Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
~ ~5~305
This invention relate~ to room temperature
rechargeable nonaqueous cells having active metal anodes such
as lithium and more particularly to such cells having sulfur
dioxide electrolyte svlvent/cathode depolarizers.
In the past a considerable amount of effort has been
expended in the development of a viable, practical and
commercially acceptable room temperature operable rechargeable
lithium cell which would have the advantages over the co~mon
rechargeable lead-acid and nickel-cadmium batteries of higher
efficiency, lower weight and greater primary lifetimes thereby
allowing for more time between charging cycles. Such efforts
have met with varying degrees of success. However, such cells
have rarely achieved greatex than 80% recycling efficiencies
over extended charge and discharge cycles. These cy~ling
efficiencie~ are to be differentiated from the very high
lithium plating efficiencies since lithium plates out of
electrolyte solutions, commonly used in lithium cells, with
close to 10~ efficiency (the exchange current of the reaction
Li~,Lit + e is very high-on the order of 1 x 10 3A/cm2).
The recycling efficiencies are instead related to subse~uent
anodic oxidation of the plated lithium wherein generally the
effectiveness of the plated lithium as an anode material
decreases rapidly upon cycling despite the high plating
efficiencies.
1 ~5~5
M-3531
Several reasons have been postu]ated ~or the inefficlency of
plated lithium for repeated anodlc oxidation and of cells containing
such anodes. One reason given is that lithium is dendritically deFosited
and becomes coated,particularly at the narrow point of contact with
the anode substrate with an insulating film which electrically insulates
it from the anode substrate despite its physical presence on the anode.
The plated lithium dendrites as they become electrically insulated
become unavailable for efficient anodic oxidation during discharge.
Furthermore, the plated lithium dendrites are fragi~e and may be easily
mechanically dislodged from the anode substrate. The dislodged lithium,
because it is insoluble in the electrolyte is thereafter lost from discharge
and further replating. Efficiency is therefore reduced by depletion of
available lithium for repeated cycling.
The electrically insulated lithium generally results from the
interaction of the lithium with the organic solvent or solvents utilized
in the cells. As the cell is recycled, the lithium metal is deposited
in the form cf dendrites having high surface areas which therefore react
to an increasingly greater e~tent ~ith the electrolyte solvent particularly
at the Flating site to form the insulating surface films of increasingly
greater area whereby such plated ]ithium becomes increasingly electrically
isolated from the anode. These films, when extensively formed, also
tend to reduce the rate a~ which lithium cations enter solution and
thereEore may also reduce cell capability. Furth~rmore, reaction products
of lithium with commonly used electrolyte solvents are irreversible
(particularly with respect to the solvent) in nature. Accordingly,
during repeated cycling the electrolyte solvent itself becomes depleted
with loss of conductivity and cell performance. The reaction products
generated from the solvents also tend to act as detrimertal impurities
further destroying cellcapability. Additionally, even lithium contained
in such reaction products may also be lost thereby causing increasing
reduction of available lithium for recycling.
~ 2 --
1 ~ 5~3~ 3531
As an example, propylene carbonate will react with lithium to form an
insulating film of lithium carbona~e and propylene gas which cannot be
reversed to obtain the original solvent. Though recovery of some of the
lithium is possible under charging, the lithium carbonate is however
not efficiently reversible into its component elements. Some lithium
is therefore lost from further cycling. Similarly other solvents such
as tetrabydrofur~n and acetonitrile form complex reactior. products
with lithium which are also irreversible. In fact the organic solvents
by their very nature must react with the lithium anode. In order to
dissolve the electrolyte salts needed for conductivity, the organic
electrolytes must be somewhat polar, and it is this very chcracteristic
which causes such solvents to react with the lithium in the formation of
; the irreversible reaction products.
