NASA Technical Reports Server (NTRS) 20000056911: Development of High Conductivity Lithium-Ion Electrolytes for Low Temperature Cell Applications

5 Pages·0.35 MB·English

by  NASA Technical Reports Server (NTRS)

,T
Development of High Conductivity Lithium-ion Electrolytes for Low
Temperature Cell Applications
M.C. Smart, B.V. Ratnakumar, S. Surampudi
Jet Propulsion Laboratory
California Institute of Technology
4800 Oak Grove Drive
Pasadena, CA 91109
Introduction
and reported earlier.(5) Other groups have also
NASA has continued interest in
reported electrolytes based on mixtures of
developing power sources which are capable of
carbonates and acetates.(6) Inthe present study,
r operating at low temperatures (-20°C and below) to electrolytes which have been identified to have
enable future missions, such as the Mars Rover
good low temperature conductivity and stability
and Lander. Thus, under aprogram sponsored by
were incorporated into lithium-graphite cells for
the Mars Exploration Program, we have been
evaluation. Using various electrochemical
involved in developing Li-ion batteries with
methods, including ac impedance and DC
improved low temperature performance. To
micropolarization techniques, the film formation
accomplish this task, the focus of the research has
characteristics of graphite electrodes in contact
been upon the development of advanced
with various electrolyte formulations was
electrolyte systems with improved low temperature investigated.
properties. This had led to the identification of a
carbonate-based electrolyte, consisting of 1.0 M
LiPF6 in EC + DEC + DMC (33:33:34), which has Experimental
been shown to have excellent performance at The specific conductivity of a number of
-20°C inLi-ion AA-size prototype cells.(1,2) Other electrolyte solutions was measured over the
groups are also actively engaged in developing temperature range of -60°C to 25°C using a
electrolytes which can result in improved low conductivity cell which consists of two platinized
temperature performance of Li-ion cells, including platinum electrodes which are immobilized in a
Polystor (3), Yardney, and Covalent. (4)
glassapparatus andseparated by afixed distance.
In addition to developing cells capable of The cellconstantofthe apparatus wasdetermined
operation at -20°C, there is continued interest in bycomparing the resistivityofa 0.1M KCI solution
systems which can successfully operate at even with the value reported in the literature. The
lower temperatures (< -30oC) and at high temperature wascontrolled in these experiments
discharge rates (> C/2). Thus, we are currently by utilizing a Tenney environmental low
focusing upon developing advanced electrolytes temperature chamber (+/- 1°C).
which are highly conductive at low temperatures
Half-cellstudies were carded out using O-
and will result in cells capable of operation at
ring-sealed, glasscellswhichcontained jelly rollsof
-40°C. One approach to improve the low graphite (KS44) and LiCoO2(Alfa) electrodes with
temperature conductivity of ethylene carbonate-
porous polypropylene separator (Celgard 2500)
based electrolytes involves adding co-solvents
and Lireference electrodes. Allelectrolyte were of
which will decrease the viscosity and extend the battery-grade purity. Electrochemical
liquid range. Candidate solvent additives include
measurements were made using an EG&G
formates, acetates, cyclic and aliphatic ethers, Potentiostat/Galvanostat (and Solartron 1255
lactones, as well as other carbonates. Using this Frequency Response Analyzer for impedance
approach, we have prepared a number of
measurements) interfaced with an IBM PC, using
electrolytes which contain methyl formate (MF), Softcorr 352 (and M388 for impedance) software.
methyl acetate (MA), ethyl acetate (EA), ethyl Cycling Data were collected using an Arbin battery
proprionate (EP), and 1,2-dimethoxyethane test system.
