PRACTICAL
ELECTRO-CHEMISTRY
FIRST EDITION, JAN. 1901.
REPRINTED, AUG. 1903.
SECOND EDITION, JAN. 1906.
515" p
Practical
Electro-Chemistry
By
BERTRAM BLOUNT
/JJ
F.I.C., Assoc.Inst.C.E
CONSULTING CHEMIST TO THE CROWN AGENTS FOR THE COLONIES
FULLY ILLUSTRATED
SECOND EDITION REVISED
AND BROUGHT UP TO DATE
V. 4 .33,
LONDON
ARCHIBALD CONSTABLE y CO LTD
NEW YORK
THE MACMILLAN COMPANY
1906
BUTLER & TANNER.
THE SELWOOD PRINTING WORKS,
FROME, AND LONDON.
Preface
r^HE intention of this book is to give an account of those
electro-chemical processes which have been already
or are likely to be turned to industrial use. Historical
matter has generally been omitted for the sake of concise-
ness. For the same reason, comparison of electro-chemical
processes with chemical or metallurgical methods accom-
plishing the same results has been confined to the indication
of their relative advantages, a knowledge of the older pro-
cesses being assumed. The relation between the output of
a given process and the energy necessary for that output
has been dealt with somewhat fully, and in like manner the
practical advantages to be gained by the use of electro-
chemical methods in certain cases have been indicated.
I venture to hope that the book may be found useful by some
of those interested in one of the youngest and most promising
of modern industries.
B. B.
London, 1900.
Preface to Second Edition
r "T v HE first edition of this book has been received with
such discernment that I am encouraged to revise it.
Both here and in the United States it has been found accept-
able. Since its publication many new processes have been
devised, and various improvements of old methods have been
made. I have endeavoured to embody in the present
volume what is essential of these.
My acknowledgment and thanks are due and are sincerely
tendered to Dr. Moll wo Perkin for his kind and valuable
aid in the revision of the section on organic electro-chem-
istry.
B. B.
London, 1906.
VI
Table of Contents
SECTION I
INTRODUCTION GENERAL PRINCIPLES . . . 1-28
Definitions Nature of Electrolysis Constitution of
Electrolytes Theory of Solution Ionic Theory
Energy and Electrolytic Output Conversion
of Electrical Energy into Heat.
SECTION II
WINNING AND REFINING OF METALS BY ELECTROLYTIC
MEANS IN AQUEOUS SOLUTION .... 29-154
Electrolytic Refining and Winning of Copper, Lead,
Gold, Silver, Nickel, Cobalt, Tin, Antimony, Zinc.
SECTION III
WINNING AND REFINING METALS IN IGNEOUS SOLUTION 155-190
Aluminium Magnesium Sodium.
SECTION IV
WINNING AND REFINING METALS AND THEIR ALLOYS IN
THE ELECTRIC FURNACE CARBIDES, BORIDES, AND
SILICIDES 191-238
The Electric Furnace Moissan's Researches
Chromium Molybdenum Tungsten Calcium
Carbide Silicon Carbide Carbon Boride
Silicides.
vii
TABLE OF CONTENTS
PAGES
SECTION V
IRON AND STEEL 239-250
SECTION VI
ELECTRO-DEPOSITION ...... 251-286
Electrotyping Coppering Silver Plating Electro-
gilding Nickel Plating Electro-zincing
Aciertype Electro-deposition of Alloys.
SECTION VII
ALKALI, CHLORINE, AND THEIR PRODUCTS . . . 287-340
General Considerations Processes Using a Fused
Electrolyte Processes Using Dissolved Salt as
an Electrolyte Caustic Potash Chlorates
Hypochlorites Perchlorates.
SECTION VIII
ELECTROLYTIC MANUFACTURE OF ORGANIC COMPOUNDS
AND FINE CHEMICALS 341-358
Resolution of Organic Salts Electrolytic Oxidation
and Reduction Direct Production of Dye-stuffs,
Aniline, Vanillin, lodoform, Chloroform Puri-
fication of Sugar JuiceElectric Tanning.
SECTION IX
POWER 359-378
Efficiency of Existing Methods The Carbon Cell
The Gas Cell Water Power.
Vlll
List of Illustrations
FIG - PAGE
1 VAT FOR COPPER REFINING ..... 33
2 DITTO DITTO ...... 33
3 DITTO DITTO ...... 34
4 VATS FOR COPPER REFINING, SHOWING CIRCULATION . 43
5 DITTO DITTO ...... 43
6 DITTO DITTO ...... 44
7 DITTO DITTO ...... 45
8 VAT FOR COPPER REFINING, SHOWING SYPHON . . 46
9 DITTO, SHOWING CIRCULATING PIPES ... 46
10 DITTO DITTO .47
11 DOLPHIN METHOD ....... 49
12 COPPER ELECTRODES IN SERIES . . . . .51
13 "AciD EGG" . . 53
14 COPPER DEPOSITING, ARRANGEMENT OF CELLS . . 68
15 DITTO DITTO AUTOMATIC SYPHON . . 69
16 SIEMENS-HALSKE CELL FOR COPPER DEPOSITING . . 72
17 COHEN'S DITTO DITTO ...... 83
18 LEAD WINNING METHOD 85
19 TOMMASI'S CELL FOR LEAD REFINING ... 89
20 BORCHERS' APPARATUS FOR LEAD REFINING . . 93
21 CATHODE FOR SILVER REFINING . . . . .105
22 DITTO DITTO 105
23 ANODE DITTO . 106
24 MOEBIUS APPARATUS FOR SILVER REFINING . .107
ix
LIST OF ILLUSTRATIONS
FIG.
1'AGE
25 COWLES ZINC FURNACE . . . . . .133
26 MOND PROCESS . . . . 148
27 BORCHERS' APPARATUS FOR ELECTROLYSIS OF ZINC
CHLORIDE . . . 149
28 PH(ENIX PROCESS . .150
29 HEROULT' s APPARATUS FOR REDUCTION OF ALUMINIUM 160
30 APPARATUS FOR REDUCTION OF ALUMINIUM . .162
31 HEROULT PROCESS ,.,..( . . . . . 163
32 HALL'S APPARATUS FOR DITTO . . . . .168
33 GRAETZEL'S APPARATUS FOR REDUCTION OF MAGNESIUM 181
34 CASTNER'S APPARATUS FOR REDUCTION OF SODIUM . 187
35 ELECTRIC FURNACE . . . . . . . 194
36 COWLES' ELECTRIC FURNACE . . . . .195
37 MOISSAN'S DITTO . . 198
38 WILLSON'S CARBIDE FURNACE ..... 210
39 DITTO DITTO . ... . . 211
40 DITTO DITTO . . . . . . .213
41 DITTO DITTO . . . . . 214
42A KING'S DITTO 215
42s DITTO DITTO 215
43 HORRY FURNACE .. . . . . . . 216
44 FURNACE FOR CALCIUM CARBIDE . . . .218
45 DITTO DITTO 219
46 FURNACE FOR CARBORUNDUM . . . . . 227
47 DIAGRAM OF DITTO . . . . . . . 228
48 KELLER FURNACE WITH FOUR HEARTHS . . . 243
49 HEROULT STEEL FURNACE . . . . . 244
50 GENERAL VIEW OF HEROULT FURNACE . . . 245
51 KJETLLIN FURNACE . . . ... .. . 247
52 GIN ELECTRIC FURNACE . . . . . . 249
53 DITTO DITTO . . . . ... 249
53A TRAY ILLUSTRATING CORROSION OF GALVANISED IRON . 276
53B DITTO DITTO . . . **' . . . 277
54 CELL FOR DECOMPOSITION OF FUSED SALT . 292
LIST OF ILLUSTRATIONS
FIG. PAGE
55 VAUTIN'S APPARATUS FOB ELECTROLYSING OF FUSED
SALT 294
56 HULIN'S APPARATUS FOR ELECTROLYSING OF FUSED SALT 296
57 ACKER'S PROCESS ....... 298
58 BORCHERS' DITTO ....... 300
59A ELECTRO-CHEMICAL Co.'s APPARATUS FOR PRODUCTION OF
ALKALI AND CHLORINE 302
302
303
303
305
60A HARGREAVES-BIRD CELL 306
60s DITTO DITTO 306
61 DIAGRAM OF DITTO ....... 307
62 CASTNER-KELLNER PROCESS . . . . .313
63 DITTO DITTO 319
64 LE SUEUR PROCESS ....... 320
65 OUTHENIN CHALANDRE PROCESS 322
66 SIEMENS PROCESS ....... 330
67 SCHUCKERT PROCESS 331
68 NATIONAL ELECTROLYTIC Co.'s CELL . . . .337
69 JACQUES CARBON CELL 366
59s
DITTO
DITTO .
59c
DITTO
DITTO .
59o
DITTO
DITTO
59E
DITTO
DITTO
XI
SECTION I
Introduction General Principles
SECTION I. INTRODUCTORY
The Principles of Electro-Chemistry
ALL electro-chemical operations are performed either
by the analytical property of electrical energy when
passed through an electrolyte or by the heat which is pro-
duced when a current of electricity is passed through a
conductor which is not an electrolyte. Numerous applica-
tions of both have been made, arid the principles involved
in these applications must be understood before the applica-
tions themselves can be considered intelligently.
ELECTROLYSIS
IT must be assumed that the reader is familiar with the
general principles of chemistry and electricity. This being
granted, it is necessary merely to state the meaning of certain
special terms in order to make possible the communication
of an intelligent idea of the nature ot electrolysis.
Electrolysis itself may be defined as the course of chemical
changes induced by the passage of a current of electricity
through a chemical compound in solution or in the fused
state. 1
An electrolyte is a compound substance capable of under-
going resolution into its constituent elements or radicles by
the passage of a current of electricity.
1 Resolution of a compound by the mere heating effect of the
current is not contemplated in this definition.
3
PRACTICAL ELECTRO-CHEMISTRY
An electrode is a conductor of the metallic class serving
to convey a current of electricity into or out of an electrolyte.
The electrode by which the current enters the electrolyte
is called the anode ; that by which the current leaves the
electrolyte is called the cathode.
Those constituent elements or radicles of an electrolyte
which are believed to be the material carriers of a current
of electricity through an electrolyte are called ions.
Ions which appear as such or as their products at the
anode are termed anions ; those appearing at the cathode
are called cations.
The ions appearing at the anode are negative to those
appearing at the cathode ; thus, in general, metals or their
oxides or hydroxides appear at the cathode, and non-
inetallic elements or their oxides or hydroxides (bodies of
the class of acids) appear at the anode.
It is useful to remember that, just as one generates hydro-
gen by the action of zinc on a dilute acid, so when a dilute
acid is electrolysed hydrogen is evolved at the electrode
connected with the zinc plate of the battery. Seeing that
hydrogen stands to the other constituents of the acid in the
relation of a metal, and is thus the positive element, it is
clear that the electrode to which it is attracted must be the
negative, i.e. the hydrogen or its equivalent metal appears
at the negative electrode or cathode. Such notions, based
on elementary chemical facts, make it easy, when the learner
is confronted by two poles labelled + and , to couple them
aright to the apparatus he intends to use. Having thus
cleared the ground, we may return to the consideration of
the nature of electrolysis.
Many substances, notably metallic salts, when fused or
dissolved by a suitable solvent (most commonly water),
suffer chemical change when a current of electricity is
passed through them. Thus, when zinc chloride is fused,
and two platinum plates (electrodes) are immersed therein,
one being connected with one pole and one with the other
of a sufficiently powerful source of electricity, a current
passes through the liquid zinc chloride, and that body is
4
INTRODUCTORY
separated into its constituents, zinc appearing at the negative
electrode, which is called the cathode, and chlorine at the
positive electrode, which is called the anode. Similarly,
when zinc chloride is dissolved in water and electrolysed,
the same products appear at the same electrodes. In each
case the products appear at the surface of the electrodes,
and there is no indication of change in the liquid between
the two electrodes. But there is reason to believe that
many of the molecules of zinc chloride occupying the space
between the electrodes undergo change during the passage
of the current. It is supposed that each atom of chlorine
separated from a molecule of zinc chloride at the anode
is immediately replaced by another atom from an adjacent
molecule of zinc chloride, and that the atom of zinc thus
left, in its turn requires an atom of chlorine from its neigh-
bour ; this process continues until, at the end of a string
of molecules, an atom of zinc is left, robbed of its chlorine
and without an available neighbour to borrow from. This
atom of zinc appears as the free metal at the cathode.
This is a simple case of electrolysis in which the products
are the same from the fused salt as from the salt dissolved in
water, and where there is little tendency towards the for-
mation of products other than the two primary substances,
zinc and chlorine. It must not be supposed that the process
of electrolysis is always as simple as this. In many in-
stances the actual products are not those which would be
formed by the splitting of the salt into its metallic and non-
metallic constituent, but include substances formed by the
action of these primary constituents on the solvent. Thus,
when sodium chloride is fused and electrolysed, the products
are sodium and chlorine. When its aqueous solution is
electrolysed, the products are sodium hydroxide, hydro-
gen, and chlorine. Seeing that free sodium acts spontane-
ously on water, liberating hydrogen and forming sodium
hydroxide (caustic soda), it is convenient to regard the
process of electrolysis in this instance as separating sodium
from its union with chlorine, and the sodium thus liberated
in the midst of an ample supply of water molecules at once
5
PRACTICAL ELECTRO-CHEMISTRY
reacting with them in precisely the same manner as it does
when a piece of the metal is placed in contact with water.
But further complexity may arise in this seemingly simple
case of the electrolysis of an aqueous solution of sodium
chloride. It is true that the products appear at separate
electrodes, but as the process of electrolysis goes on the
products will encounter each other unless some means of
mechanical separation be devised. Considering the pro-
ducts one by one, it is clear that the greater part of the
chlorine will escape as gas, but that a small portion will
remain in solution. The hydrogen will escape as gas almost
entirely ; the whole of the caustic soda will remain in the
liquid. The portion of the chlorine which is dissolved will
eventually encounter the caustic soda and form sodium
hypochlorite and sodium chlorate, according to the tempera-
ture of the solution ; should it chance to meet the hydrogen
before the latter escapes from the cathode, hydrochloric acid
will be formed. This reacting with the caustic soda will
regenerate sodium chloride and water, or encountering
sodium hypochlorite will give sodium chloride and chlorine,
or meeting sodium chlorate will yield sodium chloride and
chloric acid, a body itself always on the verge of splitting
up. 1
All these reactions may be proceeding at once, according
to the local conditions of the liquid in contact with the
electrodes. Thus a mixture of considerable complexity
may result from a resolution which appears simple enough
at first sight.
Among all these various possibilities, there is one truth
which has the force of a canon. It is that the energy im-
pressed on the solution serving as an electrolyte can have as
its outcome only its strict equivalent in the substances which
1 The substances thus formed themselves act as electrolytes,
carrying part of the current and yielding characteristic products ;
thus caustic soda will yield oxygen water and sodium, the last-
named promptly reacting to liberate hydrogen and produce caustic
soda. The net result of this bye-reaction is the electrolysis of water
induced by the presence of primary product caustic soda.
6
INTRODUCTORY
are produced. There may be (and usually is) a waste in
transforming the electrical energy of the current used for
the decomposition into the chemical energy represented by
the products of the decomposition, but there can never be a
surplus.
Part of this truth is involved in the hypothesis known as
Faraday's law. By a number of well-conceived researches,
executed with that skill in experiment which was native in
this great chemist and physicist, Faraday established that,
for a large number of electrolytes which he examined, the
same current produced equivalent quantities of products at
the anode and the cathode. Now, as it is known that in a
given electrical circuit the current passing between any pair
of points is the same as that passing between any other pair
of points, it follows that, when any number of electrolytic
cells are coupled in series, the products separated at their
anodes and cathodes are in all cases equivalent. 1 Thus, a
current sent through a cell containing a solution of copper
sulphate, and then through one containing fused zinc
chloride, will liberate at the cathode of the first 63-5 grammes
of Cu for every 65 grammes of Zn liberated at the cathode
of the second. Corresponding with each of these quantities
there will be produced at the anodes 96 grammes of the
hypothetical radicle S0 4 and 71 grammes of chlorine re-
spectively.
The metals copper and zinc, being divalent, are set free
atom for atom in their respective electrolytes. The radicle
S0 4 , being also divalent, is strictly equivalent to one atom
of either of the metals. The element chlorine, being mono-
valent, is set free in quantity equivalent to each of the
others, that is, two atoms become free for one atom of zinc
or copper. The radicle S0 4 has no objective existence, but
what may be termed its natural products appear in strictly
equivalent amount, viz., and S0 3 , the latter of course
1 For the discussion of matters purely electrical the reader is
referred to any good text-book dealing with the branch of physics
known as electricity.
PRACTICAL ELECTRO-CHEMISTRY
combining with the water of the copper sulphate solution
to form H 2 SO 4 .
But though the quantities of the elements or radicles
liberated at the electrodes are all equivalent, yet the energy
expended in each cell is not necessarily equal to that ex-
pended in any other. The current flowing through each cell
is equal, but the fall of voltage from anode to cathode in
each cell will vary with the chemical energy represented by
the union of the anion and cation. It is convenient to
measure the energy evolved by the chemical union of two
bodies in thermal units. Thus the heat of combination of
23 grammes of sodium and 35-5 grammes of chlorine is 97-69
Cal. 1 To liberate 23 grammes of sodium and 35-5 grammes
of chlorine from 58-5 grammes of sodium chloride, 97-69
Cal. or its equivalent in electrical units of energy must be
expended. The electrical unit of energy or joule is 0-7375
foot-pound, or 0-00024 Cal. Therefore, for the decom-
position of 58-5 grammes of NaCl, assuming no waste to
occur, 407,042 joules must be expended. But a joule,
being a unit of energy, can be expressed as the product of
two values one of the nature of a quantity, the other of a
pressure. Thus 1 joule is the product of 1 coulomb and 1
volt. The coulomb, being the unit of the quantity of
electricity, involves in its flow the separation of a definite
and equivalent amount of any electrolyte. One coulomb
can decompose 0-0006045 gramme of NaCl, and 58-5
grammes of NaCl require 96,540 coulombs for their decom-
position. But in order that the passage of 96,540 coulombs
should represent the expenditure of 407,042 joules, they
407 042
must be delivered at a pressure of volts, i.e. at
96,540
4-22 volts.
It will be observed that here no mention is made of time ;
1 Throughout this book the unit of heat energy used is the kilo-
gramme-calorie (represented by Cal.) unless a direct statement to
the contrary is made. The kilogramme-calorie is the quantity of
heat needed to raise 1 kilogramme of water from" 15 C. to 16 C.
INTRODUCTORY
the work may be done in any time provided the requisite
number of coulombs are caused to flow at a pressure not
lower than 4-22 volts. For the purpose of this argument
the figure which is usually accepted for the heat of combina-
tion of sodium and chlorine has been taken ; seeing, however,
that the electrolysis of sodium chloride into its cation sodium
and its anion chlorine cannot be effected when the salt is in
the solid state (because it is then almost non-conducting),
but is carried out with the salt in a state of igneous fusion,
a condition which it attains at a moderate red heat, it is
certain that this value is too high, for at this working tem-
perature sodium chloride is already approaching its tempera-
ture of dissociation, i.e. its constituent atoms are less firmly
united than they are at the ordinary temperature, and, there-
fore, the total energy needed to dissever them is smaller
than that which would be requisite at the ordinary tempera-
ture. In a word, part of the work of disconnecting sodium
and chlorine has been performed by the heat needed to fuse
sodium chloride, and the electrical energy which has now to
be impressed on it is correspondingly smaller in amount.
Now, by Faraday's law, each unit of electrical quantity
liberates one equivalent of sodium and one of chlorine, and
thus the number of coulombs necessary for decomposing
58-5 grammes of sodium chloride at a red heat is the same as
that which would be needed at the ordinary temperature ;
therefore, the factor which suffers change is the electrical
pressure or voltage. Thus in this case the minimum voltage
necessary to decompose fused sodium chloride at a red heat
is smaller than 4-2 volts. Its precise value has not been
determined (see p. 18).
In the foregoing argument all consideration of the possi-
bility of a portion of the heat energy impressed on the fused
sodium chloride being converted into electrical energy and
thereby providing a voltage auxiliary or opposed to the
voltage of the external source of electrical energy has been
purposely omitted. The one fundamental fact to be thor-
oughly grasped is that the energy necessary to decompose
a given substance at a given temperature is a constant
9
PRACTICAL ELECTRO-CHEMISTRY
quantity, and that, if Faraday's law be true for that sub-
stance, there is a fixed minimum voltage necessary for its
decomposition. Any accurately made experiment which
shows that a given electrolyte can be decomposed by a volt-
age smaller than that calculated from the heat of formation
of the electrolyte at that temperature will invalidate Fara-
day's law.
There is no need to shrink from such an overthrow, but
the experiments needed to accomplish it must be less open
to criticism than any which have yet been published.
Perfect understanding of these principles is necessary for
intelligent study of any practical process of electrolysis.
The efficiency of a process is frequently stated in terms of
current alone, i.e. the efficiency is stated as the ratio which
the weight of product actually obtained bears to the weight
of product which should be obtained by the passage of
the number of coulombs known to have passed. But this
method of statement ignores the equally important factor
of voltage, i.e. it fails to take into account the pressure at
which the coulombs have been delivered. Therefore it is
necessary, in addition to giving the current efficiency for a
process, to give also the energy efficiency, i.e. the ratio of
the weight of the product actually obtained to the weight
which should be obtained by the theoretically perfect ex-
penditure of the total number of electrical energy units
(joules) which have passed through the cell. Thus a solution
of sodium chloride may be electrolysed with a voltage of
2-3 volts. In practice the voltage required is as high as 4
volts ; the current efficiency may be 90 per cent., but the
2-3
energy efficiency under these conditions is only 90 x -
4
-- 51 1 per cent.
10
INTRODUCTORY
THE CONSTITUTION OF ELECTROLYTES
AND THE MECHANISM OF ELECTROLYSIS
SOME aid to clear thought as to the way in which reactions
are brought about by electrolysis is afforded by considering
the ultimate structure of a typical electrolyte and the mole-
cular mechanism by which electrolysis is effected. A full
discussion of this subject is a proper matter for a text-book
on chemical physics, but certain of the more important
theoretical conceptions and their consequences may be set
down here.
NORMAL CONDITION OF A DISSOLVED NON-
ELECTROLYTE
This may first be considered as a simpler case, before
passing to the discussion of the condition of a dissolved
electrolyte. At the present time it is generally held that
the molecules of a substance, such as sugar, which is not
an electrolyte, are, when dissolved in a solvent capable of
no appreciable chemical action on the dissolved substance,
in a condition comparable with that of the molecules of a
substance existing as a gas. A solution of such a non-
electrolyte exercises a pressure proportional to the number
of molecules per unit volume occupied, thus behaving
precisely in the same manner as a gas.
This pressure, which is termed the osmotic pressure of
the dissolved substance ; is detected and measured by
a device which will be understood from the following
concrete case. Suppose the osmotic pressure of a sugar
solution is to be determined, a " semi-permeable " mem-
brane is prepared by depositing within the pores of an
ordinary porous pot a precipitate of cupric ferrocyanide, a
body which is found to allow the diffusion of water but not
of sugar. The formation of the semi-permeable membrane
ii
PRACTICAL ELECTRO-CHEMISTRY
is effected by filling the pot with a solution of potassium
ferrocyanide and surrounding it by one of copper sulphate.
The two liquids, meeting in the interstices of the pot, form
there a layer of cupric ferrocyanide which has the property
enunciated above.
After removing the pot from the liquids and washing out
traces of soluble salts the membrane is ready for use. The
pot is filled with sugar solution, the top is closed by a cork
carrying a manometer and the pot is then immersed in pure
water. 1 On each side of the semi-permeable membrane
in the pores of the pot molecules of water are constantly
impinging. Those impinging on the inside in contact with
the sugar solution are, however, fewer per unit of time than
those on the outside in contact with pure water, because a
certain part of the volume of the sugar solution is occupied
by sugar molecules instead of water molecules. Now those
molecules of water from the outside which do not collide
with water molecules on the inside pass through the mem-
brane ; no corresponding efflux of sugar molecules is possible
because the sugar molecules cannot pass through the semi-
permeable membrane. The influx of water molecules from
the outside goes on until those on the inside are sufficiently
crowded together to make the same number of impacts on
the inside of the membrane as do those on the outside.
That is, the pressure due to water molecules is equal on
each side of the membrane. But the pressure of the sugar
molecules on the inside is over and above this pressure of
the water molecules, and the total pressure of the sugar
solution is thereby increased. The amount of the increase
is indicated by the manometer. A simple calculation will
show the order of magnitude of such osmotic pressures. A
porous pot of a capacity of 100 c.c. is filled with sugar
solution containing gramme molecule of sugar per litre,
1 In laboratory practice the construction of an apparatus with a
strong and perfect membrane and absolutely tight closure and con-
nection with the manometer is very difficult, and indeed taxes the
best resources of the instrument maker.
