Solid State Ionics 134 (2000) 139–158

www.elsevier.com / locate / ssi

Batteries
fifty years of materials development

R.M. Dell*
2 Tullis Close, Sutton Courtenay, Abingdon, Oxfordshire OX14 4 BD, UK

Abstract

This paper reviews developments in batteries over the past 50 years with particular reference to advances in materials
science and engineering technology. The emphasis is on commercially available batteries such as zinc / manganese dioxide
primary cells, and lead / acid and nickel / alkaline secondary cells, although specialist batteries and those in an advanced stage
of development are also described briefly. Rechargeable lithium-ion batteries will be covered in another paper in this
symposium.  2000 Ronald M. Dell. Published by Elsevier Science B.V.

Keywords: Batteries; Development; Materials science

1. Introduction Leclanche´ cells wired in series to supply the high DC

voltage. The radio weighed several kilograms, had to
Fifty years is a long time in the history of be maintained in an upright position to avoid spillage
materials science, and also in the design, develop- of acid and was anything but ‘portable’. No doubt
ment and application of primary and secondary there were other applications for batteries in industry
batteries. In the 1940s, the principal domestic uses and commerce, for example emergency lighting and
for batteries were in torches (flashlamps), in a few back-up for the primitive telephone system of the
toys, in vehicles (for starting, lighting and ignition) day.
and in radios. Before the days of transistors and solid Over the past 50 years the applications for small
state electronics, when thermionic valves (tubes) batteries in the home (consumer batteries) have
were used in radios, it was necessary to have a expanded phenomenally. Today small primary or
source of low voltage electricity to heat the filament rechargeable batteries are employed in a huge num-
and a DC high voltage to accelerate electrons ber of appliances. Some examples are as follows.
between the valve cathode and anode. Many radios
in those days were remote from the electricity supply • Household: telephones, clock-radios, security
and then two batteries were required, a 2 V lead / acid alarms, smoke detectors, portable fluorescent
accumulator to supply the filament current and a lamps, torches and lanterns, car central locking
‘high tension’ battery consisting of 100 or 120 activators, door-chimes.
• Workshop and garden: portable tools (e.g. screw-
*Tel.: 144-1235-848-339; fax: 144-1235-848-296. drivers, drills, sanders), portable test meters,
E-mail address: ronalddell@compuserve.com (R.M. Dell). hedge trimmers, lawnmowers.

0167-2738 / 00 / $ – see front matter  2000 Ronald M. Dell. Published by Elsevier Science B.V.
PII: S0167-2738( 00 )00722-0
140 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158

• Entertainment: portable radios and TVs, compact the military and space fields, and in the past decade
disc players, tape recorders, keyboards, electronic new rechargeable batteries have been introduced into
games and powered toys. civilian and consumer markets.
• Personal hygiene and health: toothbrushes, bat- This article is intended as an introductory over-
hroom scales, hair trimmers, shavers, blood pres- view of improved batteries and of the advances in
sure monitors, hearing aids, heart pacemakers. materials science and development which have led to
• Portable electronic devices: watches, cameras, them. The emphasis is on commercially available
camcorders, calculators, organisers, mobile batteries, such as zinc / manganese dioxide primary
phones, laptop and notebook computers, bar code batteries and lead / acid secondary batteries. These
readers. long-established electrochemical couples still com-
mand major market shares. However, materials
With all these new applications, which have science advances have resulted in new battery types,
developed over the past few decades, the market for such as nickel / metal hydride, and these are de-
consumer batteries has grown correspondingly. scribed in outline. Lithium rechargeable batteries are
Today the average family may have as many as not covered in any detail as they are the subject of
40–60 consumer batteries at any one time in and another paper at this symposium.
around the home. Many of these will be batteries of
advanced design and construction, giving greatly
improved performance as a result of developments in 2. Consumer batteries
materials science and technology. Although most
small consumer batteries are still of the primary 2.1. Zinc /manganese dioxide (1.5 V) primary cells
(throw-away) variety, there is a growing trend to
adopt secondary (rechargeable) batteries as being Fifty years ago most primary batteries were of the
more economical. Leclanche´ type in which the zinc negative electrode
There has been a similar mushrooming in demand was in the form of a metal can which acted as the
for large size, installed battery packs. Almost every container for the other cell components. There was a
public building (hotels, supermarkets, multiple central carbon rod which served as the positive
stores, railway stations, airports, hospitals etc.) must electrode and this was surrounded by a mixture of
have an uninterruptable power supply and this crude MnO 2 (often the mineral pyrolusite) and
requires a battery pack to take over seamlessly when carbon powder, intimately mixed together. The func-
the mains supply fails, until such time as a local tion of the carbon powder was to increase the
generator can be started. Most factories, office conductivity of the positive active mass and so to
blocks, ships, aircraft, telephone exchanges and even reduce the internal resistance of the cell. The elec-
power stations have their own uninterruptable sup- trolyte, an aqueous solution of ammonium chloride
ply. Other applications for large primary or recharge- and zinc chloride, was absorbed into the pores of a
able batteries are in the defence field (submarine paste type separator (e.g. starch) and the MnO 2 / C
traction batteries, torpedoes, missiles), in space ve- mixture, and for this reason these cells became
hicles (satellites, space probes), in solar energy known as ‘dry cells’. The cells had a seal and vent at
storage (remote area power supplies, navigation their upper end and a non-conducting board cap
beacons) and in electric and hybrid vehicles (tugs which served to insulate the positive carbon rod from
and tractors, forklift trucks, invalid conveyances, the negative zinc can. Finally, the zinc can was
golf carts, milk floats and other road vehicles). surrounded by a cardboard jacket, on which the
Although many of these applications utilise long- manufacturer’s name and information was printed.
established lead / acid or nickel / cadmium recharge- Over the years some improvements have been
able batteries, these have been transformed over made in the design of Leclanche´ cells and their
recent decades by advances in materials and design. materials of construction. The metallurgy of the zinc
Moreover, there are now specialised batteries with can has been improved through alloying additions to
unconventional chemistries, particularly for use in facilitate deep drawing. Better designs of seals have
 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158 141

