Journal of Power Sources, 4 (1979) 1 - 9 1

0 Elsevier Sequoia S. A., Lausanne - Printed in the Netherlands

THE LITHIUM-THIONYL CHLORIDE BAlTERY - A REVIEW

R. GANGADHARAN, P. N. N. NAMBOODIRI, K. V. PRASAD and R. VISWANATHAN

Central Electrochemical Research Institute, Karaikudi-6, T.N. (India)
(Received February 10.1978; in revised form December 7,1978)

summary
The lithium-thionyl chloride nonaqueous cell system is a recent devel-
opment. Low atomic weight and high electrode potential make lithium unique
as an anode in energy devices. Preparation of the anode, cathode, and the
electrolyte is reviewed. The construction of the cell, its performance charac-
teristics, uses, etc., are also surveyed.

Introduction

In the area of power sources, the lithium-thionyl chloride battery system

is one of the major research developmentsin recent years [ 1 -61. This battery
has evoked considerable interest because of its ability to operate at extremes
of temperature. The lithium-thionyl chloride battery system has the advan-
tage of long shelf life and possesses high energy density, power density, and
cell voltage.
In the initial stages of development, lithium-thionyl chloride battery
research suffered because of the battery’s inability to discharge at high rates.
However, this difficulty was overcome by suitable modifications to the design
of the battery and by using better conducting electrolytes. These results have
been achieved after more than a decade of research work in the area of high-
energy density nonaqueous battery technology.
Lithium metal has certain unique properties. Of these, two important
physical properties are made use of for constructing the lithium-thionyl
chloride cell. They are the atomic weight and the single electrode potential.
The former is the lowest among the known metals, and the latter is the highest.
However, lithium metal reacts vigorously with water, as do other alkali metals,
and therefore new problems and challenges have to be faced when this metal
is used in the fabrication of a battery.
In order to overcome the problem of the high reactivity of lithium with
water, anhydrous solvents are used. Typical organic and inorganic solvents
have some additional special advantages; their freezing points are much lower
2

than that of water, and often their boiling points are higher. Hence, batteries
with solvents other than water will have low-temperature operational capa-
bilities. However, these advantages are offset by the inability of the non-
aqueous solvents to conduct ions in a manner similar to water even when
salts are dissolved in them. The dielectric constants of the nonaqueous sol-
vents am, typically, much lower than water, and the conductivity of non-
aqueous solutions is much less than aqueous solutions.
Despite the reactivity of lithium and water, lithium batteries have been
constructed to operate in aqueous solutions [3]. However, for long term use,
nonaqueous batteries are much preferred and this review will be restricted to
that field of endeavour.
Thionyl chloride has been used as both a solvent and as an oxidant
(depolariser) in the lithium-thionyl chloride battery. Compatibility studies
with lithium metal reveal that the solvent, when pure, does not undergo any
change and is reasonably inert. Some of the properties of thionyl chloride
are compared with other inorganic solvents such as sulfuryl chloride and
phosphorus oxychloride in Table 1.

Anode

According to the available literature [7], the anode is prepared by press-

ing lithium ribbon of the desired size on to an expanded nickel grid which
acts as a current collector. For the well known commercial D size cell, the
dimensions of the anode are reported as 20 X 1.5 X 0.015 in. Dey [7] has
reported that the size of the anode has a great influence on the performance
of the cell.