12
INTRODUCTORY
i.e. 34'2 grammes per litre, or in the 100 c.c. 3' 42 grammes
of sugar. Now, if it were possible to gasify sugar by heat
without decomposition, 342 grammes of sugar-gas would
occupy a volume which, corrected to C. and 760 mm.
pressure,would equal 2 x 1T2 litres ; therefore, 3*42 grammes
would occupy '224 litre, i.e. 224 c.c. Regarding the dis-
solved sugar as being in the same condition as if it were
gasified, it is evident that the pressure above that of the
224
atmosphere which the sugar is capable of exerting is -
100
x 760 mm. = 1,702 mm. of mercury. Direct experiment
in the manner described above gives figures closely corre-
sponding with this. Additional evidence in favour of
the belief that a non-electrolyte dissolved in a neutral
solvent has its molecules in the same condition as those
of a gas is afforded by a variety of other chemico-physical
measurements.
Thus, when a solution of sugar in water is frozen, water
free from sugar is first separated as ice, and this at a lower
temperature than the freezing-point of pure water, viz.
C. Now as pure water (in the form of ice) is abstracted
from the solution the volume available for the molecules
of the dissolved substance is diminished ; if the molecules
of the dissolved substance be in the same condition as those
of a gas, the diminution of the volume which they occupy
(pressure being constant) can be effected by the abstraction
of a quantity of heat readily calculable. This quantity of
heat is found to be measured jointly by the lowering of
temperature needed to bring about the diminution of volume
and the latent heat of the solvent. The latter being known
(e.g. 80 Cal. for water), the former can be directly compared
with the lowering of temperature experimentally observed.
They are found to agree, and it may thus fairly be deduced
that the molecules of the dissolved substance are in the same
condition as those of a gas. Other means of judging the
condition of the molecules of a dissolved non-electrolyte in
a neutral solvent, such as the lowering of the vapour pressure
of a given solvent by the addition of a soluble substance,
13
PRACTICAL ELECTRO-CHEMISTRY
lead to the same result. Therefore it may be provisionally
accepted as consonant with experiment that the molecules
of a non-electrolyte dissolved in a solvent on which it does
not act chemically behave in many respects similarly to
the molecules of a gas. 1
CONDITION OF A DISSOLVED ELECTROLYTE
When the methods briefly described above, of examining
the condition of a substance which is not an electrolyte
dissolved in a solvent on which it does not act chemically,
are applied to the examination of solutions of electrolytes,
it is found that such solutions give indications of abnormal
behaviour. Electrolytes behave in manner similar to that
of a compound gas, the molecules of which are dissociated
into a larger number of simpler units. Thus a dilute solution
of NaCl in water behaves as if it contains nearly twice as
many ultimate particles as it does molecules. From this
it is assumed that most of the molecules NaCl must be split
up into their ions Na and Cl. 2 Evidence of complete
ionisation is forthcoming only when the solution of sodium
chloride is exceedingly dilute, e.g. contains y OTTO 77 f a
gramme molecule per litre, i.e. has a strength of 0-000585
per cent. Increasing dilution gives increasing ionisation,
and it is assumed that at infinite dilution ionisation would
be complete. Solutions of moderate strength, such as those
containing 1 gramme equivalent per litre (5-85 per cent.
NaCl), behave as if a portion of the molecules were ionised
and a portion were present as ordinary molecules.
This ionisation occurs with all electrolytes, and approaches
completeness more nearly with substances whose solutions
1 When a solution is somewhat strong, the molecules of the dis-
solved substance do not conform perfectly to gaseous laws. This
divergence is comparable with that of gases themselves when highly
compressed or near their liquefying point.
2 An ion is not necessarily an atom ; thus the ions of potassium
nitrate are K and NO 3 .
INTRODUCTORY
are good conductors than with those which are indifferent
conductors. Certain substances give evidence of being split
up into more than two ions. Thus the osmotic pressure,
depression of freezing-point, etc., of dilute barium chloride
solution point to the salt being split up into the three ions
Ba, Cl, Cl ; similarly, according to its dilution, sulphuric
acid may be split up into the two ions H and HS0 4 , or into
the three ions. H, H, and S0.j.
It is evident that if the ions Na and Cl exist free in a
solution of NaCl, they must be endowed with properties
very different from those of the elements sodium and
chlorine in ponderable masses as we know them. Certainly
a solution of sodium chloride gives no indication of containing
free chlorine, while free metallic sodium could not exist as
such for a moment in the presence of a large quantity of
water.
Still more conclusive is the consideration that the sever-
ing of NaCl into Naand Cl needs the expenditure of 97-69
Cal. per gramme molecule, and no such energy is impressed
on it by the mere act of dissolving the salt in water. There-
fore it is clearly impossible to regard the ions Na and Cl as
free sodium and chlorine in the ordinary sense. To meet
these objections to the belief in the existence of free ions, it
is assumed that each ion carries a charge of electricity, the
cations a charge of positive electricity and the anions one of
negative electricity, and that their properties are profoundly
modified by the possession of these charges, the total number
and amount of which are equal and opposite and counter-
balance each other, so that the solution as a whole gives no
indication of possessing any charge at all.
This conception is a mode of thought and not an objective
reality, and may eventually be replaced by an hypothesis
involving fewer and less sweeping assumptions.
Further, it is believed that these ions in solutions of
moderate concentration are at times free, and at times
united to form an ordinary molecule, and that they move
through the solution forming and breaking unions with ions
of the opposite kind. It is also considered that each kind
15
PEACTICAL ELECTRO-CHEMISTRY
of ion moves at its own pace, and that an ion may remain
free in its solution for an appreciable time.
The mechanism of electrolysis, according to this theory,
is as follows : On a current being passed through an elec-
trolyte, such as the aqueous solution of a metallic salt between
two unattackable electrodes, the cations carrying positive
charges flow to the cathode and there give up their charges,
becoming ipso facto ordinary molecules and appearing at
the surface of the cathode as metal, or as the products of
the action of this metal on water, viz. hydrogen and a
metallic hydroxide. Similarly, the anions carrying negative
charges flow to the anode and there give up their charges,
appearing as ordinary molecules of the same composition
as the ions themselves, or as the products of their action
on water. The function of the current is to neutralise
the charges thus given up at each electrode, and to allow
the ions to assume the ordinary molecular condition. The
conception of the existence of an ion as carrying a charge of
electricity, and of the transference of electricity through an
electrolyte being dependent on the flow of charged ions, has
been extended to form a theory of the primary cell. Thus, in
a Daniell cell, consisting of zinc in zinc sulphate and copper
in copper sulphate, it is considered that the zinc possesses a
" dissolution pressure" whereby its molecules tend to become
ions in the solution of zinc sulphate with which it is in con-
tact. In order to attain this ionised state an equal number
of ions already existing in the solution must be changed from
the ionised to the molecular state. Such a transformation
happens to the copper ions in the other compartment of the
cell, because the dissolution pressure of the zinc is greater
than that of the copper. The zinc ions require to be
positively charged, and equally the copper ions in the act
of becoming ordinary molecules give up their positive
charges, which are transmitted through the exterior cir-
cuit to the zinc plate. The difference between the dis-
solution pressures of copper and zinc is a measure of the
voltage of the cell.
In the foregoing sketch I have endeavoured to state
16
INTRODUCTORY
fairly and clearly the chief ideas embodied in the ionic
theory of electrolysis. The theory at present serves to
correlate facts rather than to explain the real mechanism
of electrolysis. As now expounded it is not completely
convincing, involving as it does a good many assumptions
not very probable nor even wholly consonant with experi-
ment. Fortunately all practical applications of electrolysis
can be satisfactorily considered without having recourse to
this hypothesis, and the practician who is equipped with a
sound knowledge of the principles of chemistry and electricity
can grapple successfully with any problem in electrolysis
which is likely to present itself, irrespective of the precise
explanation which may be at the moment most agreeable
with the teachings of the ionic hypothesis. 1
METHOD OF CALCULATING OUTPUT IN
ELECTROLYTIC PROCESSES
IN the foregoing sections sufficient has been said to give
some idea of the nature of electrolytic changes, the mech-
anism by which they are possibly brought about, and the
quantitative relations of the electrical energy used and the
products obtained. This last-named subject has only
been touched on lightly and incidentally, merely as far
as was necessary to illustrate the other two, and seeing
that it is all-important in practical work, a special section
may conveniently be devoted to its consideration, even
though it involve occasional repetition of what has already
been said.
The best way to understand the quantitative relations
of any process of electrolysis is to consider the process on
the basis of the amount of energy which it requires. To
bring about a given chemical change which is endothermic
the expenditure of a definite quantity of energy is requisite,
and the electrical energy supplied to cause this change by
1 A good deal of experimental work has been done tending to
show that gases may be ionised, but it has not yet been correlated
with the electrolysis of liquids.
17 c
PRACTICAL ELECTRO-CHEMISTRY
electrolysis must be equal to or greater than this quantity.
It matters not what cunning arrangement for conducting
the electrolysis may be devised, this fundamental law
cannot be circumvented.
Thus, if a process be schemed for the electrolytic de-
composition of sodium chloride into sodium and chlorine,
the amount of electrical energy which will be needed for
the decomposition of 1 gramme molecule (58' 5 grammes)
is not smaller than that equivalent to 97 '69 CaL, assuming
this to be the heat of combination of sodium and chlorine.
The fact that all electrolytic processes for the direct de-
composition of sodium chloride require the salt to be fused,
and are therefore carried out at a red heat, in no way in-
validates the general truth of this statement. At a red
heat the heat of combination of sodium and chlorine is not
97'69 CaL, but a smaller value, e.g. 88'21 Cal. 1 Accepting
this value, it is certain that a quantity of electrical energy
not smaller than 88'21 Cal., when translated from electrical
into heat units, must be impressed on the salt kept just
at its fusing-point by extraneous heat. The quantity
needed may be larger than this for a number of reasons,
which will become evident when this particular electrolytic
process and others cognate with it are discussed in their
proper place, but it will certainly not be smaller unless the
sensible heat supplied from without is capable of conver-
sion in the decomposing cell into electrical energy, and
of this we have no evidence. Now the unit of electrical
energy is the joule, and is equal to 0'00024 CaL, i.e. 0*00024
kilo of water raised from 15 C. to 16 C. 2 Therefore the
1 This figure may be approximately arrived at by deducting from
the heat of combination the quantity of heat needed to raise 58-5
grammes of salt from 15 C. to its melting-point, 772 C., taking
the specific heat of salt as 0-214, and neglecting its latent heat of
fusion, which is probably small.
2 Formerly the calorie was taken to be the quantity of heat
needed to raise the temperature of 1 kilo of water from C. to 1 C.,
but of late years it has been found convenient to choose a somewhat
higher temperature, because the specific heat of water exhibits cer-
tain irregularities near its point of maximum density (4 C.) and its
18
INTRODUCTORY
decomposition of 58' 5 grammes of sodium chloride at its
fusing-point into sodium and chlorine requires 367,542
joules. But according to Faraday's law (see p. 7) the
isolation by electrolysis of a chemical equivalent of any
substance requires the passage of the same number of
units of current. For ] gramme equivalent of any sub-
stance this number of units of current is 96,540 coulombs.
Now 1 unit of electrical energy may be expressed as the
product of 1 unit of electrical quantity x by 1 unit of
electrical pressure, i.e. 1 joule = 1 coulomb x 1 volt.
In order to represent 367,542 joules, 96,540 coulombs
must be delivered at a certain electrical pressure,
i.e. 367,542 = 96,540 x x volts.
x = 3-807 volts.
This means that, accepting the heat of formation of
sodium chloride at its fusing-point as 88-21 Cal. and as-
suming the truth of Faraday's law, the minimum voltage
necessary to effect the electrolytic decomposition of fused
sodium chloride into sodium and chlorine is 3,- 807 volts.
It means neither more nor less than this. It does not
mean that sodium chloride at any temperature requires
this voltage, and it does mean that no smaller voltage
will decompose sodium chloride under the conditions given.
Such steadfast data are continually available in electrolytic
work, and in those cases where they appear not available
it is our knowledge either of the precise heat of combination
of the constituents of the electrolyte at the temperature
chosen or of the precise products obtained by electrolysis
which is at fault, and not the truth of the doctrine of the
conservation of energy or of Faraday's law.
Thus in practice he who is firmly grounded in these
primary principles can deal with each particular case as
it arises, not experimenting blindly, but with certain definite
and exact generalisations to guide him.
freezing-point (0 C.), and first becomes approximately constant for
a considerable range of temperature at 15 C. The unit, like most
others, is essentially arbitrary, and the precise value chosen is merely
a matter of convenience and convention.
19
PRACTICAL ELECTRO-CHEMISTRY
It is evident from this that the output of any given sub-
stance for a given current can be calculated from the single
datum, One chemical equivalent of any electrolyte expressed
in grammes requires the passage of 96,540 coulombs for its
liberation or decomposition, and that the critical pressure
for the decomposition of any given electrolyte can be cal-
culated from the single datum, The energy of combination
expressed in joules of 1 gramme equivalent of any electrolyte,
divided by 96,540 coulombs, equals the minimum pressure
in volts necessary to bring about the electrolysis.
It is unnecessary to give an elaborate table of the chem-
ical equivalents of a long list of substances, together with
their calculated output per coulomb or per ampere second
(the same thing as a coulomb) or their ampere hour (a con-
venient commercial unit). It will suffice to set down a
few, both to give some idea of the order of magnitude of
the quantities obtained and for convenience of reference
to those numbers which are constantly occurring in electro-
chemical work.
>>
Output per
coulomb,
Output per
3,600
Element.
Atomic
weight.
o
a
<D
13
K^.
Equivalent
weight.
i.e. per
ampere
second.
coulombs,
i.e. per
ampere
Milli-
hour.
grammes.
Grammes.
. .
Aluminium (Al) .
Chlorine (Cl) . . .
Copper (Cu) (Cupric)
(Cuprous)
Hydrogen (H) . .
Iron (Fe) (Ferrous) .
Lead (Pb) (as cation)
(as anion)
Nickel (Ni)
Oxgyen (O)
Silver (Ag)
Sodium (Na)
Zinc (Zn)
,
27
35-5
63-5
63-5
1
56
207
207
58-5
16
108
23
65
3
1
2
1
1
2
2
4
2
2
1
1
2
35-5
31-75
63-5
1
28
103-5
51-75
29-25
8
108
23
32-5
0-093
0-368
0-329
0-658
0-0103584
0-290
1-072
0-536
0-303
0-083
1-119
0-238
0-337
0-3356
1-3237
1-1837
2-3674
0-03729
1-0441
3-8595
1-9297
1-0907
0-2983
4-0273
0-8577
1-2119
20
INTRODUCTORY
THE CONVERSION OF
ELECTRICALENERGY INTO HEAT FOR
ELECTRO-CHEMICAL PROCESSES
UP to this point the electro-chemical principles which have
been discussed are those which relate to electrolysis. There
is another method of applying electrical energy to chemical
and metallurgical processes, and that is by direct conversion
into heat. In an electrolytic operation every unit of heat
appearing means waste, for in one working ideally all energy
impressed on the electrolyte should appear in the form of
chemical energy of the products of electrolysis. Where
heating alone is to be accomplished any electrolytic decom-
position means waste ; a complete conversion of electrical
energy into heat in the vessel in which the process is to be
carried out should be achieved. There are, however, certain
electro-chemical processes in which electrical energy is used
both for heating and for effecting electrolytic resolution ;
the most noteworthy instance is in the manufacture of
aluminium (q.v.) by the electrolysis of alumina dissolved
in a double fluoride of aluminium and sodium. The bath
is not only decomposed electrolytically, but is also kept
fused by heat obtained at the expense of electrical energy
passing between the electrodes. The principles which have
been laid down for electrolysis without heating and those
about to be enunciated for heating without electrolysis may
be applied to these mixed cases ; any special considerations
for particular instances will be discussed under the individual
heads of processes of this type.
The principles of electrical heating as far as they concern
the electro-chemist may be very briefly dealt with.
The passage of a current of electricity through a con-
ductor of the metallic class is always attended by the pro-
duction of heat. When the conductor is of large section
21
PRACTICAL ELECTRO-CHEMISTRY
nd of a material which conducts well the quantity _ of
hea* produced is small compared with the total quantity
of energy transmitted; when the conductor is of small
section and conducts badly, the quantity of heat produced
^relatively great. These qualitative statements are made
by saying that the heat produced by the passage
of electrical energy through a metallic conductor is measured
by the drop of voltage from one end of the conductor to
the other, multiplied by the total current passing Thus
if a current of 1,000 amperes be passed through a conductor,
and there is a drop of voltage from one end of the conductor
to the other of 50 volts, energy is being expended at the
rate of 50,000 watts, i.e. 50,000 joules per second. There-
fore in one second the number of units of heat generated m
that conductor will be 50,000 x -00024 Cal. - 12 Cal.
If the conductor be of small mass, and if the specific heat
of the conductor be also small, the temperature of the con-
ductor will be greatly raised. Further, if the conductor
be enveloped in a sheath which will not conduct heat, every
unit of heat generated is retained, and the temperature of
the conductor is a direct measure of the energy expended.
Under theoretical conditions, therefore, any conductor
contained in a sheath perfectly non-conducting for heat and
conveying any current however small, will ultimately attain
an infinite temperature. In practice this goal is approached
by encasing the conductor in a mass of material as nearly
non-conducting for heat as possible, and passing through
it so large a current as to generate heat more rapidly than
the heat can be conveyed away until the temperature (heat
potential) becomes extremely high.
These requirements are fulfilled by the apparatus known
as the electric furnace. Two forms are commonly used.
In one an arc is produced in the midst of the mass to be
heated, and in the other a current is passed through a con-
tinuous core of refractory material (usually a thin rod of
carbon) which has a small section, a small heat capacity,
and a high electrical resistance. When a large current is
passed through such a core, its temperature rises until it
INTRODUCTORY
becomes so high that losses by radiation and conduction
balance the energy impressed upon it. But besides this
limiting condition there may be in electric furnace opera-
tions a direct demand on the electrical energy supplied in
the shape of energy needed to effect some chemical change in
the mass to be heated. Thus if a mixture of lime and coke
is to be converted into calcium carbide (q.v.), energy is
required to bring about the formation of calcium carbide
irrespective of that necessary to heat the charge. Therefore
it is clear that every electro-chemical furnace process must
be considered individually, and that no general principles
can be usefully laid down other than that already enunciated,
viz. that the total heating effect of a current passing through
a given conductor is measured by the current and the drop
of voltage between the ends of the conductor.
RELATIVE VALUE OF ELECTRO-CHEMICAL
PROCESSES AND PURELY CHEMICAL
PROCESSES'
FROM the point of view of the practical worker, choice
between an electro-chemical process and one which does
not depend on the application of electrical energy from
without depends solely on the relative monetary advan-
tages, of the two methods.
In almost all cases a given product which can be obtained
by electrolytic means can be prepared equally well by
purely chemical methods.
For example, pure copper can be prepared by precipi-
tation of cuprous oxide with glucose in alkaline solution
and reduction of the cuprous oxide in hydrogen, as well as
by the electrolysis of copper sulphate. Pure zinc can be
prepared by fractional distillation of the commercial metal
in vacuo at least as well as it can be obtained by any electro-
lytic process.
23
PRACTICAL ELECTRO-CHEMISTRY
Chlorine and caustic soda have been made for nearly a
century successfully, on a large scale and of excellent
quality, without having recourse to electrical processes.
Sodium, although originally prepared by the electrolysis
of caustic soda, has been manufactured for more than fifty
years by reduction of sodium carbonate with carbon ;
quite recently the electrolytic method has again been used,
and is now the only remunerative process. On the other
hand, there are certain products which have only come
into existence (actual or industrial) since electrical methods
have been developed. The most noteworthy instance of
such a product is silicon carbide (carborundum) (q.v.).
This body, as far as we know, does not exist naturally in
the earth's crust, and has certainly not been prepared by
ordinary chemical methods. It has been created by the
electric furnace, and there is no question as to what process
must be used in preparing it. Calcium carbide (q.v.) stands
on a similar, but not identical, footing. It can be pre-
pared by chemical methods, but they are commercially
impracticable, whereas it is already made on a large scale
at a low cost by an electrical process. In like manner
persulphates, which find a somewhat limited use as oxidising
agents, can, as far as we know, be prepared only by electro-
lytic means.
In all such cases there is no difficulty in choosing a pro-
cess ; the selection is made on the principle of Hobson's
choice. But in many other instances several possible
processes are available ; adoption of one rather than another
depends on many considerations which must be taken
into account in each individual case. Here it will suffice
to indicate the chief conditions which make an electrical
process preferable to a chemical process having the same
product, or vice versa.
When these are thoroughly understood, it is possible to
decide what method should be adopted under any given
local circumstances. The following generalisations may
be found of utility.
When an operation requires a large quantity of heat at
24
INTRODUCTORY
a temperature not exceeding 2,000 C. - 3,632 F., there
is strong primd facie ground for choosing a chemical rather
than an electrical method. This is because heat obtained
from electrical energy is greatly more costly unit for unit
than heat obtained directly by combustion. Electrical
energy, if obtained from the heat energy of coal through
the agency of the usual intermediaries boiler, engine,
and dynamo does not represent more than one-tenth of
the energy given out by the original combustion of the
fuel under the boiler. Its costliness in money as distinct
from energy is higher still, because an expensive plant,
representing heavy interest and upkeep charges, is required
for the conversion of heat energy into electrical energy.
Where water power is available, these charges still make
electrical energy much dearer than heat energy directly
obtained by the combustion of fuel. Thus, the energy
from a water power represented by 1 H.P. acting for a year
of 365 days, each of 24 hours, costs in interest and main-
tenance not less than 2 under the most favourable con-
dition. One H.P. year -= 5,646-2 Cal. = 705-8 kilos of
coal of calorific value 8,000, which, at 10s. per ton, costs
7s. Competition is therefore out of the question if the
object to be obtained is merely heating.
But when the heating has to be conducted by trans-
mission through a refractory envelope, as in the reduc-
tion and distillation of zinc, the aspect of affairs is alto-
gether changed. Much loss of heat takes place in such
transmission, and the cost of renewal of the envelope, e.g.
a fireclay retort, is extremely heavy. Heat can be generated
electrically in the interior of a refractory vessel, and loss
of heat and cost of renewal of the receptacle can be reduced
to a small value. Thus the more costly form of heat (electri-
cal energy) may enter into successful competition with the
intrinsically cheaper method of direct heating by the com-
bustion of fuel.
Again, when the temperature necessary for a given
operation exceeds 2,000 C. - 3,632 F., the electrical
method stands unrivalled, because no other means exists
25
PRACTICAL ELECTRO-CHEMISTRY
of obtaining so high a temperature. Into a box made of
refractory and non-conducting material electrical energy
can be poured, so that the attainable temperature is limited
only by the stability of the materials composing the hearth
and the conductors. All known substances can be fused, and
in most cases volatilised, under these conditions, and opera-
tions needing a temperature ranging between 2,000 C.
= 3,632 F. and 3,500 C. = 6,332 F. can be performed
thus and only thus. The ultimate reason for the impos-
sibility of attaining these high temperatures by chemical
means is that all reactions which generate heat are annulled
at these temperatures, only those reactions which absorb
heat occurring. Hydrogen and oxygen co-exist at 2,000 C.
without uniting, and carbides, e.g. those of the type of
acetylene, which, it is reasonable to suppose, are endo-
thermic compounds, are freely produced. 1 In fact, at
such temperatures certain borides, carbides and silicides
are almost the only substances which are stable. Thus,
all ordinary fuel ceases to act as such, and the electric fur-
nace is the only effective apparatus.
When an operation requires the application of energy
in the form of heat, and the product is liable to deterioration
by contact with fuel and the substances generated by its
combustion, electrical methods of heating possess an advan-
tage over chemical methods which may more than com-
pensate for their greater cost per unit of heat.
For example, the fusion of steel of the highest grade in
an ordinary furnace is attended by some risk of change by
oxidation or by absorption of sulphur from the fuel ; fusion
by the direct application of electrical energy can evidently be
effected without incurring such risks ; this is now practised.
Regarding processes which can be carried out almost
as well chemically or metallurgical^ as electrolytically,
such as the refining of copper, the precipitation of gold
from cyanide solutions, the parting of gold and silver alloys,
1 It has lately been stated that a temperature as high as 2 000 C.
>e produced by the combustion of acetylene in oxygen.
26
INTRODUCTORY
no better or more informative generalisation can be made
than that given at the head of this section, viz. that the
whole matter is one of cost. For a product of equal quality,
electrolytic copper is cheaper than copper refined by the
ordinary methods of metallurgy ; therefore the greater
part of the industry is already electrolytic. But lead of
surpassing purity can be prepared quite as easily by metal-
lurgical as by electrolytic methods ; therefore no displace-
ment of existing processes is probable. Silver and gold
can be separated effectively by parting with nitric acid or
sulphuric acid, but electrolytic " parting " is believed to
be somewhat cheaper, and is accordingly making headway.