been developed. The use of mercury to increase the necessary for cell discharge. The resulting fall in pH
over-potential for hydrogen evolution at the zinc near the zinc anode leads to enhanced chemical
electrode has been largely phased out for environ- corrosion, with liberation of hydrogen. As the inter-
mental reasons. The cardboard jacket is now re- nal pressure in the cell rises, the seals ultimately
placed by a steel container coated with polymer and leak, forcing electrolyte to the exterior of the cell.
a polyester film label. This steel outer case is Many of these problems are overcome in the zinc
insulated from the zinc can. The insulating cell cap is chloride cell which has a simpler chemistry as the
made of a hard polymer rather than board. These ammonium chloride electrolyte is replaced entirely
improvements greatly reduce the tendency of the by ZnCl 2 solution. The reduced tendency for the
cells to leak electrolyte when fully discharged, which electrodes to be blocked by product solids permits
used to be a serious problem. faster diffusion and higher rate discharge. A better
Leclanche´ cells are in declining use today. Al- quality grade of MnO 2 , prepared electrolytically, is
though cheap to manufacture and purchase, they employed and a greater proportion of carbon black to
suffer from a number of limitations: facilitate the cathodic reaction. The cells also have a
more sophisticated seal design and these factors,
• they are not suitable for high-drain applications taken together, result in a higher cost of manufacture
(such as driving electric motors) as they readily to offset the improved performance. Zinc chloride
polarise and the available capacity falls sharply cells fall between Leclanche´ and alkaline manganese
with increased discharge rate; cells as regards both performance and cost. They
• their shelf life is not especially long; perform better than Leclanche´ for high drain applica-
• the optimum temperature range of operation is tions and, particularly, for use at low temperature (to
20–408C; outside of this temperature range the 2208). These two types of cell together are known
performance deteriorates markedly. generically as zinc /carbon cells.
Alkaline manganese, the premium form of Zn /
Leclanche´ cells are best suited to low drain MnO 2 cell, differs from zinc / carbon in a number of
intermittent use, with rest periods for recuperation important respects. A principal difference lies in the
(depolarisation) to take place, and for use within 1–2 use of a different electrolyte, viz. concentrated
years. A good example is a domestic flashlamp (|30%) potassium hydroxide solution. Another sig-
battery. They may also be used in low-drain applica- nificant difference is in the nature of the zinc
tions such as door-chimes or smoke detectors, but electrode which consists of finely divided zinc
then their relatively short operational life becomes a powder packed around a brass current collector at the
nuisance as they require periodic replacement. centre of the cell. A third difference lies in the cell
Superficially, the electrochemistry of the Le- configuration; the cathode mix, of electrolytic MnO 2
clanche´ cell is simple. At the negative electrode zinc and fine graphite powder, is packed around the
metal is oxidised to Zn 21 ions outside of the zinc anode and separator and is in
electrical contact with the nickel-plated steel can
Zn ⇒ Zn 21 1 2e (Fig. 1). This inversion of the cell configuration
while at the positive electrode manganese dioxide is might be expected to lead to an inversion of polarity
reduced to trivalent manganese with resultant rise in of the terminals. As this would be unacceptably
pH confusing to the consumer, the problem is overcome
by bringing the central current collector (negative)
2MnO 2 1 2H 2 O 1 2e ⇒ 2MnO.OH 1 2OH 2 . into contact with the cell base, rather than the cap,
and having the insulating seal at the bottom of the
In reality, the cell reactions are much more cell rather than the top, with the steel can contacting
complex than this, with the intermediate formation of the top (positive) terminal.
sparingly soluble species such as Zn(NH 3 ) 2 Cl 2 and Because concentrated potassium hydroxide elec-
ZnCl 2 .4Zn(OH) 2 , particularly at high discharge trolyte is chemically reactive, and is also prone to
rates, and these solids tend to impede ionic diffusion flow (creep) along surfaces, the development of
142 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158

where the effective capacity delivered is many times

that of a Leclanche´ cell under similar conditions.
Table 1 makes a comparison of the approximate
service life of alkaline manganese and zinc / carbon
cells at various discharge rates. It will be seen that
the ratio of hours service (alkaline manganese / Le-
clanche)´ is greater for the large D size cells than for
the smaller AA cells and also greater at high
discharge rates (small loads). At very low discharge
rates the ratio falls to about 2 at 218C, but rises again
sharply at temperatures below 08C. It will be clear
that major advances have been made in the design
and construction of Zn / MnO 2 cells over the past 40
years, with new materials playing a significant role
in this success.
Recently at least two companies have introduced a
premium grade of alkaline battery for high drain
applications. Materials science advances have been
Fig. 1. Construction of an alkaline manganese primary cell. responsible for these developments through reducing
the internal resistance of the cells. In one case this is
said to be as a result of coatings deposited on the
satisfactory seals for the alkaline manganese cell was anode and cathode current collectors; in the other the
a technical challenge which had to be solved before result of using a finer grade of graphite powder and
the cells could be commercialised. Modern cells packing more MnO 2 powder into the available
have effective seals which are resistant to leakage cathode space. The claims for the improvements in
and corrosion. discharge power and service life are quite impres-
Alkaline manganese cells, which retail at a pre- sive.
mium price, are more suitable than zinc / carbon cells Zn / MnO 2 cells are made in a number of sizes and
for continuous high drain applications and for use at configurations. The most popular sizes are AAA,
low temperatures. They also have a longer shelf life AA, C and D, although both smaller and larger cells
of .4 years at 208, making them attractive for low are available. Another popular item is the prismatic 9
drain ‘fit and forget’ applications such as smoke V (PP3) battery. This consists of six small cells lying
detectors. However, their principal economic advan- flat (i.e. perpendicular to the long axis of the case)
tage is seen in high drain applications such as and joined in series. These 9 V batteries are ideal for
children’s toys, cassette players, electric shavers, etc. low drain applications such as smoke detectors.

Table 1
Comparison of approximate service lives of alkaline manganese and zinc / carbon cells at different discharge rates a
Cell size AA Load D
(ohms)
Cell type Zn / C Alk.Man. Ratio Zn / C Alk.Man Ratio
Load Hours Hours Hours Hours
(ohms)
62 35 140 4 10 13 130 10
24 12 50 4 4.7 6 50 8
3.9 0.9 9 10 2.2 1.5 20 13
a
Data published by Duracell for discharge to 0.8 V at 218C.
 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158 143

2.2. Zinc /manganese dioxide secondary cells are now available in the shops, along with the
custom designed chargers, and it seems likely that
Traditionally, the Zn / MnO 2 cell has always been this will be a rapidly growing market in the near
seen as a primary (throw-away) unit. In the past few future for those who are prepared to take the trouble
years developments in materials, and also in elec- of recharging their cells. Correspondingly, it seems
tronic charge / discharge control, have led to the likely that rechargeable Ni / Cd consumer cells will
Rechargeable Alkaline Manganese (RAM) cell. The decline in popularity.
incentive for recharging alkaline manganese cells is
considerable as shown by the following comparative 2.3. Button cells and miniature batteries
figures for the capacity of an AA sized cell:
A major advance of the past 30 years has been the
Alkaline manganese 2–3 Ah development of button and coin cells, and miniature
Nickel / cadmium 0.5–1.0 Ah batteries, for use in clocks, watches, hearing aids, car
Nickel / metal hydride 1–1.5 Ah alarms, key locks etc. These may be based on any
one of several different chemistries. The first button
When recharged after 100% depth of discharge cell to be introduced in the 1940s was the zinc /
(DOD), RAM cells are capable of providing at least mercury oxide cell. Later, in the 1960s, the zinc /
10 times the service hours of single use alkaline silver oxide cell was developed to meet the new
cells, while if discharge is restricted to about 20% market of electric watches (Fig. 2). Both cells
many more hours of service may be obtained. Under employ alkaline electrolytes. Attractive features of
conditions of low DOD cycling, up to 200 Ah these batteries are their high volumetric energy
cumulative charge is possible from a 2.5 Ah RAM density of 500–550 mWh / cm 3 (cf. alkaline MnO 2
cell. Bearing in mind that these cells are cheaper 350 mWh / cm 3 ) and their particularly flat discharge
than Ni / Cd, and do not suffer from the memory curves, even under continuous drain at low currents
effect or from recycling problems with toxic cad- (mA–mA). The anode in both cell types is an
mium, the attraction of RAM cells is obvious, even amalgamated pellet of zinc powder and the cathode a
though their cycle life is less than Ni / Cd. pellet of HgO or Ag 2 O 2 mixed with carbon. Provi-
What are the advances which have made possible sion has to be made in the cell design for the venting
the recharging of alkaline manganese cells? Firstly, of hydrogen, which may arise through corrosion, and
new separators were developed which are highly there is a safety sleeve filled with absorbent material
stable in the strongly alkaline environment and to take up any displaced electrolyte. These zinc-
which prevent zinc dendrite formation (leading to based button cells are particularly suited to applica-
internal short circuiting). Measures were also taken tions requiring constant discharge voltage, but they
to prevent swelling of the cathode. The manganese are gradually being replaced by lithium cells on
oxide electrode was made rechargeable by limiting environmental and cost grounds.
the capacity of the zinc electrode so that discharge Another option for the alkaline zinc cell is to
does not take place beyond the first electron employ an air cathode, with technology similar to
(Mn 41 →Mn 31 ), corresponding to a cut-off voltage that in the alkaline fuel cell. The primary zinc /air
of 0.9 V/ cell. To allow for hydrogen formed by cell is half battery, half fuel cell and has been
corrosion, catalysts are added to the cathode mix so developed in button form particularly for the hear-
as to facilitate the recombination of H 2 gas and also ing-aid market where the attraction lies in the high
to provide overcharge protection via the oxygen volumetric energy density of the cell (|650 mWh /
recombination cycle. Special chargers have been cm 3 ). The development of a satisfactory air cathode
designed to taper charge RAM cells to a maximum provided the greatest technical challenge. For the
of 1.7 V/ cell, so as to prevent overcharge and electrode reaction to take place, it is necessary to
gassing. ensure three-phase contact between the air, the KOH
Industry has been slow to introduce RAM cells for electrolyte and the solid electrode and also to employ
fear of upsetting the primary battery market, but they a good catalyst for oxygen reduction. This was
144 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158

