Poggendorff Bichromate Cell. This cell resembles the above in
using a liquid depolariser, but differs from it in containing only one
liquid and in dispensing with a diaphragm. The electrolyte is a solution
containing H 2 S0 4 and either potassium or sodium bichromate. It of
course evolves no fumes. The total reaction at the cathode can be
Cr t 7 " + 14H* > 2Cr'" + 7H 2 + 6 ,
and that in the whole cell by
3Zn + Cr 2 0/ + HH' > 2Cr'" + 3Zn" + 7H 2 0.
The electrolyte very slowly attacks the zinc electrode chemically,
and, in order completely to avoid local action during long idle periods,
the latter is usually arranged so that it can be lifted out of the solution
or immersed in it at will. The use of Na 2 Cr 2 7 in the electrolyte is
preferable to that of the potassium salt, owing to its greater solubility.
This means a higher capacity, and a lesser tendency to depolarisation.
If potassium bichromate be used, the solution contains 8-10 parts
of this salt and 10-18 parts H 2 S0 4 per 100 parts water ; and with
sodium bichromate, 17-20 parts of salt and 20-24 parts of H 2 S0 4 per
100 parts water. The resistance of the cell is about 0'3 ohm low,
though higher than that of the Bunsen cell. The cathodes are of
We have seen * that the potential of a chromic acid solution in H 2 S0 4
is higher than the potential of 66 per cent. HN0 3 . It follows that the
E.M.F. of the Poggendorff cell should exceed that of the Bunsen cell,
which is the case. It is usually 2*0 volts. On the other hand, it
polarises more easily. This is partly due to sluggishness in oxidising
1 Ihle, Zeitoch. Elektrochem. 2, 174 (1895). " Table, p. 196.
xv.] PRIMAKY CELLS 203
the discharged hydrogen, but also to concentration polarisation. The
Cr 2 7 " ions around the cathodes and in their pores become exhausted
and are replaced by Or'" ions. 1 On standing, these concentration
changes are neutralised by diffusion and the cell recovers. The polar-
isation due to slow action between the chromic acid and hydrogen can
be largely removed by adding some soluble chloride, which acts as a
catalyst. It unfortunately also increases the rate of chemical solution
of the zinc by the electrolyte. Owing to the high voltage, and its com-
paratively ready polarisability, the bichromate cell is particularly
suitable for furnishing large currents for short periods, between which
it is allowed to recover.
Lalande Cells. These cells also have a soluble zinc anode. But,
unlike those already considered, the electrolyte is alkaline, not acid,
and the cathodic depolariser a solid, viz. CuO. The system is
Copper | Oxides of copper . Alkali | Zinc.
First invented by Lalande, they were worked on by Edison and others,
and successful forms developed by Messrs. Umbreit and Matthes of
Leipzig (Cupron element) and by Wedekind (in England known as
the Neotherm cell). These latest types are undoubtedly the best and
most efficient primary cells we possess.
In construction the Cupron element consists of a rectangular glass
trough, with an ebonite lid through which pass the nickel terminals
of the plates inside. These number three, and hang parallel to one
another. The two outer are of amalgamated zinc, the inner is the
porous copper oxide plate. This is constructed by rilling a flat cage
of copper gauze with cupric hydroxide, compressing strongly, and
subsequently baking. 15 per cent, to 18 per cent. NaOH solution is used.
The Cupron cells have an E.M.F. (open circuit) of about 1*0-1 'I volts.
This very rapidly falls when current is furnished to about 0*9 volt,
and then slowly decreases to 0'75 volt during the discharge. At this
point it commences to again rapidly fall and the cell polarises. As long,
however, as the active material is still not all reduced, very considerable
currents can be given without causing the working voltage to appre-
ciably fall. The depolariser acts quickly, and the internal resistance
is very low 0*03 to 0*05 ohm. When run down, fresh zinc plates and
NaOH are put in, and the cupric oxide plate regenerated by heating
it to 150 in an oven for example.