Where fuel is dear, water power abundant, raw materials
weighty, but close at hand, and the finished material rela-
tively small in weight and of relatively high price, it may
be feasible to carry out an electrolytic or electro- metallurgical
operation rather than attempt one requiring either fuel
brought to the distant seat of the works or raw materials
conveyed to a remote centre where fuel is cheap.
To sum up, there is no magic in electrical or electrolytic
methods. With a few exceptions they are simply alterna-
tive processes, and choice between them and chemical or
metallurgical operations capable of arriving at the same
goal can be made only when all the circumstances proper
to each case are competently considered and their influence
computed.
27
SECTION II
Winning and Refining of Metals by
Electrolytic Means in Aqueous
Solution
COPPER
THE ELECTROLYTIC REFINING OF COPPER
THIS is the largest of all electrolytic industries. It is
practised in this country and on the Continent, but
the place of its greatest development is the United States ;
the output there is said to be at least ten times that of all
European factories put together. The electrolytic winning
of copper, as distinct from its refining, is even now scarcely
beyond the experimental stage ; an account of the most
hopeful processes will be found in a separate section.
PRINCIPLES OF THE ELECTROLYTIC REFINING OF
COPPER
The copper to be refined is already as metal, although
the metal is crude, containing not more than 98 per cent,
of Cu, and sometimes a smaller percentage. The rationale
of electrolytic refining is to transfer this copper, by the
selective action of the current, from the anode to the cathode,
and to leave the impurities behind as a sludge or dissolved
in the electrolyte preferably in the sludge. On first prin-
ciples it is evident that this mere transference of copper
should require no expenditure of energy, because metallic
copper is both the raw material and the product ; the
energy needed to precipitate it from its solution is precisely
balanced by the energy set free by its dissolution. This
theoretical deduction is entirely consonant with experiment,
but there are limitations commercial and industrial rather
than technical which prevent the full advantage of this
economy of energy being reaped in practice. They will be
discussed in the ensuing section.
PRACTICAL ELECTRO-CHEMISTRY
The only other principles needing mention are that the
electrolyte should be maintained sufficiently rich in copper
to ensure the presence of an ample supply of copper ions
at the cathode ; so that the electrical energy passing may
not be expended on any work but that of depositing the
copper, that the copper should be deposited in a coherent
and manageable form, and that the conditions of electrolysis
should be so adjusted that -copper and copper only is the
product at the cathode.
THE PRACTICE OF ELECTROLYTIC COPPER REFINING
Various methods of carrying out the process of electrolytic
copper refining are in use. In the most usual arrangement
the anodes are plates of crude copper, the cathodes are
thin sheets of pure electrolytic copper, and the electrolyte
is a solution of copper sulphate acidulated with sulphuric
acid.
The details of the process, such as the composition of
the crude copper, the strength and acidity of the electrolyte
and the arrangement of the electrodes, will differ in different
works according to local conditions, but a general state-
ment may usefully be made before proceeding to the descrip-
tion of any particular works.
In this typical works the anodes are of crude copper
similar in grade to the commercial product known as Chili
bars, and having a composition such as that given below :
Per Ont.
Cu . 98-60
As 0-80
Sb 0-10
i 'Pb. 0-10
Bl 0-05
Fe .
0-10
0-10
S 0-10
A 8 0-05
100-00
32
COPPER
This metal is cast into plates about 3 feet long, 18 inches
wide, and f inch thick. The cathodes are of similar length
and width, but about -$ inch in thickness. The anodes
and cathodes are suspended opposite to each other at a
r ,
FIG. 1.
A
distance of 2 inches and connected in the manner shown
diagrammatic ally in the accompanying sketches.
Fig. 1 shows the length of the vat (which is of wood,
lined with lead), and Fig. 2 shows the top of the vat in plan.
The anodes, marked A, are suspended from the positive
conductor, and the cathodes, c, from the negative. The
FIG. 2.
vats are arranged in series, as is shown in Fig. 3, and for
convenience the conductors are coupled #o that the
same serves for the anodes of one vat and for the
cathodes of the next ; by this arrangement no crossing of
33 D
PRACTICAL ELECTRO-CHEMISTRY
connections from side to side of the row of vats is
necessary.
The drop of pressure in each vat is about 0-2 volt, anc
the current passing is such as to be equivalent to about 10
amperes per square foot. The farther side of the end anode
is not faced by a cathode, and so its surface is not fully
effective. Assuming that it has about half the efficiency
of a surface directly opposed to a cathode, and that the
full surface of the remaining four anodes in each vat is
utilised, the total available anode surface will be
4 (3 x 1-5) 2 + 3 x 1*5 square feet = 40' 5 square feet
(neglecting the area of the edges of the anodes). Each
vat will therefore require 405 amperes, and, assuming theore-
tical current efficiency and that 1-1827 grammes of copper
is deposited by one ampere in one hour, there will be deposi-
ted 479 grammes of copper per hour, or in the 24 hours
very approximately 25 J pounds in each vat. This simple
calculation shows how small is the yield of electrolytic
FIG. 3.
copper per unit of plant, and makes clear that in a works
of any magnitude the vats must be very numerous. Great
increase of the size of the vats or of the number of plates
which they contain is not feasible for practical reasons,
such as the difficulty of maintaining a uniform current
density over very large surfaces and the difficulty of ensur-
ing a rapid and thorough circulation of the electrolyte
throughout a large vat. It is also obvious that if a larger
current density can be employed, a proportionally larger
output per unit of area of the electrodes will be obtained.
In practice 10 amperes per square foot is sometimes
exceeded, as much as 20 amperes per square foot having
34
COPPER
been used in the United States ; it is found, however, that
with a high current density the copper tends to be deposited
in warty and cauliflower-like masses, to be of inferior purity,
of feeble coherence and to tend to grow across to the
anode and form a short circuit.
As regards the consumption of energy for the deposition
of this copper, it is evidently directly proportional to the.
voltage necessary for each bath. In theory with an infinite
electrode surface and an infinitely small internal resistance
this is nil. In practice this cannot be approached because
the current density would thereby be so far reduced that
the output of copper per unit of plant would be unduly
small. The interest on the capital represented by a huge
plant would 'be too heavy a charge, and yet more the interest
on the capital represented by the value of the copper tempo-
rarily locked up as anodes would be prohibitorily great.
Moreover, seeing that the price of copper fluctuates consi-
derably, every electrolytic copper refiner would be in the
position of an enforced large holder of a gambling stock
which he could handle more slowly and with more restric-
tions than those affecting other holders ; thus he would
always be financially at the mercy of a less burdened
operator.
Therefore a very appreciable voltage must be used to
get a reasonable output on a given stock of copper, i.e. to
obtain a fairly rapid turnover. 0-2 volt is not a high esti-
mate of what the voltage would be in our typical works.
Accepting this, the watts necessary for each vat are 405 x
81
0-2 = 81 watts, i.e. = 0-109 H.P. Each horse power
746
hour in a set of vats identical with that which has been
described would deposit 4,394 grammes of copper, i.e. a
horse power acting for 24 hours would deposit 232 pounds
of copper. A plant of 1,000 H.P. would deposit 232,000
pounds of copper per 24 hours, or 37,803 tons per year of
365 days of 24 hours each. With a drop of voltage for each
vat of 0-5 volt, which approximates more closely to what
would be expended in practice, all the above figures repre-
35
PRACTICAL ELECTRO-CHEMISTRY
senting output must be multiplied by f , t'.e. the output per
horse power for 24 hours would be 93 pounds, and for 1,000
H.P. for a year of 365 days 15,121 tons.
Even this reduced amount corresponds with a turnover
of 750,000 worth of copper per year ; at a moderate esti-
mate the weight of copper permanently present in each
vat will be one-twelfth of the output per year of that vat ;
so that the cost of the stock of copper alone which is neces-
sary for carrying on the business is 62,500, representing
an interest charge of 3,125. It is clear from this that any
saving in the stock of copper relative to the output will be
worth achieving, even if the cost of the energy expended
be somewhat increased. In other words, in order to get
the maximum output per vat it will pay to drive each vat
at a sufficiently high voltage to obtain the maximum cur-
rent density which will still permit of the deposition of
coherent copper of good quality. Where power is cheap
a considerable waste of energy can be permitted with pecu-
niary profit in order to make the turnover large compared
with the stock of copper permanently in the vats.
In our typical works the electrolyte will be circulated
throughout the vats so as to replace that part of the solu-
tion which has passed over the surface of the cathodes and
has been thereby impoverished in copper, with liquor which
has passed over the surface of the anodes and has been
enriched in copper. Any stagnation would result in lack
of copper at the cathodes and separation of hydrogen and
possibly of metallic impurities together with the copper,
and would also cause a superfluity of copper sulphate at
the anodes, upon which the salt would crystallise, hindering
their dissolution.
At intervals the electrolyte will become inconveniently
impure, and will have to be purified or replaced by fresh
sulphate of copper. Save for this the work will proceed
continuously, crude copper being used up and pure copper
obtained in a merchantable form.
The black mud which comes from the anodes and repre-
sents the insoluble impurities of the crude copper, contains
36
COPPER
gold and silver, and is worked up for the recovery of these
metals.
Such being a general outline of the essential parts of
an electrolytic copper refinery and of the process of refin-
ing, the various portions of the plant may be considered
in detail.
SOURCE OF POWER
Water or steam power is used according to the situation
of the works. The former is less advantageous than would
appear at first sight, because its cost per unit of energy is
by no means negligible; being represented by the interest
on the plant and the upkeep of the plant ; the former is a
heavy item. In general, water power is utilised by choosing
a river (which may have to be impounded so as to equalise
its flow) at a point where its bed makes a considerable fall,
and conveying the water through an artificial channel to
turbines which are coupled direct to dynamos. The con-
struction of a reservoir, channel, turbine pit and tail race
involves a large expenditure of money, and the turbines
and dynamos are costly machines. The precise capital
sum expended per horse power will differ in each case, and
fts consideration is a purely engineering matter. It is
sufficient to say here that the capital is so large that the
lowest probable estimate which has been arrived at for
the cost of the power obtained is 2 per horse power
year of 365 days of 24 hours each, i.e. 0-0548d. per horse
power hour. Under less favourable conditions the cost
would probably be double this, i.e. 4 per horse power year
or 0-1096d. per horse power hour. The lowest probable
cost of steam power is 0-25d. per horse power hour, or
9 16s. per horse power year with fuel at 8s. per ton, and
under less favourable conditions it may reach 15 per horse
power year or 0-4 lie?, per horse power hour. In each case
it is assumed that the horse power used is large, l,OOOi.H.P.
or more. A modern gas engine plant of large size might
produce power at a rate approaching that of water power
e.g. 0- 15d. per horse power hour. The difference in favour of
37
PRACTICAL ELECTRO-CHEMISTRY
water power is considerable, but its monetary advantage is
smaller than would be supposed, because the cost of the
energy required in copper refining is not the chief item of
expenditure, as is evident from the appended table :-
COST OF ENERGY IN COPPER REFINING
Cost per horse
power hour.
Cost per horse
power year. '
Cost per ton of
copper refined.
Water power
0-0548(1
0-1096d.
2 05.
4 Os.
2s. 8d.
5s. 4d.
Steam power |
0-25d.
0-411d
9 165.
15 Os.
12s. Od.
20s. Od.
Even the highest of these sums is not much more than
1 % on the selling price of the pure metal.
It is in processes such as the manufacture of calcium
carbide (q.v.), in which the quantity of energy expended is
extremely large, that its cost becomes great relatively to
the total cost of working, and the economy affected by the
use of water power becomes sensible.
The power, however obtained, is used to drive dynamos
which in modern installations are generally coupled direct
to the turbines or steam engines. For practical reasons it
would be inconvenient to work with a single large engine
and dynamo, because a breakdown would necessitate the
stoppage of the whole works. Therefore a unit of 200-300
H.P. is chosen, and each dynamo coupled direct will yield
130-200 kilo watts. These may be delivered at any conve-
nient pressure, say 100 volts ; the current will therefore
be 1,300-2,000 amperes. The voltage for which the dyna-
mos are wound of course depends on the number of vats
it is proposed to run in series and the voltage necessary for
each. These will vary with each installation, and it can
only be said that a fairly high voltage is desirable as dimin-
38
COPPER
ishing the requisite cross section, and therefore cost of the
leads, while any very high voltage should be avoided as
likely to cause loss by leakage from bare conductors con-
stantly liable to be accidentally wetted from the baths.
RAW MATERIAL
In all cases the raw material is crude copper containing
about 98 per cent, of Cu. It will vary in composition
according to the character of the ore and the method of
dry refining adopted. The following figures will indicate
the nature and extent of these variations :
I.
II.
III.
IV.
V.
VI.
Per cent.
Per cent.
Per cent.
Per cent.
Per cent.
Per cent.
Cu . .
96-35
97-19
98-32
98-53
94-06
98-24
As . .
0-08
2-68
0-19
0-62
4-36
0-94
Sb . .
0-10
0-01
0-06
0-57
0-40
Sn . .
0-22
Trace
Pb . .
1-19
0-16
0-13
0-02
Bi . .
0-05
0-08
0-07
0-06
0-02
0-04
Fe . .
0-61
0-02
0-01
0-01
0-02
Trace
Ni . .
0-02
0-25
0-14
0-37
0-28
S . . .
0-69
0-68
0-01
0-37
0-03
O and loss
0-71
0-26
0-63
0-10
0-05
100-00
100-00
100.00
100-00
100-00
100-00
Besides these, the crude copper will contain a little silver
and a much smaller quantity of gold, e.g. 30 ounces of silver
and T T ounce of gold per ton. The value of these at 2s.
and 3 155. per ounce respectively is 3 7s. 6d.
Sometimes a better grade of metal is available ; thus
Titus Ulke 1 states that the following may be regarded as
typical anode copper. The composition is shown side by
side with that of the refined copper produced from it.
1 Electrical Review (New York), 1901, p. 85.
39
PRACTICAL ELECTRO-CHEMISTRY
Crude %.
Refined %.
Cu . .
As . .
Sb . .
Pb . .
Bi . .
Fe . .
Ni . .
Se and Te
o . .
Ag . .
Au
. 99-250
0-033
. 0-054
0-009
0-002
Trace
0-002
0-008
0-300
0-338
0-001
Cu
. . . . 99-925
As
. . . . 0-001
Sb
0-002
Pb
. . . . 0-001
Aff
0001
o
. . . 0-070
99-997
100-000
The quantities of silver and gold in this crude copper
correspond with 110 and J oz. per ton respectively. Thus
the copper is not only comparatively pure but remarkably
rich in precious metals. It cannot be accepted as repre-
sentative of ordinary crude copper.
Before electrolytic refining had become as well estab-
lished an industry as it is to-day, copper containing these
small proportions of precious metals was sold simply as
crude copper. Nowadays the natural effect of competition
among refineries has caused sellers to exact a fair price for
the gold and silver found by analysis to be present in the
crude copper. But even under these conditions it is
advantageous for a copper-smelting works to have its
own electrolytic refinery.
It may be stated that in general it is not remunerative
to refine very crude metal. It is better to bring it to a
content of about 98 per cent, of Cu, because a cruder copper
speedily causes the electrolyte to contain such large quan-
tities of soluble impurities that the deposition of pure copper
is hindered, and the electrolyte has to be renewed. The
behaviour during refining of the various impurities com-
monly present in crude copper may be summarised thus :
Using a copper anode in a solution of copper sulphate
acid with sulphuric acid, silver and gold remain undissolved
40
COPPER
in the anode sludge as metals. Lead also remains as sul-
phate. Antimony, tin and bismuth dissolve partly to form
unstable sulphates, from which oxides or basic sulphates
are deposited on standing ; the larger part, however, of
each remains with the anode sludge. Arsenic, iron and
nickel dissolve and are not redeposited ; thus they con-
taminate the electrolyte, but do not contaminate the puri-
fied copper under ordinary working conditions. Cuprous
oxide remains partly in the sludge and partly dissolves
according to the acidity of the electrolyte. Its only evil
effect is to neutralize a portion of the free sulphuric acid
which is essential to clean working. Copper sulphide
distributes itself similarly. Tellurium and selenium are
sometimes found in the anode sludge, but their quantity is
naturally small. It might be supposed from this that it
would be possible to purify very crude copper by electro-
lysis, and, indeed, numerous attempts in this direction have
been made. They have all failed, not because it is impossi-
ble to separate the bulk of the impurities and obtain a pure
copper at a single operation, but because anodes of even a
moderate crudeness are dissolved unevenly and wastefully,
the electrolyte penetrating into the interior of the plate,
causing local corrosion and eventually detaching portions
of the anode, still rich in copper, bodily. Besides this,
the electrolyte has to be purified or renewed more often
than when working with a fairly pure raw material.
The composition of the anode sludge will evidently vary
with the composition of the crude copper. Thus the
various elaborate analyses which have been published from
time to time are of purely local interest. It may be taken
that an ordinary sample rich in silver will contain about 30
per cent, of copper (partly as oxide, antimoniate, sulphide,
etc.), 30 per cent, of silver, and 30 per cent, of lead sulphate,
oxides of antimony and tin, and the various small impuri-
ties, such as bismuth, sulphur, selenium, tellurium, and
gold. The working up of the anode sludge will be dealt
with elsewhere. The composition will vary enormously
according to the richness of the alloy in silver (and gold),
PRACTICAL ELECTRO-CHEMISTRY
but in general it may be said that copper, silver, and lead
are the three chief metals commonly present.
COMPOSITION OF THE ELECTROLYTE
In all cases the electrolyte consists of copper sulphate
acid with sulphuric acid ; a usual strength is 1 J pounds of
crystallised copper sulphate and J pound of sulphuric acid
per gallon. Much mystery is sometimes made about the
precise composition of the electrolyte, but the only princi-
ples to be observed are : (1) That there should be plenty
of copper, short of saturating the solution and causing risk
of crystallisation ; (2) That there should be sufficient sul-
phuric acid to prevent the separation at the cathode of
hydrated cupric oxide ; (3) That the quantity of sulphuric
acid should not be so great as to cause hydrogen instead of
copper to be separated at the cathode.
These conditions are fulfilled within a fairly large range
of composition, and thus secret recipes are of small import-
ance. Perhaps the only effective addition is a small quan-
tity of salt or of hydrochloric acid to ensure the precipita-
tion of any silver which may find its way into solution.
The electrolyte may be kept warm, e.g. at a temperature
of 35 C. = 95 F., whereby circulation is aided and the
use of a high current density with the production of sound
coherent copper is facilitated.
ARRANGEMENT OF VATS
The vats themselves are of wood, strong, well- jointed
and lined with pitch or sheet lead autogenously soldered.
Like all tanks for chemical purposes, they should be ar-
ranged with a clear space round each, so that any leak may
be at once detected and remedied. The vats are some-
times placed in steps in order that the electrolyte may
flow from one to the other throughout the whole series, and
be finally collected at the end of the series and returned to
the beginning by the aid of a pump. The appended figure
42
COPPER
(Fig. 4) illustrates this method. The overflow from the
vat A passes through the pipe E to the bottom of the vat B ;
in like manner the overflow from B passes through the pipe
F to the bottom of the vat c. From the last vat of a series
,_-._ J F
FIG. 4.
such as this the liquor flows into a collecting tank, whence r it
is pumped to an overhead distributing tank at the upper
end of the series. When the tanks are not arranged ter-
race-wise, circulation is effected as it were in parallel instead
of in series ; Fig. 5 illustrates the method, A, B, and c being
the tanks, D the supply pipe, and E the exit pipe ; both are
connected by branches to each vat.
It is evident that in strictness the circulation of the
electrolyte should be so arranged that the liquor never passes
FIG. 5.
from one vat to another when the two are coupled in series,
but only when the two vats are in parallel. Otherwise a
leakage of current along the stream of liquid flowing from
vat to vat will occur. That it is possible to fulfil this con-
43
PRACTICAL ELECTRO-CHEMISTRY
dition is clear from the appended diagram (Fig. 6). In the
figure only two plates (one anode and one cathode) are
shown in each tank, for the sake of simplicity, but the same
scheme holds good whatever the number of plates in each
vat. In the group of vats shown there are sixteen individual
vats arranged in groups of four coupled in parallel and four
in series. The members of the group, A, A, A, A, are in
parallel ; similarly the members of the group B, B, B, B,
are in parallel ; the same is true of c, c, c, c, and D, D, D, D.
But the whole of the group A is in series with the remaining
FIG. 6.
groups B. c, and D. Therefore, the electrolyte is circulated
through the group A by the connecting pipes p, p, p, p, and
similarly through each of the three remaining groups. It
does not, however, pass from group A to group B. A cir-
cuitous connection through the tank whence all the distri-
buting pipes start may exist, but the resistance of each
long narrow column of electrolyte would be so large that
no appreciable leakage of current could occur. Though
this is the best method of circulation, it does not follow,
44
COPPER
however, that one in which the circulating pipe connects
the tanks which are in series would necessarily fail.
Leakage of current, though inevitable with such an arrange-
ment as is shown in Fig. 7, would be small. Thus in the
vat A the anode E has not only its legitimate cathode F
opposed to it, but also the plate G of the vat B, because the
connecting pipe between the vats makes A and B electrically
one cell. This current passing from A to B through the
connecting pipe will tend to make G a cathode and to deposit
copper on it. But G is the anode of the vat B, and therefore
loses copper instead of receiving it. This does not neces-
sarily involve a loss of energy, save that fraction spent by
the current traversing the connecting pipe of small sectional
area and high resistance, but it does involve a smaller out-
B C
FIG. 7.
put of refined copper per unit of plant, and is to that extent
objectionable.
Devices for regulating the flow of the electrolyte, such
as cocks on the pipes or screw clamps on rubber distributing
tubes, are necessary, to ensure that every vat may receive
its quota of liquid and that there shall be no risk of over-
flow. In like manner the exit tubes may take the form of
syphons of sufficient bore to take the maximum quantity
of liquid which is likely to flow into the vats, and at the
same time to avoid any chance of the outflow being so free
as to empty the vats. The principle of such arrangements
may be gathered from Fig. 8, which illustrates a common
laboratory apparatus for maintaining a constant circulation
and a constant level of any liquid. The tank A is kept
filled to the level shown by a constant or approximately
45
PRACTICAL ELECTRO-CHEMISTRY
constant small flow through 'the delivery tube B, any surplus
beyond that level being carried off by the intermittent
syphon c . This syphon has equal limbs , and the flow through
it is therefore determined by the height of the liquid
FIG. 8.
in the tank above the level of the end of the limb in the
tank. Both its limbs being upturned, and its head per se
being nil, the syphon is incapable of emptying itself, and
remains full under all conditions, ready to come into action
C
FIG. 9.
immediately the level in the tank rises. Nothing but a
supply so inordinate as to be beyond the capacity of the
syphon to carry off can derange the working of this device.
An equivalent design is shown in Fig. 9. Here the syphon
46
COPPER
is dispensed with and an exit tube is provided, passing
through the bottom of the tank, and of such width that with
a very small head above its upper end it can discharge the
whole of the liquid supplied through the pipe B. It is
obvious that many such contrivances can be adopted for or
a-dapted to the circulation of an electrolyte ; their use is
not peculiar to the electrolytic refining of copper, but is a
matter of ordinary engineering.
A method of circulating the electrolyte in copper refining
has been worked out by Messrs. K. and H. Borchers, of
FIG. 10.
Goslar, and may be briefly described. The circulation of
liquid from vat to vat is abolished, and, as a substitute for
this, the liquid in each vat is caused to circulate in such a
way as not to stir up the sludge from the anodes and make
the liquid muddy. The accompanying figure (Fig. 10)
shows the chief features of the method.
The vat A has a leaden pipe D passing down through the
false bottom c, carrying the leaden tray B (which is ordin-
arily used in electrolytic refining vats for the collection of
the anode sludge). Inside the pipe D is a narrow glass tube
E, drawn out to a point at the lower end. Through this
air is blown and is distributed in fine bubbles, which, by
47
PRACTICAL ELECTRO-CHEMISTRY
giving an upward motion to the liquid in the leaden pipe,
cause the liquid in the pipe to flow over at the upper end
and to be replaced by fresh liquid from without at the lower
end, thereby securing a gentle and continuous circulation
of the liquid in the vat. The contents of the vat may be
drawn off when the proportion of impurities becomes ex-
cessive, by the cock F, but during the whole time that the
electrolyte remains useable the same liquid remains in any
given tank, and there is no need to provide the ordinary
system of circulation from tank to tank, which is compara-
tively complicated. It is claimed as a collateral advantage
that the air blown in primarily to cause circulation acts in
addition as an oxidising agent, and purifies the electrolyte
to a great extent by precipitating iron and arsenic jointly
as ferric arseniate. The correctness of this claim has been
contested, and having regard to the fact that the electrolyte
is kept fairly strongly acid, it is intrinsically improbable.
The merit of the invention consists rather in the employment
of air, which is a convenient agent for agitating the liquid
in such a way as to induce circulation of the electrolyte
without stirring up the anode sludge.