Fig. 2. Cutaway view of zinc–silver oxide button cell.

achieved by using a porous nickel or carbon matrix, MoO 3, V2 O 5 etc. Commercially, the most widely
coated with catalyst, and with a water repellant outer adopted positive-electrode materials are CF x , and
surface so as to maintain a stable liquid meniscus MnO 2 . The electrolyte employed in lithium cells is a
within the thin cathode and also to minimise loss of solution of a lithium salt in an organic solvent. The
water by evaporation. To prevent absorption of CO 2 number of possible combinations of salt and solvent
by the electrolyte during the shelf life of the cell, the is large, and most lithium cell manufacturers have
air inlet is sealed and the seal has to be broken their own preferred solution.
before bringing the cell into use. Lithium /manganese dioxide primary cells are
manufactured in a standard range of cell sizes for
consumer applications, as well as button cells and
3. Lithium batteries special thin cells sized to customers’ requirements.
These cells are noted for their high specific energy
For 15–20 years there has been enormous interna- (260 Wh / kg), flat discharge curves, wide operating
tional activity in the development of lithium bat- temperature range (240 to 1608C), and long shelf
teries, both primary and secondary. The interest life (up to 10 years). It is the latter factor which is
stems from the low atomic mass of lithium (6.94), its causing them to replace alkaline primary cells in
high specific capacity (3.86 Ah / g) and its high smoke detectors in the USA. The standard bobbin
electrochemical reduction potential (23.045 V), all cells are only suited to low rate applications, but
of which contribute to a high specific energy for there is also available a cell constructed on the ‘jelly
lithium cells. roll’ principle (Fig. 3) which may be discharged at
Primary cells normally employ lithium metal foil the 2–4 h rate. This is finding widespread use in
as anodes, whereas secondary lithium cells are not portable radio communications.
usually based on lithium metal since experience has There are several other types of lithium primary
shown that repeated recharging of lithium metal cell which find specialised application.
anodes can be dangerous. Rather, most rechargeable
lithium cells are of the ‘lithium ion’ type in which 1. Lithium / iodine cells: these are all-solid-state cells
Li 1 ions are intercalated into a carbon anode in the comprising a lithium foil anode and a positive
charged state and into an oxide cathode in the made from a chemical complex of iodine and
discharged state. poly-2-vinyl pyridine. When these two solids are
Numerous cathode compounds have been used in brought into contact a thin film of lithium iodide
lithium primary cells, e.g. CuO, CuS, CF x , MnO 2 , is formed and this serves as a solid electrolyte.
 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158 145

Fig. 3. Construction of a high rate lithium / manganese dioxide cell.

Although this electrolyte has a low ionic con- and have an indefinite shelf life prior to activa-
ductivity, the film is very thin and the conduct- tion.
ance is adequate for low drain (mA) service. 4. Lithium polymer cells: in the 1970s it was shown
These solid-state cells have no electrolyte to leak that certain polymers, such as polyethylene oxide,
and are ideal for use where this could be a major could dissolve lithium salts, giving an electrolyte
concern. They are widely employed in heart with a reasonable ionic conductivity at moderate
pacemakers and, to some extent, in watches and temperatures (|1008C). This research gave rise to
pocket calculators. the concept of the polymer battery. Later work
2. Lithium / thionyl chloride cells: these are large, has employed different polymer matrices and
high rate primary cells which are used in military dissolved various organic solvents in them, along
applications such as naval sonar buoys and space with the lithium salts, to produce electrolytes with
power. The discharge reaction is an acceptable conductivity at room temperature
and below. From this extensive materials re-
4Li 1 2SOCl 2 → 4LiCl 1 SO 2 1 S V 0 5 3.6 V search, carried out in many countries over a
20-year period, has evolved the lithium / polymer
The cell has a high specific energy (330 Wh / kg), battery which is now becoming commercially
a high energy density (.700 Wh / dm 3 ), a good available in rechargeable as well as primary form.
low temperature performance and a long shelf- Individual cells are of thickness ca. 0.5 mm and
life. are prepared as large sheets of laminate which
3. Lithium / iron sulphide cells: these cells are used may be contoured into cells of any desired size or
as power sources in missiles and rockets where shape. The manufacture of such cells has involved
the discharge time is short (|minutes). They have the development of much new battery production
a molten salt electrolyte (a eutectic of lithium technology, based largely upon that used in the
halides) and operate at high temperatures (typical- paper, coatings and laminates industries.
ly 400–4508C). A great deal of materials research
has gone into developing this concept. The anode
is a Li /Al alloy and the cathode FeS 2 . The cell is 4. Lead / acid batteries
heated to operating temperature by a pyrotechnic
charge. The characteristics of these cells, which The lead / acid battery was invented in 1859 by
match to the application, is that they are able to Plante´ and developed further by Faure in 1881. In
provide enormous power output for short intervals the original Plante´ cell the active materials of the
146 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158

electrodes were obtained by passing a current be- 1 2H 2 O (2.041 V)