The Neotherm element consists of a rectangular iron containing
vessel, closed with an indiarubber gland and an iron lid. Through
the lid pass ebonite bushes, insulating the terminals. There is one
zinc electrode, a vertical plate suspended in the middle of the cell.
1 In the Bunsen cell, although the HN0 3 is consumed, there is no accumulation
of other products, as the gases formed stream away.
204 PKINCIPLES OF APPLIED ELECTROCHEMISTRY [CHAP.
The copper oxide depolariser is caused by a special process to adhere
to the cell wall, and this is directly connected with the positive terminal.
Like the Cupron cell, the E.M.F. of the Neotherm cell is l'0-l'l volts,
and its discharge voltage 0'90-0-75, though it is customary to take it
down to 0'5-0*6 volt when working. Below that limit it drops very
rapidly. The internal resistance is even smaller than that of the Cupron
cell. The copper oxide is regenerated by heat, the whole cell body
being warmed up.
In spite of its low voltage, the Lalande cell, now that difficulties
inherent in the CuO plate have been overcome, has won considerable
favour, and is undoubtedly to be preferred to any other type of primary
cell. It is compact in structure, and has a high capacity per unit-
weight. (A Neotherm cell weighing 12 Ibs. will give 150 ampere-hours
if discharged at 1 amp.) The zinc consumption is practically theore-
tical during discharge, and local action whilst standing is negligible,
as the solubilities of CuO and Cu 2 (particularly the latter) in the
electrolyte are very small. The alkali is well protected against atmo-
spheric C0 2 . The positive plate depolarises rapidly, and admits of
large currents, whilst the regeneration is quick and convenient. There
are no diaphragms or similar complications, and the resistance is very
In the Lalande element the zinc dissolves as a complex anioii,
probably according to the equation
Zn + 30H' > Zn0 2 H' + H 2 + 20.
Owing to the exceedingly low Zn" concentration in equilibrium with
this complex, the anode potential will have a high negative value.
Measured against a normal calomel electrode, Lorenz * found a voltage
difference of T55 volts, which would correspond to a single potential
of about 1'3 volts. We know that such an electrode is 0-4 volt more
negative than a hydrogen electrode in the same solution, and about
- 1'24 volts is a more probable average figure for this single potential
(assuming Zn | n. Zn" = 0*77 volt).
The change of the copper oxide plate potential during working
lias been studied by Johnson, 2 and his results are expressed in Fig. 54.
The initial potential is 1'4 volts positive to hydrogen in the same
solution (this is probably too high), but very rapidly falls to 0'6 volt.
At this value it keeps constant for some time, then falls quickly to
0-4 volt, and remains nearly constant during the greater part of the
discharge. Towards the end there is another rapid drop in potential,
and the point of reversible hydrogen evolution is passed. But no
Zctoch. EUktrochem. 4, 308 f
Tran*. Amrr. Klrrlrnrhnn. Nor. 1, 187 (W<K). The 'charge' curve refers to
the attempted use of the copper oxide j>lai< in MOODcUzy oeHs, S p. 221.
hydrogen is evolved until a potential corresponding to an overvoltage
of about 0'3 volt l is reached. During the first horizontal stage of the
discharge, CuO is reduced to Cu 2 0. Before all the CuO has disappeared,
however, polarisation has lowered the potential to the point at which
Cu 2 is reduced to copper. Along the main horizontal part of the
curve this is the chief process, though the further reduction of CuO to
Cu 2 occurs simultaneously. 2 When the Cu 2 in the plate approaches
C 40 80 120 160 200 240 280 320 360
Time in Hours.