An ingenious method of circulating the electrolyte has
been devised by H. E. Dolphin. It is in use on a large
scale at Lewis & Sons' works at Widnes, and is being tried
by the Amalgamated Copper Co. in America. The principle
of the method is to use a jet of the electrolyte as an injector
to pull in air and to drive liquid and air together to the
bottom of the cell agitating its contents ; the liquid displaced
by that injected continuously overflows and is returned to
the distributing tank. The appended figures show one form
of arrangement used. In Fig. 11 A is the distributing reser-
voir, i the injecting pipes leading to the vats D, G the overflow
pipes, and H the collecting reservoir. ,The details of the
injecting pipes are given at the side of the figure.
The nozzle B has an opening of about ^ in. in diameter,
and, acting as an injector, pulls down air through the side
hole E, and discharges both air and liquid at the lower end
the pipe c. As advantages of this method may be reckoned
48
COPPER
the fact that, on account of the small diameter of the jet,
electrical connection between the vats by means of the electro-
lyte itself is practically severed ; also it is stated that the
air will tend to oxidise the impurities in the electrolyte ;
further, and most important, it is claimed that by reason
of the brisk circulation a higher current density than usual
can be employed without impairing the quantity or coherence
of the deposited metal.
The following description of the working of the process
at Widnes, based on eighteen months' experience, embodies
many points of interest. A typical installation will con-
sist of 30 cells, each 6 ft. 6 in. x 3 ft. x 2 ft. 2 in. deep, and each
containing 38 electrodes -f in. thick arranged in series as de-
scribed on p. 51 ; the cells themselves are in parallel. A current
FIG. 11.
density of 20 amperes per square foot is used, and with a
current efficiency of 87.6 per cent, the output per vat is
1 ton 5 cwt. in fourteen days. Usually the series method
of arranging the electrodes has the disadvantage of pro-
ducing much scrap e.g. about 22 per cent.; in the present
49 E
PRACTICAL ELECTRO-CHEMISTRY
with a better circulation of the electrolyte not
mo tC 15 per cent, is made. It has also been found that
The Ion torn the anodes does not remain in solution this
Sect being assigned to oxidation by the injected air (com-
the remarks g on this point on p. 48), and that the anti-
mony precipitated almost entirely, not more than
16 grains per cubic foot of electrolyte remaining in
solution.
ARRANGEMENT OF THE ELECTRODES
Usually all the electrodes in a single vat are connected
in parallel. There may be many electrodes in each vat,
but electrically all the anodes constitute a single electrode,
and all the cathodes another. This is the most sensible and
effective method, but other systems have been put forward,
concerning which a few words must be said. In these the ob-
ject has been to do away with all connections except two for
each vat. This can be easily arranged by relying on the
moderately regular drop of voltage that occurs from one
electrode to its fellow at the other end of a long vat. Plates
placed between the anode and cathode in an electrolyte,
and unconnected with either, will act as intermediate elec-
trodes, the side of each facing the anode functioning as a
cathode, and that facing the cathode as an anode. The
simplest and at first glance the most plausible of these
arrangements is shown in Fig. 12 ; it is due to Hayden,
and is said to have been largely used in America and to
be even now in use by the Baltimore Electrical Refining
Company. The anode A is a plate of crude cast copper ;
the cathode B is a thin sheet of pure copper ; the in-
termediate electrodes c, D, E, r, G are of crude cast
copper.
During the passage of a current from A to B, copper is
dissolved from A and precipitated on the side of c facing
A. At the same time, the other side of c (remote from A)
acts as an anode, and copper is dissolved therefrom and
50
COPPER
deposited on the side of D facing c, which acts as a cathode.
This proceeds throughout the series of immersed intermedi-
ate electrodes until B is reached ; this receives copper from
G and acts purely as a cathode. It will be seen that a con-
tinuation of this process will gradually convert these inter-
mediate electrodes into plates of pure copper, and, suppos-
ing the change to have proceeded with perfect regularity,
each intermediate plate will have become shifted towards the
anode A by a distance equal to the thickness of an intermediate
plate. Nothing could well be neater than this arrangement,
provided everything would go smoothly. Many connec-
tions are abolished, the whole of the intermediate plates
may be immersed so that there should be no waste anode
ends to melt down and re-cast, and the need for separate
A
i
B
+
C
i
)
E
F
(
*
m
-
1
' ' V
{ -
f~^
+ "=
* -
+ -
+~ "="
4- -
-\
'/.
<
%
1,
\
V.
'/
I
\_-
:
^j
:
-
.-=--
~~ _L
;-_
- \
W////7/,
47//W/,
7 ///W//
W///W,
'//////
'//$&
FIG. 12.
cathodes of pure copper sheet is done away with. In prac-
tice, however, the disadvantages are numerous and serious.
The dissolution of the intermediate electrodes does not pro-
ceed regularly. Cavities appear on the anode side and
continue to form until they reach the pure copper already
deposited on the cathode side, which is corroded in its turn.
Complete clearance of the half of the plate acting as an
anode does not always occur, and the resulting plate is still
partly composed of crude copper, and has to be scrapped
and melted down again for anode making.
PRACTICAL ELECTRO-CHEMISTRY
There are several other systems using intermediate
electrodes, differing chiefly in the arrangement of the plates,
whether vertical or horizontal, with the cathode faces look-
ing up or down, with a single plate for each intermediate
electrode, or a composite plate made by attaching electrically
a thin plate of some conducting material, e.g. pure copper,
to one side of a thick plate of crude copper. They are all of
doubtful utility. Their genesis is probably to be traced to
the inherent belief in many minds that in some way the law
of the conservation of energy may be evaded. Inventors
proceeding on this principle are ignorant of the fact (stated
.above) that the mere transference of copper from anode to
cathode requires no expenditure of energy ; that the need for
a considerable expenditure of energy in practice arises from
the necessity of keeping the size of the plant and the stock
of copper for a given output within reasonable limits. Such
inventors have therefore striven to force an open door, and
have gone the wrong way to work to do so. Assuming
smooth- working conditions, the total energy necessary to
refine a given weight of copper is the same whether the
electrodes are arranged in the ordinary manner or are of the
intermediate class. Choice between the two methods is to be
arrived at purely from considerations of convenience, and
experience has shown that the so-called " series " system,
i.e. the method of using intermediate electrodes unconnected
directly with the terminals of the dynamo, is not the most
convenient.
MODE OF WORKING THE PROCESS
With an installation arranged on the lines given above,
the running becomes a matter of simple routine. A
switch-board in the works manager's office should enable
him to read the current and voltage for each vat at pleasure.
In large works an automatic recorder put into action peri-
odically by a commutator driven by clock-work allows a
regular record of the conditions obtaining in each vat
throughout the twenty-four hours to be secured. The
52
COPPER
anodes and cathodes are hoisted into and out of the vats by
overhead travelling cranes or some equivalent device. In
short, the methods of handling the raw material and finished
product are precisely those which would be adopted by any
competent engineer to whom the matter was submitted. It
is not proposed to give detailed descriptions of devices which
are well known and in constant use in many industries ;
the whole plant is of a per-
fectly simple character, its f
only peculiarities arising from
the large number of identical
units necessary for an instal-
lation of any considerable size.
The circulation of the elec-
trolyte may be effected by
compressed air in an " acid
egg," such as is used in vitriol
making. This apparatus,
which is designed to save
moving parts in contact with
corrosive liquids, consists es-
sentially of a closed chamber ,
with an exit pipe from its
lowest part for the liquid to
be conveyed, and another
pipe in its upper part for the
entrance of compressed air.
The whole arrangement is
represented diagrammatically
in the annexed figure (Fig.
13), where A is the pressure vessel, B the inlet for compressed
air, and c the exit for the liquid to be conveyed and distri-
buted. The distribution may be most conveniently effected
by gravitational flow from an overhead tank, and the acid
egg or its equivalent used to return the electrolyte to this
tank after its passage through the vats.
The conduct of an electrolytic copper refinery may be
gathered from the following description of a large plant
53
FIG. 13.
PRACTICAL ELECTRO-CHEMISTRY
belonging to the Baltimore Copper Co. at Baltimore. The
raw material is copper of the grade of Chili bars, and is
generally obtained from the Anaconda Co. The tanks are
about 18x4x4ft., and the electrodes are arranged in the
manner described on page 51 ; that is, the crude copper is cast
into plates, one side of each acting as cathode and the other
as anode. The electrodes are carried in wooden frames, and
are divided horizontally thus :
the division being probably designed for convenience of
handling and to decrease the waste which must occur when
a plate is spoiled by irregular dissolution. The circulation
of the electrolyte is by gravity, the liquid being collected in
a common trough and pumped back to a distributing tank.
The tanks are of wood, with a pitch lining. No artificial
heat appears to be used to warm the electrolyte, in which
respect, as well as in the employment of plates serving as
cathode-anodes, the installation differs from other modern
plants. In a part of the works the ordinary multiple system
with individual cathodes and anodes is used, and it seems
that the two methods are regarded as equally efficient. It
may be taken that the modern practice of electrolytic
copper refining is represented by that of this works, with the
exception that in general the use of plates serving both as
anodes and cathodes is less frequent. The reasons for the
method being discarded have been already stated.
QUALITY OF THE PRODUCT
In all well-conducted electrolytic refineries the copper is
very approximately chemically pure. The following anal-
54
COPPER
yses, by the author, of copper deposited by the Elmore
process (see below) indicate the very small quantity of
foreign matter present :
I.
II.
Copper
Per cent.
99-961
Per cent.
99-938
Arsenic
Nil
Nil
Antimony
0-002
0-002
Tin . . .
Nil
A7V7
Lead
Nil
Nil
Bismuth
Nil
A7V7
Iron
0-005
0-004
Nickel
Nil
Nil
Sulphur
Trace
Trace
Oxygen and loss ....
0-032
0-056
100-000
100-000
In this case the copper was deposited in the form in which
it was to be used, viz. in that of a tube, and had not been
subjected to any operation after it had left the deposit-
ing vat. Ordinary electrolytic copper, however, which is
deposited in the form of plates, has to be fused and cast into
ingots before it can be worked in any way, e.g. be drawn
into wire for electrical purposes. In this fusion there is a
risk of the metal absorbing oxygen (cuprous oxide being
soluble in metallic copper), to its detriment in conductivity,
and probably in tensile strength. This can be guarded
against by conducting the fusion in a neutral or reducing
atmosphere, and in practice some such precaution seems
to be adopted, inasmuch as wire made of electrolytic copper
is usually of excellent quality, having a conductivity nearly
or quite as high as that of the best metal that can be pre-
pared in the laboratory.
The following are analyses by the author of copper electro-
lytically refined and of high conductivity :
55
PRACTICAL ELECTRO-CHEMISTRY
I.
II.
III.
Copper .
Arsenic .
Antimony
Lead . .
Bismuth .
Iron .
Nickel .
Oxygen .
99-977
Nil
Nil
0-008
Trace
Nil
Nil
0-015
99-85
Nil
Trace
Trace
Trace
0-01
Trace
0-14
99-92
Nil
Trace
0-01
Trace
0-01
Trace
0-03
100-000
100-00
99-97
WORKING UP THE ANODE SLUDGE
A typical anode sludge contains (as stated above) copper
and insoluble compounds thereof, 1 silver, and sulphate of
lead as its principal ingredients, as well as small quantities
of numerous impurities varying with the nature of the crude
copper used as a raw material.
The chief valuable constituents are silver, and a little
gold. Their recovery may be effected, if they are present
in sufficient quantity, by cupellation with lead, the silver
and gold being left and parted by boiling with sulphuric
acid or by electrolysis (see the section on the electrolytic
refining of silver) in the ordinary way. Direct treatment
of the sludge with boiling sulphuric acid is also practicable,
the silver being converted into silver sulphate and dissolved
by diluting the acid liquid with hot water, running off the
silver sulphate solution from the lead sulphate, and preci-
pitating the silver with copper. The gold and lead sulphate
can be reduced by fusion with charcoal to an ingot of auri-
ferous lead, which can then be cupelled, leaving the gold
fairly pure. There is little else worth recovering, except
1 It has lately been observed by F. Foerster and O. Liedel that
the quantity of copper in anode sludge is smaller when working at
a fairly high current density, e.g. 10 amperes per sq. ft., than when
using a current density of about 3 amperes per sq. ft. This con-
dition is observed in modern refineries.
56
COPPER
perhaps selenium and tellurium, the trade in which is very
small. Special wet methods, involving the reduction of these
elements with sulphur dioxide, are necessary tor their
recovery, and the working up of the silver and gold would
then be carried out on the lines given above. The processes
of working up the anode sludge must obviously vary with
the composition of the sludge,in its turn ultimately dependent
on the character of the crude copper. A suitable method
for any given case can be devised and worked out by any
competent chemist. The question, though of great im-
portance, presents no special electrolytic interest, and
cannot be dealt with here.
The vast growth of the process of electrolytically refining
copper in the United States may be understood from a very
clear historical statement given in The Mineral Industry
for 1896. The first plant of any considerable size was
worked successfully in 1890 by the Baltimore Copper Com-
pany ; a Hayden plant (v.s.) was then put up in 1891 by the
Baltimore Electric Refining Company. The next year the
capacity of this plant was doubled, and thus the great
Baltimore Copper Works was developed, which now refines
two-thirds of the Anaconda output, viz. about 100 tons
daily. The world's production of electrolytic copper in
1892 was 32,000 tons, produced in 30 refineries.
In 1893 the production in the States alone was 37,500
tons, i.e. a quarter of the whole output in the States ; in
1894 it was 57,500 tons, or one-third ; in 1895, 87,000, or
a half ; in 1896, 124,000, or three-fifths ; this amounts to
one-third of the whole world's production. This very large
quantity is turned out by eleven refineries, which jointly
yield 14,000,000 ounces of silver and 68,000 ounces of gold
per year. The process of expansion has continued, and in
1902 278,860 tons of 2,000 Ibs were produced, yielding
27,000,000 ounces of silver and 346,020 ounces of gold.
The cost of refining has been considerably reduced of
late years. It was 20 dollars (say 4) per ton in 1892, and
about 8 dollars (1 12s.) per ton in 1896. At the present
time it is not greater than 4-5 dollars (16s.-l) per ton.
57
PRACTICAL ELECTRO-CHEMISTRY
This sum is the manufacturing as distinct from the com-
mercial cost, and does not include the expense of manage-
ment Comparing these figures with those given above for
the cost of power, it will be seen that the latter, although a
large item, is by no means the largest ; interest on plant
and that on copper locked up in the process are heavy
charges. The cost in Europe is put down at 13-18 dollars
per ton (2 12s.-3 125.). The reason for this difference is
that many of the European plants are antique and almost
obsolete, and, working on a smaller scale without the
mechanical labour-saving devices characteristic of the
American industry, are operated at a disadvantage. The
largest works in Europe is that of Bolton & Sons, at
Widnes, which turns out about 7,000 tons of copper per
year. Elliott's Metal Company's works at Penibry,
South Wales, is credited with 3,120 tons.
SPECIAL METHODS OF DEPOSITING REFINED COPPER
Owing to the fact that electrolytic copper is usually
deposited in rough plates, and has to be rused before it can
be formed into ingots suitable for rolling into rods (for
drawing into wire) or plates, or for drawing into tubes, there
is an extra cost incurred in thus bringing it into a workable
form, and there is also a risk of contaminating it, especially
with oxygen, during the process. Thus it conies about that
any process capable of depositing the metal in the form in
which it is to be used presents obvious advantages. It
would seem at first sight simple to deposit copper in the most
complicated shapes, and the fact that electrotyping (see
below) was successfully practised long before copper re-
fining became an industry lends colour to the view. But it
is quite impracticable in the ordinary vat to cause the
deposition of the metal to take place regularly enough to
give a uniform thick coating on a mould even of a simple
shape. Moreover, the metal as usually deposited is not
particularly homogeneous, and the strength of a plate is by
58
COPPER
no means great. Special means must therefore be adopted
to deposit the copper in a coherent form.
One of these methods is that devised by Elmore. Crude
copper of the grade of Chili bars is granulated and placed
on trays at the bottom of a vat, where it serves as an anode;
The electrolyte is a solution of copper sulphate acidulated
with sulphuric acid. The cathode is a roller of metal, or
wood coated with plumbago so as to be conductive ; this
roller must not, however, be so perfectly conductive as to
allow the copper deposited on it to adhere, as the copper must
afterwards be stripped from it. The roller revolves in
bearings, which also serve to convey current to it. On a
carriage like that of a screw-cutting lathe is mounted a rod
tipped with agate, which is pressed against the surface of
the roller and traverses its length, being automatically
reversed when it comes to the end of the roller and sent bac k
again. By this means the copper, as it is deposited, is
subjected to a continuous burnishing action, and small
rugosities are planished down. If once a visible excrescence
forms, it is almost impossible to prevent its growing, because
ipso facto it increases the current density at that point ; the
burnisher suffices to keep down microscopic eminences and to
maintain a smooth surface under ordinary working con-
ditions. Tubes of great regularity of shape and closeness
of structure may be thus prepared. The metal is, of course
almost perfectly pure, 1 and may have a tensile strength as
high as 20 tons per square inch, ordinary " tough pitch "
copper made by dry processes having a tensile strength
of about 14 tons per square inch. The tubes, being seamless
and very strong, are well adapted for use as steam pipes ;
it is, however, not easy to make bends by the Elmore process.
Another application of the method is the manufacture
of wire. For this purpose the metal is deposited in the form
of a tube, which is then cut spirally from end to end into a
strip of square section capable of being drawn down into
wire in the usual way. Technically, the Elmore process is
1 For analysis of Elmore copper by the author, see p. 55.
59
PRACTICAL ELECTRO-CHEMISTRY
a success ; commercially, it has been in most cases a failure
owing to reckless financing.
A modification of the Elmore process consists in the use
of a small hammer continuously tapping the metal as it ia
deposited, and consolidating it much as does the agate
burnisher.
A different method is that of Thofern, who causes the
electrolyte to play on the surtace of the cathode in jets. By
this means it is said that a current density of 50-100 amperes-
per square foot can be used in place of 10-20, common in
ordinary copper refining ; also it is stated that the copper is-
consolidated, and is deposited in felted microscopic filaments.
Details of a similar process are given in a patent by
Graham (Eng. Pat. 986 of 1896). In this specification it
is proposed to deliver the electrolyte under a head of 1-2
feet in jets | inch in diameter, at a distance of 1J inches
from the surface of the cathode. It is alleged that a current
density of 300 amperes per square foot may be used within
the area influenced by each jet, which is found to have an
effective radius of about 5 inches. The Dumoulin process,
in which the cathode rotates pressing against sheepskin
rubbers, belongs to the same class.
The Cowper-Coles process is one of the most successful
attempts to solve the problem of depositing copper in a
smooth continuous sheet so that it can be used at once with-
out fusing or reworking. It consists in depositing copper
on a cathode rotating with a peripheral speed of about 1,000
feet per minute in a hot solution of copper sulphate fairly
concentrated and rapidly circulated. Under these conditions
a current density (e.g. 200 amperes per sq. ft.) far greater
than that ordinarily used can be employed. With a stationary
cathode the copper deposited by so dense a current will be
loose, porous and mechanically worthless ; with a rapidly
rotating cathode, the other conditions being maintained,
a firm coherent sheet of copper is produced, pure and with
excellent mechanical properties.
A similar improvement in the current density permissible
has been observed by Dr. F. M. Perkin in small scale experi-
60
COPPER
merits on the deposition of iron, nickel and cobalt, the
cathode being rotated at a high velocity.
It must be noted that no authentic information is forth-
coming as to whether these plans have actually been worked
successfully on a manufacturing scale, but they merit
attention because a rush of liquid directed against the
surface on which the metal is being deposited is more likely
to prevent local impoverishment of the electrolyte in copper
than is any ordinary method of circulation ; similarly, a
slight but constant pressure and attrition may tend to keep
the metal smooth ; a relatively small pressure is certainly
effective in the Elmore process, and it would be rash to deny
that the same result may be attained by the use of a jet
of liquid. In like manner the friction of the revolving
cathode against the electrolyte in the Cowper-Coles process
may attain the same end. 1 Thus there is a primd facie case
for methods of this kind which warrants further experiment.
Quite apart from the consolidation of the copper, any device
which allows a high current density to be used is worthy
consideration, because the output of copper for a given stock
carried and for a given number of cells is proportionately
increased, and the money advantage thus secured (cf . p. 36)
is evident enough.
COST OF ELECTROLYTIC COPPER REFINING
This is a matter of ordinary calculation when the site,
material, cost of labour and of power are considered. But
certain of the factors are interdependent, and a very notable
attempt has been made to correlate them by Mr. Arnold Philip
in the latest edition of Electro-plating and Electro-refining
by Watt and Philip. The data are not altogether sufficient
for this purpose, but taking them as they are the attempt is
interesting, and may best be studied in the book cited. As an
instance of what has been done in a special case Badt's
estimate may be quoted. It is rather old (published 1892),
1 The Cowper-Coles process has already been tried on a considerable
scale ; the product is of good quality.
61
PRACTICAL ELECTRO-CHEMISTRY
but is worthy of attention as being probably based on the
results of actual manufacture.
Output of 5,357 tons of copper per year-
Buildings .
Pipes
Shtt lead lining ' .00 f 8,400
Lead burning
Steam injector .
Dynamos
Steam engine and shafting
Electrolyte
Conductors .
28,900
The estimate is approximate, and is given here merely as
a cruide ; it relates simply to the cost of the plant, and does
not touch the question of running expenses. These can be
readily computed from the ordinary data for cost of power,
management and the like which are common to many
industries. To discuss such matters, which are purely
subsidiary and can be worked out by any intelligent clerk,
would be foreign to my purpose.
THE ELECTROLYTIC WINNING OF COPPER
The electrolytic winning of copper stands on a very
different footing from its electrolytic refining. Some twenty
years ago the great success which even then could be seen
to be attainable in the refining of copper by electrolytic
means led to efforts being made to use a product much
cruder than ordinary crude copper as a raw material. In
the usual process of copper smelting the metal is separated
from* the gangue accompanying its ores by taking advantage
of the ease with which copper sulphide is formed, and of the
comparative stability of that sulphide and of its insolubility
in a siliceous slag. These properties are utilised by smelting
ores containing copper in such a manner as to form a matte
62
COPPER
containing approximately equal parts by weight of copper r
iron, and sulphur, corresponding nearly in composition with
pure copper pyrites (Cu 2 SFe 2 S 3 ). This matte, called " coarse
metal," is sufficiently coherent and conductive to permit it
to be cast into plates and used as the anode of an electroly-
tic cell. The quantity of impurities (iron and sulphur) is r
however, so great that the uniform dissolution of the anode
soon ceases, its surface becomes protected by a coating of
sulphur, and the electrolyte is rapidly contaminated with
iron. Coarse metal being unsuitable, a more advanced
product of the dry smelting of copper was tried, viz. " white
metal," which is essentially cuprous sulphide (Cu 2 S). This has
also been found wanting, the attack being irregular and the
quantity of separated sulphur excessive. Ultimately, after
the expenditure of much time and money, all these attempts
have been abandoned, and I do not propose to occupy
space with their description and discussion. 1
More recent and more nearly successful methods have
been devised on different lines. Instead of smelting copper
ore to a matte and using this as an anode, the ore itself is
extracted by a suitable solvent and the solution containing
copper is electrolysed with an insoluble anode. It must be
observed that in this case the electrical energy is not used
merely to transfer metallic copper already existing at the
anode to the cathode, and there deposit it precisely in the
same condition (save for the absence of impurities) as that
in which it was at the anode. This operation, as has been
already shown (p. 36), requires an indefinitely small amount
of energy. The reduction of copper from its salts, however,
needs a very appreciable quantity of energy, which must be
furnished by the current. Thus the ultimate products
of a solution of copper sulphate, electrolysed with insoluble
anodes, are copper, oxygen and dilute sulphuric acid ; the
1 The Marchese process, using anodes of copper matte, was tried
on a considerable scale and with great ingenuity. It failed at Casarza
utterly, but is said to be used in a modified form by Nicolajew at
Nishni-Novgorod. If this be true, the modifications must be
radical, because the original process was faulty in principle.
63
PRACTICAL ELECTRO-CHEMISTRY
requisite energy is therefore that represented by the heat of
combination of Cu and to form CuO, and of CuO and
H 2 S0 4 Aq to form CuS0 4 Aq. That is 63-5 grammes of copper
uniting with 16 grammes of oxygen liberate 37-16 Cal., and
the resulting CuO dissolved in dilute sulphuric acid liberates
18-80 Cal. To perform the decomposition into Cu. and
dilute H 2 S0 4 , 37-16+ 18-80 - 55-96 Cal. are needed.