tween soft lead plates. This led to the conversion of
the surface of the positive plate to PbO 2 and that of The sulphuric acid is consumed during discharge and
the negative plate to spongy metallic lead. Modified liberated during charge, allowing the state of charge
Plante´ cells are still used in many standby batteries. to be determined by measuring the specific gravity.
The innovation introduced by Faure was the pasted The theoretical specific energy of the lead / acid
plate wherein a sheet of lead is coated with a paste of battery, counting reactants only, is 170 Wh / kg,
lead oxide and sulphuric acid. By passing a current whereas in practice most real batteries deliver just
through the cell (‘forming the plates’) the positive is 30–50 Wh / kg. This discrepancy is in part accounted
converted to PbO 2 and the negative to spongy lead. for by the mass of the inactive components (cell
Because of the greater mass of active material, the case, grids, current collector bars, separators, termi-
Faure cell has greater capacity than the Plante´ cell. nals, etc.) and in part by poor utilisation of the active
Although invented more than 100 years ago, the materials. The latter stems from slow diffusion
lead / acid battery still has the overwhelming share of processes in the porous reactants and also from the
the secondary battery market. low conductivity of the discharge product (PbSO 4 )
During this century the battery has been subject to which coats individual particles and tends to isolate
continuous improvement in materials of construction, them electrically.
design and manufacture, and this process is still
ongoing. Many of these improvements have been of 4.1. Positive plate alloys
an engineering rather than scientific nature. Some
examples are: Conventional grid alloys are based on lead / an-
timony alloys containing up to 8% Sb. The role of
• the replacement of early glass containers by hard the Sb is to strengthen and harden the Pb through the
rubber and later by high impact polypropylene; formation of a finely dispersed eutectic phase at the
• the development of through-the-wall cell inter- grain boundaries. The tensile strength, yield strength
connects, rather than external cell connectors; and creep resistance of the alloy all increase with Sb
• heat-sealed plastic case-to-cover assemblies; content. In addition, the alloying agent improves
• automatic central watering systems for traction castability and the adherence of the paste to the
batteries; plate. Unfortunately, the addition of antimony has
• the development of sealed cells with safety vents. attendant disadvantages also. During battery charge,
the Sb dissolves (corrodes) progressively from the
Other improvements such as low or zero mainte- positive grid, diffuses through the electrolyte and
nance batteries, new alloys and better designs for the deposits on the negative electrode where it reduces
plates, and better separator materials have all in- the over-potential for hydrogen evolution. This re-
volved a substantial scientific input. The result of sults in greater gassing rates, loss of water and the
this programme of work is that lead / acid batteries need for more frequent maintenance, as well as
have shown a continuous improvement in perform- enhanced self-discharge.
ance and life at relatively lower costs. In order to overcome this problem, a range of
The lead / acid battery is unusual in that the low-Sb alloys has been developed. Early attempts to
electrolyte, sulphuric acid, takes part in the cell use low-Sb alloys led to cracking on casting, but this
charge / discharge reactions as shown: was overcome by the addition of grain refining
agents such as As, Cu, Se, S and Te. These enhance
negative electrode: Pb 1 H 2 SO 4 áPbSO 4
the stiffness, the tensile strength and the corrosion
1 2
1 2H 1 2e (20.356 V) resistance of the grid alloys. Castability is improved
positive electrode: PbO 2 1 H 2 SO 4 1 2H 1 by the addition of 0.1 wt.% Sn which increases the
fluidity of lead alloys. Low-Sb alloys, as used in
1 2e 2 áPbSO 4 1 2H 2 O (1.685 V) ‘low-maintenance’ automotive batteries, contain |2
overall: Pb 1 PbO 2 1 2H 2 SO 4 á2PbSO 4 wt.% Sb.
 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158 147

Many automotive batteries today are of the so- Traditionally, grids are cast from molten alloy,
called ‘maintenance-free’ variety in which the grids either singly or in pairs joined by their lugs, and are
are constructed of Pb–Ca alloy (up to 0.1 wt.% Ca), quench-cooled. The active material is pasted into the
sometimes with the addition of Sn or Sr. On quench- rectangular holes in the grid (Fig. 4). In recent years
ing the grids during manufacture, a dispersion of advanced techniques have been developed for the
Pb 3 Ca forms around the pure lead grains resulting in continuous casting of low-Sb grids, thereby increas-
precipitation hardening. These alloys exhibit a higher ing throughput and productivity. Calcium-containing
hydrogen over-potential than antimonal types and grids may be made by direct casting, although a
water loss during charging is largely eliminated, newer process is to cast a continuous strip of alloy
provided that the top-of-charge voltage is controlled. on a water-cooled drum and subsequently to make
Also they have improved electrical conductivity, but the grid by slitting and expanding into a diamond-
are substantially weaker than their Pb / Sb counter- shaped mesh. The cast strip may be rolled (wrought)
parts. Batteries with lead / calcium grids are normally before slitting and expanding. Much research has
used in ‘float’ conditions (e.g. automotive or stand- gone into optimising the design of grids so as to
by applications). They are unsuited to regular deep maximise their conductivity and their strength for
discharge as the cycle life under these conditions is minimum mass and cost. Computer-aided design has
strictly limited. Explanations advanced for this are been extensively employed.
complex, involving the build-up and cracking of Tubular plates consist of a row of parallel, vertical
corrosion layers on the Pb / Ca plate and structural tubes, made of braided glass-fibre or woven poly-
changes in the porous active material. ester-fibre in the form of a gauntlet. Into each tube,
along the central axis, is placed a thin lead alloy rod
(or ‘spine’) which acts as current collector. These
4.2. Plate design and manufacture spines are cast as a single set with a common
‘header’ (Fig. 4) and the active material is packed
The Plante´ version of the positive plate, as used in into the tubes around the spines. The principal
large stand-by batteries, is made from a casting of advantage of this design is that the gauntlet tube
pure lead. It has a surface of vertical grooves to hold holds the active material in place, which is necessary
the active material, which is formed by electro- for repeated deep discharge. The disadvantages are
chemical oxidation of the lead itself. There are two that the specific power of the battery tends to be
further designs of positive plate: lower than with grids and the manufacturing cost
higher.
• grids, as used in automotive and leisure batteries; From this brief review it will be clear that
• tubular plates, as used in traction batteries. fundamental metallurgical research has played a

Fig. 4. Lead / acid battery plates with (a) flat plates and (b) tubular plates.
148 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158