` In order to attain a highly efficient rechargeable cell in
accordance with the present invention, plated lithium should have reduced
dendrite character and should not be coated with a non-reversib]e insulative
film. It is also essential that there must be a complete cycling of
substantially all of the active cell components without the introduction
or formation of additional reaction by-products which are irreversible in
nature. Accordingly, free organic solvents or co-solvents are excluded
from the cells of the present invention. Furthermore, though the problem
of recharging efficiency has been described as being inherent in organic
solvent lithium cells, cells having only inorganic components such as a
lithium cell with an inorganic thionyl chloride solvent/cathode depolarizer
may have s:imilar problems of inefficient recycling. Reaction of the
thionyl chloride with the lithium produces reaction products of lithium
chloride and an unstable 'S~' species which cannot be effectively recombined
to the original starting materials. Thus the electrolyte solvent and
cathode depolarizer, even if inorganic, must only react with the anode
metal only to the extent of formation of totally reversible reaction products.
-- 3 --
~s~3a~
M-3531
Cathode depolarizing materials have recently been discovered
which are in themselves highly rechargeable. Examples of such materials
include the layered metal chalcogenide compounds described by Whittingham
in the ~.S. Patent No. 4,009,052 which intercalate lithium ions within
the spacing between the layers without undergoing full reactions. This
property makes them effectively reversible and rechargeable. However,
such materials ~re rtilized in ambient temperature lithium cells with
organic electrolytes whereby the cell as a whole remains inefficiently
rechargeable.
Various attempts have been made to improve the efficiency and
the rechargeability of the lithium anodes in nonaqueous cells. Such
expedients generally attempted to minimize the dendri~ic plating of the
lithium with the use of various means such as additives, alloying of the
lithium anode, utilization of specific electrolyte salts and solvents
etc. U.S. Patents Nos. 3,953,302 and 4,091,152 describe the use of
metal salt additives comprised of metals which are reducible by lithium
and which coplate with the lithium on charging to form lithium-rich
metallics or alloys. The use of polyalkylene glycol ethers in ~.S.
Patent No. 3,928,067 was described as improving the recycling characteristics
of lithium ce]ls by improving the morphology of the plated lithium.
Though such expedients improve rechargeabiligy, such cells still contain
organic elements which preclude truly efficient rechargeable cells as
described. Dendritic plating of lithium in secondary cells is described
in U.S. Patent No. 4,139,fi80 as being effectively prevented ~ith the use
of clovoborate electrolyte salts. ~lowever, such electrolyte salts are
difficult to synthesize and are accordingly very expensive. U.S. Patent
No. 3,580,828 describes specific electrolyte salt concentra~ions and
current density limlts which, if observed, improve lithium deposition
characteristics. Other methods for improving rechargeability of plated
lithium include the initial utilization of lithium allo;~ anodes particu]arly
with aluminum as described in U.S. Patent No. 4,002,492.
-- 4 --
~ 15~3~5
M-8531
General improvements in rechargeable lithium cells include the
use of complexed inorganlc lithium salts as charge transfer agents (U.S.
Patent No. 3,7~,6,385~ and the judicious use of organic cosolvents with
S2 in order to improve solubility of the electrolyte salts (U.S. Patent
No. 3,953,234). The use of solvents which are relatively stable with
respect to the lithium anode was in fact recognized in patents such as
U.S. Patent No. 3,540,988 as being required in order to provide enhanced
rechargeability of the cells.
Various systems requiring external mechanical components
include molten lithium cells (not inherently sub~ect to dendritic plating)
which require extensive heating and shielding components but which are
the most feasible efficient rechargeable lithium cells since there are
no dendrites or films on molten lithiumO Other systems include cycling
electrolytes such as in ~.S. Patent No, 4,154,902 which require complex
circulating mechanisms.
Electrolytic processes for lithium deposition, however, generally
require organic solvent carriers of the electrolyte salt for high conductivity
and efficient plating out of the lithium metal. Exemplary of such
lithium deposition procedures are ~.S. Patents Nos. 3,791,945 and 3,580,828.
Similarly, cells as described above (e~cept for those containing clovoborate
electrolyte salts) require organic solvents for high conductivity and
efficient lithium plating.