Now assuming that the decomposition of copper sulphate
takes place (as it does) in accordance with Faraday's law,
63-5 1
each gramme equivalent of copper, i.e. grammes, needs
96,540 coulombs for its liberation, i.e. 63-5 grammes of
copper require 2 x 96,540 (= 193,080) coulombs. But the
heat representing the energy necessary to liberate by elec-
trolysis 63-5 grammes of copper from an aqueous solution
of its sulphate is 55-96 Cal. ; this is equivalent to 233,167
joules ; therefore, in order to yield this amount of electrical
energy, 193,080 coulombs must be delivered at a pressure
of 1-2 volts. 2 The maximum possible output of copper per
horse power hour is therefore 735 grammes. This is equiva-
lent to 38-9 pounds per horse power per 24 hours. Thus the
process differs radically from copper refining, in which, as
has been shown on p. 36, any desired output can theoretically
be obtained with an indefinitely small expenditure of energy,
and in which as much as 93 pounds per horse power per
twenty-four hours may be obtained in practice. To this
calculated minimum expenditure of energy for reducing the
1 Confusion constantly arises from the fact that the number of
units of electrical quantity (coulombs) needed for the liberation of
an element is always reckoned on the gramme equivalent of that
element, whereas the heat of combination of that element is
reckoned on its gramme atom. For a monovalent element these
are identical, but for a divalent element, such as copper in the
cupric state, the gramme atom represents two gramme equivalents
of the metal.
2 By actual experiment in my laboratory the minimum pressure
necessary for the deposition of copper from copper sulphate, using
an insoluble anode, is 1-375 volts.
64
COPPER
copper there must be added certain extra quantities common
to all electrolytic processes, which are needed for overcoming
the resistance of the leads and that of the electrolyte (as
distinct from that corresponding with the heat of combina-
tion of the substances separated). It follows that the mini-
mum working voltage of a copper-reducing plant will be
about 1-5 volts, and the output per horse power hour 585- &
grammes of copper, i.e. 30-9 pounds per horse power acting
for twenty-four hours ; hence a plant of 1,000 H.P. would
deposit 5,040 tons of copper per year if run day and night
for 365 days. Given water power at a cost of 2 per
horse power year, the cost for power alone for winning one
ton of copper is 7s. lid. ; and if steam power be used at
9 16s. per horse power year, each ton of copper will cost
1 19s. Wd. to win. (These figures may be compared with
those for the refining of copper given on p. 38.) This very
moderate expense warrants the idea that an electrolytic
process for winning copper from its ores should be exceed-
ingly remunerative.
But the cost of the power required is not the largest
part of the expense. The roasting of an ore containing
the copper as pyrites is necessary in most processes, and
in all the need for leaching out the ore occurs. The solvent
usually becomes charged with matter other than copper
extracted in the leaching process, and has to be purified or
renewed at fairly frequent intervals. The upkeep of the
depositing vats, electrodes and diaphragms is a heavy item,
and the risk of obtaining impure copper or bad and non-
coherent deposits, is considerable. Hence the cost of the
energy required, though important, is not of such extreme
moment as to give a water-power plant an overwhelming
advantage over one using coal.
The processes giving greatest promise of commercial
success in the electrolytic winning of copper from its ores
are as follows :
PRACTICAL ELECTRO-CHEMISTRY
THE SIEMENS-HALSKE PROCESS
This process depends on the extraction of copper from
its ore by a solution of ferric sulphate, which is thereby
reduced to ferrous sulphate, the deposition of the copper
thus dissolved by passage of the liquor through the cathode
compartment of an electrolytic cell, and the oxidation of
the ferrous sulphate by subsequent passage of the liquor
through the anode compartment. The regenerated liquor
is sent back to extract a further quantity of copper from a
fresh portion of ore.
The details of the scheme of working first proposed may
be stated. An ore containing copper as pyrites is roasted
at a low temperature so as to oxidise the sulphide of iron
which it contains to ferric oxide, and to free the cuprous
sulphide originally forming a constituent of copper pyrites
(CuaSFesSs) in the ore. In the course of this roasting,
part of the cuprous sulphide is oxidised to cupric sulphate
(CuS0 4 ) ; this is no disadvantage, as that part of the copper
is at once rendered soluble in water, irrespective of the
solvent action of the ferric sulphate subsequently used for
leaching. The sulphur dioxide (S0 2 ) given off in roasting
may be used for making vitriol, which is needed for acidula-
ting the leaching liquor. In this case the roasting is effected
in Gerstenhofer kilns, which are narrow vertical structures
down which the ore passes, meeting a limited supply of air
on its way, and thus generating gases sufficiently rich in
S0 2 to be practically available for vitriol making. The
roasted ore is placed in leaching tanks and extracted sys-
tematically ; by this is meant that fresh liquor always
conies in contact with nearly exhausted ore, and nearly
saturated liquor with fresh ore containing its full percentage
of copper. The copper already existing in the roasted ore
as sulphate dissolves as such ; copper existing as cuprous
sulphide is also dissolved by the action of the ferric sulphate,
which may be represented thus
Cu 2 S + 2 Fe 2 (S0 4 ) 3 = 2 CuS0 4 + 4 FeS0 4 + S
Cuprous Ferric Cupric Ferrous Sulphur
sulphide sulphate sulphate sulphate
66
COPPER
When a solution containing cupric sulphate and ferrous
sulphate and acid with sulphuric acid is electrolysed, copper
is deposited to the exclusion of iron. If this electrolysis
be performed in a cell without a porous diaphragm, the
ferrous sulphate is oxidised at the anode to ferric sulphate,
and reduced again at the cathode to ferrous sulphate. The
energy represented by these changes is provided by the
current, and appears as heat, which is lost. Thus it is
desirable to keep the liquor at the anode separate from
that at the cathode, and it is also necessary on account of
the fact that the liquor to be returned to the leaching vats
must contain its iron as ferric sulphate.
The process as thus described seems satisfactory enough,
but in working serious difficulties are encountered. Selec-
tive roasting of the ore is not an easy matter ; it must be
done slowly, at a low temperature, and with constant
stirring ; these are somewhat expensive conditions of work-
ing. The leaching needs much attention, and the leached
liquors may be muddy with basic iron salts and require
filtration ; an ordinary iron filter press is not adapted for
liquors containing copper and iron salts, as the frames and
plates are attacked ; wooden presses are needed, and these
wear rapidly. The anodes must be insoluble and with-
stand the disintegrating action of the current. This point
is of great importance in many electrolytic operations, and
it cannot be said that complete success has yet been attained
in devising a permanent anode. Platinum is too costly
for any ordinarj^ process. All other commercial metals
are attacked. Ferro-silicon, which is a difficultly attackable
substance, has been suggested, but does not appear to
have proved successful in practice. In almost all cases
carbon is the only substance which can be employed with
fair results. The quality of carbons prepared for electrical
and electrolytical purposes varies considerably, but even
the best are eventually destroyed. The choice of a dia-
phragm is even more difficult than that of an anode.
In the original arrangement a porous cell or membrane
was employed, the disposition of the various parts being
67
PRACTICAL ELECTRO-CHEMISTRY
such as is shown diagrammatically in the appended figure
(Fig 14). c, c, c are the cathode compartments of the
three cells shown; they are separated from the anode
compartments A, A, A by the porous partitions B, B, E.
Each of the cathode compartments is fed with a solution
of cupric sulphate and ferrous sulphate supplied by pipes
B, B, B, conveying the liquor through the series of cells. A
portion of the copper in the liquor is deposited on each of
the cathodes K, K, K. Seeing that the liquor as it enters the
first cell contains more copper than when it leaves it, its
specific gravity is higher at the point of entrance than at
that of exit, and thus the decrease of the content of copper
corresponds pari passu with the alteration of specific gravity,
C E A
FIG. 14.
and the lighter liquor, poorer in copper, flows out through
the U end of the second tube B in the first cell down to the
bottom of the cathode compartment of the second cell c,
where the process of elimination of copper and specific
lightening of the liquid recurs. Therefore throughout the
series of cathode compartments the deposition of copper
proceeds step by step, the heavier, richer liquor always
entering at the bottom of the cell, and the poorer, lighter
liquor flowing away at the top.
Precisely the converse holds good with the anode com-
partments A, A, A. The liquor from the last of the cathode
compartments, nearly exhausted of copper, but containing
all its iron as ferrous sulphate, flows into the first of the
68
COPPER
anode compartments by the pipe D, and is there oxidised
at the anode L. The ferrous sulphate is converted into
ferric sulphate, the solution of which is specifically heavier
than that of the ferrous sulphate, and sinks in the anode
compartment, increasing in its content of ferric sulphate
and in specific gravity until it reaches the bottom, whence
it flows by the pipe D into the next anode compartment.
Thus the oxidation of the liquor is as systematic as is the
reduction of copper from it, and the ultimate product on
one side is a solution of ferrous sulphate containing a small
residuum of cupric sulphate, and on the other a solution of
u
B
FIG. 15.
ferric sulphate (still containing a small quantity of cupric
sulphate) ready for extracting a fresh portion of roasted
ore. The arrangement of pipes shown having upturned
ends is merely a device, such as those which are shown in
Figs. 8 and 9, p. 46, for allowing the flow of liquid
through the tanks to be irregular, or to be stopped alto-
gether, and started again without risk of any tank over-
flowing, or any syphon becoming empty and therefore
unable to perform its functions when the flow of liquid
begins again. The appended figure (Fig. 15) shows the
arrangement on a somewhat exaggerated scale for the sake
of clearness.
69
PRACTICAL ELECTRO-CHEMISTRY
The tank B is on a lower level than the tank A, and thus
liquid can flow through the syphon c. The original levels
of the liquid in the tanks are represented by the lines L, L.
The liquid in A flows into B until the level in each is altered,
and becomes that represented by the lines L 1 , L 1 . On reaching
these levels the syphon ceases to act, but the U-shaped
bend remains full, and the syphon again begins to work
when the level in A is raised by any fresh influx of liquid. Now
suppose by some irregularity or accident the level in B falls
again to L without there being any compensating influx
from A. The liquid only falls in the short upturned limb
of the syphon to an equal extent, and on the resumption
of a regular flow the short limb fills up again and the syphon
resumes its office. The longer the IT of the syphon the
greater may be the irregularities of flow without throwing
the syphon permanently out of action. This assumes that
the pipe forming the syphon is made wide enough to allow
the upturned end to fill quietly without enclosing air spaces,
which, when the syphon started again, might cause a pocket
of air at the top of the syphon and stop its working.
It will be noted that in the Siemens-Halske process the
energy necessary to deposit copper from copper sulphate
at the cathode is diminished by that afforded by the oxida-
tion of ferrous sulphate to ferric sulphate at the anode.
This saving of energy is secured by taking advantage of
the fact that the ore, even when roasted, is not a completely
oxidised body (for it contains copper as cuprous sulphide)
and is capable of effecting the reduction of ferric sulphate
to ferrous sulphate, thus providing a body capable of
oxidation with the production of energy at the anode. A
similar case is fully discussed and its quantitative relations
are computed in the description of the Hoepfner process
which is given in succeeding pages.
An estimate of the cost of a small plant for the Siemens-
Halske process has been published (J.S.C.I., 1892, 534).
It may be given as an example of the items to be considered
in calculations of this sort rather than as being of any
intrinsic value, for, as will be seen presently, the Siemens-
70
I
COPPER
Halske process has not hitherto proved commercially
successful.
The quantity of copper to be won is taken as one ton
per 24 hours, using an ore containing 4-4-5 per cent, of
Cu. The cost of the plant exclusive of buildings is reckoned
at 5,765 ; crushing machinery, 1,557 ; leaching plant,
3,057 total, 10,379.
The cost of working per 24 hours is calculated thus :
Interest on plant (10,379) at 5 per cent. 1-42
Depreciation at 10 per cent. ... 2-84
130 H.P. . 3-12
Labour (15 men at 2s.) 1 . . . 1-50
Interest on copper in baths . . . 0*50
Fuel for heating extracting solution . . 0*50
General expenses and supervision . . . 2-00
11-88
Thus the winning of one ton of copper cost, exclusive of
the cost of the ore, nearly 12. This expenditure is not
immoderate, and would be smaller if a larger plant were
employed. Nevertheless the process has not achieved
success, for the reasons stated below.
The difficulties experienced in obtaining suitable per-
manent anodes and diaphragms have led to several modi-
fications of the Siemens-Halske process.
In these the arrangement of electrodes and diaphragm
has been horizontal instead of vertical, and the diaphragm
has served not only as a separating membrane, but as a
slow filter. 'This alteration is exemplified by the accom-
panying sketch.
The vat A is separated into two parts by the horizontal
filter B, of felt or asbestos. In the lower part is the anode
c, and in the upper division is the cathode D. The anode
may be built up of carbon plates or rods, while the cathode
is a piece of copper sheet supported by a wooden frame-
work (not shown). The leached liquor is fed in at F and
1 The estimate is German, wherefore the low labour charge.
71
PRACTICAL ELECTRO-CHEMISTRY
drawn off at B, the rate of flow being so adjusted that
it passes continuously through the filtering partition B,
and is in contact with each electrode successively for a
time sufficient to allow of the deposition of the bulk of
the copper in the upper division and of the oxidation of
the ferrous sulphate in the lower division, whence it is led
back to the leaching tanks.
The circulation is thus from cathode to anode compart-
ment of a single electrolytic cell, and not through all the
cathode compartments of a number of cells and then through
FIG. 16.
all the anode compartments of the same cells, as in the
arrangement shown in Fig. 14, p. 68.
Several forms of apparatus having these characteristics,
viz. the horizontal electrodes and the completion of the
treatment of a given quantity of leaching liquor in a single
cell, have been patented, but in spite of all these attempts
no authentic account of a successful installation on a large
scale has been published, and if a process of the kind is
being worked it is kept secret.
THE HOEPFNER PROCESS
Two chief underlying ideas may be traced in this pro-
cess. The first is to extract copper from its ores in which
the metal exists as sulphide by a solvent which shall extract
the copper from the unroasted ore. The second is to deposit
72
COPPER
copper from its cuprous salts instead of from its cupric
salts. This latter idea may be profitably considered irre-
spective of any particular process. In the first place it is
evident that cuprous chloride (Cu 2 CU) in which the copper
is monovalent contains twice as much copper per unit
weight of chlorine as does cupric chloride (CuCl 2 ). There-
fore the number of coulombs necessary to decompose
134-5 grammes of CuCl 2 and yield 63-5 grammes of copper
will decompose 198 grammes of Cu 2 Cl 2 and will yield 127
grammes of copper. In other words, a current of one
ampere acting for one hour will deposit 1-1827 grammes
of copper from cupric chloride, and 1-1827 x 2 x 2-3654
grammes of copper from cuprous chloride.
It has been shown above that there is a substantial
commercial advantage to be gained by using a high current
density, because the quantity of copper turned out per
unit of copper locked up and per unit of plant is thereby
increased. The limiting current density is set by the diffi-
culty of obtaining copper in a sound, coherent and pure
state when the current density exceeds a certain modest
value, e.g. 10 amperes per square foot. Now assuming
that ceteris paribus the same current density can be used
with a cuprous as with a cupric solution, 1 it follows that
with a given stock of copper, and with a given plant, twice
as much copper can be reduced from the cuprous as from
the cupric state with the same current. But it must not
be assumed that twice as much copper can be reduced
with the expenditure of the same amount of energy. This
needs separate inquiry. Thus the heat of formation of one
gramme molecule (134-5 grammes) of cupric chloride (CuCl 2 )
is 51-63 Cal. Hence to liberate 63-5 grammes of copper
from cupric chloride requires 51-63 Cal., i.e. 215,125 joules.
But the flow of 2 x 96,540 coulombs will deposit 63-5
grammes of copper from a cupric salt. Therefore these
1 This is an assumption, not a demonstrated fact. Like many
other questions in the electrolytic winning of copper, this point is
in need of experimental investigation.
73
PRACTICAL ELECTRO-CHEMISTRY
215 125
coulombs must be delivered at a pressure of - volts =
. x y
1-114 volts.
But the heat of formation of one gramme molecule (198
grammes) of cuprous chloride (Cu 2 Cl 2 ) is 65-75 Cal. Hence
to liberate 2 x 63-5 grammes of copper from cuprous
chloride requires 65-75 Cal., i.e. 273,958 joules. But the
flow of 96,540 coulombs will deposit 63*5 grammes of
copper from a cuprous salt and 2 x 96,540 coulombs must
flow to deposit 2 x 63' 5 grammes of copper. Therefore
273,958
the coulombs must be delivered at a pressure of -
2x 96,540
volts - 1-419 volts.
Thus, although it is true that a given current deposits
twice as much copper from a cuprous as from a cupric
solution, yet it requires per molecule of salt decomposed
a higher voltage in the proportion of 1-419 volts to 1-114
volts. That is, the total energy required per unit weight
of copper liberated from cuprous chloride is - of that
2x1-114
needed per unit weight of copper liberated from cupric
chloride, i.e. approximately . Of course the same result
25
is arrived at by considering directly the heats of formation
of cuprous and cupric chloride, remembering that in the
former each molecule contains twice the weight of copper
present in a molecule of the latter. The foregoing calcula-
tion serves, however, to show the method by which compu-
tations of this kind may be made, and also to illustrate the
fallacy of referring the efficiency of a given process solely
to its output per coulomb (or, if over a given time, per
ampere), ignoring the true efficiency, i.e. the output per
unit of energy, this being stated in calories, joules, foot
pounds or other convenient unit.
In the particular case now under discussion, the mere
statement of the output per coulomb would imply that a
process using a solution of cuprous chloride would "be twice
74
COPPER
as efficient as a process using a cupric solution. In reality,
however, it is about one and a half times as efficient, taking
as a criterion the minimum possible consumption of energy.
Its real claim to consideration (assuming practical diffi-
culties to be overcome) is in the greater output of copper
per unit of plant and of copper locked up, always provided
that the maximum current density at which good coherent
copper can be deposited is as high as that attainable with
the use of cupric solutions.
These principles having being discussed, we may return
to a consideration of the process illustrating them.
The Hoepfner process, as originally devised, was described
by the inventor in a paper read before the Upper Silesian
Society of Applied Chemistry, and transcribed into the
Zeits. /. angewandte Chemie, 1891, p. 160. The gist of this
description, together with any necessary comments, may
be given briefly here.
The cells are divided by a porous partition into anode
and cathode compartments. Through all the cathode
compartments of a given group of cells flows a solution
containing cuprous chloride dissolved in a solution of sodium
chloride or calcium chloride. Copper is deposited from
this solution in double the quantity that would be deposited
from a cupric solution by the same current. The liquor,
having passed through the whole set of cathode compart-
ments, floAvs away nearly free from copper. In similar
manner a solution of cuprous chloride is supplied to the
anode compartments. Now at the anodes chlorine appears
in quantity corresponding with the copper deposited in
the cathode compartments. 'This chlorine, however, does
not become free, but combines with the cuprous chloride
in the anode compartments to form cupric chloride. This
reaction in itself tends to produce a current in the same
direction as the current used for electrolysis, and thus the
necessary minimum voltage is diminished. The minimum
voltage for a cell having cuprous chloride in both anode
and cathode compartments (the two being separated by, a
Vporous diaphragm) may be calculated. The calculation
75
PRACTICAL ELECTRO-CHEMISTRY
resolves itself into reckoning the voltage corresponding
with the heat of combination of copper and chlorine to
form cuprous chloride, minus that of cuprous chloride and
chlorine to form cupric chloride ;
i.e. Cu 2 + a a = Cu 2 Cl 2 65-75 Cal.
and Cu 2 Cl 2 + C1 2 = 2 CuCl 2 32 Cat.
33-75 Cal.
Therefore the total energy to be provided from with-
out is 33-75 Cal. - 140,625 joules for 2 x 63-5 grammes
of copper deposited from the cuprous chloride solution.
Seeing that 2 x 96,540 coulombs must flow in order to
deposit 2 x 63-5 grammes of copper from a cuprous solu-
tion, it follows that the current must have a voltage of
140,625
volt =0-73 volt.
2 x 96,540
In the foregoing calculation such thermal changes as
attend the removal of cuprous chloride at the cathode
from its solution in brine or calcium chloride solution, and
the production of cupric chloride (having a high heat of
dissolution in water) in solution at the anode, have been
intentionally neglected. Thus the main point stands out
clearly, viz. that by taking advantage of the power of
copper to form two chlorides the chlorination of cuprous
chloride can be caused to yield energy in the cell, and
thereby diminish substantially the quantity of energy
necessary to be impressed from without.
The energy required to reduce again the cupric chloride
to cuprous chloride, and by this means to economise the
electrical energy which has to be expended in the cell, is
afforded by the ore, which, being an unoxidised copper
sulphide, is capable of acting thus. Therefore the saving
of energy effected by taking advantage of the existence of
two chlorides of copper comes ultimately from the ore itself.
Just as a sulphide ore can be roasted in heaps by its own
heat of combustion and without the aid of extraneous fuel,
so can the same ore serve in great measure to go towards
76
COPPER
reducing copper which it contains to the metallic state.
These energy considerations are quite elementary, but are
often neglected or slurred over in dealing with electro-
metallurgical questions.
The cupric chloride formed in the anode compartments
during the systematic flow of a portion of the cuprous
extract from the ore through these compartments is returned
to the leaching tanks for extracting a fresh portion of the
ore ; there it acts on the cuprous sulphide in the ore accord-
ing to the equation
Cu 2 S + 2 CuCl 2 - 2 Cu 2 Cl 2 + S. 1
It will be remembered that the liquor which has passed
through the cathode compartments, though robbed of
its copper, contains untouched the sodium chloride or cal-
cium chloride used to keep the cuprous chloride in solution.
Now, if complete reduction to cuprous chloride occurs (as
it should) in the leaching vats, a quantity of cuprous chloride
equal to that originally starting from the leaching vats will
be regenerated. This will need the same quantity of
sodium chloride or calcium chloride to retain it in
solution as was requisite when the first solution was pre-
pared. Therefore the liquor from the cathode compart-
ments must be mixed with that from the anode compartments
in order to provide sufficient sodium chloride or calcium
chloride to hold the whole of the cuprous chloride in solution.
To take a concrete case for the sake of clearness : Suppose
a solution having a volume of 1 litre contains 2 gramme
molecules of Cu 2 Cl 2 and that this is kept in solution by 4
gramme molecules of NaCl. 2 Let half this solution pass
1 This equation has been disputed. Experiments in the author's
laboratory have, however, shown it to be substantially correct. It
must not be assumed, however, that a practicable process of leach-
ing on these lines can necessarily be realised. Completeness of
extraction depends largely on the fineness of the ore, the proportion
of solvent to ore, and the temperature at which the extraction is
conducted.
2 Whether these solubilities are possible or not is immaterial as.
far as the argument is concerned.
77
PRACTICAL ELECTRO-CHEMISTRY
through a cathode compartment and there deposit all its
copper. The half litre of solution then contains 2 gramme
molecules of sodium chloride. The other half of the solution
passing through the anode compartments is there chlorinated
and after this change contains 2 gramme molecules of
CuCl 2 and 2 gramme molecules of NaCl. Then passing to
the extracting tanks, it is reduced to Cu 2 Cl 2 , fresh copper
going into solution, and forms 2 gramme molecules of Cu 2 Cl 2 ,
which require ex hypothesi 4 gramme molecules of NaCl
for their solution ; but in the solution itself are only 2,
hence the 2 bereft of copper in the cathode liquor must be
supplied to make up the deficit.
It is evident that the process possesses some elements
of elasticity of working. If it were found, as is likely, that
impurities accumulated in the leaching solution to an incon-
venient extent, the liquor from the cathode compartments,
thoroughly freed from copper, could be thrown away and
replaced by clean water in which the requisite quantity of
salt to make an effective solvent for the cuprous chloride
had been dissolved. In this way purification could be
attained with the expenditure only of the sodium chloride,
and there need be no waste of copper, or necessity for work-
ing up a crude solution.
A subsidiary advantage claimed for the process is that
cupric chloride is a solvent for silver contained in the copper
ore ; thus
Ag 2 S + 2 CuCl 2 - Cu,Cl 2 + 2 AgCl + S.
The resulting silver chloride is fairly soluble in the solution
of cuprous chloride in sodium chloride or calcium chloride,
and from the solution the silver can be precipitated by well-
known means, e.g. treatment with metallic copper, before
the solution goes to the cathode or anode compartments.
When the silver has been separated, removal of other
impurities can be effected by precipitation with a limited
quantity of lime. This, which is a common operation in
wet metallurgical processes, can be easily carried out,
because cuprous oxide is a strong base, and all ordinary
78
COPPER
impurities are precipitated before its salts are decomposed,
when a base such as lime is added gradually.
The foregoing description is based on the facts set forth
in Hoepfner's original paper. In the same document he
proceeds to give an estimate of the cost of the plant and of
the fuel required in a works using this method. These are
here pretermitted, as they have not been realised in practice.
The nature of the difficulties encountered may be gathered
from the following abstract, appearing in the J. Soc. Chem.
Ind., 1895, p. 279, of a paper by E. Jensch (Chem. Zeit.,
1894, p. 1906).