FIG. 54. Discharge Curve of Lalande Cell Positive^Electrode.
exhaustion, the potential rapidly falls to the value necessary for
hydrogen evolution. During this stage Johnson noticed that the small
quantities of Cu 2 left suffered further reduction to green copper
quadrantoxide Cu 4 0. The potentials of the two stages of the
discharge (0'6 and O'l volt respectively, referred to a hydrogen electrode
in the same solution) lie somewhat below the equilibrium values. These
are 3 0'75 O66 volt for the CuO Cu 2 mixture, depending on the
method of preparation of the CuO, and O47 volt for the Cu 2 Cu
electrode. The discrepancies up to 0'15 volt are due to concen-
tration polarisation, as the amounts of the oxides dissolved in the
electrolyte at any moment are exceedingly small.
Johnson found for the system Cu [ CuO KOH Pt | H 2 the E.M.F.
0-73 volt. This value is probably too high, owing to the presence of
dissolved oxygen. The most probable figure is -f- 0'68 volt referred
to hydrogen in the same solution. If we subtract from this 0*41
volt, the potential of the zinc anode, we get 1'09 volt as the E.M.F. of
the cell, which is the value actually obtained. Its working voltage
1 Ci. values for copper on pp. 118-119.
- The Lalande cell cannot be strictly regarded as a form of the Daniell cell,
as is sometimes done. The process Cu * > Cu does not take place, but instead
we have the two successive reactions Cu" > Cu* and Cu* > Cu.
3 Allmand, Trans. Chem. Soc. 95, 2151 (1909) ; 97, 603 (1910).
206 PRINCIPLES OF APPLIED ELECTROCHEMISTRY [CHAP.
during the first stages is somewhat lower, and during the chief part
of the discharge is made up of the difference of the potentials of the
CuO plate + 0*4 volt and the zinc plate 0*4 volt, i.e. 0'8 volt. Lorenz x
has shown that the zinc electrode behaves nearly reversibly except at
very high current densities. Finally the initial high voltage stages of
the discharge are due to depolarisation by air dissolved in the elec-
trolyte and adsorbed on the positive plate. Its presence keeps the
Cu' concentration low and the potential high. The highest possible
initial voltage thus obtainable would correspond to the air potential
(-}- 1*22 volts with respect to hydrogen in the same solution) and
would be about 1'6 volts. In practice figures above T2 volts are never
Leclanche' Cell. Like the Lalande element, the Leclanche cell 2
uses a zinc anode and a solid cathodic depolariser, but the efficiencies
and capabilities of the two cells are very different. The electrolyte
in the Leclanche is a strong NH 4 C1 solution (perhaps 20 per cent.).
Various hygroscopic substances, such as glycerine, zinc chloride or
calcium chloride, may also be added, thus lessening the tendency of the
cell to lose water on standing. The depolariser is Mn0 2 , which becomes
reduced to manganese sesquioxide, OH' ions being liberated. We can
suppose the cathodic processes to be successively
Mn0 2 + 2H 2 Mn"" + 40H'
Mn"" * Mn"* -f
Mn- + 30H' > JMn 2 3 + fH 2 0.
The total result is
Mn0 2 + JH 2 > JMnA + OH' + .
The exact form in which the Mn 2 3 is precipitated is not known. This
depolariser has several disadvantages. It does not react as rapidly as
the CuO plate of the Lalande cell. One or other of the above reactions
takes place comparatively slowly, and, unless the current density
be kept low, the polarisation increases to such an extent that the
potential necessary for H* discharge is reached. The cathode potential 3
also becomes more negative owing to the formation of OH 7 ions. This
formation occurs in all the primary cells we have considered with the
exception of the Daniell cell. In elements with acid depolarisers it is
neutralised, and in the Lalande cell its effect is small in comparison with
1 Loc. cit.
Friedrich, EUktrochem. Zeitsch. 16, 219, 252, 287 (1W9-10).
3 In the present case
g = (E.P, + 0.058 .o
= constant 0'058 log [OH']
See p. 102.
xv.] PRIMARY CELLS 207
the OH' ions already present. But in the Leclanche element, with a
neutral electrolyte, the effect on the cathode potential can be marked.