Electrodeposition of lithium in an electrolyte comprised of
lithium and sodium tetrachloroaluminate or lithium and sodium tetrabromoaluminate
dissolved in pure sulfur dioxlde (without organic cosolvents) is described
in ~.S. Patent No. 3,493,433. However discharge performance of such
cells is severely llmited with a discharge capacity substantially ]ess
than theoretical capacity. Since such enumerated salts are described
therein as being the only salts having sufficient solubility and conductivity
in pure liquid S02 for plating efficiency, cells having other salts in
-- 5 --
~ ~5~3~
liquid SO 2 have as a rule re~uired the further utilization of
organic cosolvents as described in U.S. Patent No. 3,953,234 and
as discussed above.
It is an object of the present invention to provide a
highly efficient roorn temperature rechargeable inorganic lithium
or other alkali or alkaline earth metal cell having substantially
only reversible reaction products which is both efficiently
dischargeable and substantially completely rechargeable over
extended periods of cycling.
This and other objects, features and advantages of the
present invention will become more evident from the following
discussion and drawing.
The sole figure of the drawing is a voltage profile of
charge and discharge cycles of a cell of the present invention.
Generally the present invention comprises a totally
inorganic non-aqueous efficiently rechargeable cell having a
lithium or other active metal (generally alkali or alkaline earth
metal or alloys thereof) anode, a totally inorganic electrolyte
solvent consisting essentially of sulfur dioxide which may also
function as cathode depolarizer (with an inert generally carbon
cathode) and an inorganic gallium salt such as gallium halides
having anode metal cations and LiGaCl4 (with a lithium cation
and a GaC1 4 anion in a lithium anode cell) in particular,
dissolved in said sulfur dioxide electrolyte solvent~ Other
gallium salts include Li 2 O(GaC13 ) 2 and Li2 S(GaC13 )2 (with
lithium cations and O(GaCl3 )2 and S(GaC1 3 )2 anions
respectively) described in co-pending application (M-31l65).
A completely reversible solid cathode depolarizer such as an
intercalation compound may optionally be utilized with the S02
electrolyte solvent. Examples of such solid cathode depolarizers
include chromium oxide (SELOXCETTE, a trade mark of Ventron
Division of Thiokol Corp. - Alfa Products.), titanium disulfide,
manganese dioxide, etc. ~he cell is efficiently rechargeable
since all reactions therein including internal reactions between
cell compounds such as between the lithium anode and the S02
solvent and the electrochemical cell reactions produce
substantially
:~ ~$~3~3~
~1-3531
/
only reversible products of, for example, lithium dithionite which is
100% reversible on recharge or intercalated or similarly reacted lithium
which is also cc,mpletely reversible. Additionally, use of the gallium
salts appreciably reduces dendritic plating as well~
Generally, a]l of the organic co-solvents commonly used in
non-aqueous lithium cells/S02 cells such as propylene carbonate, acetonitrile,
tetrahydrofuran, dioxolane, gamma-butrolactone and the like are detrimental
as co-solvents in the present invention since they tend to form complex
non-reversible reaction products with the active metal anodes such as
lithium. Thus it is a re~uirement of the present invention that the
electrolyte solvent be entirely inorganic. It is, however, not sufficient
that the electrolyte solvent be entirely inorganic since the most common
inorganic solvent ~sed in completely inorganic cells, thionyl chloride,
as described above, also forms irreversible reaction products with an
active metal anode such as lithium. Accordingly, the inor~anic solvent
of the present invention is specifically S02 which reacts with lithium
in the formation of the completely reversible lithium dithionite.
However, sulfur dioxide is a relatively poor solvent for lithium salts
since it is an acceptor solvent which interacts primarily with the
electrolyte salt anion rather than cation. Accordingly, in order to
promote solubility and conductivity, organic solvents (normally donor
solvents) have been invariably coupled therewith in order to complete
solvation, with the organic solvents solvating the electrolyte salt
cations as fully described in U.S. Patent ~o. 3,953,234. The only salts
generally described as being sufficiently soluble in the S02 alone are
the aforementioned lithium and sodlum tetrachloroaluminates and borates,
and clovoborates. However, while a salt may be soluble in S02 it must
also provide a cationically conductive solution (greater than 1 x 10 3
~1 --1
ohm cm ) for it to be considered as providinL an acceptible electrolvte.
