transference number by electrochemical methods and electrophoretic NMR.

s Assess polymer purity and stability towards Li metal by half-cell cycling, impedancespectroscopy and post-mortem chemical analysis.

Accomplishments

● Measurements on a variety of comb-branch and linear polymers containing segments ofethylene oxide units at least 5 units long show that the conductivity is the same regardless ofthe polymer architecture. The conductivity limit of 5 x 10-s S/cm at 25°C was reached withethylene oxide units as the solvating polymer structure.

● Thermal and transport property studies on Parel (high-MW PPO)/Li(CF3S0,)2N and

amorphous PEO/ Li(CF3SOz)2N and LiCF~SOq were completed.

● Polyelectrolytes with unit transference numbers were prepared and tested.

Future Directions

● Complete transport measurements of selected comb-branch polymersethers, polyvinyl and polyepoxide ethers and a variety of salts.

● Prepare comb-branch polymers with pendant crown ether units of

containing polyacrylate

different ring size andflexibility of the crown ether units. Prepare single-ion conductor materials with practicalconductivities and confkrn predicted benefits of these materials.

● Develop methods to measure side reactions at electrodes and provide rate data to validateelectrochemical system models designed to predict failure behavior of batteries.

27

. .—--- .- —-..—— ..-— — —.. .— — ———- . . . . ..— —,—-- .. ... . ----— -- .—~- -

Synthesis. Synthesis of vinyl-ethermonomers using the Williamson reaction withdiethyleneglycolvinyl ether and monomethylether of polyethylene glycol ethylchlorideyielded a variety of chain lengths from 2 to 7ethylene oxide units. Several initiation systemswere employed to produce living polymerconditions. The non-stoichiometric Et,,,AICl 1.5gave polymers with increased MW from thosereported in the literature but the PDI ratio of1.7- 1.8 shows that the living polymer was notachieved. Attempts were made to increase theMW by using more-controlled conditions,varying the initiator/monomer ratios and the useof chain-extension agents. The structure ofthese agents influences the MW of the polymerand excess cross-linker leads to the formation ofintractable rubbers. We could not produce high-molecular-weight polymers under livingpolymer conditions.

Polyepoxide ethers were successfullyprepared by anionic initiation conditions and,polyacrylate ethers were formed under radicalconditions. It was noted that the radical‘conditions could not be completely quenchedand uncontrolled cross-linking continued tooccur indefinitely leading to intractable rubbers.An alternative method of in situ cross-linkingwas developed which provides self-standingpolymer films with excellent mechanicalproperties.

Characterization of Comb-branch (CB)and Comb-branch/PPO Blended Electrolyte.All of the PVEX(polyvinylether),PEPEX(Polyepoxide ether) and PAEX(Polyacrylate ether) materials where x=2-7 werecompletely amorphous, as indicated by DSCmeasurements. The T~ is generally below -60”Cbefore addition of LI salts, and increases withincreasing salt concentration, particularly whenthe ratio of ether oxygen to Li ions (0/Li) is lessthan 15:1. Although the tractable materialsexhibited flow and did not form self-standingfilms, electrochemical measurements could be

made by solution casting directly on theelectrodes.

An alternative method to form self-standingfilms is to blend the CBX materials with high-MW PPO. This material was used to avoid thecrystallinity problems that exist with PEO atlow temperatures that lead to phase separation.The PPO was Parel Elastomer (Zeon Chemical,MW - 5X105) which has a conductivity that isgenerally two orders lower than that of the CBXpolymers, indicating that the majority of the iontransport in the blends is via the CBX material.However, severe phase separation also occurredwhich was dependent upon the concentrationand identity of the Li salt. This result limits theuse of polymer blends in binary salt electrolytesdue to concentration polarization.

Lithium-Ion Transference Number (t+(’)Measurements. The thermal and transportstudies on Parel (high-MW PPO)/Li(CF$02)2Nat 85°C, and amorphous PEO/LiCF$Oq andLi(CF,S02),N systems at 40”C, were completed.Polarization experiments were carried out torelate the transference number measurements tocycling behavior. As predicted by theory,materials with low or negative transferencenumbers exhibit rapid polarization at the Lielectrode.

Conductivity Limit of Ethylene Oxide-Based Polymers. Conductivity measurementson comb-branch polymers with side chainlengths of 5 ethylene oxide units or longer gavevery similar conductivities. The linearamorphous PEO which has ethylene oxide chainlengths of about eight units also had the sameconductivity, about 5 x 10-5 S/cm at 25”C,regardless of the polymer architecture withLi(CF,SO,),N as the salt. Intuitively this makessense as all systems possess the same solvationstructures for the Li+ ions, and the polymerarchitecture has only a secondary influence.This indicates that new polymers must beprepared that have different structures thanethylene oxide to increase the conductivity to10”3S/cm at 25”C.

28

Composite Polymer Electrolytes for Use in Lithium and Lithium-Ion Batteries ,

SaadA. Khan and Peter S. FedkiwDepartment of Chemical Engineering, North Carolina State University, P. O. Box 7905, Raleigh NC27695-7905(919) 515-4519; fm:(919) 515-3465; e-mail: khan@ eos.ncsu.edu

Objectives

● Develop solid composite electrolytes utilizing synthesized fumed-silica fillers with tailoredsurface chemistries.

● Investigate the electrochemical and theological characteristics of novel composite polymerelectrolytes.

Approach

“ Utilize a combination of electrochemical and theological techniques and chemical synthesisto study the effects of silica surface chemistry on the electrochemical and mechanicalproperties of composite polymer electrolytes.

Accomplishments

● Identified solid electrolytes with high yield stresses, high modulus and room-temperatureconductivity of -103 S/cm by varying the amount of polymerizable monomer (butylmethacrylate) in the composite system.

● Established experimental techniques and protocols to measure self-diffusion and Litransference number of the composite electrolytes using electrophoretic NMR.

c Developed a scaling relationship to predict a priori the solid electrolyte modulus in terms ofcomponent properties.

Future Directions

● Determine the transference number in candidate electrolyte systems using the newlydeveloped protocol for electrophoretic NMR.

“ Explore interracial stability of composite electrolytes.

● Develop a comprehensive understanding of the mechanism of crosslinking to produce solidelectrolytes with tailored mechanical behavior.

The objective of this research is to developnew composite solid polymer electrolytes forrechargeable Li and Li-ion batteries. Inparticular, the goal is to develop highlyconductive electrolytes that exhibit goodmechanical properties, and at the same timeshow good compatibility with typical electrodematerials. The unique feature of our approach isthe use of surface-functionalized fimed-silicafillers to control the mechanical properties of theelectrolytes. A low-molecular-weight liquid

29

polyether is used as the matrix polymer, therebyensuring high conductivities, and the fumedsilica serves to provide mechanical support.