On standing, the greater part of the alkali of course diffuses away.
Finally the specific resistance of the Mn0 2 is high. Not only does
this mean an increased cell resistance, but it renders it difficult to
make the mass so porous that the full capacity of the depolariser can
be utilised. It is always necessary to mix some highly-conducting
powdered graphite or carbon with the Mn0 2 on this account.
At the anode zinc dissolves, forming ZnCl 2 . The OH 7 ions produced
at the cathode give undissociated NH 4 OH with the NH 4 * ions, which
in its turn furnishes ammonia gas. Some of this escapes into the atmo-
sphere. But the greater part remains dissolved and sets itself into
equilibrium with the ZnCl 2 as follows :
ZnCl 2 + 2NH 3 ^ Zn(NH 3 ) 2 Cl 2 .
This last salt, the chloride of a complex zinc-ammonium cation, is
sparingly soluble and gradually separates out. The equation expressing
the main reaction in a Leclanche cell is therefore Zn -f- 2NH4C1 -f- 2
Mn0 2 > Mn 2 3 -f Zn(NH 3 ) 2 Cl 2 -f H 2 0. Whether water actually
separates or not depends on the form in which the Mn 2 3 is precipitated.
Besides Zn(NH 3 ) 2 Cl 2 , basic insoluble zinc ammonium salts can be
formed. Another effect of the cathodic production of OH" ions is to
precipitate zinc hydroxide where their concentration is greatest, i.e.
in the porous Mn0 2 mass. This precipitation naturally still further
raises th^ resistance of the depolariser. The conductivity of the
cathodic mass of a-n exhausted element is exceedingly low, due to this
cause and to the production of the manganese sesquioxide.
There is one further complication. In a working Leclanche cell it
is usually observed that the upper end of the zinc electrode is far more
attacked than the lower end. The cause is as follows. The ZnCl 2 solu-
tion formed anodically is heavier than the rest of the electrolyte, and
falls to the bottom of the vessel. We have then a zinc rod, its two ends
dipping into solutions of Zn01 2 of different strengths, in other words a
concentration cell. 1 This cell will furnish a current that will tend to
neutralise the concentration difference. Zinc is deposited at the
bottom end of the rod and dissolves at the upper end, whilst positive
electricity flows along the rod from bottom to top. The result is the
The actual construction of the Leclanche cell shows many variations,
designed to increase capacity and lessen resistance. A common type
is a glass vessel containing the NKtCl solution. The zinc anode is an
amalgamated rod, whilst the cathode is a carbon rod surrounded by a
cylindrical mass of the depolariser. This consists of powdered pyrolusite
intimately mixed with excess of graphite or carbon. Various additions
1 P. 103.
208 PRINCIPLES OF APPLIED ELECTROCHEMISTRY [CHAP.
are sometimes made, such as MgO, CaO, Fe 2 3 , but their function
is doubtful. The mass is held together by a suitable cement. In other
types the depolarising mixture is packed round the carbon into a
The voltage of the Leclanche on open circuit is T4-1-65, but, owing
to polarisation, drops to ri-1'2 volts when furnishing even small
currents (0'1-0'2 amperes for an ordinary cell). The value 1-4 volts
corresponds to the Mn0 2 -Mn 2 3 electrode ; the higher figure, 1-65
volts, is due to depolarisation by air dissolved and adsorbed in the
active mass, and is only given for a very short time by fresh cells. The
air potential is about + O81 volt, and the potential of the zinc elec-
trode at the prevailing low Zn" concentration can be taken as 0*84
volt. The current furnished depends on the active area of the depo-
lariser. Generally (to avoid hydrogen evolution) it should not exceed
0*1 amps, /d.m. 2 , i.e. O'l 0*2 amps, for an average cell. The resistance
varies enormously with the construction (0*05-10 ohms), but averages
0*4-2 ohms. Despite its many disadvantages, the Leclanche is
widely used for purposes requiring small intermittent currents (bells,
telephones, etc.). The reasons are that it is readily set up, requires
but little attention, and is cheap. For larger continuous currents it is
Dry Cells. These cells 1 are Leclanche elements in which the
electrolyte is contained in some porous material with which the cell is
filled. Such cells can be placed in any position without losing liquid,
often a great convenience. Millions are made yearly, all of small
capacity, and used for small handlamps, telephones, door bells and
motor ignition. The outer containing vessel can be of zinc and form
the anode, or of impregnated cardboard, enamelled iron, celluloid, etc.