Thus, materials such as LiAlC14 which are soluble and electrically
conductive in ~2 are nevertheless generally unsuitable for the cells
1 ~ 5~3~ 3531
of the present invention because of the low caticnic conductivity of
the electrolyte. The clovoborate salts which are soluble and cationically
conductive are, however, very expensive. The electrolyte salts of the
present invention which are specifically gallium salts such as halides
with anode metal cations have been discovered to be soluble and highly
cationically conductive in pure S02. Furthermore, when compared to
clovoborate salts such salts are relatively inexpensive.
It is preferred that the anode metal be supported on a metal
foil substrate. A preferred substrate for a lit~ium anode is a copper
foil with the lithium being applied to both sides of the copper in tbe
form of a sen~ cll.
With an S02 cathode depolarizer, the cathode is an inert material
such as carbon supported on a usually metallic substrate such as an
expanded metal, for example aluminum.
The most preferred gallium halide salt having the requisite
conductivity in S02 without the need for orgaric cosolvents is one
having a gallium tetrachloride anion such as a LiGaC14 salt for use in
a lithium anode cell as described in ~.S. Patent ~o. 4,177,329. In
said patent the gallium salts are, however, specificE~lly described as
being utili~ed in an inorganic SOC12 contEining cell which is not rechargeable
in accordance with the present invention. lt has been discovered that
such salt~ may be formed in situ in a pure S02 solvert by reaction
between, for examrle~ LiCl and GaCl~ to form LiGaCl~. This is in acldition
to such in situ salt formation in an SOCl2 solvent as described in
said patent.
The LiGaCl4 salt may also be prepared by direct fusion of LiCl
and GaC13 by melting such materials tcgether in stoichiometric amounts and
allowing the melt to crystalli7e.
As shown in Table I, LiGaC14 electrolyte salt provides a high
cationic conductivity over a wide range of temperature even when dissolved
1~5~3~5
M-3531
in a relatively poor electrolyte solvent of pure S02:
TABLE I
TE~ERATURE CONDUCTIV~TY (lM LiGaC14-S02)
(ohm cm)
~0C 5.32 x 10
30C 5.27 x 10 2
20C 5.17 x 10 2
10C 5.05 x 10 2
0C 4.76 x 10 2
-10C 4.45 x 10
-16.8C ' 4.19 x 10-2
It has also been discovered that the aforementiGned LiGaCl4
salt reduces dendritic plating of anode materials such as lithium since
the electrolyte solutions remain clear of li~hium particles and limited in~
soluble precipitates even under repeated cycling. It is also postulated that
because the electrolyte solution remains clear of such particles, the
LiGaC14 also forms scavenging species during charging which scavenge
disconnected lithium dendrites, and lithium from lithium reaction products
both at the anode and the cathode. As a result, cell reaction generated
products even if isolated from the anc,de or cathode substrates are returned
to solution to reform both anode metal and electrolyte solven~:.
In order to illustrate the efficacy of the present invention
in providing an efficiently rechargeable cell, the following examples and
comparative data (relative to other materials in t'he prior art) are
presented. It is understood that such examples are for comparative
purposes and that any enumeration of detail should not be construed as a
limitation on the present invention. Unless otherwise indicated all
parts are parts by weight.
EXAMPLE 1
A 'D' siæe c,ell is constructed with convolute wound lithium
foil (20"(50.8cm) x 18 (4.13cm) x 0.012"(0.03cm)) and porous carbon
_ 9 _
3 ~ 5
M-3531
(an aluminum expanded metal substrate) (20'(50.8cm) x 4 (4.4cm) x
0.025" ~0.063cm)) electrodes with a polypropylene separator therebetween.
The cell is filled with a so]ution of 1~ LiGaC14 in pure S02 (about 40
grams). The theoretical capacity of the lithium anode is about 13 Ahr
and the capacity of the S02 is about 14 Ahr. The cell is cycled at 0.5A
OD 2 hour discharge followed by 2 hour charge~ The voltage profiles of
the second and sixty ninth cycles (the cell failed abruptly after the
69th cycle) are shown in the figure. There is almost a negligible
change in the discharge and charging voltages thereby indicating an
almost 100% cycling efficiency.