major role in lead / acid battery development in recent borosilicate glass microfibres processed into sheet
years, both in terms of improving the performance of form by a wet process on a conventional paper-
batteries and in facilitating their manufacture. making machine. Individual fibres are of diameter
0.25–4 mm and length of |1 mm and mostly lie in
4.3. Separators the plane of the paper. Glass-microfibre separators
meet all of the above requirements for a separator
Separators are interleaved between the plates of and also have high compressibility, so that they
batteries so as to prevent the plates touching and conform to irregularities in the plate surface with
giving rise to internal short circuits and also to help consequent good acid transfer and low internal
retain the active material in firm contact with the resistance.
grids. Originally ‘leaf’ separators were made of
natural materials such as wood veneers and paper. A 4.4. Valve-regulated lead /acid ( VRLA) batteries
more recent trend has been towards ‘envelope’
separators which enclose the electrode and so pre- In addition to the requirement for low-mainte-
vent dendritic growth around the edge which tends to nance or maintenance-free lead / acid batteries, there
occur with a leaf separator. Modern separators are is also a demand for sealed cells which may be used
generally made from a microporous polymer, such as in any orientation. The problem lies in eliminating or
polyethylene or polyvinyl chloride, loaded with an accommodating the gas which is liberated on charg-
inorganic filler, or from glass-microfibre. These are ing. Over the years many attempts have been made
highly sophisticated materials, into which a great to develop sealed lead / acid batteries using the
deal of development work has gone. The require- principle of catalytic H 2 / O 2 converters sealed into
ments for a satisfactory separator are: the top of the cell. The catalyst was usually platinum
or palladium deposited on a high-area support such
• that it should have a uniform, microporous struc- as pelleted alumina. These devices were found never
ture (|60% porosity, pore size ,1 mm); to be totally reliable, as they tend to flood with
• adequate strength and flexibility during storage electrolyte in the pores, thereby isolating the catalyst
and use; and leading to the build-up of gas pressure.
• good chemical stability in battery strength sul- The solution to the problem lay in electrochemical
phuric acid; rather than catalytic recombination, with oxygen
• high oxidation resistance to maximum operating liberated at the positive electrode on charge being
temperature; recombined electrochemically at the negative elec-
• low electrical resistance; trode. This process moves the potential of both
• high purity. electrodes in a positive direction and thereby inhibits
hydrogen liberation at the negative. The problem
Most polymeric separators utilise a purified, pre- then reduces to one of ensuring that the liberated
cipitated silica as the carrier for a pore-forming agent oxygen has unfailing access to the negative elec-
such as starch or oil. The precipitated silica has a trode, not a simple matter in a flooded cell. This was
skeletal structure and is highly absorptive and acid- solved by operating in a starved electrolyte mode
resistant. It is also hydrophilic and easily wetted, whereby the acid is absorbed in the interstices of the
contributing to a low electrical resistance. The separator, leaving sufficient voidage for the oxygen
internal resistance of lead / acid batteries is measured to pass through the separator to the negative plate.
in milliohms. The separator materials are designed Two types of separator have been employed in
with vertical ribs which provide added strength and VRLA batteries viz: absorptive glass-microfibre
also facilitate the convective flow of acid over the (AGM) separators and gel type separators. The latter
plates as its specific gravity changes during charge / are prepared by mixing very fine ‘fumed’ silica
discharge. (particle size |10 nm) with sulphuric acid and
Glass-microfibre separators are used particularly in allowing to stand, when it undergoes gelation. When
valve regulated lead / acid batteries (see Section 4.4, the gel has partially dried on standing, it develops
below). These separators are made from high purity cracks and fissures through which oxygen passes to
 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158 149

the negative electrode. Both types of separator have development over the years, especially as regards the
been employed successfully in sealed batteries. common nickel oxide positive electrode.
These batteries are always fitted with pressure re-
lease safety valves which accounts for their name 5.1. Nickel oxide electrodes
(valve regulated lead / acid batteries). There is still
considerable debate and discussion over the precise There are four basic methods for preparing nickel
mechanisms by which these cells operate and also oxide electrodes.
their modes of failure.
The first VRLA sealed cells were made in the 1. Pocket plate electrodes: this is the traditional
1970s as small (D size) cells by the Gates Corpora- method. The positive active material is contained
tion in the USA. The approach adopted was to take a in flat, rectangular channels (or ‘pockets’) made
single pair of positive and negative plates interleaved from perforated, nickel-plated steel strip (Fig. 5).
with an AGM separator and to spirally wind the The channels are then interlinked horizontally to
assembly to form a cell (as in Fig. 3). VRLA form a plate. The active material consists of a
batteries are now produced in a range of larger sizes mixture of Ni(OH) 2 and either nickel flake or
in rectangular battery housings. graphite. The latter component is added to pro-
In summary, the advances which have been made vide a conducting matrix. Often a few % of cobalt
in lead / acid battery technology over the past 50 hydroxide is added to improve cell capacity and
years have been largely of a materials and design cycle life.
nature, involving the metallurgy of lead alloys, the 2. Plastic bonded electrodes: the electro-active ma-
manufacture of grids, the science of separators (both terials are simply bonded into electrodes using a
polymeric and glass-microfibre), the development of polymer binder. This is the lowest cost electrode,
safety vents, high impact polypropylene battery which finds widespread application in small re-
housings and their sealing, and the internal cell-to- chargeable cells for consumer products.
cell connectors. 3. Sintered plate electrodes: a porous nickel plaque
is first prepared by lightly compacting nickel
powder onto a perforated nickel screen or gauze
and sintering in hydrogen at high temperature.
5. Alkaline electrolyte secondary batteries
The plaque should have |80% porosity. It is
charged with active material by chemical or
Rechargeable alkaline electrolyte batteries were
electrochemical impregnation. These electrodes
invented at the end of the 19th century by Jungner in
have a lower internal resistance than the above
Sweden and Edison in the USA. These were based
types and are capable of delivering higher specific
upon nickel oxide positive electrodes and either iron
energy and power.
or cadmium negatives, and are popularly known as
4. Fibre electrodes: these latest electrodes are based
the ‘nickel / iron’ and ‘nickel / cadmium’ batteries,
upon compressed nickel fibres, or nickel felts or
respectively. The electrolyte is concentrated KOH
foams, which again are impregnated with the
solution. The overall chemistry of each cell is
active material.
analogous:

Fe12NiOOH14H 2 OáFe(OH) 2 12Ni(OH) 2 .H 2 O The latter two electrode types have been de-
veloped over the last 20–30 years and depend upon
ocv51.37 V having appropriate starting materials, which are of
fairly recent origin.
Cd12NiOOH14H 2 OáCd(OH) 2 12Ni(OH) 2 .H 2 O
5.2. Nickel /iron batteries
ocv51.30 V
Nickel / iron batteries found early industrial use as
Both batteries were commercialised at an early traction batteries in forklift trucks, mine and railway
date and have been subject to parallel research and locomotives, etc. Usually, both the positive and
150 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158

Fig. 5. Construction of pocket plate nickel oxide electrodes.

negative electrodes were of the pocket plate type.

The attraction of nickel / iron batteries is that they
have 1.5–2 times the specific energy of lead / acid
Table 2
batteries and, relatively, are particularly good at high Comparative specific energies of Ni / Fe and Pb / acid batteries at
discharge rates (Table 2). two discharge rates
The battery is also noted for its ruggedness and
Rate (W/ kg) Ni / Fe (Wh / kg) Pb / acid (Wh / kg)
long cycle life at deep discharge (2000 cycles at 80%
DOD). However, compared to lead / acid it has 20 54 36
40 50 26
several disadvantages:
 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158 151