The zinc is best amalgamated, taking the form of a rod or a cylindrical
piece of foil. The carbon rod (cathode) is surrounded with the depo-
lariser, a mixture of graphite or carbon powder and Mn0 2 , the latter as
pure as possible. The electrodes are placed closely together.
As porous material are used paper-pulp, sawdust, cotton-wool,
cocoanut charcoal, clay, infusorial earth, etc. Gypsum, magnesia, and
ZnO have the disadvantage of setting into a solid mass under the
action of the electrolyte and increasing the cell resistance. The electro-
lyte contains usually NH 4 C1 (25 per cent.) and ZnCl 2 . The latter
decreases the local action at the zinc electrode and also the polarisa-
tion. As in the Leclanche, hygroscopic substances such as CaCl 2 and
glycerine are also added, it being essential that the cell does not
become too dry.
The cells being intended for intermittent use, and having to stand
idle for long periods, great care must be taken in their construction
Trans. Amer. Electrochem. Soc. 16, 97 (1M9) ; 17, 341 (1910).
xv.j PRIMARY CELLS 209
to avoid defects that may lead to local action, 1 and the materials used
must be very pure. A trace of copper is fatal. The cells are usually
closed up with hard pitch, an opening being left for the escape of the
ammonia when working.
On open circuit the voltage of a dry cell is equal to that of a Le-
clanche, perhaps 1*6 volts. But if discharged at an average rate only
about one volt is obtained. The capacity depends largely on the rate of
discharge. A normal cell under normal conditions will furnish perhaps
30 ampere-hours before its voltage falls to 0*5 volt. The resistance
increases continually during working, the evaporation of water and
the deposition of basic zinc compounds being responsible.
3. Fuel Cells 2
The Problem. When mechanical energy (or electrical energy) is
produced from fuel by means of boiler and steam engine (and dynamo),
only some 10 per cent, to 15 per cent, of the total amount of energy
liberated by the complete combustion of the fuel is obtained, 3 and with a
gas engine some 25 per cent. One of the most important technical
problems is a better utilisation of this available chemical energy of
oxidation of fuel, and many attempts have been made to convert it
directly into electrical energy without the intermediate production of
heat, i.e. to bring about the combination of the fuel and atmospheric
oxygen electrochemically. If it were possible to do this, and reversibly,
not merely a small fraction of the free energy of combustion of the
fuel would become available as mechanical energy, but the whole of it,
and this naturally would mean an industrial revolution only comparable
with the first introduction of steam engines. Besides the enormous
drop in the cost of power, resulting in a great extension of its use and in
the unquestioned supremacy of coal as against water-power (this fact
would be of particular economic importance to industrial electro-
chemistry), a successful fuel cell would mean a great mitigation of the
smoke nuisance and vast changes in the engineering and chemical
industries. Obviously all schemes so far proposed have been fruitless,
but in view of the extreme importance of the subject we must consider
it at some length here.
The chief constituent of our fuels is carbon. If burnt to C0 2 at
room temperature, one gram-atom (12 grams) liberates 97,650 cals*
This is U, 4 the decrease of total energy. A, the decrease of free energy,
the measure of the maximum amount of useful work obtainable from
1 In some types the water necessary is only added immediately before use.
Such cells can be stored indefinitely till required without fear of deterioration.