TABLE 1
Physical properties of inorganic solvents

Properties Solvents

sCC12 SO2Clg POCl3

1. Physical nature Liquid Liquid Liquid Liquid

2. Density (g/cm3) 1.655 1.667 1.675 1.0
3. Boiling point (“C) 76.1 78.0 105.3 100
4. Melting point (“c) -104.4 -53.8 1.5 0
5. Viscosity (cP) 0.6 0.92 i.ii 1.0
6. Dielectric constant 9.05 10.0 13.9 80.0
7. Conductivity (ohm-’ cm-’ ) 2x 1o-e 3x lO+j 2 x 1o-8 0.8 x 1o-6
8. Li-Depolariser (cell voltage) V 3.65 3.90 3.10 2.20
9. Theoretical energy density (W h/kg) 1540 - 1540 2 200
 3

Cathode

Different varieties of carbon [8] are used to construct the cathode on a

nickel matrix which acts as the current collector. According to the literature
[7,9] colloidal Teflon is used as the binding material for fabricating the
carbon electrode. The composition of the paste is given as: carbon; 81% by
weight; Teflon binder; 19% by weight.
The paste is spread over an expanded nickel grid. It is then pressed
against two stainless steel plates. The pressed electrode is dried under vacuum
at a temperature of 110 “C. Reported literature reveals that the cathode of
the commercial D size cell has dimensions of 20 X 1.5 X 0.19 in., and that of
the AA size cell 2 X 1 X 0.03 in. [lo]. Although Teflon is commonly used
as a binder for the cathode construction, polyethylene, paper pulp, etc., are
also used as binding materials [7 - 91. The performance of the battery is
significantly affected by the quantity and the nature of the binder [ll].

Separators

Different types of separator are made use of in the construction of the

lithium-thionyl chloride battery. They are generally of nonwoven glass sheet
[ 10,111, nonwoven rayon [9], glass filter paper [9] or porous polyethylene
sheet [2].

Electrolyte

The complex electrolyte, lithium chloroaluminate, or lithium aluminium

chloride, is used in the lithium-thionyl chloride battery. The preparation of
this compound was first reported by Kendall [12] in 1923. It is prepared by
mixing equimolecular proportions of anhydrous lithium chloride and alu-
minium chloride in a mortar and the pulverised sample is kept at a tempera-
ture of 180 “C, under a helium [13] argon [ 141 or nitrogen [ 151 atmosphere.
During the preparation of this compound, a small quantity of aluminium
metal is formed as an impurity [ 161. This can be removed from the electro-
lyte by skimming, since aluminium is insoluble in molten lithium ahuninium
chloride.
Lithium aluminium chloride can also be prepared by other methods. In
one of these [9], dry lithium chloride and aluminium chloride are mixed
together in equimolecular quantities and the mixture is heated at 100 “C for
4 - 5 h under dry hydrogen chloride gas. In order to overcome the tedious
methods of preparation of lithium aluminium chloride, Aubom et al. [ 171
used equimolecular quantities of anhydrous lithium chloride and aluminium
chloride dissolved in distilled thionyl chloride under a nitrogen atmosphere.
Lithium aluminium chloride solution can be prepared by reacting aluminium
chloride and lithium chloride in thionyl chloride solution, the excess of alu-
4

minium chloride is precipitated with a quaternary ammonium salt and the

precipitate is filtered from, the solution [ 181.
Recent X-ray examinations have not revealed the structure of the com-
pound [13], and a comparison of the crystal structure of lithium aluminium
chloride shows that it is not isomorphous with other known alkali alurninium
chlorides [ 7,191. Weppner et al. [ 131 studied the conductivity of lithium
aluminium chloride over a temperature range of 25 - 200 “C. Its conductivity
at 25 “C was found to be 1 X lo-’ ohm-’ cm-’ [17], and at 174 “C, 0.35 X
10m6 ohm-’ cm-’ [20]. The phase diagram of this compound has been
studied in detail [12, 19, 211, and it was shown that lithium aluminium
chloride is in the ternary phase of lithium chloride and aluminium chloride.
Campanella et al. [ 221 have reported on the method of preparation of
anhydrous electrolytes in anhydrous solvents. Lithium aluminium chloride
solution can be prepared by dissolving known quantities of the electrolyte
in anhydrous thionyl chloride under an inert, dry atmosphere using nitrogen
gas. The conductivity of this compound in thionyl chloride is reported as
[ll, 231:
u LiAlC& = 2.04 X lo-’ ohm-’ cm-’ at 1.73M;
u LiAlCl* = 7.8 X 10m3 ohm-’ cm-’ at 0.9M.
The highest solubility of lithium aluminium chloride in thionyl chloride is
4.5M at 25 “C. Traces of water in the solution can be revealed using infrared
spectra [ 241. The conductivity of lithium chloroahuninate is reasonably
satisfactory for a practicable battery. In nonaqueous solutions, the solubility
of inorganic electrolytes largely depends on the size of the anion. The larger
the size of the anion the greater will be the solubility of the electrolyte. The
reason for the conductivity of lithium aluminium chloride in thionyl chloride
is attributed to the reaction between thionyl chloride and aluminium chloride
and, thereby, the formation of associate pairs. These associated pairs dissociate
and the process of dissociation is accelerated in the presence of lithium
chloride.