(DME), some of which have been characterized
ResultsandDiscussion
solution is a good candidate for ultra-low
ConductivityMeasurements temperature applications, due to the fact that the
conductivity was observed to have greater than 1
A numberof EC-basedelectrolytes
mS/cm at -60°C. We also initiated investigations
containing lowviscosityaliphatic and cyclic edgers
into the effect of mixed salt solutions upon the low
have been prepared and the conductivity
temperature conductivity and compared 1.0M
measured over a temperature range of -60°C to
LiPFe solutions with 0.50M LiPF6 + 0.50M LiAsF6
25°C. The electrolytes assessed consisted of
solutions, although little benefit was observed
baseline formulations, 1.0M LiPF8 EC + DEC +
when the conductivity was measured at low
DMC (1:1:1), to which lowviscositysolvents were
temperatures.
added, including: methyl acetate, ethyl acetate,
ethyl butyrate, ethyl propionate, and methyl Lithium-Graphite Half-Cells
formate. Of the electrolytes investigated, the
A number of lithium-graphite cells have
formulations that displayed the highest
been fabricated to study the effect of different
conductivity at low temperatures were ones with
electrolytes upon the film formation characteristics
lowermolecularweight acetates and displayed the
on graphite electrolytes and the irreversible and
following trend: > 1.0M LiPF6EC + DEC + DMC +
reversible capacities of graphite electrodes. The
MA (1:1:1:1) > 1.0M LiPF6 EC + DEC + DMC + EA
lithium-graphite half-cells can serve as an additional
(1:1:1:1) > 1.0M LiPF8 EC + DEC + DMC + EP
screening test to identify the compatibility and
(1:1:1:1). The conductivity values for these
stability of candidate electrolytes with
solutionsisshown inFig. 1.
carbonaceous electrodes. The electrolytes
selected in the first group of cells include: 0.75 M
LiPFsEC + DEC + DMC (1:1:1), 0.75 M LiPF6EC +
1oo DEC (30:70), 0.75 M LiPF6 EC + DMC (30:70),
,oI _MEA,A._.m.e'_l,he___K._:eettamme'e . 0.75 M LiPF6EC + DEC +DMC + MA (1:1:1:1), and
0.75 M LiPF6EC + DEC + DMC + EA (1:1:1:1), and
0.75 M LiPF6 EC+ DEC+ DMC + DME (1:1:1:1).
As shown in Fig. 3., electrolytes based
J
upon the addition of aliphatic esters, such as
methyl acetate and ethyl acetate, were observed
,,L 1.0M IIPF6 EC+DEC+DMC+EB (1:1:1:1 to have small irreversible capacities (5th cycle).
•.e. 1,0M LiPF6 EC+DEC+DMC (1:1:1)
This suggests that desirable protective interfacial
oo60 -40
2O 40 films are formed on the electrode surface with
these electrolytes. The exclusively carbonate-
based electrolytes showed similar behavior, with
Fig.1. Theeffectoflowviscosityadditivesuponthelow
higher DMC content resulting in more protective
temperature conductivity of EC-based electrolyte
solutions. surface films. Incontrast, the electrolyte with DME
showed the highest irreversible and lowest
reversible capacity, suggesting that the solvent is
In addition to investigating low viscosity
more reactive with the lithiated carbon and/or the
additives, we have initiated research into the effect
electrolyte salt and forms less desirable surface
of salt type and salt concentration upon the
films. Overall, the reversible capacities obtained
conductivity of electrolyte solutions at low
for the lithium-graphite cells studied, as shown in
temperature. It was determined that solutions
Fig. 4, were somewhat lower than that observed in
consisting of 0.75M LiPFeEC + DEC + DMC (1:1:1)
experimental and prototype LiCoO2-graphite cells
showed an improvement in conductivity at
which can be attributed to the selected charge
temperatures below -20°C compared with 1.0M
voltage (0.025 V vs Li/Li+) and the cell design
LiPF6 EC + DEC + DMC (1:1:1). In addition, the
which has excess electrolyte. Some correlation
electrolyte formulation 0.SM LiPF6 EC + DEC +
between the extent of irreversible capacity lost in
DMC + MA (1:1:1:1) displayed better conductivity
the first cycle and how quickly the cell reflects
at low temperatures below -40°C compared with
potentials indicative of the lithium intercalation
the 1.0M EC + DEC + DMC + MA (1:1:1:1)
reaction (>150 mV vs. Li/Li*) was observed as
solution, with little loss in the room temperature
shown in Fig. 3.