" The Hoepfner process was used at Schwarzenburg
from August, 1891, to March, 1892, and in the Giessen and
Weidenau works. It was applied both to rich ores and
mattes, and to cuperiferous pyrites from the Sulitjelma mines
in Northern Norway, in which the copper percentage ranged
from 9-5 to 12-25, and that of iron from 32-6 to 34-5. The
ore was very finely crushed, so that 85 per cent, of the sam-
ple passed through a No. 200 and 96 per cent, passed a No.
100 sieve ; but some little trouble was caused by the block-
ing of the meshes by the fine powder. The leaching was
effected by means of a solution of cupric chloride in calcium
chloride, which latter (instead of brine) becomes the solvent
of the resulting cuprous chloride, the mixture being placed
in revolving wooden drums of 900 to 6,600 litres capacity.
The drums caused considerable difficulty by leakage, which
began when the temperature of the liquid was raised by
the admission of steam to hasten the reaction, and increased
with the rise of temperature and the growing percentage
of cuprous chloride, yet for obvious reasons lead and iron
vessels could not be used. With the rich materials three
or four extractions sufficed, but with the Sulitjelma ore,
although the first extraction removed half of the copper,
even ten or twelve teachings failed to extract the whole of
the remainder, partly on account of the large percentage
of iron present, partly owing to the increasing dilution of
the liquid. At the temperature of the reaction, magnetic
pyrites reacts with cupric chloride, giving equivalents of
79
PRACTICAL ELECTRO-CHEMISTRY
ferrous chloride, cuprous chloride, iron bisulphide and sul-
phur, while the resulting ferrous chloride reacts with another
quantity of cupric chloride to give ferric and cuprous chlo-
rides ; and iron pyrites reacts directly with cupric chloride
to give ferrous and cuprous chlorides and sulphur. For
this reason an excess of cupric chloride must be used in the
leaching solution. The slimes were filter-pressed at a tem-
perature of 40 to 50 C. in order to avoid the retention of
copper by them. The anodes were of paraffined carbon,
the cathodes thin copper plates, experiments with coppered
carbon cathodes having proved unsuccessful. Difficulties
with the parchment paper diaphragms were also met with."
The copper obtained by the Hoepfner process is said to
be of good quality, in spite of the fact that it is precipitated
from a somewhat impure solution. A published analysis
shows only traces of iron, arsenic, antimony and lead, nickel
and cobalt amounting to 0-0012 per cent, and molybdenum
0-0023 per cent.
One of the most serious difficulties of the Hoepfner pro-
cess has been the provision of refractory anodes and dia-
phragms. The patents taken out by Hoepfner in the years
immediately succeeding the original promulgation of his
process indicate this. He has suggested the use for anodes
of ferro-silicon, i.e. iron containing sufficient silicon (10-15
per cent.) to constitute a silicide which is less readily attacked
than iron and is still sufficiently conductive ; for diaphragms
he has advocated the use of mica plates joined together
(this being necessary because the price of fairly large pieces
of mica is high, and any piece over one foot square is practi-
cally unattainable) and perforated with numerous fine
holes so that the liquids to be separated may be in electro-
lytic contact and yet be prevented from commingling freely.
These almost desperate expedients indicate the heavy
mechanical difficulties with which the process has had to
contend. Having regard to all these things, the Hoepfner
process, in spite of its ingenuity and the soundness of the
principles on which it rests, must be pronounced a failure
up to the present.
80
COPPER
A process for obtaining copper from its ores electrolyti-
cally has been described by Keith in a paper read before
the American Institute of Electrical Engineers in 1902.
The ore, containing about 2 per cent, of copper, is roasted
and extracted with sulphuric acid (5-15 per cent, strength).
The solution is passed through a series of vats in which it
is electrolysed, and as the liquor is robbed of its copper on
its passage, so is the current density decreased, not by
diminishing the amperage of each vat, but by increasing
the surface of the electrodes. This process is strictly
scientific ; with a high and constant current density a liquor
poor in copper will be decomposed holus bolus, hydrogen as
well as all metals electropositive to copper appearing pro-
miscuously at the cathode ; with a diminished current
density the proper selective deposition of copper which
makes it possible to precipitate that metal pure and with a.
good current efficiency even from an impure and weak
solution will be maintained.
The pressure corresponding with that necessary for the
reduction of the copper salt to metallic copper is given by
the author at 1/6 volts, somewhat greater than the calcu-
lated figure (1*2 volts) and than that observed by the author
(1-375 volts). Both anodes and cathodes are of lead ; the
anodes naturally become covered with lead peroxide in the
course of electrolysis. When the cathodes have received
a film of copper the latter is stripped and serves as a cathode*
on which copper can be deposited until a merchantable
thickness has been attained. In the operation of roasting;
referred to above, some iron present in the ore is left im
a soluble condition, and this dissolving yields ferrous or
ferric sulphate. Either salt is a source of loss, because
each will suffer alternate oxidation and reduction at the
electrodes with corresponding useless expenditure of
energy. This process is rational, but has not yet been
made a commercial success ; the conditions under which it
was tried appear to have been unfortunate, because the
total content of copper in the ore was low, 2 per cent.
The author, in the light of present experience and of
81 G
PRACTICAL ELECTRO-CHEMISTRY
his own observations, is of opinion that there is no particu-
lar difficulty in extracting copper from its ores electroly-
tically. The obstacle to success has been that inventors,
fascinated with the beautiful flexibility of electrolytic
methods have been apt to overlook practical considerations,
and in endeavouring to obtain at a stroke and with ideal ex-
actness very difficult metallurgical separations, have ignored
more simple and trustworthy methods. There has been
some delay in consequence, but of the ultimate success of
the extractions of copper by electrolytic means no reasonable
doubt can be entertained.
A case in which some success has already been reached is
'Cited by Coroda, who states that at Papenburg a Rio Tinto ore
containing 3-4 per cent. Cu has been successfully worked.
A process presenting some novelty of idea has been pa-
tented by the Illinois Reduction Co., by which a sulphide
ore is treated with manganese dioxide and sulphuric acid
under heat and pressure. The sulphate solution is electro-
lysed and the sulphuric acid used for the next operation.
It is evident that in order to make the process commercially
practicable the manganese must be recovered in some way.
There is no evidence that the method has actually been
worked. The Carmichael process may also be mentioned.
The ore is leached with acid in the ordinary way and the
electrolyte is treated with S0 2 , which serves to agitate the
liquid, to prevent the peroxidation of the anodes which are
of lead, and by its oxidation to contribute a small amount
of energy which reduces that which has to be supplied
electrically for winning the copper. The sulphurous acid
also serves to neutralise lime and other bases present in
the ore more cheaply than can be effected by sulphuric acid.
Before dismissing the subject of winning copper directly
from its ores by extraction with some solvent which can
be regenerated and by electrolytic treatment of the resulting
solution, a brief description must be given of an ingenious
device due to Cohen (who has described it in the Zeitschrift
fur Elektrochemie, 1895, p. 25), by which he seeks to avoid
the necessity for a diaphragm. The arrangement is shown
82
COPPER
in Fig. 17. There is no porous diaphragm ; the cathode
K is about half the length of the anode A, and the latter at
its lower end is separated from the rest of the tank by the
short vertical partition c. Cuprous chloride solution is
fed in by the pipe B, and flowing down is partly robbed of
its copper in passing over the cathode K. On reaching the
anode A the cuprous chloride still remaining in solution is
oxidised to cupric chloride, and its specific gravity is thus
increased, wherefore it slides down the anode and collects
in the sump E formed by the partition c. From this it is
syphoned off by the pipe D, and is available for extracting
another portion of the ore. The weak point of this arrange-
! 'Cud.
FIG. 17.
ment is that the more completely the cuprous chloride is
robbed of its copper (as is desirable) at the cathode, the
smaller quantity of cuprous chloride remains in solution
to be oxidised at the anode ; under the best conceivable
conditions only half the copper is deposited at the cathode,
leaving an equal quantity to be oxidised from the cuprous
to the cupric state at the anode. But, seeing that the
upper parts of the two electrodes are not separated, more
than half the cuprous chloride is likely to escape decompo-
sition at the cathode and pass directly to the anode. It
cannot be oxidised there by the action of the current,
because the amount of chemical action at the anode is
83
PRACTICAL ELECTRO-CHEMISTRY
equivalent to that at the cathode. Therefore a considerable
proportion of the cuprous chloride circulates idly through
the extracting vats and electrolytic tanks. Moreover, it
is highly doubtful whether the difference of specific gravity
of the two solutions is large enough to ensure the cathode
and anode liquids remaining fairly separate. Altogether
the device is more ingenious than practicable.
The example of Hoepfner in using cuprous salts from
which to deposit copper has been followed by Douglas, who
proposes roasting sulphide ores to sulphate, extracting
with a solution of sodium chloride, reducing the resulting
cupric chloride to cuprous chloride by means of sulphur
dioxide, and electrolysing the cuprous chloride (which
may be as a paste if the quantity of sodium chloride is
insufficient to keep it in solution) depositing copper at the
cathode, and collecting the chlorine given off at the anode
for use as such. There is no reason to suppose that this
is more than a paper process.
From the foregoing description of the Siemens-Halske,
the Hoepfner and the Keith processes, the only methods
which have been fairly tried on a manufacturing scale, it
is evident that the electrolytic winning of copper, as distinct
from its refining, has not yet been remuneratively accom-
plished. That it will be achieved in the near future is
probable enough ; meanwhile it presents an excellent field
for invention. What is wanted is not so much a totally
new device as a well-schemed plant, embodying perhaps
nothing but what is common knowledge, but planned so
as to be fairly permanent as a whole, and with its perishable
parts easily and cheaply renewable.
84
LEAD
IT is not probable that a successful method of winning
lead from its ores by means of electrolysis will be devised.
An attempt in this direction has been made at Niagara
Falls, where a process is at work in which galena, separated
mechanically from, gangue, is reduced electrolytically to
spongy lead. The galena is about 75 per cent, pure, and is as
free as possible from silver. The cells consist of a number of
shallow saucers made of antimonial lead, and piled one upon
the other to form a column. Each cell is insulated from
its neighbour by a rubber ring, which also serves to make
the joint between them. The crushed galena is placed on
Galena. " Lead Trays.
^ :;; ^^^^^^^^a^ Eubber Rings<
2L*
^
FIG. 18.
the bottom of each pan or saucer, and the whole set is run in
series, the outer surface of the bottom of each pan being a
cathode, and the inner surface with its charge of galena
being an anode. The electrolyte is dilute sulphuric acid.
The whole arrangement is represented diagrammatically
in the figure.
The sulphur appears as H 2 S, which is not utilised at pre-
sent. The cathode product, spongy lead, is washed free
from residual gangue, and either used for accumulator plates
or is roasted for the production of red lead or litharge.
85
PRACTICAL ELECTRO-CHEMISTRY
The prospect of any such process proving successful is
remote, because lead is an easily reducible metal, fusible
at a low temperature, and of low price. Certain attempts
have been made to refine crude lead, and these have met
with a qualified success. It happens that the refining of
lead by ordinary metallurgical processes has been brought
so nearly to perfection that commercial lead, such as is
used for the commonest purposes plumbing, roofing and
the like is almost chemically pure, as may be seen from
the following typical analysis :
Per cent.
Lead 99-9837
Copper 0-0014
Antimony ... 0-0037
Zinc 0-0016
Iron 0-0016
Silver . 0-0080
100-0000
Not only is the lead all but absolutely free from the com-
moner metals, but it contains only a small quantity less
than y^j. of 1 per cent. of the most characteristic and
valuable impurity, namely silver. In the sample, the
composition of which is quoted above, there is 0-008 per
cent, of silver, i.e. 5 ounces per ton. In many commercial
samples of lead there is even a smaller amount, e.g. 2-3
ounces per ton. Thus it is evident that by existing methods
of refining lead can be obtained of a quality good enough
for all ordinary purposes, and at the same time free from
the chief foreign constituent worth recovering. From this
it follows that any electrolytic process is not likely to achieve
better results, and its only chance of adoption lies in the
possibility of its being cheaper than the usual processes.
It will be seen that the refining of lead stands on a totally
different footing from that of copper (p. 31 et seq.). There
a product (copper almost absolutely pure) is obtained which
is procurable in no other practicable way and for which
there exists a large demand ; with lead, on the other hand,
86
LEAD
the product can be at best only insignificantly more nearly
pure, and can fulfil no demand not already fully satisfied
by the ordinary metal of commerce. Therefore, whereas the
extension of electrolytic copper refining and the ultimate
extinction of dry processes are certain, the future adoption
of electrolytic lead refining on any considerable scale is
inherently improbable, unless an appreciable saving in cost
of refining can be proved.
KEITH'S PROCESS
This process, although no longer in use, is worthy of
brief description in that it illustrates the lines on which a
refining process may be worked, provided the cost can be
kept within reasonable limits. Crude lead containing
96-97 per cent, of Pb was used as the raw material.
The following analysis will serve to show the composition
of lead of this class :
Per cent.
Lead 96-36
Antimony . . . . . . T07
Arsenic 1-22
Copper 0-31
Silver 0-55
Zinc, iron, etc. 0-49
100-00
This crude lead is cast into plates to serve as anodes.
These are enclosed in bags of muslin to retain the anode
sludge containing the silver. The electrolyte consists of
a solution of lead acetate or of lead sulphate dissolved in
sodium acetate. The cathodes are thin sheets of pure lead,
and on them the lead is deposited as loosely adherent crystals
which fall to the bottom of the depositing cell and are re-
moved from time to time. The anode sludge remains in
the muslin bags and is worked up for its silver. The lead
crystals have to be squeezed into blocks and fused in the
presence of a little charcoal and run into ingots. A certain
87
PRACTICAL ELECTRO-CHEMISTRY
amount of slagging and loss is apt to occur in this operation.
Such loss can be minimised by adding the lead sponge to
lead already molten, instead of fusing it per se. There is
a certain limited demand for spongy lead for accumulator
plates, and for this purpose the lead deposited electroly-
tically in a mass of loose crystals is well adapted. If, how-
ever, electrolytic lead refining is ever to be established on
a large scale, this outlet would be much too small to take
any considerable fraction of the lead produced, and some
plan of fusing the metal and running it into ingots must be
adopted.
One of the best attempts which have been made to refine
lead electrolytically is that due to Tommasi. Like other
methods, it has not yet reached a manufacturing status,
but is nevertheless worthy of a brief description.
The electrolytic cell a, shown in vertical section (Fig.
19), contains two lead anodes b, 6, which may be either
cast plates or powdered lead packed in a perforated case 1
Between the anodes is a large thin disc c (shown in vertical
section in the figure, and therefore appearing as a line),
made of copper or aluminium bronze and having its centre
above the top of the cell. It is mounted on a spindle pro-
vided with a rubbing contact, and is made the cathode.
The disc is rotated, and is alternately immersed in and
withdrawn from the electrolyte. On each side of the disc
is a scraper, which detaches the loose lead crystals deposited
during the passage of the disc through the electrolyte and
also aids in depolarising it. The finely divided lead falls
into gutters, by which it is conveyed to a sieve. Here it
is drained and washed. The lead is compressed and fused
into ingots, a little charcoal being used to hinder oxidation.
Ihe electrolyte is said to be a solution of lead acetate and
dmni or potassium acetate, to which certain substances
e added to prevent the formation of lead peroxide at the
The nature of these substances is kept secret It
that they are fairly cheap and easily oxidisable
1 A device of doubtful utility.
88
LEAD
organic substances which serve to reduce any lead peroxide
that may tend to be deposited. This tendency to deposit
a portion of the lead as peroxide at the anode is a standing
difficulty in lead refining. It is objectionable, not only
because the lead is deposited in the wrong form at the wrong
place, but also because it gives rise to a back E.M.F. which
increases the voltage needed for the decomposition of the
electrolyte. 1 Supposing the formation of lead peroxide
FIG. 19.
to be prevented from occurring by means other than the
addition of reducing substances to the electrolyte, there
will be no necessary consumption of energy in the transfer-
ence of lead from the anode to the cathode. The only expen-
diture of energy required will be that needed to overcome
the resistance of the electrolyte. In fact, the general condi-
tions are identical with those obtaining in copper refining,
1 It must not be supposed that this back pressure, which is of a
static character, increases the consumption of energy necessary for
the deposition of a given quantity of lead ; rather its occurrence
deranges the adjustment of the pressure necessary for electrolysis.
89
PRACTICAL ELECTRO-CHEMISTRY
and the remarks made on p. 36 apply equally here. Now
it is evidently advantageous to keep the electrodes as close
together as possible, so as to reduce the resistance of the
cell. This is feasible with a revolving cathode, because
the deposited lead is continually removed by the scrapers,
and is thus prevented from forming irregular crystalline
growths which would bridge the space between the elec-
trodes. Tommasi calculates that with a distance of 2 cm.
between anode and cathode, and using a current density
of about 3 amperes per square decimetre (say 27 amperes
per square foot), the drop of pressure in each cell would be
0-75 volt, and that tor an output of 84 tons of lead per day
of 24 hours an installation of about 1,000 H.P. would be
required. Making all the usual charges for labour, fuel,
depreciation and the like, the cost of the process per ton
of refined lead is about 7s. By using water power this sum
may be reduced to about 4s. Taking the cost of casting
the anodes and working up the anode sludge for the recovery
of silver and bringing the precipitated lead into marketable
form at 2s. 6d. per ton of crude lead, the total cost will be
6s. Qd.-9s. 6d. per ton. This is to be compared with a working
cost of 24s., said to be incurred by the ordinary dry method
of refining and desilverising. The low cost estimated for
the Tommasi process can be covered by the value of even
a small amount of silver, e.g. 4 ounces per ton, extracted.
It must not be assumed, however, that the present dry pro-
cess, even if requiring an expenditure of 24s. per ton of
crude lead, cannot deal profitably with metal containing less
than 12 ounces of silver per ton. The average content of
silver in commercial refined lead, 2-3 ounces per ton, dis-
proves this at once. The reason, of course, is that for most
purposes the lead must in any case be refined, and the desil-
verising is an incident in the refining. Thus, the value of
the silver need not be so large as to cover the cost of refin-
ing ; the enhanced value of the refined lead is also to be
reckoned when considering whether a lead poor in silver
can be profitably treated. It is evident that the question
is wholly one of cost, and, in deciding for or against the
90
LEAD
Tommasi process, detailed estimates, based on large scale
experiments, would have to be compared with the actual
works cost of a modern refining plant working on the Parkes
or Pattinson system. As regards the production of spongy
lead, there is, as stated above, some prospect of useful
application of the Tommasi or some similar process. The
cost of 1 ton of spongy lead will not be greater (assuming
that the electrolytic process costs about as much as the dry
method) than that of 1 ton of ordinary refined lead say
11 10^. The cost of spongy lead obtained by any method
of chemical precipitation, such as with zinc, which is some-
times employed, will be considerably greater, e.g. 50 per
ton, both because the comparatively expensive acetate
of lead is used and because the zinc acetate formed is of
small commercial value. Any direct method of precipi-
tation will include the impurities of the zinc in the spongy
lead an undesirable outcome when the lead is to be used
for accumulator work, in which it is needed to be as pure
as possible. To avoid the inclusion of these impurities it
would be necessary to dissolve the zinc out of direct contact
with the lead salt in fact, in one compartment of a single
voltaic cell, short-circuited. The lead would be deposited
on the negative plate precisely as copper is deposited in a
Daniell cell. In short, the lead would be produced electroly-
tically in the most expensive way. Its cost would make
its use quite impracticable for all but very special purposes.
On the other hand,, at a price of about 11 105. per ton
there is no reason why spongy lead should not be used as
the starting-point in manufacturing oxides of lead (litharge
and red lead), and perhaps white lead (basic carbonate of
lead). Should such an outlet be found, electrolytic lead
may be profitably manufactured, irrespective of its utilisa-
tion as ordinary massive metal.
PRACTICAL ELECTRO-CHEMISTRY
PROCESSES USING A FUSED ELECTROLYTE
Lead may be refined by electrolysis conducted -with a
fused salt of lead, instead of an aqueous solution of a lead
salt, as the electrolyte. Difficulties caused by the deposition
of the lead in a spongy state from an aqueous electrolyte
naturally disappear when the electrolyte is fused and kept
at a temperature above the melting-point of lead. It also
appears that electrolysis can be conducted successfully with
a far higher current density in a fused electrolyte than in one
which is aqueous. This allows the apparatus to be smaller
for a given output of lead a matter of considerable import-
ance. Of ordinary salts of lead, the chloride is most gener-
ally suited for use as a fused electrolyte. It melts at 498 C.
= 928 F. and does not vaporise largely until considerably
above this temperature. It is relatively cheap, not unduly
corrosive and is a good conductor. The use of a fused
electrolyte for lead refining must, of course, be so arranged
that both the crude lead acting as anode and the refined
lead collecting on the cathode may be kept fused, that fresh
crude lead may be added, and the separated pure lead may
be removed continuously or from time to time, preferably
without interrupting the working of the plant. An ingeni-
ous apparatus has been designed by Borchers to meet these
requirements. It does not appear that this apparatus has
ever been tried on a manufacturing scale ; nevertheless, it
illustrates certain principles and ideas sufficiently well to
warrant a brief description. The chief features of the
apparatus are shown in the diagrammatic sketch given on
the opposite page.
A cast-iron vessel A, shown in cross section, is divided
electrically by the insulating joint B. The left-hand side
f the vessel (which serves as the anode) is not vertical, but
has a slope sufficient to "allow a series of groves in its face
) retain melted lead and to allow this lead to flow down the
aide of the vessel terrace-fashion in a continuous stream.
The lead is put in through a hopper (not shown) at the top,
rawn off by an inverted syphon (not shown) at the
92
LEAD
bottom. The electrolyte filling the vessel is a mixture of
potassium chloride and sodium chloride in molecular pro-
portions, to which has been added lead oxy chloride. The
vessel is set in the flue of a furnace, so that its contents may
be kept fused. The only sensitive part is the insulating
joint, and this is water-cooled, so that it is protected from
the electrolyte by a crust of solidified salt. The part of the
vessel on the side of the insulating joint opposite the anode
serves, as the cathode, and in it the refined lead is deposited
and collected. This lead is drawn off by an inverted syphon
in manner similar to that used for the anode side. Using
this apparatus, Borchers has been able to employ as high a
current density as 10 amperes per square decimetre ( = 9&
amperes per square foot) even when the lead was rich in bis-
FIG. 20.
muth and it was sought to refine it and separate the bulk of it
from this valuable impurity. If the quantity of bismuth be
small, the enormous current density of 60 amperes per square
decimetre can, it is said, be adopted without impairing the
purity of the lead produced or endangering the apparatus.
Borchers also states that even with these high current
densities the requisite voltage is only 0-5 volt, and that thus
10 pounds of lead can be refined by an expenditure of energy
of 1 H.P. hour. Taking this as costing Id. for comparatively
small installations, one finds that the cost of refining is about
93
PRACTICAL ELECTRO-CHEMISTRY
1 per ton as far as the requisite energy is concerned. It
must be noted that it is by no means clear that this process
is adapted for dealing with argentiferous lead. Probably
with a moderate current density the silver would remain
unattacked and be concentrated in the residual anode lead.
The distribution of other metallic impurities common in crude
lead, and having to be provided for by any scheme of refining,
is also uncertain. Thus, speaking generally, it may be justly
said that, ingenious as is the apparatus, it and its action
require close and extensive study before it can be considered
as an improvement on existing methods of lead refining.
I am not aware that it has yet been put to practical use.
94
GOLD AND SILVER
THE ELECTROLYTIC EXTRACTION OF GOLD FROM
ITS ORES
GOLD almost always occurs as free metal in ordinary ores ;
its extraction, therefore, consists in acting on it with an
appropriate solvent which will not attack appreciably the
non-auriferous part of the ore. It is on this account that
the oldest of all extraction processes, that of amalgamation,
has been, and still is, largely and successfully employed.
Mercury is an excellent solvent for gold, and does not dis-
solve the oxides and sulphides of base metals or the earthly
gangue which accompany the gold. The reason why a
process of amalgamation is not always the best that can be
used for extracting gold is that the gold may be covered
with a film of sulphide or oxide of some other metal, which
may prevent its being brought into full contact with the
mercury, or it may be so finely divided that it may float in
the water carrying the powdered ore, and may thus equally
escape contact with the mercury. Further, devices to mix
the mercury intimately with the pulp of ore and water so
as to catch this finally divided gold are very apt to convert
the mercury itself into a " flour " so difficult to restore to
its normal condition that it is carried away and lost. These
and other difficulties make the use of mercury as a solvent
less ideal for the extraction of gold than would at first sight
appear. Free chlorine will dissolve gold, and is used in
a number of processes of " chlorination," which consist
essentially in treating the finely powdered ore with water into
which chlorine gas is led or in which it is generated by the
action of sulphuric acid on bleaching powder. The objections
to these processes are that other metals than gold are dis-
solved, and that the quantity of chlorine required is hugely
in excess of that strictly necessary to dissolve the gold.