A key element in the research is to developprotocols to measure transport properties oflithium, anion diffusion and transferencenumbers using electrophoretic NMR (ENMR).

We have successfully determined the self-diffusion coefficient of Li+ ions in the compositepolymer electrolyte at 30”C. We have recentlydeveloped the ability to observe “F, thus

enabling self-diffiwion measurements on theanion. The lithium irnide/poly(ethylene glycol)dimethyl ether (PEG-DME)/R805 (octyl-modified) fumed-silica composite electrolytesystem is the most studied to date. For theconditions of a Li:O ratio of 1:20, 10 wt% R805and 30”C, Dti is 3.0M3.2 x 10-7 cm2/sec. Thisself-diffusion coefficient for of Li+ ions iscomparable to DU in polymer electrolytes. Thefumed silica appears to have only a small effecton the diffusion of Li ions given that for thesame system without R805 fumed silica, Dti isabout 3.3 x 10-7cm2/sec. Thus, we conclude thatfor 10 wt% fumed silica, the mobility of the Liions is only slightly impeded by the fumed-silica network, a very encouraging result.

The lithium triflate/poly(ethylene glycol)dimethyl ether(PEG-DME)/R805 system withLi:O=l :20, 10 wt% fumed silica, and 30°C wasmore amenable to ENMR. This is probably dueto the much lower anion concentration (fluorineconcentration is 3.7 M compared to 6 M). TheD,, is 3.2 x 10-7 cm2/sec and D, is 2.8 x 10-7cm2/see, which is comparable to literaturevalues. Preliminary transference numbers forthe 590 R805/PEG-DME composite electrolytescontaining either lithium imide or lithiumtriflate salts were higher (0.29) for the irnidecompared to the triflate (O.19) salt. Thecalculation and measurement need to beverified, but we are encouraged that we canobtain the behavior expected by ENMRmeasurements.

Effort was also focused on determining if wecan a priori predict theological behavior withelectrolytes containing the liquid medium andfumed silica. Octyl-modified fumed silica(R805) in a range of liquids of different polaritywas used. We found that the degree offlocculation of the fumed silica to form solidswas dictated by the mismatch in volubilityparameter between the liquid medium (8.) andthe surface layer of the fumed silica (5,). Ascaling relationship between the elastic modulus(G’) of the system and the volubility parametersgiven by G’ -(5, - b.)’ was successfully derived.The practical implication of this result is that itpermits us to predict and control one of theimportant theological parameter of a compositesystem.

Studies on the crosslinking mechanismfocused on determining the yield behavior ofsolid polymer electrolytes. While we weresuccessful in obtaining high-modulus solids, theyield stress in these systems were found to berather low (-1 Pa). A high yield stress (>102 Pa)is critical for the electrolyte to perform as adimensionally stable solid and avoid creep. Theyield stress could potentially be enhanced byadding a small amount of polymerizablemonomer, such as butyl methacrylate, to theelectrolyte that would not detrimentally affectthe conductivity (-1x103 S/cm at 25°C). Byvarying the amount of butyl methacrylate in acrosslinkable fumed silica/polyethylene glycoldimethyl etherflithium imide system, we couldnot only tailor the modulus of the system butalso obtain solids with much higher yieldstresses (Fig. 15). In Fig. 15, solids with moduliranging between 4x104 and 4x105 Pa areobtained by varying the amount of butylmethacrylate. More importantly, the yieldstress, defined as the product of G’ and strain atonset of sharp decrease (shown by arrow),exceeds 103 Pa in both cases. While theseresults are preliminary, they are veryencouraging and work in this direction willcontinue to understand how composition andcrosslinking time affect mechanical properties.

‘“~0.01 0.1 1 10 100

Strain Amplitude (%yO)

Figure 15. Elastic Modulus (G’) as a functionof strain amplitude for solid compositeelectrolytes containing crosslinkable fumedsilica fillers. BMA refers to butyl methacrylatecontained in the composite.

L I

30

PUBLICATIONS

J. Fan, S.R. Raghavan, X-Y Yu, J. Hou, G.L.Baker, S.A. Khan, P.S. Fedkiw, “CompositePolymer Electrolytes Using Surface-Modified Fumed Silicas: Conductivity andRheology,” Solid State Ionics, 111, 117-123(1998).

S.R. Raghavan, M.W. Riley, P.S. Fedkiw, S.A.Khan, “Composite Polymer ElectrolytesBased on Polyethylene Glycol andHydrophobic Fumed Silica: DynamicRheology & Microstructure;’ Chemistry ofMaterials 10(1), 244-251 (1998).

S.A. Khan, “Shear-induced MicrostructuralChanges in Filled Polymers,” (invited

speaker) NIST Workshop on Fillers andNanocomposites, Gaithersburg, MD, July.1998.

S.R. Raghavan, J. Hou, G.L. Baker, S.A. Khan,“Colloidal Silica Gels as Novel Electrolytes:Synthesis, Microstructure and Rheology.”ACS National Meeting, Boston, MA, August1998.

S.R. Raghawm, J. Hou, G.L. Baker, P.S.Fedkiw, S.A. Khan, “A Novel PolymerElectrolyte Incorporating CrosslinkableParticles into a Composite PolymericFramework,” AIChE Annual Meeting,Miami, FL, November 1998.

Polymer Electrolyte for Ambient-Temperature Traction Batteries: Molecular-Level

Modeling for Conductivity Optimization

Mark A. RatnerDeparttnentof Chemistry,NorthwesternUniversity,EvanstonIL 60208-3133(847)491-5371,fa: (847)491-7713;e-mail: ratner@mercury.ehem.nwu.edu

Objectives

● Explore conductivity mechanisms in polymer/salt complex Li-based soft electrolytes toidentify the role of salt, anion, dynamics of the polymer host solvent, and coupling betweenrelaxation and transport on conductivity.

● Develop a predictive capability based on electronic structure calculations and formal sitemodels to optimize polyelectrolyte-based soft electrolytes for Li batteries.

● Identify optimized composite structuresthat utilize inorganic fillers in the polymeric hosts to

obtain high cation transference number and ion mobility.

Approach

● Use theoretical models, both of electronic structure type and of dynamical transport type,

to understand and predict the effects of specific variation of variables (temperature,density, salt choice, filler size, plasticizer, thermal history) on conductivity of polymer-based electrolytes.

Accomplishments

● Predicted and analyzed anionic electron density that produce higher numbers of mobile ionsin polyelectrolytes, and identified optimized aluminosilicate structures for simplepolyelectrolytes and composites

● Completed ab-initio calculations on polyelectrolyte structures, identi~ing mechanistically theion-dipole interactions that control free carrier number, and how to optimize them by polymerdesign.