Rechargeability alone is not~ however, a sufficient criterion
of cell utility. The cell must also have good primary cell characteristics.
The following examples 2-4 illustrate such capabili~y of the cells of the
present invention.
EXAMPLE 2
A 'D' cell is constructed as in Example 1 and is polarized at
room temperature (25C) and-30C with the results shown in Table II:
TABI.E II
i (Am ) Volts at 25~Volts at-30
P
open circuit 2.91 2.97
.026 2.o9 2.85
.050 2.89 2.80
0.10 2.87 2.7S
0.25 2.84 ----
0 35 ---- 2.58
0.5G 2.79 2.53
1.0 2.63 ----
1.5 2.58 2.32
2.0 2.53 2.27
3.0 2.47 2.15
1.85
- -- 10 --
1~583~5 M-3531
The relatively high voltages obtained at currents over l
ampere show the high conductivity of the electrolyte salt and the good
rate capability of the cell at both amhient and low temperatures.
EXAMPLE 3-4
Two 'D' cells are constructed as in Example 1 and are discharged
at various rates and temperatures with the ~onditions and results given
in Table III:
TABKE III
Ex.#Temperature Discharge rate or load Capacity to 2v
3 -30C 4.4 ohm 3.5 Ahr
4 25C 0.25 A 9.0 Ahr
Since the cell has a theoretical capacity of 10 Ahr ( C limited)
the room temperature capacity obtained (90%) at a rate of 0.25 A indicates
very good primary cell performance. The 3.5 Ahr obtained at -30C is
also excellent for low temperature performance.
EXAMPLE 5
A cell made in accordance with Example 1 but with an anode of
two 0.010" (0.025cm) foils of lithium sandwiching a copper foil substrate,
is cycled at 0.10 A for 10 hours discharge and charge. The cell pro~ides
about 104 Ahrs about S times original lithium capacity.
EXA~IPLE 6
A glass cell is made using a 2 cm x 0.5 cm Li anode and a
2.0 cm x 0.5 cm catalytic carbon cathode with about 20 cc of lM LiGaC14
in pure S02 electrolyte. The cell is discharged at 1 ma/cm for 5 hours
then recharged at 1 ma/cm2 for 5 hours. AEter 15 cycles the lithium
surface is clean and dendrite free and the e]ectrolyte remains clear.
EXA~IPLE 7
Two cells are made in accordance w th Example 1 but with one
having an electrolyte of lM LiAlC14 in pure S02 as in ~.S. Patent ~o.
3,493,433. The cells are discharged with the results given in Table IV:
-- 11 --
~L ~583~5
M-3531
TABLE IV
_ectrolyte OCV Cap (0.25A discharge rate)
1 M LiGaCl4 2.94V 9.0 A.hr
1 M LiAlC14(PRIOR ART) 3.26 0.38
It is evident from the above comparison that the LiAlC14 electrolyte
slat described in the prior art provides only a minimal primary cell
capability in pure SO2 and is certainly unsuitable for the secondary or
rechargeable cells of the present invention. It may also be noted that
LiAlCl4 is the electrolyte salt of choice in totally inorganic cells having
lithium anodes and thionyl chloride electrolyte solvent/cathode depolarizers.
The following examples are presented as further illustrating
the rechargeable efficacy of the cells of the present invention under
varying cycling conditions.
EXAMPLES 8-10
Three cells are made in accordance with Example 5 and are
cycled under conditions and with results shown in Table V:
TA~LE V
l)tili7ation
Ex. # Test cond. (disch. ~ ch.) Capacity (Ahr) Li S02
8 0.5 A x 2 hrs D/C 73 360% 57n%
9 0.25A x 4 hrs 66 330~O 500%
0.5A (discharge to 2.5v 80 400% 600
charge at 3.5v)
The above examples are for illustrative purposes only with
changes in cell structure and components being within the scope of th~
present invent;on as defined by the following claims.
- 12 -