values (> 10 mS/cm @ 23oc). This electrolyte
Electrolyte Formulation R_C,_ _'_ _,C_ !_C=p Electrolytes incorporating acetate additives, such
I==¢y¢ 1=1¢yc. !sin C_. !m Cy{
075ML*PF I EC.0MC*OEC(1 1I) 227.2 106.0 240.4 12711 as MA or EA show similarbehavior to the ternary
0.75M _ptr. EC. OMC (3070) 302.0 94.3 312.6 122.9 carbonate mixture,as shown in Fig. 6. Whereas,
the electrolyte possessing DME displayed the
07SM L_F eEC •DEC (30:70) 268.1 1068 27S.4 136.9'
largest film resistance consistent with the high
075M LiPF6 EC *OMC _OEC .MA {1;1:1:1) 201.5 _+9 23_.5 S6.g
irreversiblecapacityobserved withthiscell. The ac
07SMLiPF4 EC+OMC*0EC+EA (I 11,1) 210.4 49.9 214.2 68+5/ ; impedance measurements were repeated after the
07SM bPF4 EC+0MC -OEC .OME (111:1) 147.5 138,2 IM.3 192-_ cells had been subjected to a number of cycles
4_
and aself-discharge study, asshown in Figs.7 -8.
Fig.2. Reversible and irreversible capacities of lithium- In the case of the solely carbonate-based
graphite cells possessing different EC-based electrolytes, little variation was observed in the
electrolyte formulations. Nyquist plots generated. However, a dramatic
increase in the overall cell resistance and the
charge transferresistance wasobserved withcells
410.75M LJPF6 EC.¢.OEC+It_MC (1:1:1} !
possessing electrolytes with MA or EA, as shown
.0.75M LiPF6 EC+OMC (30:70)
e.le in Fig. 7. These results suggest that these
e0.75M LJPI=6 EC+OEC (30.70) i
g! II0,7SM LiPF6 EC.,.OEC+-DMC+_ (I:1:1:1 ) i electrolytes continue to react with the lithiated
• 0.75M I.JPF6 EC+OEC+-O_+E.4 (1:1:1:1}
"0.75M LIPF6 EC+DEC+OMC+OME (1:1:1) i carbon electrode and/or other in situ generated
! byproducts after the initial "formation" cycles. The
impact of temperature upon the film resistance was
also investigated for all of the samples and
displayed similarbehavior to thatshown inFig.9.
(l,04
e.ooe I,I0O 4.|44 4.)e* 0.400 o+iee
Cho+ge Cap=cRy (Ah) O+7
Fig. 3 First lithiumintercalation cycle of lithium-graphite
0.41
cells possessing different EC-based electrolyte ,0.75M LIPF6 EC +DEC (30:70)
formulations.
0.$ •0.75M L;PF6 EC+OMC÷DEC (1:1:1)
0.4 10.TSM UPFI; EC +OMC (30:70)
!.40 _1, 0.3
o.eo 0.2
oo..Tlet I 0,I
e+go
0.0
oseJ 6 5 4 I 3 2 0 0.1 0.2 0.3 0.4 O.S 0.1 0.7
0.40 Z' (Ohms)
]
Fig. 5. AC impedance (Nyquist) plots of lithium-graphite
cells possessing carbonate-based electrolytes after
formation cycles.
e.ee
ILOO O,Ol O.ll 0.11 l+:lo 0.21 11,34
1,2
F-mj.4. Reversible capacity (Ah/g) of lithium-graphite
cells possessing different EC-based electrolyte
;1.0.75M LJPF6 EC+DMC+OEC (1:1:1)
formulations (1= EC+ DEC + DMC, 2=EC+ DMC, 3=EC +
_' 1.o I-0.75M LJPF6 EC+OMC +DEC,.MA (1:1:1:1)
DEC,4= EC+DEC + DMC+ MA, 5= EC+ DEC + DMC +
,_ o, -j,o_mupF+scoo,c.oec._A (+:1:1:I>
EA,6=EC+ DEC + DME).