Cell

Three sizes of lithium-thionyl chloride batteries, viz., AA, C, and D

sizes, are being commercialised. These cells are designed to discharge at low-
and high rates. The “bobbin” type [17] cells are designed to discharge at
low rates and the “spiral” or “wound” or “jelly roll” type [17] at high rates.
The spiral type.is a high-powerdensity battery, and the bobbin type is a high-
energydensity battery. The lithium anode is separated from the cathode by
a suitable separator and the insulation of the electrodes is provided by Teflon.
The electrolyte is carefully introduced into a nickel-coated iron can under a
dry, inert atmosphere and the can is then sealed under vacuum using glass as
a sealing material. In order to reduce the explosion hazards, specially designed
vents are used. The performance characteristics of these cells are given in
 5

TABLE 2
Discharge characteristics of C [ 281 and D [ ‘7] size cells

Temperature c cell D cell

(“C)
Energy density (W h/kg) Energy density (W h/kg)

100 mA 1A 3A 100mA 1A 3A

25 366.4 316.8 262.2 387.2 308.0 220.0

-30 299.2 233.2 - 220.0 88.0 -
-55 - - - 187.0 - -

Table 2. Designs for the lithium-thionyl chloride batteries are discussed by

Marincic [25 - 271.
Commercial AA, C, and D size batteries [7,17,28 - 301 were discharged
at different rates and temperatures. Typical discharge characteristics of C
[28] andD [7] size cells at various discharge rates and temperatures are
given in Tables 2 and 3.

Comparison of the lithium-thionyl chloride cell with other primary cells

It is worthwhile to compare the lithium-thionyl chloride cell with other

well known primary cells such as the zinc-carbon cell, the Mallory cell, etc.
Table 3 gives the available data from the literature [31,32] and from these,
it is clear that the lithium-thionyl chloride battery is superior to many of
the existing aqueous primary cells, especially from the point of view of energy
density, shelf life, and low-temperature performance.

Explosion hazard

Dey [ 7,331 discovered that the lithium-thionyl chloride cell tended to

explode on accidental short circuiting, heating or overdischarge. It was ob-
served that after 28.5 min of discharge at the 10 A rate the cell voltage dipped
to a very low level. At that point the current was turned off and, after a
lapse of 1.5 min, the cell exploded. It was also noted that when the cells
were shorted, explosion occurred, in spite of the operational vents which were
designed to open at 500 - 550 psig (3 4 - 308 MP). In order to understand
l

and avoid the cause of this phenomenon, two possibilities were explored.
The first approach was to use a vent which opened at very low pressures
and allowed the electrolyte to escape prior to thermal runaway. When the
cells containing electrolyte were shorted, the temperature of the cells initially
increased and consequently the electrolyte began to escape from the cells.
No explosions were noticed. The pressures developed inside the cells were
between 100 and 130 psig (O-7 - O-9 MP).
TABLE 3
Comparison of the lithium-thionyl chloride cell with qther (aqueous) primary cells [31,32 ]