31

● Completed a formal, doubly harmonic analytical model for analyzing and predicting effectsof polymer stiffness, site density, temperature and ionic radius on ion nobilities.

Future Directions

● Develop new methods to identify nanocomposite structures and hard polymer/salt complexes.

● Investigate the nature of the softening mechanism (how does the nanocomposite inorganicfiller lower the glass transition behavior of the polymeric host), and to optimize the freecarrier density based on this same local electronic structure in polyelectrolytes.

The effort is based on the idea thatappropriate theoretical models, closely coupledwith experimental advances and measurements,can identi$ improved polymer electrolytes forLi batteries. Polymer electrolytes werediscovered about 25 years ago, but most of theideas (coupling of relaxation to transport, low-glass-transition hosts, charge delocalizedcounter-ions) have not been meaningfullyextended in the last decade. The aim of thisresearch is to focus more explicitly onexperimental modifications to optimize theconductivity properties (and the complianceproperties and stability properties) of polymerelectrolytes.

The polymer/salt neat complexes wereextensively investigated, but they suffer fromlow conductivity (at best, ambient-temperatureconductivities near 10-4 S/cm are reported), andcation transference numbers are <0.5, leading toanion motion, cell polarization and inefficientcurrent flow. One solution to the problems is todevelop a stiff electrolyte with sufficientavailable free volume that decoupling can beattained. Theoretical analysis of these structureshas suggested both that decoupling occurs, andthat the conductivity should be of simpleArrhenius rather than bent WLF type. This isobserved, and ambient-temperatureconductivities are equal to those of the bestprevious polymer/salt complexes.

Polyelectrolytes of polymer-based structuresin which the anionic charge is bonded to thepolymer backbone was investigated. Thesematerials have unit transference number for Li+ions, therefore minimizing cell polarization.The difficulty to date has been that theirconductivities are even lower (often by a factorof 100) than the polymer/salt complexes. Two

research efforts aimed at improving thepolyelectrolyte conductivity were completed.Using ab-initio calculations (in collaborationwith Larry Curtiss at Argonne NationalLaboratory), we have shown thatalurninosilicates, which provide lower local-bond dipole densities, show weaker Lewisbasicities than aluminates, silicates, orborosilicates. This suggests that aluminosilicatestructures will be ideal polyelectrolyte hosts,both in simple structures and in composites.This is one of the first uses of ab-initiomethodology in polymer electrolytes, and itprovides an important conceptual link betweendesign of materials and transport properties.

A simple, doubly harmonic analytical modelfor understanding how host softness and ionsize, as well as carrier density, affect ionicconductivity in Li polyelectrolytes wascompleted. While this model is simple, itprovides analytic (as opposed to simplynumerical) predictions, and therefore can beused to design host polyelectrolyte stmctureswith optimized conductivity properties.

Polymer-based electrolytes that comprisecomposites, particularly organic hosts withinorganic fillers were analyzed. Two aspectswere studied: (i) ab-initio calculations suggestthat the silicate structures are not ideal, becausethere are insufficient free ions; (ii) thesubstantial variations in the glass transitiontemperature in nanocomposite materials, asopposed to macrocomposite materials, stronglysuggests that one can optimize the coupledtransport (as opposed to the decouplingtransport discussed above), and therefore thisnanocomposite polyelectrolyte structure wouldhave unit transference number, high mobility,

32

high conductivity and stable complianceproperties. M.

While the modeling in this area has justbegun, it will constitute the major effort in thefuture.

PUBLICATIONS

E. Sire, A.Z. Patashinski and M.A. Ratner, J.“Glass Formation and Local Disorder:Amorphization in Planar Clusters:’ J. Chem.Phys. 109,7901-7906 (1998).

Ratner, “Mathematics, ComputationalChemistry, and Battery Building: SomeNotions, in Modeling and Computation forApplications in Mathematics, Science, andEngineering,” J.W. Jerome, cd., OxfordUniversity Press, 1-24 (1998).

Kolafa and M. Ratner, “O1igomers ofPolyethylene Oxide): Molecular Dynamicswith a Polarizable Force Field,” MolecularSimulation 21, 1-26 (1998).

Corrosion of Current Collectors in Rechargeable Lithium Batteries

James W. Evans (Lawrence Berkeley National Laboratory)585 EvansHall, MC 1760, Universi~of cal~ornia, BerkeleyCA 94720(510) 642-3807,fu: (510) 642-9164; e-mail: evans@socrates. berkeley.edu

Objectives

● Examine corrosion of current collectors in Li-polymer and Li-ion batteries.

● Develop corrosion-resistant collectors and/or corrosion inhibition approaches.

“ Develop the electrochemical quartz crystal microbalance as a tool for investigating reactionsin Li batteries.

Approach

● Use electrochemical and electrochemical quartz crystal microbalance (EQCM) techniques todevelop understanding of the corrosion behaviour of current collectors and corrosivity ofvarious electrolytes, along with other reactions limiting the performanceflifetime of Li cells.

c Measure rates and identi~ corrosion processes of current collectors in Li batteries underdifferent charge conditions and systems by SEM analyses.

Accomplishments

● Identified the most serious pitting corrosion of Al current collectors in Li-polymer batteriesamong the combination of four salt candidates (LiCIOd, LiCF~SO~, Li imide and Li methide)and four cathode candidates (LiMn20J, LiCo02, TiS2 and VcO1~)during overcharging in thesystem Li / PEO + LiCF~SO~ / TiSz..

“ Observed from EQCM data that solvent oxidation and salt oxidation are more severe thancorrosion of Al (when the Al is protected by a passivating film) at high oxidation potentialsin contact with PC + Li imide.

Future Directions

c Identify corrosion products on the surface of Al current collectors using the EQCMtechnique.

● Develop a new solvent stable up to high oxidation potential such as ethylmethylsulfone andevaluate its performance.

33

.. . . . .

Potentiodynamic scans were used toinvestigate the corrosion behavior of currentcollectors in the presence of Li salt (LiCIO,,LiCF,SO,, Li imide and Li methide) inconventional electrochemical cells (consistingof Al working electrode, a counter electrode anda reference electrode with a Li salt/PEOelectrolyte). Pitting corrosion was previouslyidentified as the most severe form of corrosionexperienced by Al in these cells. Only a fewpits were observed on the Al working electrodeafter cyclic potentiodynamic anodic polarizationup to 5 V (vs. Li/Li+) with three Li salts (LiCIO,,Li imide and Li methide). However LiCF,SO,showed an increase in current density with cyclenumber and the collector displayed severepitting corrosion. The current density in thissystem increases to over O.lrnA/cm2, which isequivalent to a penetration rate of 1 mm/year, inthe potential region 4.2 V. Therefore, LiCF,SO,cannot be used when the operating voltageabove 4.2 V are encountered due to thecorrosion of Al current collectors.