[,0.75M UPF6 EC +DMC +OEC +OM[ (1:1:1:1)
_1 0.1
AC Impedance Measurements
0.4 -
In addition to studying the o../...ii .II ,iiI,i,n.l ,Iii ii iiiIx,.,,
charge/discharge characteristics of these cells, a.c.
impedance was used to probe the interfacial 0.0 ___ __
properties of the anode passivating films on the • o.:, o+, o.i o+e I +._,
Z" (Ohm=)
graphite electrodes. As shown in Fig. 5, the
Fig.6. AC impedance (Nyquist) plots of lithium-graphite
carbonate based electrolytes with high DMC
cells possessing carbonate-based electrolytes with
content result in cells with lower charge transfer different low viscosity additives, MA, EA, and DME,
resistance compared with DEC rich electrolytes. after formation cycles.
Fig. 10 and 11, the cells that had electrolytes
07O
which resulted in higher irreversible capacity
• I 07_Mt.JPF41 EC*OEC*I_t*4C(I;tl)
losses, such as DME and DEC, displayed the
OeO a 075ML_'6 ECoOMC{3O:?0)
• :1 07_MUPF6 EC*DEC(30:70) lowest polarization resistance. The improved
0SO _!
311 kinetics of Li intercalation observed with these
• A = I_lial IIA_lsummil_l
040 cells may be due to the production of porous, non-
v_ 030 : 8 • Aft,, Cyclk_ protective films which facilitate Li ion diffusion
_o20 and/or charge transfer. In contrast, the cells
containing electrolytes with MA or EA produced
o=°'° films which resulted in high polarization resistance
of the electrode. This is undesirable for low
o(x} 010 020 03o 040 0S0 0_0 070 temperature applications where the effect is
Z' (Ohms)
magnified and high rates become difficult to
Fig. 7. AC impedance (Nyquist) plots of lithium-graphite sustain.
•cells possessing carbonate-based electrolytes before
The limiting current densities were also
and after cycling.
F determined for the lithium deintercalation process
from the graphite electrodes by conducting Tafel
1zoo polarization measurements to evaluate the rate
capability of the electrodes in contact with the
10,00 • 07SMI.JPFe EC*OEC_DMC+MA(t:I:t:t)
• 07_3MI._FO EC_DEC÷OMC+EA(tt:I:I) various electrolytes. The results obtained
correlate well with the DC micro-polarization
measurements in that the cells possessing high
_ 6_OO
polarization resistance, i.e. with electrolytes
4oo After Cycling
Ilnitisl Measurement possessing MA or EA, have lower diffusion limiting
currents (measured at an overpotential of 250 mV).
0`00 200 dl00 0`00 100 10.00 1100
Z' (Ohms)
Fig.8. AC impedance (Nyquist) plots oflithium-graphite
cells possessing carbonate-based electrolytes with
different lowviscosity additives, MA and EA, before and
after cycling.
22.$0I 075 M I.JPF6 EC +DEC _.DMC ÷MA(It1:l:f)
ic.u¢ T..*._e T.,. |e.1_¢ Tw..llJ T_*.|A
_ 1s.oo
Elect_ryte "i'yl_
1 11.._
Fig. 10 Linear polarization resistance calculated from
IX; micropolarJzation plots of graphite electrodes with
different electrolytes atroom temperature.