Properties Lithium-thionyl Zinc-carbon Mallory Alkaline Magnesium Silver oxide-

chloride Zn-HgO Zn-MnOz cell ZiIlC

O.C.V. 3.65 V 1.5 v 1.35 v 1.5 v 1.8 V 1.85 V

Efficiency at 85% 0% 0% 3% 23% 50%
-28.5 “C
Efficiency at 96% 5% 3% 15% 58% 70%
-6.6 “C
Energy density 209.0 42.9 88.0 57.2 58.0 110.0
(W h/kg)
Shelf life 10 1-2 3-4 2-3 5-7 1 approx.
(Yea@
 7

However, when the cells with low pressure vents were discharged at low
rates, explosion occurred. It was also noted that the low pressure vents did
not prevent the explosion of the cells on over discharge.
The second approach to alleviate the explosion hazard was to reduce
the rate capability of the cell by reducing the size of the electrode. It was
expected that this method of approach would reduce the heating effect of
the cell when shorted. With the D size cell, evaluation was made with 7 and
11.5 in. long lithium anodes. The cells were fabricated with vents to with-
stand 500 psig (3.4 MP) pressure. It was noted that the cell with the 11.5 in.
lithium anode exploded when short-circuited at room temperature. The cell
with the 7 in. lithium anode did not explode at room temperature. However,
at the storage temperature of 72 “C, the 7 in. lithium electrode cell exploded
when short-circuited. Hence, from these experiments, it was concluded that
cell explosions could not be prevented under all circumstances by decreasing
the rate capabilities.

Cost comparison

Since lithium cells are a new addition to the available energy sources, it
is of interest to know their cost. It is difficult to compare the cost of the
lithium cells with existing alkaline cells because of the differences in cell
manufacture processes. We can, however, compare to some extent the cost
of the active materials for each type of cell. Table 4 [ 341 gives an approximate
comparison between three alkaline cells and lithium cells.
It can be seen from the Table that the cost per ampere hour is less for
the lithium cell than for the alkaline cells, except for the zinc-manganese
dioxide cell. Even though the cost of the materials is less for lithium cells,
the sophistication of the manufacturing technology may result in a higher
cost. Here, one should note the following advantages of the lithium cells
[ 351: (a) more payload may be possible with lithium cells due to their smaller
size and light weight; (b) greater shelf life; (c) good sub-zero temperature
performance; (d) lower maintenance cost associated with battery replace-
ments, especially in remote areas or in applications where replacement of the
cell is a laborious process. Considering the above advantages, we may con-
clude that the cost of the cells may be comparable with other alkaline cells.

uses
Future prospects for lithium batteries are exceedingly bright. Lithium
batteries find applications in military and civilian equipment [31,36,37].
TABLE 4
Comparison of the cost of the lithium-thionyl chloride cell with other primary cells component-wise [ 34 ]

Components used Cell system

Zinc-manganese dioxide Zinc-mercuric oxide Zinc-silver oxide Lithium-thionyl chloride

cost (cents/A h) cost (cents/A h) cost (cents/A h ) cost (cents/Ah)

Zinc 0.113 0.113 0.113

Mercury 0.128 0.128 0.128
Potassium hydroxide 0.085 0.060 0.060
Carbon 0.042 0.053
Zinc oxide 0.008 0.006 0.006
Manganese dioxide 0.177 -
Mercuric oxide - 8.548 -
Silver oxide - 45.23 -
Lithium 0.583
Thionyl chloride 0.118
Aluminium chloride - - 0.023
Lithium chloride - - 0.047
Total components cost 0.553 9.908 45.537 0.814
Total cost of the
components per W h 0.369 7.339 30.358 0.226
 9

Acknowledgement

The authors thank Dr. H. V. K. Udupa, Director, Central Electrochemi-

cal Research Institute, Karaikudi, India, for his keen interest and for per-
mission to publish this review pacer.