To investigate the corrosion behavior of thecurrent collectors for the composite cathode(LiM~O,, LiCoO,, TiS, and V,O1,), cellsconsisting of a positive electrode, a negativeelectrode, and the appropriate current collectorswere employed. From the selection of LiMnzOd,LiCo02, TiS, and V,O,,, pitting corrosion of Alcurrent collectors is the most severe with TiS2.Direct SEM showed that pitting corrosion of Alinitiates below 4 V. Efforts to elucidate the

cause of corrosion of Al current collectorsattached to composite cathodes were made byvarying the amount of cathode material orcarbon. However the composition change hadlittle effect on Al corrosion.

Corrosion measurements with the EQCMwere conducted in carbonate-based propylenecarbonate (PC) or ether-based polyethyleneglycol) dimethyl ether (PEGDME) solutionscontaining LiPFc or Li imide. The EQCMenables us to observe very small mass change ofan electrode such as Al by measuring thefrequency change of a quartz crystal on whichthe Al is placed as a film. From the masschange, the species that take part in corrosionand passivation is revealed. There was noevidence of severe corrosion of as-received Al,which is protected by A120~ produced byatmospheric oxidation, in either LiPFb or Liimide. Even though the oxidation current washigh, the mass change of Al was very small,indicating that solvent oxidation and saltoxidation is more severe than Al corrosion at thehigh oxidation potentials used. However, thesituation is different for Al polished in the Aratmosphere in a glove box. It appears that Alreacts with the anion of Li imide and formssoluble corrosion products. On the other hand,Al is likely to form insoluble corrosion productsin LiPFc system. These investigations suggestthat darnage of the Al during manufacture ofbatteries in protective atmospheres may lead topitting corrosion.

Development of Novel Chloroaluminate Electrolytes for High-Energy-Density

Rechargeable Lithium Batteries

Kraig A. Wheeler

Department of Chemistry, Delaware State University, Dover DE 19901

(302) 739-4934, fa: (302) 739-3979; e-mail: kwheeler@dsc.edu

Objectives

“ Develop electrochemically stable sulfone electrolytes.

● Characterize and investigate the electrochemical and physical properties of these electrolyticmaterials for lithium rechargeable battery (LRB) applications.

34

Approach

“ Synthesize sulfones by direct akylation of mercaptan, followed by further oxidation

● Evaluate electrochemical and spectroscopic properties of electrolytes.

Accomplishment

● Reproducibly synthesized three aryl sulfones using nonaqueous, inert atmosphere techniques.

Future Directions

● Develop and characterize organic-based electrolytes with improved electrochemical stabilityfor LRB.

● Investigate the electrochemical properties of sulfones in collaboration with Case WesternReserve University.

The program focuses on the synthesis anddemonstration sulfone electrolytes withimproved electrochemical stability inrechargeable Li batteries. Sulfone-basedelectrolytes represent materials with improvedstability towards electrochemical processes aswell as reduced chemical toxicity. Thisprogram is directed towards the design andsynthesis of stabilized sulfone electrolytes.

The major thrust of the program is tosynthesize a family of structurally relatedsulfone-based compounds (1 - 4). Thesesolvents were systematically chosen toinvestigate the effect of varying the electron-donating/withdrawing ability of R and R’.Because of the electron-donating ability of thephenyl group, phenyl sulfones possess astronger C-S bond resulting in greater stabilitytowards oxidation.heduction.

o-1: —//CH3—

w o-2: —// CH&H2—

ii~— —~

o 0-3: —// CF3—

o-4—

// C6Hc$H2—

Aryl sulfones 1,2, and 4 have been successfullysynthesized in reasonable quantities (Fig. 16) bydirect alkylation of the phenyl mercaptan, which

formed the sulfide, followed by furtheroxidation with meta-chloroperbenzoic acid(MCPBA). These materials have been fullycharacterized by lH-NMR spectroscopy. Thesuccessful syntheses and characterization of arylsulfones 1 – 4 nearly completes phase one ofthis project.

The second phase of this project centers onthe purification and electrochemical studies ofaryl sulfones 1 – 4. Through a grant from DOD,we have recently been able to obtain an inertatmosphere glove box and a bipotentiostat toconduct electrochemical studies of theelectrolyte. The acquisition of this equipmentwill permit us to freely study and manipulatethese aryl sulfone compounds.

The glove box (Vacuum Atmospheres) wasinstalled in January 1999 and the purification ofsamples 1, 2, and 4 in an inert atmospherefollowed. The project’s capacity and strength atDSC centers on organic methodologies andcharacterization. As a result, in order to alsobecome proficient in electrochemical studiesselect personnel at DSU will travel to CWRU inApril 1999 to be trained on the operation andapplication of the newly purchasedbipotentiostat equipment. Such a collaborationwill not only provide the essential skills toinvestigate the electrochemical properties ofsulfones 1 – 4, but also strengthen the overallprogram at DSU in the area of electrochemicalmethods and applications.

35

. . . .. . ..—— ..--— --- -—— ----- ------- .-...-.7--., .—-, . . .. . . ----- .. . .,

o- 1. NaOEt—

o-

— MCPBA — j

\/sH~ \/s–R~ e\ / ;–R2. R-X o

Figure 16. Synthetic scheme for the preparation of aryl sulfones 1 – 4.

CROSS-CUTTING RESEARCH

Cross-cutting research is carried out to address fundamental problems in electrochemistry, current-density distribution and phenomenological processes, solution of which will lead to improved electrodestructures and performance in batteries and fuel cells.

Analysis and Simulation of Electrochemical Systems

John Newman (Lawrence Berkeley National Laboratory)201 Gilman Hall, MC 1462, UniversiQ of Calijomia, Berkeley CA 94720(510) 642-4063, fa: (510) 642-4778; e-mail: newman@newman.cchem. berkeley.edu

Objectives

● Improve the performance of rechargeable electrochemical cells by identifying the controllingphenomena.

● Identifj important parameters which are crucial in the operation of an advanced secondarybattery.

● Determine transport and other properties for electrochemical applications.

Approach

● Develop and implement mathematical models, computer programs, and characterizationexperiments to describe phenomenologically Li batteries and components.

Accomplishments

● Developed and demonstrated a Monte Carlo computer program to describe the concentrationdependence of the diffusion coefficient in a Li-intercalation electrode.

● Simulated a hybrid electric vehicle using a detailed battery model to guide design andoperation.

Future Directions

● Assess quantitatively a transport model for a plasticized, polymer electrolyte.