• its low temperature performance is inferior; applications on railways, for marine duties, on board
• it has a comparatively high corrosion and self aircraft, and as traction batteries for mine locomo-
discharge rate; tives, industrial trucks, etc.
• its overall electrical efficiency is poor because of Over the past 40–50 years small, sealed Ni / Cd
a low over-potential for hydrogen evolution at the batteries have been introduced which can be used in
iron electrode. Considerable gassing on charge any orientation. The capacity of these ranges from 10
leads to a need for frequent maintenance. mAh to 15 Ah. The KOH electrolyte is absorbed in
several layers of reservoir separator, often a cel-
For these reasons, the applications for Ni / Fe lulosic material which allows free passage of oxygen
batteries have been limited in scope. Many attempts liberated on charge to pass to the negative electrode
at solving the gassing problem have not been suc- where it is reduced. These sealed cells are analogous
cessful. The Ni / Cd battery has enjoyed greater to the valve-regulated type of lead / acid battery as
success commercially as it is not subject to these regards their recombination technology and also in
limitations. having safety valves to relieve excess pressure. They
are manufactured both as prismatic cells and as
5.3. Nickel /cadmium batteries cylindrical cells, the latter as replacements for pri-
mary Zn / MnO 2 cells in standard sizes. Because of
After lead / acid, nickel / cadmium is the most the lower voltage, a five cell Ni / Cd battery (6 V)
widely used rechargeable battery. The cell operating replaces a four cell Zn / MnO 2 battery. The cylindri-
voltage (1.20 V) is slightly less than that of Ni / Fe cal cell may have either a spiral (‘jelly-roll’) con-
(1.25 V) and the mass of cadmium is higher than struction, for high power output, or a bobbin con-
that of iron. Taken together, these two factors lead to struction, for higher volumetric capacity. Sealed Ni /
a specific energy for the battery (30–40 Wh / kg) Cd cells find their primary use in consumer applica-
which is nearer to that of lead / acid than Ni / Fe. tions, although special versions have been designed
However, the high rate and low temperature per- and used in telecommunications, and also in satel-
formance is better than that of lead / acid. Other lites and spacecraft. Nickel / hydrogen batteries are
positive features of Ni / Cd batteries are a flat dis- progressively replacing Ni / Cd for the latter applica-
charge voltage, long cycle life (|2000), continuous tion.
overcharge capability, low maintenance requirement
and excellent reliability. Cells and batteries are
available in many different sizes and with pocket- 5.4. Nickel /hydrogen and nickel /metal hydride
plate, plastic-bonded, or sintered electrodes. The cells
principal disadvantages of Ni / Cd batteries are their
high cost (up to 10 times that of lead / acid) and the There are relatively few negative electrode materi-
environmental concerns associated with the disposal als which can be used in strong alkaline solution, the
of batteries containing toxic cadmium. most promising being cadmium, iron, zinc and
Apart from the introduction of sintered electrodes, hydrogen. The nickel /hydrogen battery has a sin-
the improvements associated with Ni / Cd batteries tered nickel oxide positive electrode and a platinum /
over the past 50 years have been more in the area of hydrogen negative electrode. This battery was de-
cell design than in basic materials science. Tradition- veloped specifically to replace Ni / Cd in space
al cells are of the vented variety and these are sold in applications on account of its higher specific energy
large sizes for various applications. The reliability content (ca. 50 Wh / kg) coupled to a very long life
and low maintenance of the battery makes it ideal for when subjected to deep discharge cycling.
stand-by power applications, whilst its high power The Pt / H 2 electrode is fully reversible and, when
output, maintained at low temperatures, makes it combined with a nickel oxide positive, forms a cell
suited to the starting of large engines where ‘cold- of potential 1.25 V. The cell reactions are as follows:
cranking’ currents of 5000–10 000 amps may be
required. Vented nickel / cadmium batteries also find H 2 1 2OH 2 á2H 2 O 1 2e 2 negative
152 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158

2NiOOH 1 2H 2 O 1 2e 2 á2Ni(OH) 2 1 trolyte. Ni / H 2 cells may be overcharged since

2 liberated oxygen recombines at the negative elec-
2OH positive
trode. Nickel / hydrogen space-cells are a sophisti-
cated product which are exceedingly expensive to
Note that there is a counterflow of water mole- buy, but which nevertheless have largely usurped
cules and OH 2 ions across the separator. The Ni / Cd for use in satellites.
hydrogen gas liberated on charging is stored under Nickel /metal hydride cells are a later development
pressure (|40 bar) within the cell itself. This necessi- of nickel / hydrogen for terrestrial applications. They
tated considerable development work on the design depend on the principle that hydrogen may be stored
and construction of the cell vessel, the cell stack and reversibly in the form of a metal hydride which
its supports, and the electrical leads through the cell forms the negative electrode of the cell. The positive
wall. The pressure vessel is cylindrical in shape, with electrode is a standard nickel oxide electrode. Early
hemispherical end caps, and is constructed of thin work, in the 1970s, utilised LaNi 5 alloy as the
gauge inconel alloy (Fig. 6). As the cell is cycled, negative; this takes up hydrogen reversibly at am-
the hydrogen pressure cycles also, from |40 bar in bient temperature to form LaNi 5 H 6.5. This alloy has
the charged state to |2 bar in the discharged state, a dissociation pressure of 1 bar at 158C. Later
and the pressure vessel must withstand this treatment commercial development work has focused on two
without embrittling, corroding or cracking. new alloys to store hydrogen.
The cell stack is built up from back-to-back
sintered nickel electrodes, electrochemically impreg- 1. A complex alloy based upon the rare earth
nated, and standard hydrogen fuel cell electrodes of ‘misch-metall’ with various additives to adjust the
platinum black dispersed on carbon paper. The dissociation pressure to the desired value and also
separators are formed from a woven cloth of yttria- to form a surface oxide film which acts as a
stabilised zirconia which absorbs the KOH elec- barrier to prevent oxidation of the metal hydride.

Fig. 6. 40 Ah nickel / hydrogen cell as used in satellites.

 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158 153

This is a proprietary formulation of the AB 5 -type such as electric vehicle traction, storage in the
alloy, where A is a mixture of rare earths and B is electricity supply system, etc. The materials prob-
partially substituted nickel. lems associated with these are outlined below.
2. A similarly complex, multi-component, alloy of
the type AB 2 , where A5Ti or Zr. 7.1. Zinc /air batteries

The latter is claimed to have higher hydrogen Zinc / air primary batteries are well-known in small
storage capacity, superior oxidation and corrosion sizes for use in hearing aids and similar applications.
resistance and to be less costly. However, all attempts to develop electrically re-
The operating voltage of a Ni / MH cell is almost chargeable zinc / air cells have met with little success,
the same as that of Ni / Cd (1.2–1.3 V), making for primarily on account of the growth of zinc dendrites
ready interchangeability, and the discharge curve is leading to internal short circuits. An alternative
flat. The specific energy of Ni / MH batteries (60–70 approach, now under serious development for large
Wh / kg) is 1.5–2 times higher than that of Ni / Cd batteries, is to produce a mechanically rechargeable
and their specific power may be as high as 250 W/ kg battery. This is essentially a large primary battery in
Batteries are resilient to overcharge and overdis- which, at the end of discharge, the spent zinc
charge and may be operated from 230 to 1458C. hydroxide and KOH electrolyte are removed from
Cells of both cylindrical and prismatic design are the cell stack and returned to a central plant for
now manufactured in a range of sizes; small cells are recycling. There the slurry is electrolysed to produce
used in portable electronic devices (e.g. mobile fresh zinc powder which is returned to the battery
phones), while prismatic cells of 100 Ah capacity are stack. Such mechanical recharging is attractive be-
available for assembly into 12–14 V modules (e.g. cause the positive reactant (atmospheric oxygen) is
for use as traction batteries). The new materials available in continuous supply and it is not necessary
technology involved in the development of Ni / MH to develop a bifunctional oxygen electrode. Another
batteries is almost entirely associated with the hy- advantage of mechanical recharging is that it may be
dride negative electrode. accomplished in a few minutes, rather than the hours
required for most electrical recharging.
The mechanically rechargeable zinc / air battery
6. Lithium ion batteries has been developed in Israel for use in electric
vehicle traction. It has a high specific energy (200
Rechargeable lithium ion batteries represent an Wh / kg), but only a modest specific power (100
entirely new class of battery, developed and com- W/ kg at 80% discharge state). Modules comprise a
mercialised in the 1990s. They are now rapidly stack of 66 cells and an eight module battery (150
superseding alkaline batteries for a wide range of kWh) has been built and tested in Germany in a
applications in the portable electronics field, and Mercedes Benz 410 postal van. The vehicle had a
have potential for much larger cells and modules. range of 300 km between recharges.
The development of lithium ion batteries involves a In order to succeed with this mechanically re-
challenging set of issues in materials science and chargeable system, it was necessary first to develop
technology, and research in this area is extremely three distinct units: the cell stack, the factory-based
active. Professor C.A. Vincent will be covering this fuel regeneration plant, and the stack dismantling
topic in his lecture. and refuelling machine. The first two posed both
electrochemical and mechanical engineering prob-
lems, while the latter was a purely mechanical
7. Future concepts problem. The general approach adopted was to have
a bed of zinc powder particles compacted on to a
There are a number of other battery chemistries in current collector frame and inserted in a separator
an advanced stage of development, but not yet envelope, flanked on both sides by air electrodes.
commercialised, for use in large-scale applications This electrode assembly unit forms a cassette which
154 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158