References

1 J. J. Auborn, U.S. Pat. 3926669 (1975).

2 G. E. Blomgren and M. L. Kronberg, Ger. Pat. 2262256 (1973).
3 E. L. Littauer and K. C. Tsai, Power Sources Symp., Proc., 26 (1974) 57.
4 M. Eisenberg, U.S. Pat. 3891458 (1975).
5 J. J. Auborn, U.S. Pat. 3897264 (1975).
6 N. Marincic, U.S. Pat. 3907593 (1975).
7 A. N. Dey, Electrochim. Acta, 21 (1976) 855.
8 W. A. Behl, J. A. Christopulous, M. Ramierz and S. Gihnan, Tech. Rep. ECOM-4101,
April, 1973.
9 W. A. Behl, J. A. Christopulous, M. Ramierz and S. Gilman, J. Electrochem. Sot.,
120 (1973) 1619.
10 N. Marincic, J. Appl. Electrochem., 6 (1976) 54.
11 A. N. Dey, J. Electrochem. Sot., 123 (1976) 1262.
12 J. Kendall, E. D. C&tender and H. K. Miller, J. Am. Chem. Sot., 45 (1923) 963.
13 W. Weppner and R. A. Huggins, J. Electrochem. Sot., 124 (1977) 35.
14 B. Di Pietro and B. Scrosati, J. Electrochem. Sot., 124 (1977) 161.
15 M. L. B. Rao, J. Eiectrochem. Sot., 114 (1967) 13.
16 V. N. Storozhenko, Zh. Fiz. Khim., 48 (1974) 1709.
17 J. J. Auborn and N. Marincic, in D. H. Collins (ea.), Power Sources 5, Academic Press,
London and New York, 1976, p. 683.
18 J. P. Gabano, French Pat. 2079744 (1971).
19 K. N. Semeneko, J. Inorg. Chem., 14 (1969) 481.
20 Y. Yamaguti and S. Sisido, J. Chem. Sot. Jpn., 62 (1941) 304.
21 E. M. Levin, C. R. Robinson and H. F. McMurdie, Am. Ceram. Sot., Inc., Ohio, 1969,
p. 300.
22 L. Campanella and G. Pistoia, J. Electrochem. Sot., 118 (1971) 1905.
23 J. J. Auborn, A. Heller and K. W. French, Power Sources Symp., Proc., 25 (1972) 6.
24 K. French, P. Cukor, C. Persiani and J. J. Auborn, J. Electrochem. Sot., 121 (1974)
1045.
25 N. Marincic, J. Appl. Electrochem., 5 (1975) 313.
26 N. Marincic, J. Appl. Electrochem., 6 (1976) 51.
27 N. Marincic and D. B. Koffman, J. Appl. Electrochem., 6 (1976) 263.
28 A. B. Dey and C. R. Schiaikjer, Power Sources Symp., Proc., 26 (1974) 47.
29 J. J. Auborn, K. W. French, S. I. Lieberman, V. K. Shah and A. Heller, J. Electrochem.
Sot., 120 (1973) 1613.
30 N. Marincic, J. Epstein and F. GoebaI. Power Sources Symp., Proc., 26 (1974) 51.
31 R. Hill, Pop. Sci., 202 (1973) 18.
32 H. R. Grady, Foote Prints, 21 (1974) 9.
33 A. N. Dey and P. Bro, in D. H. Collins (ed.), Power Sources, 6, Academic Press, London
and New York, 1977, pp. 493,509.
34 N. Marincic, J. Epstein and F. GoebaI, Power Sources Symp., Proc., 26 (1974) 153.
35 Lithium Booklet, Honeywell Power Source Centre, Pennsylvania, U.S.A., 1976, p. 8.
36 B. B. Owens, Electrochem. Sot. Ext. Abst., 76 (1976) 9.
37 N. Marincic and J. Epstein, Electrochem. Sot. Ext. Abst., 76 (1976) 127.