● Continue development of molecular dynamics program to determine diffusionelectrolyte systems.

coefficients in

● Develop improved experimental and theoretical model for thermal management of Libatteries.

36

This program rigorously researchesfundamental electrochemical phenomena andthe impact on rechargeable batteries for electricand hybrid vehicles. Mathematical models withminimal assumptions are developed to createcomputer programs and guide experimentalefforts. In-house numerical codes solve derivedsystems of equations to study coupled transport,kinetic, and thermodynamic problems ofelectrochemical interest. Experimental studiesseek to provide information crucial for modelimplementation. The overall effort results indesign and operational guidelines forelectrochemical devices and components.

Cathode studies. The dependence of thediffusion coefficient of Li+ ions in the LiJvf~OJlattice on concentration was modeled. Astatistical, Monte Carlo approach wasimplemented to model the short-time diffusionof Li. This simulation method determines thevariation of diffusionconcentration.

A 3-dimensional latticevariation of the diffusionmodel system consists of

coefficient with

model captures thecoefficient. The

a region of fixedtemperature, volume, and Li intercalantoccupancy in the host lattice. Using adiscretized grid, a block (high concentration) ofparticles is permitted to move instantaneouslyinto the surrounding, void (low concentration)space. Monitoring the movement of Li atomsover simulation time enables the calculation ofoverall and “jump” diffusion coefficients. Thelatter is the component limited purely byoccupancy (concentration). A thermodynamiccomponent is tractable from repeatedcomputations. Canonical ensembles areequilibrated over a range of occupancy andtemperature.

The results qualitatively match the measureddiffusion coefficients over a range ofconcentrations. Only trends are calculable bythe implemented Monte Carlo technique.Absolute values cannot be determined, givingrise to one fitting (scaling) parameter.Simulation results agree particularly well withexperiments at low Li state of charge.

Hybrid vehicles. Another modeling effortinvolved the load leveling of hybrid vehiclesusing dual Li intercalation electrodes. The

hybrid configuration consisted of a drive trainthat derived steady power from a conventionalengine, with a Li battery for instantaneouspower requirements. Parameters adjusted foroptimal performance included saltconcentration, electrode porosity, electrodethickness, and porous phase particle sizes.Batteries were optimized over a six-minute,partially urban driving cycle.

One study compared optimization by fieleconomy vs. minimal battery size. The systemresults with minimum fuel consumption werecompared with output for minimum electrodearea. The area-minimized case yielded a fueleconomy close to the maximum. In otherwords, no significant advantage in gas mileagewas gained in optimizing with respect to fueleconomy vs. size.

An analysis of the benefit of regenerativebraking was also completed. Batteries wereoptimized over the previously mentioneddriving cycle with three different percentages ofbrake energy recove~. A promising result wasthat the zero-emission range (or distanceaccessible without conventional power) wasnearly 40 miles with 66910recovery. Maximumbattery power, zero-emission vehicle range, andenergy efficiency increased with available brakerecovery. Unfortunately, size and weightincreased with percent recovery as well.

Additional simulations have addressed thetradeoff between fuel economy and power.Previous optirnizations of fuel economysubjected a hybrid-vehicle model to a six-minute driving cycle. A subsequentinvestigation utilized a more stringent one-minute cycle with greater power requirements.Optimized by the same criterion as for the six-minute cycle, the system demonstrated 10910fewer miles per gallon of gasoline consumed.This result was due to the 150q0 increase inbattery weight and lower cycle efficiency. Apositive finding, however, was the three-foldincrease in zero-emission vehicle range. Thismeasure of the total capacity is the distancetravelable under battery power alone.

An additional investigation considered fueleconomy for various vehicle weights, using thesix-minute driving cycle. Heavier vehiclesrequire larger, less-efficient batteries to satisfy

the power requirements of the cycle, and themileage-per-gallon decreased roughly linearlywith an increase in vehicle weight.

Transport studies. An effort has continuedto determine the transport properties in polymerelectrolytes. Initial work entailed thedevelopment of a mathematical model to derivethe desired properties from experiments. Amodel for transport was derived to calculate thediffusion coei%cient(s), transference numberand conductivity. The molar fluxes wereexpressed linearly in terms of the driving forces.These forces were the gradients of potential,solvent concentration, and salt concentration.Experimental steady-state flux or transientconcentration profiles would then enable thecalculation of properties through the linearrelations. An understanding of the governingtransport equations proved important indeciding the necessary techniques andmeasurements.

For the acquisition of new data onplasticized polymer electrolytes and otherelectrochemical systems, a new laboratoryfacility was established. This laboratory housesan argon-atmosphere glovebox for workinvolving oxygen-sensitive or water-sensitivechemistries. Its electrochemical analysiscapabilities include impedance spectroscopy,heat conduction calorimetry, and cyclicvoltammet.ry. Currently, the verification of thegalvanostatic polarization method, a crucialexperimental technique, is sought. Thistechnique is chosen to measure the cationtransference number. Modification of thenecessary instrumentation is underway. Amodel cell (silver, potassium chloride, and silverchloride) with known properties is constructedfor diagnostic testing.

Another effort involves a molecular dynamicsimulation to obtain transport properties ofliquid electrolytes. For such nonideal systems, acomputer program is being developed todetermine the concentration dependence of thediffusion coefficient. Molecular dynamics reliesupon the knowledge of the potential functionbetween atoms, molecules, or ions. Apromising, novel approximation using theFourier transform is being implemented to

capture long-range electrostatic forces inelectrolyte systems.

Thermal management. New work hasbegun in the thermal management ofelectrochemical cells. To address heatgeneration and thermal stability issues ofintercalation systems, a small model cell hasbeen constructed. This centimeter-scale coincell includes a Li foil anode, a polymerelectrolyte separator, and a spinel cathode.Common solvent and electrolyte are used forcomparison with a commercial system. Thecathode is a mixture of lithium manganese oxideintercalant and carbon. Initial diagnostic testingof the system is underway.

Double-layer capacitors. The steadyoperation of porous carbon double-layercapacitors was evaluated. First, a resistancemodel was developed consideringmacroporosity, a large particle-to-solutionconductivity ratio, and kinetic overpotential.The performance over a range of porosities andthicknesses were considered. The results fromvarious discharge profiles were compiled intodimensionless, universal curves that can be usedin deciding the design properties and cycle timesfor a given application. Specific power andpercent recovery of available energy werecalculated for galvanostatic and constant-powerdischarges. This model quantitativelyconfirmed the reported potential applications ofdouble-layer capacitors for frequent-cycle,short-discharge energy storage. On the otherhand, batteries and other storage devices provemore effective in power delivery over minute orlonger time scales. The concentrationoverpotential and microposity were evaluated.Gradients and, thus, overpotential are small forelectrolyte concentrations on the order of 1 M orgreater. However, pores below a criticaldimension may be too small to allow ions toenter. In this case, the system cannot deliver thecalculated energy density because ofinaccessible electrode area.