95
PRACTICAL ELECTRO-CHEMISTRY
Potassium cyanide solution in the presence of oxygen will
dissolve gold, and is utilised in enormous quantity and with
the most complete practical success for treating ores,
especially those of the type found in the Rand goldfields of
South Africa. The success which has attended the use of
potassium cyanide is largely due to its property, when used
in sufficiently dilute solution, of dissolving gold rather than
other soluble matters ; this selective action tends to prevent
waste of the solvent. Even with cyanide, however, the
amount of solvent required, compared with that corres-
sponding chemically with the weight of gold to be dissolved,
is colossal.
Reflecting on the facts set forth in this preamble, inventors
have for years endeavoured to enhance the solvent powers of
the solvent which they have selected by some electrolytic
method. In many cases the methods suggested are quite
empirical and indeed wholly useless. Thus it has been
proposed to use an ordinary process of amalgamation, and
to make the mercury the cathode of an electrolytic cell. The
most that could be expected of such a proceeding is that
the surface of the mercury might be kept clean, and there-
fore in a better condition to dissolve gold ; the operation,
if effective, is similar to the addition of a little sodium
to the mercury, which is frequently practised, and tends to
prevent the mercury " sickening," i.e. becoming coated and
unfit to act as a solvent. There are obviously great diffi-
culties in devising a workable electrolytic process for the
extraction of gold from its ores. Bearing in mind the fact
that an ordinary paying gold ore may average 2 ounces per
ton, i.e. 0-005 per cent., and that many paying ores are
considerably poorer than this, it is evident that it is useless
to make the ore the anode in a suitable electrolyte (say a
chloride) and hope to cause the gold to dissolve. Such a
proceeding is impracticable, because no definite electrical
connection can be made with the minute particles of gold,
relatively very few and distributed through the whole mass
of gangue. Thus the solvent action of the current cannot
be centred on those particles which alone it is desired to
96
GOLD AND SILVER
dissolve. Therefore a practicable electrolytic process must
consist in leaching out the ore with a solvent, depositing
the gold therefrom and revivifying the solvent, and returning
the solvent to extract a fresh batch of ore. The solvent
may be actually prepared electrolytically, or it may be an
ordinary chemical bought ad hoc.
ELECTROLYTIC GOLD RECOVERY
It may be said at once that there is only one electrolytic
gold process in actual operation on a large scale, viz. the
Siemens-Halske process. Even in this the extraction of the
metal is accomplished by purely chemical means, a solution
of potassium cyanide being used. It is the recovery of the
metal from the solvent which is electrolytic. There is no
reason in the nature of things why a similar electrolytic
recovery process should not be applied to the treatment of
liquors obtained by the extraction of gold from its ores by
means of chlorine ; the gold could be deposited without doubt.
Simpler methods of chemical precipitation, e.g. with ferrous
sulphate or charcoal, are, however, generally preferable.
The process of electrolytic recovery is specially applicable
to the liquors from cyanide extraction for the following
reason : By the electrolytic process recovery can be effected
from very weak cyanide liquors which cannot be treated
equally completely with zinc the usual chemical precipi-
tant ; such weak liquors are much more economical for
extraction ; thus electrolytic recovery presents a consider-
able advantage. In short, it is the desirability of extracting
gold ores with weak cyanide liquors and the necessity of
devising some means for recovering the gold from these
liquors which have led to the invention of the Siemens-
Halske process and its modifications.
The Siemens-Halske process is carried out in a simple
form of apparatus. The cyanide liquor which has been
used for extracting gold from the ore, containing 0-05 per
cent, (or less) of potassium cyanide and about 5 to 6 penny-
weights of gold per ton, is electrolysed between iron anodes
97 H
PRACTICAL ELECTRO-CHEMISTRY
and sheet-lead cathodes. A low current density, e.g.
0-06 amperes per square foot, suffices, and even with this
the current efficiency is extremely small. This is of little
consequence, as the cost of the energy required is a mere
trifle compared with the cost of the cyanide and labour in
handling the ore. In fact, the process is simply one for the
cheap, efficient and convenient recovery of gold from its
dilute solutions in potassium cyanide, and must not be
judged by standards applicable to ordinary methods of
depositing metals electrolytically. The weak point in the
process is the difficulty of providing satisfactory anodes. It
appears that in weak alkaline liquids carbon is attacked
and disintegrated ; platinum might serve, but its cost is
excessive. Iron is used, as mentioned above, and is attacked
to some extent. By the action of the cyanide it is dis-
solved and converted into double cyanides of iron, i.e.
Prussian blue. To prevent this from contaminating the
electrolyte, the iron anodes are enclosed in linen bags ;
the Prussian blue has a small commercial value. It may
be reconverted into cyanide by treatment with alkali to
form ferrocyanide and fusion of this body with sodium to
yield cyanide, if such a series of operations be found re-
munerative at the present low price of cyanide. Whether
this recovery be practised or not, the iron goes to waste
and forms a tangible item of expense. Estimates of cost
of working the process have been made and published.
Their details are of importance in a work on gold extraction
processes, but would be out of place here. It is sufficient
to say that the cost of working the whole process of extraction
and recovery is about 85. per ton, out of which the cost of
working the electrolytic part of the plant amounts to about
8d. The chief portion of this Sd. is expended in replacing
the lead cathodes and iron anodes, the cost of power being
a minor item.
It is well to make clear, and to repeat if necessary, that
the electrolytic recovery of gold is a mere auxiliary to the
cyanide process of gold extraction a very useful auxiliary,
but still only a subsidiary part of the process. The great
GOLD AND SILVER
advantage of the electrolytic over other processes of recovery
is, as has been said above, its ability to precipitate gold
from solutions weak in cyanide. This allows extraction to
be performed with much weaker solutions, e.g. O05 per cent,
instead of 0-5 per cent., than can be effectively employed
when recovery is performed by means of zinc as the precipi-
tant of the gold. 1
Modifications of the Siemens-Halske process have been
devised. Thus Andreoli uses anodes of lead peroxide and
cathodes of iron. The lead peroxide is said to be unattacked,
and the iron cathodes are periodically stripped of their
deposit of gold by immersing them in a bath of fused lead,
the gold dissolving therein. The stripped plates are re-
turned to the bath. When the lead is sufficiently enriched
it is cupelled and the gold is recovered.
A process of combined extraction and recovery of gold
from its ores which presents certain features of interest is
that known as the Hay craft process. In this the ore is
placed in a cylindrical iron vessel filled with brine and
provided with a vertical shaft carrying arms from which
depend carbon anodes. At the bottom of the vessel is a
layer of mercury which is made the cathode. The vessel
is filled with a solution of common salt, which is heated and
the ore is mixed therewith, the whole being kept stirred by
the revolution of the agitator carrying the anodes. It is
stated that the coarser particles of gold, which are susceptible
of ready amalgamation and are not easily dissolved by
chlorination, sink through the electrolyte, arrive at the
mercury (which is kept clean and active by its being a
cathode), and are then caught. The finer gold, and that
which does not easily amalgamate, is acted on by the
chlorine liberated at the anodes and is dissolved as gold
chloride. In course of time, as the liquid is kept agitated,
this gold chloride reaches the cathode and is there decom-
posed, the gold being deposited in the mercury. When
1 Nowadays weaker cyanide solutions are precipitated by finely
divided zinc, but even when zinc dust is used the advantage lies
with the electrolytic method.
99
PRACTICAL ELECTRO-CHEMISTRY
once caught it cannot be redissolved by the chlorinated
electrolyte because it forms part of the cathode. Gradu-
ally, therefore, the ore is robbed of its gold, which is trans-
ferred to the mercury. The exhausted ore is run away
together with the solution, is allowed to settle, and the
solution is returned to be used with a fresh batch of ore.
There is nothing in this process which is absurd or obviously
impracticable, and yet it is doubtful whether it can be
successfully worked. The wear of the anodes exposed both
to the attack of chlorine and to the abrading action of the
ore is likely to be considerable. Some loss of mercury due
to the metal being mixed with the ore by the agitation
intentionally performed, and carried away with the spent
ore, might be expected. Even if these difficulties were
found not to be serious, it is not clear that the process
possesses any distinct advantage over an ordinary amalgama-
tion process, followed by chlorination or cyanide extraction
of the tailings.
Another process of combined extraction and recovery
is that known as the Pelatan-Clerici. As far as published
descriptions are intelligible, the process seems to be a kind
of blend of an amalgamation process, a chlorinating and a
cyanide extraction method. Its merits do not appear to
be commensurate with its complexity.
A method of gold recovery from cyanide solution has
recently been patented by Kendall, which, though not known
to be at work on an industrial footing, is of sufficient interest
to be worth notice. The gold is deposited on a large
cathode consisting of broken carbon, through which the
cyanide solution is caused to flow ; the anode is also carbon.
When the attenuated film of gold has been carried on the
large and irregular cathode surface the cell is reversed, the
electrolyte is changed for a concentrated cyanide solution,
and the gold is deposited on a cathode of small surface
consisting of a carbon plate previously silvered. The
method is a device first for catching the bulk of the gold from
a large volume of dilute solution and then for gathering it
on to a relatively small area.
100
GOLD AND SILVER
THE ELECTROLYTIC REFINING OF GOLD
Besides these processes for extracting gold, the electro-
lytic refining of gold is practised to a limited extent. This
term refining applies to gold already tolerably free from
impurities, and does not refer to the electrolytic parting
of gold from silver or its recovery of gold-silver-copper
alloys, which will be dealt with anon. The method of
refining gold containing platinum practised by the Nord
Deutsche Affinerie of Hamburg is said to consist in using
the crude gold as anodes in a solution of gold chloride and
receiving the deposited gold on cathodes made of thin sheets
of the pure metal. It is evident that a process thus described
would not be workable. In the first place, platinum or
palladium contained in the anodes would dissolve as well
as the gold in a bath of gold chloride. It might prove
possible to prevent their deposition on the cathode by
working with a low current density, but with a low current
density the rate of refining would be low, and the weight of
gold locked up in the baths would be so large that the
expense of interest on the value of the metal would
make the process too costly. Further, if the current
density be increased in a neutral solution of auric chloride,
chlorine is evolved at the anode without causing its equiva-
lent attack. If, however, HC1 or an alkali chloride be
present the dissolution of the anode proceeds regularly.
Apparently a chloride of the form AuCl 4 H (from AuCl 3 and
HC1) is a necessary constituent of the electrolyte, the ions
of which may be regarded as AuCl 4 and H. Applying this
observation, it is found that, when an ample supply of
hydrochloric acid is present in the electrolyte, a current
density at the anode of 10 amperes per square dm. (about
90 amperes per square foot) may be used without causing
evolution of chlorine at the anode. The electrolyte should
contain 25-30 grammes of gold per litre and the voltage
should be kept low, e.g. 1 volt, to avoid the deposition of
impurities dissolved from the anode . Under these conditions
101
PRACTICAL ELECTRO-CHEMISTRY
the gold is deposited in a crystalline adherent condition.
As in ordinary metal refining by electrolysis, certain of the
impurities dissolve and are not deposited on the cathode
and certain others remain undissolved and constitute an
anode sludge. The usual impurities in gold of the class
which is suitable for electrolytic refining are platinum,
palladium, osmium and iridium (in the form of osmiridium),
and silver. Of these the platinum is dissolved, but is not
redeposited. It can, therefore, be allowed to accumulate
in the electrolyte until the liquid contains enough to give,
with ammonium chloride, a precipitate of ammonium
platino-chloride, (NH.) 2 PtCl 6 . Palladium is also dissolved,
but is not precipitated by ammonium chloride. It can be
recovered by precipitation with potassium iodide as the
black palladous iodide Pdl . Osmiridium remains undis-
solved and unattacked in the anode sludge, and silver is
converted into silver chloride, which is slightly soluble in
the electrolyte, containing as this does both hydrochloric
acid and auric chloride. The bulk of the silver chloride
remains undissolved, but the small quantity in solution
suffices to yield a little silver at the cathode, which is de-
posited together with the gold. The proportion is, however,
quite small, so that the gold ultimately obtained is 999-8
fine. A certain amount of gold is left in the anode sludge.
As the process of electrolysis goes on the bath becomes
poorer in gold from the gradual replacement thereof by the
impurities, such as platinum and palladium, and fresh
auric chloride has to be added to maintain a proper concen-
tration of the electrolyte. It is found that the electrolysis
does not proceed smoothly with the formation solely of
auric chloride at the anode and the exact deposition of its
gold at the cathode. Besides auric chloride, aurous chloride
(AuCl) is formed at the anode. This in great measure
breaks up at once thus :
3 AuCl = AuCl 3 + 2 Au.
The gold is deposited at its place of origin, the anode,
and forms part of the anode sludge, as mentioned above.
102
GOLD AND SILVER
A part, however, of the aurous chloride escapes immediate
decomposition and diffuses through the electrolyte, ulti-
mately arriving at the cathode, where it is decomposed and
deposits its gold together with that from the auric chloride,
which forms the chief constituent of the electrolyte. It
would seem at first sight that it would be advantageous
to form as much aurous chloride as possible, because a given
current would deposit three times as much gold as it would
if auric chloride were formed. The considerable decom-
position of the aurous chloride which takes place and the
consequent appearance of two-thirds of its gold in the
anode sludge make the formation of the lower chloride
undesirable. The use of a high current density is found
to restrict the proportion of aurous chloride.
THE PARTING OF GOLD AND SILVER
Gold as obtained from its ores commonly contains a cer-
tain proportion of silver (from 10 to 50 per cent.). It may
be separated therefrom by various methods of parting.
One of the older processes is to fuse the gold-silver alloy
with enough silver to lower the proportion of gold to 33 to
25 per cent, of the whole alloy. This alloy, being compara-
tively rich in silver, can be attacked satisfactorily by nitric
acid, which dissolves the silver and leaves the gold un-
touched. A cheaper method is to part by boiling with
sulphuric acid ; in this case the gold should not exceed
one-sixth of the whole alloy to allow free and complete
attack by the acid. The same method of parting is, of
course, applicable to auriferous silver even poorer in gold.
These older chemical methods have now a formidable rival
in the shape of an electrolytic process of parting.
If an alloy of gold and silver containing two or three
times as much silver as gold is made the anode in an elec-
trolyte of nitric acid, the silver will be dissolved and the
gold left as a residual sludge at the anode. The method
is equivalent to parting with nitric acid, but has this advan-
tage, viz. that the nitric acid is not consumed. By the
103
PRACTICAL ELECTRO-CHEMISTRY
ordinary chemical method, not only is nitric acid used
(permanently) to form silver nitrate, but also another por-
tion of nitric acid is reduced in the course of the dissolution
of the silver, thus :
4 Ag + 6 HN0 3 = 4 AgN0 3 + N 2 3 + 3 H 2 0.
In this case two molecules of nitric acid over and above
those necessary to form silver nitrate are needed for every
four atoms of silver. In electrolytic parting nothing of the
kind occurs The nitric acid is only a convenient medium
serving to dissolve the silver at the anode and to provide
silver nitrate to be decomposed at the cathode where the
silver is deposited. If the electrolysis is property conducted,
and the solution kept rich enough in silver so that hydrogen
is not evolved at the cathode, no reduction and loss of nitric
acid can occur. The economy which results is sufficient
to cover the cost of power and plant, and incidentally rein-
states nitric acid as a parting menstruum preferable to
sulphuric acid, which had displaced it.
Application of this idea has been made by Moebius, whose
system is used by the Deutsche Gold- und Silber- Scheidean-
stalt vorm. H. Rossler at Frankfort-on-Main. The Moebius
apparatus consists of a set of wooden tanks containing
cast anodes of the silver-gold alloy about J to -f inch in
thickness and thin sheet silver cathodes.
The anodes are enclosed in bags of filter cloth stretched
on a wooden frame, the object of this arrangement being
to retain the finely divided gold which separates as sludge,
and to prevent it from mingling with the silver collected
at the cathode. The cathodes are placed between the
prongs of a wooden fork, which can be passed over their
surface from end to end ; the arrangement is shown in the
figures (21 and 22).
c is the cathode of thin sheet silver attached to a stout
copper rod D, which serves as its electrical connection.
E is the wooden fork made of a couple of laths connected
by a cross-piece, and carried by a roUer F, running on the
wooden rail H. The fork, one prong of which is on each
104
GOLD AND SILVER
side of the silver sheet, can thus be passed from one side
of the vat to the other, clearing off in its passage any loose
silver crystals which may be adhering to the cathode.
The silver crystals thus swept off fall into trays at the
bottom of the vat. These trays are wooden frames covered
with filter cloth, so that, when lifted from the vats, they
retain the silver crystals and let the electrolyte run through.
By these devices all risk of short-circuiting by the growth
of silver crystals from cathode to anode is avoided ; all
chance of contamination of the silver with the anode sludge
H
D
E,
C
FIG. 21.
FIG. 22.
is also removed, and the recovery of the silver in a form
easy to wash and melt into ingot form is accomplished.
The vats used are 12 feet x 2 feet, and are divided into
seven compartments, each constituting a cell in which are
three anodes and four cathodes. The anodes are not cast
in a continuous sheet extending from one side of the cell
to the other, but are composed of strips placed in the manner
shown in plan in the drawing. (Fig 23).
The anodes a are suspended on arms resting on the con-
ductors D, D, the whole contrivance being enveloped in a
bag of filter cloth as aforesaid. Fairly narrow strips of
metal serving as anodes are advantageous, because the
105
/
/
/
/
/
PRACTICAL ELECTRO-CHEMISTRY
inevitable irregular dissolution of the metal composing
the anode would be apt to break large fragments off a wide
plate, whereas from a narrow plate pieces relatively small
would be separated. Therefore the consumption of the
anode material, and consequent purity of the anode sludge,
will be greater with small anode elements than with large.
The electrolyte used is dilute nitric acid, which soon becomes
silver nitrate. It is advisable to keep the
sOi solution acid with nitric acid, so as to
avoid the deposition of copper (occurring
as an impurity in the silver-gold alloy con-
stituting the anodes) with the silver at the
cathodes. This may be done by making the
current density greater at the anodes than
at the cathodes, or by regulated addition
of nitric acid. In either case a certain
amount of nitric acid will be used up in
the production of cupric nitrate, but the
loss is infinitesimal compared with the
consumption which occurs when " parting "
with nitric acid is practised. In the sec-
tion on the refining of copper it has been
pointed out that the turnover of rnetal
should be as large as possible compared
with the stock of metal held, to minimise
the interest which must be reckoned on the
capital thus locked up. With precious metals
this necessity becomes acute. Therefore as
high a current density as possible must be
employed. In practice a current density as great as 28
amperes per square foot is used, but is diminished as the
proportion of copper to silver in the electrolyte increases.
It is obvious that any waste of current caused by the use
of a high current density is more than counterbalanced by
the reduction of interest charge. The purity of the gold
left as an anode sludge is not necessarily perfect. The
following analyses indicate the nature and amount of the
impurities :
106
D
/
D
FIG. 23.
GOLD AND SILVER
Gold
Lead
Bismuth
Per cent. Per cent. Per cent.
99-954 99-947 99-955
0-036 0-043 0-030
0-010 0-010 0-015
100-000 100-000 100-000
If the electrolysis has been carefully conducted, the pro-
portion of nitric acid maintained, and the current density
diminished as the content of the electrolyte in copper
increased, the deposited silver will be sensibly pure. The
silver and copper can be ultimately recovered by ordinary
chemical means from the electrolyte when it has become
so loaded with copper as to be no longer fit for use.
Thus the silver may be precipitated by copper plates or
FIG. 24.
as chloride, and the copper in a crude form by iron. This
process of recovery, however, will not need to be put into
use until a large amount of silver-gold alloy has been worked
up, unless, indeed, the alloy is unusually rich in copper,
and therefore the waste of electrolyte will be relatively
small.
The Moebius apparatus has been modified in the following
way. The electrodes are arranged horizontally, the anode
being separated from the cathode by a porous diaphragm.
The cathode is a thin sheet of silver travelling over rollers,
as shown in the figure. It deposits its silver on another
travelling band, from which it is scraped at a point outside
the vat. This arrangement does away with the necessity
for taking out at periodical intervals the trays containing
the silver crystals. The general scheme of the apparatus
is shown in the figure.
107
PRACTICAL ELECTRO-CHEMISTRY
The anodes A are suspended in frames covered with filter
cloth immediately above the travelling cathode c, which
runs on rollers D, D. At the right-hand end of its course
the cathode brushes against the travelling belt E, running
in the opposite direction. On this the loose silver is deposi-
ted, and by it is conveyed outside the electrolytic tank,
and is swept off by the scraper F into any suitable receptacle.
It is found of advantage to oil the cathode slightly to facilitate
the removal from it of the deposited silver. One sees again
here the care taken to build up the anode of small units
so as to prevent the wasteful breaking up which would occur
with a large plate. This point has already been dealt with
(see p. 106).
THE ELECTROLYTIC REFINING OF SILVER
The method above described is adapted to the refining
of auriferous silver as well as to the parting of gold
from silver. It stands, in fact, in a position similar
to that of the ordinary electrolytic process for refining
copper (q.v.), in that the residue of gold left as an anode
sludge goes a great way towards paying the cost of the
refining operation. So here the recovery of a little gold
will be profitable, even though much silver has to be trans-
ferred from anode to cathode in order to win it. A case
of the kind is afforded by the plant of the Pennsylvanian
Lead Company at Pittsburg, which is used for refining silver
obtained in the usual routine of refining lead. The crude
silver contains about 2 per cent, of impurities, e.g. lead,
bismuth, and copper. The plant is of the older type, with
fixed cathodes and travelling scrapers. A current density
of 18 amperes per square foot is used, and an output of 88
ounces of silver per H.P. hour is obtained ; the pressure
required is about 1-2 volts. The plant consists of fourteen
tanks, each divided into seven cells, i.e. in all 98 units,
these about 84 are usuaUy running, a certain number
laving to be left standing for cleaning and repairs. The
s capacity of the plant is 40,000 ounces of silver per
1 08
GOLD AND SILVER
day of 24 hours, and the actual output is about 33,000
ounces. The tanks are of wood, 10 x 2 feet x 22 inches
deep, each cell being 2 feet long (across the tank) and 1
foot 5 inches wide.
There are four cathodes and three anodes in each cell.
The cathodes are 22 x 13 inches and are of thin sheet silver.
The anodes are 18 x 10 inches and about \ inch thick.
It is found that such stout anodes (each weighing 13-15
kilos) are, on the whole, less advantageous than are anodes
about one-tenth this thickness, such as are used at Frank-
fort. The importance of the electrolytic refining of silver
may be gathered from the fact that in 1895 the output in
the United States was 10,000,000 ounces, or about one-
seventh of the whole. An installation of the newer form
of the Moebius process (see above) has been adopted by
the Guggenheim Smelting Company, Perth Amboy, New
Jersey. In this there are 48 tanks each 14 feet 3 inches x
16 inches wide x 7 inches deep. The material refined is
similar to that used by the Pennsylvanian Lead Company,
viz. silver containing 98 per cent, of Ag and 0-3-0-8 per
cent, of gold, the balance being casual impurities. The
electrolyte is a solution containing 0-1 per cent, of free
nitric acid, 4-5 per cent, of Cu, and about 1 per cent, of
silver. It may be assumed that the presence of the copper
is inevitable but not essential, inasmuch as this metal would
naturally dissolve from an impure anode, and could have
no sensible influence on the course of electrolysis until its
quantity became sufficient to cause it to be precipitated
with the silver. A certain amount of nitric acid is used
up, mostly for the dissolution of the copper and partly pro-
bably by reduction at the cathode. The quantity thus
consumed is 1J pounds per 1,000 ounces of silver treated
an almost negligible loss. The anodes are comparatively
small units, viz. 15 x 3} x inches, and are in separate
frames, as shown in the figure above. The silver belt consti-
tuting the cathode is 31 feet long and 15 inches in width.
Its upper side is smeared with graphite to prevent too close
an adherence of the deposited silver, so that the metal may
109
PRACTICAL ELECTRO-CHEMISTRY
be readily removed by the scrapers. These were first of
hard rubber, but are now " rush-wood brushes." A current
of 220 amperes at 90 volts suffices for the treatment of
24 ; 000 ounces of silver per 24 hours. Each tank needs a
pressure of 1J-2 volts. The cost of the process is reckoned
at -^d. per ounce of silver refined, and the capital expendi-
ture for a plant capable of dealing with 30,000 ounces of
silver per 24 hours at 1,200. It is interesting to note that
the silver which is almost pure is melted down with a little
scrap copper, because English buyers decline to recognise
a higher approximation to purity than 998 fine. This
little incident neatly illustrates the intense conservatism
of the metal trades in this country, a trait familiar to all
who have daily dealings therewith.
There is little to be said concerning the electrolytic treat-
ment of silver other than what has been given in the fore-
going descriptions. The usual wet methods of silver extrac-
tion from the ore, by which the silver is converted into
chloride and leached out by means of brine or sodium
hyposulphite, might well be found to lend themselves to
an electrolytic recovery process. At present the silver is
precipitated as metal, by bringing its solution into contact
with copper, or as silver sulphide. No attempt seems to
have been made to precipitate it electrolytically.