may be removed from the cell stack for recycling with water. When bromine is liberated during charge,
and replaced by a fresh cassette. The depleted the polybromide complex is formed as droplets,
cassettes are conveyed to a central facility for which separate from the aqueous electrolyte stream
electrochemical regeneration to produce particulate and are stored in a liquid reservoir external to the
zinc. cell stack (Fig. 7). Zinc electroplates on the negative
electrode in a satisfactory (non-dendritic) morpholo-
7.2. Zinc /bromine batteries gy, aided by the flowing electrolyte. On discharge,
bromine is returned to the cell in the form of a
The zinc / bromine battery is an example of a ‘flow dispersion of the polybromide oil in the aqueous
battery’ in which one of the reactants (bromine) is electrolyte.
stored externally to the cell stack and flows through In principle, the construction of the cell stack and
it during charge and discharge. The electrode re- housing is straightforward, although there are materi-
actions are als problems stemming from the reactivity of
bromine. A bipolar electrode is used, in which the
ZnáZn 21 1 2e 2 E 0 5 2 0.76 V conducting bipole plate is a lightweight carbon /
Br 2 1 2e 2 á2Br 2 E 0 5 1 1.065 V plastic composite formed by extrusion. The
separator, which is injection moulded, is housed in a
giving a cell voltage of 1.82 V. The electrolyte used non-conductive plastic frame which incorporates
is an acid (pH 3) solution of concentrated zinc electrolyte flow channels. These stack components
bromide from which zinc can be plated. Several are lightweight and simple to manufacture, although
different manifestations of the zinc / bromine battery the plastic employed has to be selected for stability
have been proposed, the most successful of which towards bromine. Similarly the reservoirs, connect-
stores the excess bromine external to the cell as a ing tubing, pumps and gaskets all have to withstand
complex with a quaternary ammonium bromide, such degradation by bromine.
as an alkyl morpholinium bromide or alkyl Considerable success has been achieved with this
pyrrolidium bromide. The complex of these salts battery in several countries and a 22 kWh unit has
with Br 2 is a dense, oily liquid which is immiscible been constructed and used to power an electric van.

Fig. 7. Schematic of zinc / bromine battery.

 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158 155

This 108 V battery had a specific energy of 65–70 the plate and frame principle, is being developed on
Wh / kg and a peak power density of 100 W/ kg. An a large scale for load levelling in the electrical
electric racing car has also been demonstrated po- supply industry.
wered by a Zn / Br 2 battery. Some concern has been
expressed about the safety aspects of a battery 7.4. Sodium batteries
incorporating bromine.
Sodium is most attractive as a negative electrode
7.3. Redox batteries reactant on account of its high electrochemical
reduction potential of 22.71 V (cf. Zn 20.76 V).
There is growing interest in secondary redox When coupled with an appropriate electropositive
batteries which possess a divided cell configuration, material, it is capable of giving a cell of voltage .2
like the zinc / bromine battery, but which operate on V. Moreover, sodium is abundant in nature, cheap
the principle of the oxidation and reduction of ions in and non-toxic. It is also of low atomic mass (23.0)
solution. The better known redox batteries involve and the combination of high voltage and low mass
the oxidation and reduction of cations in solutions leads to the possibility of a battery of high specific
separated by an anion exchange membrane which is energy. The disadvantages of sodium metal as a
nominally impermeable to cations. The cations in the negative reactant is that an aqueous electrolyte
two compartments may either be different (e.g. Cr 21 cannot be employed and it is a potential fire hazard.
and Fe 31 ) or the same, but in different valence states The realisation of a practical battery based upon
(e.g. V 31 and V 51 ). As with the Zn / Br 2 battery, sodium depended upon identifying a suitable non-
there are anode and cathode recirculating loops and aqueous electrolyte. This occurred in 1967 when
the bulk of the reactant is stored outside of the cell scientists working at the Ford Motor Company in
stack in four external reservoirs, two for the reactants USA showed that sodium beta alumina, a form of
in the charged state and two in the discharged state. sodium aluminium oxide, is highly conductive to-
Pumps are used to circulate the reactant solutions. wards Na 1 ions at |3008C, while being a good
The capacity of the battery is limited only by the size electronic insulator. This gave rise to the possibility
of the storage tanks and, in this regard, the system is of a solid ceramic electrolyte. They further demon-
more akin to a fuel cell than a conventional battery. strated that, using this electrolyte in the form of a
Although these batteries are bulky and have rather sintered ceramic tube, it is possible to construct a
low specific energy and energy density, they hold the cell with a liquid sodium negative electrode and a
potential for low cost storage of electricity on a fairly liquid sulphur positive which operates effectively at
large scale. The vanadium redox battery is being 300–4008C. Since sulphur is an insulator, it has to be
actively developed in Australia for the storage of absorbed in a carbon felt electrode. The construction
solar electricity. The batteries are constructed on the of the cell is shown schematically in Fig. 8. The
‘plate and frame’ principle using plastic components. practical construction is rather different as there are
As the reactant solutions are not chemically aggres- numerous materials science and engineering prob-
sive, unlike bromine, there are no serious materials lems to be faced.
problems to solve. The cell discharges in two steps as Na 1 ions pass
Recently a new type of Redox battery based on the through the beta alumina to the sulphur electrode:
oxidation and reduction of anions has been under
development in the UK (the ‘Regenesys’E battery). Step 1 2Na 1 5S ⇒ Na 2 S 5 E 0 5 2.076 V
This utilises a separator which is permeable to Step 2 2xNa 1 (5 2 x)Na 2 S 5 ⇒ 5Na 2 S 52x ( 0,x ,2)
cations but not anions. During discharge the negative
22
electrode reaction involves the oxidation of S ions E 0 5 2.076 → 1.78 V
in Na 2 S solution to sulphur, while the positive
electrode reaction is the reduction of Br 2 dissolved in In the first step, sodium polysulphide (Na 2 S 5 ) is
NaBr solution to Br 22 ions. The voltage of this cell formed as a liquid, immiscible with liquid sulphur.
is 1.57 V. The battery, which again is constructed on The open circuit voltage is therefore invariant at
156 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158

development of sodium / sulphur batteries. Among

these were the following.