PUBLICATIONS

M.L. Perry, E.J. Cairns and J. Newman, “MassTransport in Gas-Diffusion Electrodes: ADiagnostic Tool for Fuel-Cell Cathodes,” J.Electrochem. Sot. 145,5 (1998).

/.

38

R.

T.

Darling and J. Newman, “Modeling SideReactions in Composite Li~n,O,Electrodes,” J. Electrochem. Sot. 145, 990(1998).

Jones, J. Newman, and other StandingCommittee members, Review of th~Research Program of the Partnership for aNew Generation of Vehicles, Board onEnergy and Environmental Systems,Commission on Engineering and TechnicalSystems, Transportation Research Board,

National Research Council, Washington,D. C.: National Academy Press, 1998.

K.P. Ta and J. Newman, “Mass Transfer andKinetic Phenomena at the Nickel HydroxideElectrode,” J. Electrochem. Sot. 145, 3860(1998). LBNL-41467.

R.M. Darling, “Lithium Manganese OxideSpinel Electrodes,” Ph.D. Dissertation,December 1998.

Electrode Surface Layers

Frank R. A4cL.urnon90-1142, Lawrence Berkeley National tiborato~, Berkeley CA 94720(510) 486-4636, fax: (510) 486-4260; e-mail: frmclarnon@lbl.gov

Objectives

c Apply advanced in situ and ex situ characterization techniques to study the structure,composition, formation and growth of surface layers on electrodes used in rechargeablebatteries.

● Identi@ film properties that improve the rechargeability, cycle-life performance, specificpower, specific energy, stability and energy efficiency of electrochemical cells.

Approach “

● Use ellipsometry, Rarnan spectroscopy, scanning probe microscopy, impedance analysis andother methods to characterize surface layers on secondary battery electrodes.

● Determine changes in electrode structure and surface morphology which accompany charge-discharge cycling, and seek new electrode-electrolyte combinations which lead to enhancedcell cycle-life performance.

Accomplishments

● Completed spectroscopic and microscopic studies of Ni electrodes in alkaline electrolyte, andelucidated the roles of Co and Li additions.

● Discovered a novel photochromic-electrochromic material based on Ni(OH), and TiO,.

Future Directions

● Use optical, spectroscopic and microscopic techniques to characterize surface layers on Li,carbon and metal oxide electrodes in non-aqueous electrolytes.

The primary objective of this research is to energy efficiency of electrochemical cells.identify film properties that improve the Advanced in situ and ex situ techniques arerechargeability, cycle-life performance, specific employed to characterize surface layers (solidpower, specific energy, stability, safety and

39

,,, . ,. ..... >,.-..;,..>. -..,,?,m. ..- .,- .= ..--_mm-m-?. ,.- .- . ----- ~ .. .,-------- w- >-. -.7.. ---

electrolyte interfaces, SEIS) on electrodes usedin rechargeable batteries.

The behavior of Ni electrodes in alkalineelectrolyte was investigated to better understandthe cause(s) of limited initial capacity, capacityfade and poor coulombic efficiency of thisimportant electrode. Research on theelectrochemical behavior of thinelectroprecipitated Ni(OH)2 films, a geometrywhich models the regions of active material in apractical battery electrode was completed.Raman spectroscopic studies were

complemented by using in situ atomic forcemicroscopy (AFM) to investigate the effects ofcharge-discharge cycling, Co co-precipitationand the uptake of Li from the electrolyte on thefilm morphology. A Ni substrate was selected,in contrast to studies described in the literaturewhich use a highly oriented pyrolytic graphitesubstrate, because we believe that interactionsbetween the electrochemically active Ni(OH)2film and the substrate play a key role inelectrode performance of a practical batteryelectrode.

An AFM image of a freshly deposited pureNi(OH)2 film obtained at open-circuit potential(O.13 V vs. Hg/HgO) showed -700-nmaggregates with clearly defined boundaries. Asingle charge-discharge cycle transformed thefilm morphology into a structure of closelypacked 50-nm grains. Further cycling led to theformation of deep crevices and produced ahighly disordered film with non-uniformelectrochemical activity. The areas close tomechanical fissures accumulated material, mostlikely products of Ni substrate oxidation. AFMimages of the film in its reduced and oxidizedstates after 60 cycles revealed that the filmmorphology varied strongly with the state-o-charge of the electrode (Fig. 17). In situmonitoring of a specific location on the filmsurface at different potentials showed that theexpansion and contraction that accompanycharging and discharging are non-uniform andsite selective. This finding supports our Ramanspectroscopy results which revealed theformation of a new inactive phase as theelectrode was cycled and linked this phase withelectrode capacity loss.

An AFM image of a freshly precipitatedNi(OH)2 film with 11 wt% Co showed that Coinduces significant modifications in the filmmicrostructure, i.e., a compact structure ofelongated randomly oriented -200-nm grainswas formed when Co is incorporated into Nisites. A single charge-discharge cycle inducedsignificant changes in the film microstructure,transforming it into a globular structure with- 150-nm grains, similar to that of a pureNi(OH)2 film. This type of film structure waspresent in both the reduced and oxidized states,and remained almost unchanged after 60 cycles.These results agree with our correlation betweenvoltarnmetric peaks and Raman spectra, andtaken together our results provide a ratherconvincing explanation of the role Co plays instabilizing the Ni electrode capacity.

In situ AFM measurements ofelectroprecipitated Ni(OH)2 films in 0.9 MNaOH + 0.1 M LiOH electrolyte revealed thatLi+ ions have a significant effect on the filmmorphology. AFM images of pure Ni(OH)2showed a cluster structure with grain sizesranging from 50 to 150 nm, however in thepresence of Li+ ions the film morphologyremained almost unchanged during early cycles.Prolonged cycling produced a rough and highlynon-uniform surface topography, although verylittle change occurred during charging anddischarging. AFM measurements of Co-dopedNi(OH)2 films in the Li+-containing electrolytefully supported our electrochemical andspectroscopic data. AFM images revealed avery uniform compact granular structure wasformed which remained stable during potentialexcursions and cycling.