5'C &4&li&d6=_a=&ll 6&_
,_1,._.. ,,-- o_
ooIoXT"""-'_-
=e
o.oo 2_N s.oo 7.so to.oo 12.50 _.¢00 tT.IW ZEo_ 21.54 _._ 11• l¢. c1[¢.OM¢•F._
Z"(Ohms)
Fig. 9. AC impedance (Nyquist) plots of a lithium-
graphite cell possessing EC+DEC+DMC+MA electrolyte
at various temperatures. ii ,
DC Micropolarization Measurements
DC micropoladzation techniques were also 'IM!
employed to study the charge transfer behavior of
the passivating films on the graphite electrodes at ElectmMe Type
various temperatures. The polarization resistance Fig 11 Polarization resistance ofgraphite electrodes in
of the electrodes was calculated from the slopes of contact with EC-based electrolytes ofdifferent types as
the linear plots generated under potentiodynamic a function oftemperature
conditions at scan rates of 0.02 mV/s. As shown in
I
Low Temperature Discharge Capacity
these highly conducting electrolytes display low
After performing the electrochemical
irreversible capacity losses during the formation
measurements described above, the discharge
cycles, such as with MA and EA, AC impedance
capacities were determined for the various cells
and DC micro polarization methods have shown
under similarconditions. As shown inFig. 12,'the
that these cells have high polarization and charge
delivered capacities at -20°C (25 mA current to 1.5
transfer resistance. Thus, the benefits of the
V cut-off vs. Li/Li+)correlate well with the results
highly conducting medium is offset by the poor Li÷
obtained with the DC micropolarization intercalation-deintercalation kinetics of the
techniques. Due to the high polarization
graphite electrodes produced by the nature of the
resistance observed at graphite electrodes in interfacial films.
contact with MA or EA-containing electrolytes the
delivered capacities were lower than that of the References
ternary carbonate mixture or the DME-containing
electrolyte. Although differences are anticipated 1)M.C. Smart, C.-K. Huang, B.V. Ratnakumar, and
when these electrolytes are evaluated in
S. Surampudi, Proceedings of the Intersocity
r experimental or prototype LiCoO2-carbon cells due Energy Conversion Engineering Conference
to cell design, pack tightness, and electrolyte (IECEC), Honalulu, Hawaii, July, 1997.
volume, the observed trends should be
consistent.
2) M.C. Smart, B.V. Ratnakumar, C.-K. Huang, and
S. Surampudi, SAE Aerospace Power Systems
IJDO Conference Proceedings P-322, 1998, p. 7-14.
160 4b 3b 2b Ib
'''Oy
3) Juzkow, M.W., Proceedings 14th International
20 Primary and Secondary Battery Technology and
t.O0 Applications Seminar, 1997.
O.B0
060 4) Ein-Eli, Y., et al, J. Electrochem.Soc., 1996,
_ 3" ,_ 142, L273.
040 - ........... :.... _h ,j za_ ,,,,
4,
0.20 ,_. _..* ..___, .....
ooo m 5) Smart, M.C.; Huang, C.-K.; Ratnakumar, B.V.;
o.oo 0.050 0,100 0.150 0.200 0.250 0.30q and Surampudi, S., 1996, Proceedings, 37th
c._,:_ IAh)
Power Sources Conference, pp. 239-242.
Fig. 12 Discharge curves of graphite electrodes in
different electrolytes (I=EC+DEC+DMC, 2=EC+DEC
+DMC+DME, 3=EC+DEC+DMC+EA, and 4=EC+DEC+ 6) Ohta, A.H., Koshina, H.; Okuno,, and H. Mural,
J.Power Sources, 1995, 54, 6.
DMC+MA)at different temperatures (A=25°C,B=-20°C).
Acknowledgment
Conclusions The workdescribed here wascarded out at
A number of EC-based electrolytes with the Jet Propulsion Laboratory, California Institute
improved low temperature conductivity have been of Technology, for the Mars Exploration Program
and a DARPA TRP program under contract with
identified. These electrolytes are based on ternary
mixtures of organic carbonates to which solvents the National Aeronautics and Space Administration
(NASA).
possessing low viscosity and low melting points
are added, such as, MA, EA, and DME. The
compatibility of these electrolytes with graphite
anode materials was assessed in lithium-graphite
half-cells using various electrochemical
techniques. Although some cells incorporating