• Corrosion: hot sodium sulphide is highly corro-

sive to steels and other alloys. This poses a major
problem for materials of cell construction.
• Seals: it is necessary to form an insulating seal
between the two electrodes and also to seal each
compartment against air ingress; chemical stabili-
ty is a problem.
• Safety: unrestrained reaction between liquid so-
dium and liquid sulphur (in the event of elec-
trolyte fracture) leads to a fierce fire and it is
necessary to incorporate safety features in the cell
design to prevent this happening.
• Top-of-charge problems: the Na / S cell has no
Fig. 8. Schematic of sodium / sulphur cell. overcharge mechanism and a fully charged cell is
insulating. This means that in a chain of cells
connected in series the first cell to reach top of
2.076 V. Once all the sulphur has been converted to charge experiences a high voltage across it which
Na 2 S 5 , an ionically conducting liquid, further dis- can lead to ceramic breakdown. This problem can
charge to Na 2 S 3 takes place in a single phase and only be avoided by connecting cells in a series /
therefore the voltage declines linearly to 1.78 V at parallel array with many cross connections.
the composition Na 2 S 3 . These reactions are fully • Cell failure: a related problem is that when cells
reversible on recharge. fail they tend to go open circuit, either immedi-
The sodium aluminium oxide phase diagram is ately or after a while. Again, this calls for an
complex and a great deal of work has gone into automatic procedure to short circuit the failed cell
optimising sodium beta alumina ceramic for this or a small group of cells.
application. The specification for the electrolyte
tubes (Table 3) is among the most stringent ever set Although all of these problems have been ad-
for any ceramic material. dressed in detail, with some measure of success,
Many other problems were encountered in the most of the industrial development programmes have
now been terminated (with the notable exception of
that in Japan) for a mixture of technical and commer-
Table 3
Specification for beta alumina tubes a cial reasons.
In parallel with the sodium / sulphur programme,
High ionic conductivity at 3008C
there has been a development programme on a
Near theoretical density and impervious
High fracture toughness related battery system, the sodium / nickel chloride
Fine grain size and no flaws, for strength battery. This maintains many features of the Na / S
Straightness and close dimensional tolerances battery (the beta alumina electrolyte, the sodium
Good surface finish and closed end electrode, hot operation etc.) but replaces the liquid
Stable to moisture
sulphur electrode by a solid nickel chloride elec-
Good wetting by sodium
Long life at high current densities trode. The reversible cell reaction is:
Resistant to electrical breakdown
High yield with reproducible properties Na 1 NiCl 2 ⇒ Ni 1 2NaCl E 0 5 2.58 V
Rapid, low cost production
a
After many years of research and development, this spe- A second (liquid) electrolyte, sodium chloralumi-
cification has largely been met by several manufacturers. nate NaAlCl 4 , is added to the positive electrode
 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158 157

Although sodium batteries do not suffer from

parasitic side reactions, as are commonly found in
aqueous electrolyte batteries, their high temperature
of operation inevitably results in some thermal loss
which has to be made up electrically. This is
effectively equivalent to a Faradaic inefficiency in a
conventional battery. The development of vacuum
enclosures however has reduced this loss to an
acceptable level.
Development work on the sodium / nickel chloride
battery for use in electric vehicles has been in
progress for about 20 years in the UK, Germany and
South Africa. A pilot production line has been built
and full-size traction batteries manufactured for fleet
Fig. 9. Schematic of sodium / nickel chloride cell. trials in electric vehicles. The performance of these
batteries has exceeded expectation in terms of re-
liability, ease of thermal control and lifespan. Elec-
compartment so as to make good ionic contact tric cars and vans fitted with these batteries are an
between the solid NiCl 2 and the solid electrolyte attractive proposition and the project has now been
(Fig. 9). taken over by a company in Switzerland. Future
Experience with this cell showed that it has many development will require the construction of a full-
advantages over the Na / S cell: scale manufacturing plant.

• it is assembled in the discharged state and

charged only after sealing; this eliminates the 8. Conclusion
need to handle liquid sodium;
• the raw materials employed are nickel powder Over the past 50 years there has been a huge
and common salt, both commercially available increase in market demand for batteries, principally
and affordable; as a result of the many new requirements for portable
• because the sodium is produced in-situ, by elec- power in modern devices. Primary batteries are
trochemical reduction of Na 1 ions which have mostly confined to small sizes for use in the con-
passed through the beta alumina tube, it is ultra- sumer market, although large primary batteries are
pure; needed in certain defence applications. Secondary
• the cell operates at a lower temperature (250– batteries are used in both small and large applica-
3508) than sodium / sulphur (300–4008); tions.
• the cell has an overcharge and overdischarge The battery industry, ably supported by academic
mechanism which helps to balance out capacity researchers, has responded to the challenge for
inequalities between cells; series chains of cells improvements to existing batteries, as well as the
are therefore possible with no cross-connections; development of entirely new battery types. Much of
• a central positive cell design is employed and this this development work has involved solving prob-
eliminates corrosion problems; NaAlCl 4 is by no lems in materials science and fabrication technology.
means as corrosive as sodium polysulphide; As a result we have greatly improved alkaline
• hermetic seals are easier to design and make for manganese cells, both primary and rechargeable, new
this battery; types of maintenance-free lead / acid battery and
• in contrast to Na / S, there are almost no safety sealed lead / acid battery, and better performance
problems with the Na / NiCl 2 battery; nickel / cadmium batteries. Among the new types of
• the nickel powder may readily be recovered from rechargeable battery which have become commer-
a discharged battery and recycled. cially available in the past decade, the nickel / metal
158 R.M. Dell / Solid State Ionics 134 (2000) 139 – 158

hydride and lithium ion cells are outstanding. Final- 9.2. Books
ly, a number of futuristic battery types are in an
advanced stage of development particularly for large- Berndt, D., Maintenance-free Batteries (2nd Edi-
scale applications such as electric vehicle traction, tion), Research Studies Press (1997).
power storage in the electricity supply network and Falk S.U., Salkind A.J., Alkaline Storage Bat-
for use in conjunction with solar and other renewable teries, John Wiley (1969).
energy resources. The future for enterprising com- Gabano J.-P. (Ed.), Lithium Batteries, Academic
panies in the battery industry looks promising. Press (1983).
Linden D. (Ed.), Handbook of Batteries and Fuel
Cells (2nd Edition), McGraw-Hill (1995).
9. Battery bibliography McNichol B.D., Rand D.A.J. (Eds.), Power
Sources for Electric Vehicles, Elsevier (1984).
9.1. Journals Murphy D.W., Broadhead J., Steele B.C.H. (Eds.),
Materials for Advanced Batteries, NATO ASI
Batteries International (Euromoney Publications Series, Plenum Press (1980).
plc). Rand D.A.J., Woods R., Dell R.M., Batteries for
Electrochimica Acta. Electric Vehicles, Research Studies Press Ltd.
International Power Sources Symposia (USA). (1998).
Inter-Society Energy Conversion Engineering Reasbeck P., Smith J.G., Batteries for Automotive
Conferences (USA). Use, Research Studies Press Ltd. (1997).
Journal of Applied Electrochemistry. Sequeira C.A.C., Hooper, A. (Eds.), Solid State
Journal of the Electrochemical Society (USA). Batteries, NATO ASI Series, Martinus Nijhoff
Journal of Power Sources (Elsevier Press). Publishers (1985).
Power Sources Symposia, Vols. 1–20 (UK). Sudworth J.L., Tilley A.R., The Sodium Sulfur
Proceedings of the International Seminars on Battery, Chapman and Hall (1985).
Primary and Secondary Battery Technology Tuck C.D.S. (Ed.), Modern Battery Technology,
(USA). Ellis Horwood, UK (1991)
Progress in Batteries and Solar Cells. Vincent C.A., Modern Batteries (2nd Edition),
Solid State Ionics. Edward Arnold, London (1998).