The effect of Ni-film interracial processes onelectrode performance was investigated to detectthe products of these processes, determine theirformation mechanisms and assess their roles inelectrode. The goal is to develop Ni substrateprocessing steps to improve electrodeperformance. In an attempt to identify corrosionreactions, a series of chronoamperometricmeasurements was carried out. A freshlyelectroprecipitated thin-film Ni/Ni(OH),electrode was polarized at 0.0 V in lMNaOH, i.e., a potential at which the Ni(OH),phase remains in thermodynamic equilibrium

40

with pure Ni. Surprisingly, electrodepolarization was always accompanied bysignificant anodic currents that decayedexponentially in time. This current could not beattributed to charging of the active materialbecause the electrode potential was -450 mVbelow that for Ni(l[) + Nic’u oxidation. Apossible contribution might arise from theoxidation of Hz trapped in the Ni(OH)z filmduring the cathodic precipitation process,however similar measurements conducted with aPt substrate showed anodic currents one order ofmamitude smaller than at Ni. These results ledus to conclude that the Ni(0)tWbW)+. Ni(l’)cntiKiA~1~)oxidation reaction proceeds at the Ni substratein spite of the presence of the Ni(OH)z film.The product of anodic Ni substrate corrosioncan not only add capacity to the electrode, it canalso explain the frequent observation of excessanodic capacity during the fust cycle. On theother hand, too much corrosion can weaken theNi substrate and lead to poor contact with theactive material. There is also the possibility thatthe Ni corrosion products may have compositionand electrochemical properties that differ fromthose of electroprecipitated Ni(OH),.

To gain deeper insight into this problemanother series of tests was conducted. Prior toNi(OH)2 precipitation, the Ni substrate wasrepetitively cycled in 1 M NaOH between 0.0and 0.6 V at 10 mV/s for 2 h. Although thiscycling led to the formation of anelectrochemically active passive layer(documented by a pair of Ni(lI) e Ni(l’l’voltammetric peaks), a Ni(OH), filmsubsequently electroprecipitated on top of itperformed very poorly, and the electrodecapacity decayed much faster than usual.

PUBLICATIONS AND PRESENTATIONS

F, Kong, R. Kostecki, F. McLarnon and R.H.Muller, “Spectroscopic Ellipsometry ofElectrochemical Precipitation and Oxidationof Nickel Hydroxide Films,” Thin SolidFilms, 313-314,775-80 (1998).

F.

R.

F.

J.

R.

M.

R.

Kong, R. Kostecki and F. McLarnon, “h SituEllipsometric Study of theElectroprecipitation of Nickel HydroxideFilms:’ J. Electrochem. Sot. 145, 1174-78(1998).

Kostecki, T. Richardson and F. McLarnon,“Photochemical and PhotoelectrochemicalBehavior of a Novel TiOz/Ni(OH)zElectrode,” J. Electrochem. Sot., 145,2380-85 (1998).

Kong, J. Kim, X. Song, M. Inaba, K.Kinoshita and F. McLarnon, “ExploratoryStudies of the Carbon/Nonaqueous-Electrolyte Interface by Electrochemical andIn Situ Ellipsometry Measurements,”Electrochemical and Solid State Letters, 1,39-41 (1998).

Kim, F. Kong, X.Y. Song, M. Inaba, K.Kinoshita and F. McLarnon, “Studies of theCarbon/Nonaqueous Electrolyte Interface byElectrochemistry, In Situ Ellipsometry andImpedance Spectroscopy,” paper no. 72

presented by J. Kim at the 193rd Meeting ofthe Electrochemical Society, San Diego, CAMay 1998.

Kostecki and F. R. McLarnon, “In Situ

Atomic Force Microscopy Study of Co andLi-Doped Nickel Hydroxide Films,” paper

no. 36 presented by R. Kostecki at the 194rhMeeting of the Electrochemical Sociey,Boston, MA November 1998.

Inaba, R. Kostecki, K. Kinoshita and F.McLarnon, “Ti-, Zr-, Nb- and Me-DopedLithium Manganese Oxide PositiveElectrodes:’ paper no. 125 presented by M.

Inaba at the 194’h Meeting of theElectrochemical Society, Boston, MANovember 1998.

Kostecki and F. McLarnon, “NovelMonolithic Electrochromic DeviceControlled by Incident Illumination,” LBNLProvisional Patent Application (1998).

Figure 17. AFM images of a Ni(OH), film after 60 cycles in 1.0 M NaOH ~

Lithium Electrode Interracial Studies

Philip N. Ross, Jr.2-100, Lawrence Berkeley National Laboratory, Berkeley CA 94720(510) 486-6226, fm: (510) 486-5530; e-mail: pnross@lbLgov

Objectives

o Develop an understanding at the molecular level of the reactions that occur at theLi/electrolyte interface to form the solid electrolyte interracial (SEI) layer.

o Investigate the role of solute ions and solvent impuritiesprocess.

Approach

● Apply a combination of UHV surface analytical methods

Accomplishments

(such as water) on the film forming

and in situ ellipsometry.

● Studies of the reactions of clean metillic Li with the following solvents were completed:ethers - 1,3 dioxolane and 1,2 dimethoxyethane (DME); carbonates - dimethyl carbonate(DMC) and diethylcarbonate (DEC).

● A study of the reaction of clean metallic Li with CO, was completed.

Future Directions

● Study the interaction of LiC, with the solvents used in Li-ion batteries, e.g., DMC and DEC.

● Study the chemistry of intercollation of Li in graphite in UHV using photoelectronspectroscopy.

42

The reactions of dimethyl carbonate (DMC)and diethyl carbonate (DEC) with clean metallicLi in UHV were studied by X-ray photoelectronspectroscopy (XPS) using a temperature-programmed-reaction methodology. Bothmolecules are of interest as solvents in ambient-temperature Li batteries. The solvent moleculeswere condensed onto the surface of anevaporated Li film at 120 K, and spectra werecollected as the sample was warmed in ca. 25-30 K increments. The reaction of either DMCor DEC with Li was initiated at 180 K, atemperature much lower than their bulk meltingtemperatures, producing lithium methylcarbonate, methyl lithium and lithium ethylcarbonate and ethyl lithium, respectively. Attemperatures above 270-300 K, the lithium alkylcarbonates start to decompose with Li20,elemental carbon and alkyl lithium as productson the surface. Both DMC and DEC are morereactive towards metallic Li than propylenecarbonate. Since methyl and ethyl lithium arehighly soluble in the parent solvent,electrodeposited Li is predicted to have poorstability in an electrolyte composed of. eitherDMC or DEC.