- Ostwald, Elektrotech. Zeitsch. 15, 329 (1894). Haber, Grundriss der Tech-
nischen Elektrochemie, p. 178 (1898). Zeitsch. Elektrochem. 11, 264 (1905).
3 P. 10. < P. 78.
210 PRINCIPLES OF APPLIED ELECTROCHEMISTRY [CHAP.
the combustion, is rather less, 96,635 cals. Carbon is tetravalent, and
its electrochemical combustion would necessitate the passage of four
faradays per gram-atom. If E is the voltage of a C 2 cell at room
temperature, we have then
E X 96540 X 4 = 4-19 X 96635
E = 1-05 volts.
If air be used instead of oxygen, the E.M.F. will be about O'Ol volt
lower, i.e. 1'04 volt, and the reversible combustion of one kilo, of carbon
in this way will furnish about 8,940 ampere-hours at this pressure.
The coulombs are high and the voltage low.
The Difficulties. To effect the electrochemical combination of
carbon and oxygen, the obvious procedure is to construct a cell with a
carbon anode, an oxygen cathode, and a suitable electrolyte. Oxygen
or oxygen-containing substances and carbon must of course be kept
out of contact with one another. The earlier cells (Becquerel ; Jabloch-
koff) erred in this way, other considerations apart. But we at once
encounter two main difficulties. It is essential that the carbon should
ionise (not necessarily as carbon ions, for ions containing carbon and
some constituent of the electrolyte would suffice), otherwise the electro-
chemical solution of the carbon anode cannot take place. Now, carbon
stands in the middle of the first horizontal series of the Periodic System,
and one characteristic of these horizontal series is that the end members
show very marked electrochemical properties, and have high electrolytic
solution pressures, whilst this tendency steadily diminishes towards the
middle of the series. Carbon is almost electrically neutral, and has
resisted all attempts to make it ionise, whatever the electrolyte.
The second difficulty is that the forms of carbon which constitute our
fuels * are by no means pure. The presence of inorganic ash constituents
apart, the carbon itself does not really occur as elementary carbon,
but rather as highly complex substances, containing also hydrogen,
oxygen, and nitrogen. Considering that carbon itself does not ionise,
it is difficult to imagine what ions these carbon-rich complexes could
give. The anodic impurities would also mean a foul electrolyte, and
we have already seen how essentially important it is for an electrolytic
process that this fouling be minimised.
The recognition of these facts has led later experimenters to employ
a different principle. This consists in first allowing the carbon to
react chemically with other substances, forming an electromotively
active product, and using this product in the primary cell. The final
result is that the fuel is burnt to C0 2 (also water, etc.). For example,
one could imagine the carbon to be used to produce zinc from ZnO,
giving at the same time C0 2 , and the zinc subsequently electrochemically
oxidised to ZnO (assuming these reactions to be possible). Or Fe 2 8
1 These, and not electrode carbons, would form the active anodes in a fuel cell !
xv.] PRIMARY CELLS 211
could be reduced by carbon to iron and afterwards electrochemically
regenerated from the iron and air. The equations would be
(a) 2ZnO + C > C0 2 + 2Zn.
2Zn + 2 - >2ZnO.
(b) 2Fe 2 3 +3C - >4Fe+3C0 2
4Fe+30 2 - ^2Fe 2 3 .
The above examples, if feasible, would of course furnish a large
quantity of electrical energy during the second stage of the reaction,
but the first stage of the cycle is strongly endothermic, absorbing heat
and requiring a high temperature, and the wastage of fuel involved
would render the processes impracticable. To be successful, the first
(chemical) stage of the cycle must not take place with any great absorp-
tion of heat. If, on the contrary, it take place with evolution of heat,
the smaller this evolution, other things being equal, the better ; for
then a greater fraction of the original heat of combustion of the carbon
will be available for the production of electrical energy in the second
part of the process. This method of procedure frankly abandons the
attempt to obtain the whole of the available energy of combustion as