A study of the reactions of clean metallic Liwith two other prototypical ethereal solvents,1,2 dimethoxyethane (DME) and 1,3 dioxolane(1,3 -DOL) was also completed using the samemethodology. The interracial reactions weredistinctly different for the two solvents. Thereaction of Li with DME has similarities to thereaction with tetrahydrofuran (THF) studiedpreviously. Initiation of the reaction occursupon melting of the condensate layer at ca.220”K, with cleavage of the RC~-OC~ bond

and formation of C~O-Li and CI$OC~CH,-Li, and a small amount of ethylene. UnlikeTHF, however, there is no evidence ofpolymerization. In contrast, polymerization of1,3-DOL is the only reaction occurring at theinterface throughout the temperature range of180-300”K. The polymer consists of a mixtureof the repeat units –C~C~ C~-, -OCH20-,and -OC~C~- in the ratio of approximately1:1:2. The mechanism of this polymerization isunknown to us, but our spectroscopic dataclearly indicate it is probably initiated withoutthe formation of an RO-Li intermediate. Thus,1,3 DOL is the least reactive solvent of any thatwe have studied, and forms a SEI layer that is

“close to the same molecular conformation as thesolvent itself. This surface chemistry explainsthe high cycling efficiencies of Li in electrolytesbased on 1,3 DOL.

PUBLICATIONS

G.

G.

G.

Zhuang, K. Wang, Y. Chen, and P. Ross,“Study of the Reactions of Li withTetrahydrofuran and Propylene Carbonateby Photoemission Spectroscopy,” J. Vat.Sci. and Technol. A, 16,3041 (1998).

Zhuang, Y. Chen, and P. ROSS, “TheReaction of Li with Carbon Dioxide Studiedby Photoelectron Spectroscopy,” Surf Sci.,418, 139 (1998).

Zhuang, Y. Chen, and P. Ross, “TheReaction of Li with Dimethyl Carbonate andDiethyl Carbonate in UHV Studied byXPS,” Langmuir, 15,470 (1999).

43

-- .-,..- V+Tm.-. . . ,.,= ..-~,.-.--,... ....------ -. . .. . . ..-.--=-..,.-,7...- --z------- ---- ..- .-- -..= , .

ACKNOWLEDGMENTS

This work was supported by the Assistant Secretary for Energy Efficiency and Renewable Energy,

Office of Advanced Automotive Technologies of the U.S. Department of Energy under Contract No.

DE-AC03-76SFOO098. The support from DOE and the contributions by the participants in the ETR

Program are acknowledged. The assistance of Ms. Susan Lauer for coordinating the publication of this

report and Mr. Garth Burns for providing the financial data are gratefully acknowledged.

LIST OF ACRONYMS

AFM

BET

BNL

CB

DEC

DMC

DME

DOD

DOE

DOL

DSC

DTA

ECPS

EC

EIS

EMA

EMC

ENMR

EQCM

ETR

EV

EXAFS

HDPE

atomic force microscopy

Brunauer-Emmett-Teller

Brookhaven National Laboratory

comb-branch

diethyl carbonate

dimethylcarbonate

dimethoxyethane

Department of Defense

Department of Energy

dioxolane

differential scanning calorimetry

differential thermal analysis

electrochemical potential spectroscopy

ethylene carbonate

electrochemical impedance spectroscopy

effective medium approximation

ethyl methyl carbonate

electrophoretic nuclear magnetic resonance

electrochemical quartz crystal microbalance

Exploratory Technology Research

electric vehicle

extended X-ray absorption fine structure

high density polyethylene

44

HQ

ICL

LBNL

LLNL

LRB

MCMB

MCPBA

NMR

PAE

Pc

PDI

PEG-DME

PEMO

PEO

PEPE

PNGV

PPO

PVE

SEI

SEM

Soc

TCO

TEM

TGA

THF

TMA

USABC

UVIVIS

WLF

XAS

XPS

XRD

HydroQuebec

irreversible capacity loss

Lawrence Berkeley National Laboratory

Lawrence Liverrnore National Laboratory

lithium rechargeable battery

mesocarbon microbead

meta-chloroperbenzoic acid

molecular weight

nuclear magnetic resonance

polyacqdate ether

propylene carbonate

polydispersity indices

polyethylene glycol - dimethyl ether

polyethylene-methylene oxide

polyethylene oxide)

polyepoxide ether

Partnership for a New Generation of Vehicles

polypropylene oxide

polyvinylether

solid electrolyte interface

scanning electron microscopy

state-of-charge

tin-based composite oxide

transmission electron microscopy

thermogravimetric analysis

tetrahydrofuran

tetrarnethyl ammonium

United States Advanced Battery Consortium

ultraviolet-visible

Williams-Landel-Ferry

X-ray absorption spectroscopy

X-ray photoelectron spectroscopy

X-ray diffraction

45

ANNUAL REPORTS I

1. Exploratory Technology Research Program for Electrochemical Energy Storage – Annual Reportfor 1997, LBNL-41950 (June 1998).

2. Exploratory Technology Research Program for Electrochemical Energy Storage – Annual Reportfor 1996, LBNL-40267 (June 1997).

3. Exploratory Technology Research Program for Electrochemical Energy Storage – Annual Reportfor 1995, LBNL-338842 (June 1996).

4. Exploratory Technology Research Program for Electrochemical Energy Storage – Annual ReportL

for 1994, LBL-37665 (September 1995).

5. Exploratory Technology Research Program for Electrochemical Energy Storage – Annual Reportfor 1993, LBL-35567 (September 1994).

6. Exploratory Technology Research Program for Electrochemical Energy Storage – Annual Reportfor 1992, LBL-34081 (October 1993).

7. Exploratory Technology Research Program for Electrochemical Energy Storage – Annual Reportfor 1991, LBL-32212 (June 1992).

8. Technology Base Research Project for Electrochemical Energy Storage – Annual Report for 1990,LBL-30846 (June 1991).

9. Technology Base Research Project for Electrochemical Energy Storage – Annual Report for 1989,LBL-29155 (May 1990).

10. Technology Base Research Project for Electrochemical Energy Storage – Annual Report for 1988,LBL-27037 (May 1989).

11. Technology Base Research Project for Electrochemical Energy Storage – Annual Report for 1987,LBL-25507 (July 1988).

12. Technology Base Research Project for Electrochemical Energy Storage.– Annual Report for 1986,LBL-23495 (July 1987).

13. Technology Base Research Project for Electrochemical Energy Storage – Annual Report for 1985,LBL-21342 (July 1986).

14. Technology Base Research Project for Electrochemical Energy Storage – Annual Report for 1984,LBL-19545 (May 1985).

15. Annual Report for 1983 – Technology Base Research Project for Electrochemical Energy Storage,LBL-17742 (May 1984).

16. Technology Base Research Project for Electrochemical Energy Storage – Annual Report for 1982,LBL-15992 (May 1983).

17. Technology Base Research Project for Electrochemical Energy Storage – Report for 1981, LBL-14305 (June 1982).

18. Applied Battery and Electrochemical Research Program Report for 1981, LBL-14304 (June 1982).

19. Applied Battery and Electrochemical Research Program Report for Fiscal Year 1980, LBL- 12514(April 1981).