The alloy of zinc and silver obtained in the Rossler modi-
fication of the Parkes process, may be separated into its
constituent metals by electrolysis in a solution of zinc sul-
phate, the zinc being deposited and the silver remaining as
an anode sludge. This process is strictly analogous to the
refining of argentiferous copper in a sulphate solution.
REFINING OF GOLD, SILVER, AND COPPER ALLOYS
It will be understood from what has already been said
that the principle of separating copper, silver, and gold by
the selective action of the current in a nitric acid bath can
be applied generally to alloys having a large range of compo-
sition, provided the proportion of gold is moderate. With
no
GOLD AND SILVER
an alloy rich in gold, difficulty is encountered because of
the imperfect solubility of the anode and its irregular con-
sumption. Various inventors, notably Borchers and Diet-
zel, have devised apparatus in which the alloy is granulated
and is caused to move relatively to the electrolyte, so that
the anode sludge may be separated as it is formed. By
these means it is hoped to overcome the obstacles mentioned
above, but the processes in question do not appear to have
been taken into commercial use. It is probably preferable
to dilute the refractory alloy, by fusing it with copper
or silver as may be necessary, so that it may be regarded
as an auriferous copper or an auriferous silver, and may be
refined accordingly by electrolysis in a sulphuric acid or
nitric acid bath. The difficulties inseparable from the
treatment of metal in a granulated state and used as an
anode will thus be avoided. Dietzel has worked out a
process, which has now been in use for some years by the
Allgemeine Gold- und Silber- Scheideanstalt at Pforzheim,
which consists in dissolving the gold, silver and copper
alloy (about 5 per cent. Au, 35 per cent. Ag, and 50 per
cent. Cu), as anodes in an electrolyte containing free nitric
acid, prepared by the passage of a stream of copper nitrate
over the cathodes in the same cell. The silver thus dis-
solved is precipitated chemically by the action of copper
scrap in an adjacent vessel, and the regenerated copper
nitrate is returned to the electrolytic apparatus. A cur-
rent density of about 15 amperes per square foot is used,
and a pressure of 2-5-3-0 volts. For the success of this
process, it is evidently necessary to arrange a flow of copper
nitrate solution over the cathode, sufficiently copious to
prevent the diffusion of silver nitrate from the dissolution
of the anode back to the cathode, where a portion of the
silver would be deposited together with the copper.
in
NICKEL
WITHIN the last few years the refining of nickel by electro-
lytic means has become commercially practicable. The
electrolytic winning of the metal from its ores is not yet
accomplished. The metallurgy of nickel is complicated
and difficult, and the ordinary processes of obtaining it are
comparatively expensive. Thus there is a field for its direct
electrolytic production, but this field has not been cultivated
vigorously and successfully.
Regarded metallurgically, nickel stands between copper
and iron, presenting similarities to both. It is like copper
in the comparative stability of its sulphide, and like iron
in the relative difficulty of its reduction from oxide, in the
high fusing-point of the metal when reduced, and in its
tendency to unite with carbon and silicon, giving a crude
metal analogous to cast iron. Nickel, whether obtained
as a crude cast metal or from a matte of copper sulphide
and nickel sulphide or from an arsenical matte, i.e. a speiss,
is invariably impure, as is shown by the following analyses :
Per cent.
Per cent.
Nickel
98-39
98-68 i
Copper
0-76
Iron
0-10
0-30
Carbon
MO
Silicon
0-13
0-19
Sulphur
0-26
0-07
99-98 100-00
Such crude nickel may be conveniently refined electroly-
tically. On the other hand, alloys of nickel and copper
1 In these analyses the figure for nickel probably includes the
percentage of cobalt which is present in most nickel ores, the only
notable exception being the silicate ores of New Caledonia.
112
NICKEL
containing approximately equal parts of the two metals,
produced relatively easily by dead-roasting sulphide mattes
of nickel and copper and reducing the mixed oxides, are
not readily refined electrolytically. The refining of such
mixtures would be best attempted by dissolving the mixed
oxides in sulphuric acid, precipitating the copper electroly-
tically in acid solution, neutralising and depositing the
nickel in similar manner. But in both cases the process
is one of electrolytic reduction, and not merely of transfer-
ring the metal as such from anode to cathode ; the energy
required and the consequent cost would therefore be high
even were there no technical difficulties, which is not to
be lightly asserted.
No complete and authoritative account of the processes
of nickel refining as carried out in the United States and
in this country has been published. Thus it appears that
nickel, not as mere plating but in thick sheets, is being
deposited by Messrs. Thomas Bolton & Sons at Cheadle,
and that a similar operation is accomplished by the Balbach
Smelting and Refining Company in New Jersey, but in
both cases details of the process are not forthcoming. Pro-
cesses have been devised by Hoepfner, Rickets, and others,
but have not been brought into use and exhibit no idea
sufficiently novel or illustrative to warrant their description.
But, although there is a dearth of positive and detailed
information concerning plants actually at work, there exists
a considerable store of knowledge relating to the con-
ditions necessary for the successful electro-deposition of
nickel, from which can be deduced the chief precautions
which must be observed in working on a manufacturing
scale.
Before this matter is dealt with it may be said that there
is no difficulty in depositing nickel (using nickel anodes)
in thin films, as in plating. The art of nickel plating (q.v.)
is thoroughly well understood, and a good and adherent
coating of nickel can be obtained if proper care is exercised ;
bad nickel plating is common, but it need not be. But
in refining nickel the metal must be deposited in sheets of
113 I
PRACTICAL ELECTRO-CHEMISTRY
reasonable thickness, e.g. i to J inch. When it is attempted
to continue the deposition of nickel in an ordinary plating
bath, so as to produce not a mere film but a stout sheet,
it is found that as soon as a very small thickness is exceeded
the metal detaches itself from the cathode and curls up in
thin flakes. These are too thin to collect and melt to an
ingot with economy and ease, and thus it is impracticable
to work a nickel-refining plant by simply continuing the
operations of the plater. These difficulties are evident
even in the most careful work, as the following paragraphs
will show.
Pure electro-deposited nickel was prepared by Bischof
and Thiemann, as the material to be used by Winkler in
his determination of the atomic weight of the metal. In
similar manner they deposited cobalt destined for the like
purpose.
For the deposition of nickel the purest procurable nickel
sulphate was used as the raw material. 200 c.c. of a solu-
tion of this salt, containing 32-84 grammes of Ni per litre,
was mixed with 30 grammes of ammonium sulphate, 50
grammes of ammonia of specific gravity 0-905, and 250 c.c.
of water. This solution of the double sulphate of nickel
and ammonium, containing excess of ammonium sulphate
and of ammonia, was electrolysed with a current density
of 0-5 ampere per square decimetre and a pressure of 2-8
volts. An insoluble anode (of platinum) was used and the
deposited nickel was received on a nickel cathode, platinum
not being used for this electrode because of the difficulty
frequently experienced in detaching deposited nickel from
a platinum surface. When the nickel had attained a cer-
tain thickness it separated spontaneously from the cathode
and curled up in thin leaves precisely as it is observed to do
in ordinary plating, where the materials are not perfectly
pure and the same scrupulous care in manipulation is not
aimed at. The product was white, lustrous, and free from
any discoloration such as might be produced by local oxida-
tion ; on heating the metal in hydrogen its weight was
unaltered, proving the absence of oxide.
114
NICKEL
A similar experiment on the preparation of pure cobalt
was made. A solution of the sulphate was prepared con-
taining 11-64 grammes of Co per litre. 100 c.c. of this
solution was mixed with 30 grammes of ammonium sul-
phate, 30 grammes of ammonia of specific gravity 0-905,
and 500 c.c. of water. This solution was electrolysed with
a current density of 0-6 ampere per square decimetre and
a pressure of 3 volts. The cathode was of platinum, and
the cobalt formed on it a coherent and fairly stout sheet,
which was bright on the side in contact with the platinum
and had a grey matte surface on the other. The cobalt
when ignited in hydrogen lost 0-23 per cent, of its weight,
corresponding with a content of 0-55 per cent, of the hydra-
ted oxide Co 2 3 2 H 2 0. A second experiment gave similar
results, save that the deposited cobalt was received on a
nickel cathode (instead of one of platinum) and stripped
spontaneously from it precisely as did the nickel in the
former trial.
It may be noted as a point of interest that these two
metals, which may be accepted as sensibly pure specimens
of nickel and cobalt respectively, differed slightly but dis-
tinctly in colour, the nickel having a slight yellowish tint,
while the cobalt was of a bluish-white tone.
Another exact and important study of the electrolytic
deposition of nickel, which has moreover a direct bearing
on the manufacturing employment of such a process, has
been made by Dr. F. Foerster. From his researches it
appears that nickel can be deposited in thick coherent
plates if the electrolyte be kept at a temperature between
50 C. and 90 C. The electrolyte used was a solution con-
taining 145 grammes per litre of commercial nickel sulphate,
corresponding with 30 grammes per litre of metallic nickel.
The level of the liquid and its concentration were main-
tained constant throughout the experiment, and the elec-
trolyte was kept well mixed and agitated. A stout nickel
plate was used as the anode ; it was enclosed in parchment
paper to retain the anode sludge. The cathode was a thin
nickel plate from which the deposited metal could readily
PRACTICAL ELECTRO-CHEMISTRY
be detached. The preliminary experiments were made
with electrodes having an effective surface of 80-100 square
cm., and the experiment was continued until 25-40 grammes
of nickel had been deposited. It was found that with a
current density of 0-5-2-5 amperes per square decimetre
and at a temperature of 50 C. 90 C. good coherent depo-
sits, bright grey or tin white in colour, were obtained. The
higher the current density, the brighter and smoother was
the deposit. Thus with 0-5 ampere per square decimetre,
and using a solution containing 100 grammes of Ni per litre,
kept at a temperature of 80 C., the deposit had a rough
surface and was dull grey in colour ; with a current density
of 2-2-5 amperes per square decimetre the deposit was
silver white and could be obtained in plates 0-5-1 millimetre
in thickness. Frequently it was noticed that the deposit
exhibited certain rugosities, produced by the circumstance
that a stream of hydrogen had been given off for some time
at particular spots, and thus had caused a local irregularity
in the current density. This trouble could be avoided by
stirring the electrolyte so that the evolution of hydrogen
did not persist at any given point for an appreciable time.
A larger scale experiment was made under similar condi-
tions, and as much as 0-5 kilo of electrolytic nickel was pre-
pared. In this case the cathode had an area of 2 square
decimetres ; the electrolyte contained 100 grammes of Ni
per litre and was kept at 60 C. The current density em-
ployed was 1-5-2 amperes per square decimetre.
The nickel deposited was particularly tough ; the thick-
ness of deposit is not stated, but from the weight given and
the area of the cathode it can be calculated as slightly
smaller than 3 mm., say J inch. A plate of such thickness
could be melted down without serious loss, though for manu-
facturing purposes an even more substantial deposit is
desirable. Nevertheless the achievement of Dr. Foerster
is remarkable, and may well embody the only secret worth
guarding in the electrolytic refining of nickel as now prac-
tised with much mystery in this country and elsewhere.
Armed with this knowledge, an enterprising manufacturer
116
NICKEL
should have no great difficulty in refining nickel electroly-
tically with commercial success.
An important piece of collateral evidence supports the
belief that the electrolytic nickel now available as a mar-
ketable commodity is prepared by processes substantially
identical with that set forth above. Dr. Foerster found
that iron and cobalt, the characteristic impurities of com-
mercial electrolytic nickel, were also present in his own pro-
duct. The study of the degree of purification effected by
the electrolytic refining of nickel is particularly instructive,
and should suffice to dispose of, once for all, the ridiculous
belief that a metal prepared by electrolysis is necessarily
and ipso facto of unusual purity. The anodes used by Dr.
Foerster had the following composition :
Per cent.
C 0-40
Si 0-02
Cu 0-10
Fe . ...... 0-43
Co . 0-14
Mn 0-02
Nickel (by difference) 98-89
100-00
Of these all but the iron and cobalt were absent from
the electro-deposited nickel, which contained as impurities
0-3 per cent, of iron and from 0-1 to 0-3 per cent, of cobalt.
How considerable is the tendency for iron arid cobalt to be
deposited together with nickel is shown by the fact that an
electrolyte containing 0-087 gramme of iron and 0-82 gramme
of cobalt per 100 grammes of nickel contained, after it had
been used for refining, 0-034 gramme of iron and 0-064
gramme of cobalt, being thus actually impoverished in these
impurities, which were deposited in the first 100 grammes
of nickel thrown down on the cathode, the metal containing
as much as 0-38 per cent, of iron and 1-6 per cent, of cobalt.
On continuing the electrolysis a further deposit of 400
grammes of nickel was obtained, containing 0-20 per cent.
117
PRACTICAL ELECTRO-CHEMISTRY
of iron and 0-57 per cent, of cobalt, these figures correspond-
ing closely with those for the anode used in this particular
experiment, viz. 0-27 per cent, of iron and 0-60 per cent,
of cobalt. Of these two impurities the iron alone is objec-
tionable for most purposes. Both it and cobalt can be
eliminated by adding to the electrolyte an organic acid,
such as tartaric acid, and electrolysing with a low current
density (0-3-1 ampere per square decimetre), whereby the
iron is deposited, the nickel remaining in solution. On
increasing the current density above 1 ampere per square
decimetre the nickel is deposited. Such a method, although
it might be employed to purify an electrolyte periodically,
could not well be used for the continuous refining of nickel,
i.e. the transference of the metal from an anode of the crude
material to a cathode whereon it was to be deposited pure.
When a solution of nickel chloride was used instead of
the sulphate, the results were less favourable, the deposit
stripping at the ordinary temperature and a basic salt being
deposited on the cathode when the electrolyte was used
hot. A better effect was obtained by using a solution con-
taining about 2-5 grammes of free hydrochloric acid per
litre. Another trouble when using the chloride solution
is that the envelope of parchment paper round the anode
is quickly attacked ; it is better to dispense with this dia-
phragm and to trust to the natural tendency of the residue of
the anode to stick together, which it does fairly well if not
disturbed by the stirring of the solution. Regarding the
attack of the envelope round the anode an interesting obser-
vation was made. When the parchment paper was replaced
by linen, so much organic matter went into solution that
the electrolyte had a caramel-like smell, and yielded metal
containing 0-6 per cent, of C, and of dark colour, brittle,
and tending to curl off the cathode. When once the electro-
lyte was thus spoiled it continued to yield bad deposits,
even after the organic envelope had been removed ; it had
eventually to be thrown away.
The successful attempt recorded above to deposit nickel
in plates of fair thickness from solutions of its sulphate
118
NICKEL
when a nickel anode was used and the process was therefore
one of refining and not of winning, prompts the belief that
it may be practicable to deposit nickel similarly from a
sulphate solution, using an insoluble anode. Should this
be feasible, nickel could be extracted by leaching out a
roasted matte containing nickel sulphate, and, after removal
of impurities likely to be deposited together with the nickel,
electrolysing this sulphate solution with carbon anodes and
thin sheet nickel cathodes. Experiments made with a
solution of nickel chloride gave unsatisfactory results,
because the carbon anodes gradually dissolved and contami-
nated the electrolyte so considerably that the deposited
nickel soon became grey and brittle. On account of this
action, and because of the chlorine finding its way to the
cathode to some extent, the output was not more than 70
per cent, of that calculated from the current. A sulphate
solution was not tried, but it is probable, from the known
behaviour of carbon anodes in sulphuric acid, that an equally
serious attack and consequent dissolution of carbonaceous
matter would occur. Anodes of lead peroxide would possi-
bly serve, but have not yet been tried. It is of course evi-
dent that, as in the electrolysis of a solution of nickel salt
with an insoluble anode, energy must be supplied, not for
the mere transport of the nickel, but for its reduction to
metal ; the expenditure of electrical energy per unit weight
of nickel deposited will be greater than that necessary simply
for its refining. This question has been discussed fully
with regard to copper (p. 63), and need not be recapitulated
here. It may be noted in passing that in this industry,
as in other electrolytic manufactures, carbon electrodes of
high quality are much needed ; those at present made are
generally inferior to good retort carbon. Recently pure
graphite electrodes have been produced by the Acheson
process which have proved effective for many electrolytic
processes.
119
PRACTICAL ELECTRO-CHEMISTRY
COMMERCIAL ELECTROLYTIC NICKEL
In 1896 the Balbach Smelting and Refining Company
of Newark, New Jersey, began working up crude nickel
bought from the Orford Copper Company, which is engaged
in smelting the nickel ore from Sudbury, Ontario. The
composition of the crude nickel and that of three samples
of the electrolytically refined metal are given below.
CRUDE NICKEL
Ni
Cu
Fe
Si
C
s
REFINED
Per cent.
95-00
0-55
0-75
0-25
0-45
3-00
100-00
I.
a.
III.
Per cent.
Per cent.
Per cent.
Ni ....
99-48
99-17
99-20
Cu . . . .
0-10
Trace
0-14
Fe ....
0-48
0-66
0-58
S
0-29
0-03
0-03
100-35
99-86
99-95
The content of cobalt (which was probably present)
is not given. The presence of iron in considerable quantity
recalls the fact that Dr. Foerster found that metal to be
retained persistently when crude nickel containing iron is
refined electrolytically. The process used is kept secret.
It may be either electrolysis in sulphate solution, renewed
as the impurities (notably iron) accumulate, or, as suggested
by Titus Ulke, a cyanide method. In this case iron and
cobalt would tend to form complex stable cyanides, while
1 20
NICKEL
nickel would form ordinary double cyanides readily decom-
posed on electrolysis. Against this idea must be set the
fact that a cyanide bath is never used in nickel plating, and
it is doubtful whether a satisfactory deposit can be obtained
therefrom. A sample of electrolytic nickel, made by Messrs.
Gustav Menne & Co., of Siegen, Germany, was found to
contain 0-12 per cent, of lead, an impurity due to the fact
that it had been prepared* by the electrolysis of a solution
leached from a complex matte and not from a crude nickel
free from such extremely alien impurities. In quality it
was inferior to the American product.
A plant has been put down by the Canadian Copper
Company, of Cleveland, Ohio, to refine bessemerised matte
of the composition :
Per cent.
Ni 40-0
Cu 43-4
Fe 0-3
S 13-8
97-5
This matte also contains precious metals . viz. : Ag,
0-0218 per cent. ; Au, 0-0003 per cent. ; and Pb, 0-00155
per cent. The method proposed is as follows : The matte
may be used as such, or may be worked up to a copper-
nickel alloy. It may be remarked that, having regard
to general experience in the use of matte anodes, successful
refining of a matte with even relatively little sulphur is
unlikely. It may be assumed, therefore, that an alloy of
about 50 per cent, copper and 50 per cent, nickel would be the
raw material. This is cast into anodes and electrolysed in
a bath of copper sulphate acid with sulphuric acid. The
electrolyte is kept at a temperature of 30 C. = 86 F., and
is well circulated throughout the process. A current density
of 2-2 amperes per square decimetre is used at the beginning
of the operation, and is dropped to 0-8 ampere towards the
finish. Copper is dissolved and redeposited, while nickel
and a little iron remain in solution. As the electrolysis
121
PRACTICAL ELECTRO-CHEMISTRY
proceeds and the anodes are used up, the electrolyte gets
poorer in copper, and on this account the diminution of
current density becomes necessary. When the bulk of the
copper is deposited nickel tends to be thrown down if the
voltage is high enough). The rest of the copper can be
recovered, and the electrolyte thus freed from copper by
electrolysing with a small voltage and low current density,
using an anode of nickel unalloyed with copper ; but in
practice it is cheaper to precipitate the copper with sulphur-
etted hydrogen. The solution then contains nickel together
with a little iron as sulphates, and can be electrolysed with
insoluble anodes to recover nickel. It is to be noticed that
the published account, as is usually the case, stops short just
at the interesting part. The point is, how best may nickel
be deposited from a sulphate solution, using an insoluble
anode 'I The answer is not forthcoming from the description
made public.
According to the latest available information the use of a
matte is being abandoned in favour of extraction of the
mixed metals by an acid solution and the selective electrolysis
of the resulting liquid. The process is attributed to the
Canadian Copper Co., and it is said that a certain amount of
matte is used as anodes. Various accounts of processes
of this kind have been published, but they are all pleasingly
vague. The plain fact of the matter is that the separation
of nickel and copper is a simple matter when both metals
are in solution, and that a successful process must be directed
first to concentrating the ore so as to obtain the valuable
metals only ; secondly, to dissolving them ; and lastly, to
precipitating the copper in acid solution, leaving the nickel.
The mystery which has been made about the winning of
nickel is very much on all fours with that which has encom-
passed the refining of copper, the only difference being
that one is a little fresher than the other.
122
COBALT
No cobalt is prepared electrolytically on a commercial
scale. The foregoing pages contain references to such
experiments on its deposition as are likely to be of value if
its electrolytic preparation should need to be undertaken.
There is no immediate prospect of any requirement of this
kind, because metallic nickel is for most purposes as well
suited as cobalt and is greatly cheaper. With the present
relative abundance of the two metals the use of cobalt is
almost wholly confined to those purposes, e.g. the prepara-
tion of smalt and of glazes, in which the unrivalled blue of
its silicate is turned to account. The only case in which
the metal itself is preferable to nickel is in plating (q.v.), the
cobalt being stated, with some authority, to give a better
coating than does nickel.
123
TIN
ALMOST the sole source of tin is the native oxide Sn0 2 .
This body is relatively heavy, and can be separated from
the ores containing it by mechanical processes of concen-
tration. The reduction of the oxide thus separated from
the gangue can be effected without difficulty by means of
carbon. The resulting tin can also be refined to a degree of
purity sufficient for most purposes by ordinary dry methods.
Thus it conies about that there is little prospect of supersed-
ing the existing method of winning tin by any electrolytic
process. In the first place, stannic oxide is insoluble in any
agent that could be used for leaching the ore. Thus mechani-
cal concentration is inevitable. Given the concentrated
ore, its reduction to tin by carbon is by far the simplest
method of dealing with it. The only stage of the process
in which electrolytic means might be usefully employed is
in refining the crude tin. No serious attempt to do so
appears to have been made, although there is reason to
experiment in this direction, because commercial tin is often
comparatively impure (containing 0-5-1 per cent, of for-
eign metals), and because in the manufacture of certain
of the alloys of tin (notably gun metal) a pure metal would
be distinctly preferable to one containing miscellaneous alien
substances. Nevertheless, as a matter of fact, electrolytic-
ally refined tin has no industrial existence.
The case is somewhat different with scrap tin plate.
Articles such as household utensils, cans and boxes for
preserved goods and the like are usually made of what is
known colloquially as " tin," by which is meant tinned iron.
The manufacture of tin plate, i.e. sheets of iron coated with
tin, consumes the major part of the world's output of tin*
124
TIN
The metal is applied in as thin a film as possible, because it
is relatively expensive, but the aggregate quantity thus used
is very large. It has long been an object with inventors to
devise a means whereby the tin from used tin plate may be
recovered. The advantages to be derived from an efficient
process of recovery are palpable. The used tin plate (as
" tins " and the like) is a waste product ; the tin to be
recovered (amounting to about 5 per cent, of the weight of
tin plate) has a fairly high price, e.g. 60 80 per ton ; *
and the iron stripped of tin has a certain market value. The
value of the iron is smaller now than heretofore, because
ordinary tin plate is made from ingot iron (" mild steel "),
whereas puddled iron of good quality was formerly used.
In spite of this the scrap clean and free from tin would be
saleable. If imperfectly stripped and retaining some tin
its value would be smaller, because of the possible incorpora-
tion of this tin with the iron (to its detriment) on melting
the latter.
In practice the prospect of remunerative treatment of
tin scrap is less bright than would appear from this state-
ment of fact. In the first place, the raw material (old
" tins ") is hardly worth special collection, and must usually
be retrieved from dust bins and rubbish heaps. The supply
is apt to be uncertain, and the recovery is therefore somewhat
expensive. Next, the recovered tins are covered with mis-
cellaneous dirt, and have to be completely cleaned before
treatment. Thirdly, they are bulky and troublesome to
handle. Fourthly, they are extremely inconvenient to
strip electrolytically. Thus it has come about that most
of the methods which have attained even a qualified success
have been concerned with the treatment of the scrap, con-
sisting of the cuttings from new tin plate left as a waste
material from the manufacture of vessels for tinned goods.
These are clean and of such a shape as to be capable of being
packed in a space which is not excessive, and as they are
1 The fluctuations in the price of tin are large, owing chiefly to
speculative manipulations of the market.
125
PRACTICAL ELECTRO-CHEMISTRY
a factory bye-product, and do not need collection, one
cause of expense disappears.
Various methods have been proposed for treating tin
scrap. The scrap may be made the anode in an electrolyte of
dilute sulphuric acid, and the tin may be received on lead
or copper cathodes. Unfortunately, the tin dissolves less
readily than the iron, and as soon as the latter is exposed
its dissolution proceeds rapidly, and the bath becomes full
of ferrous sulphate, which is of low commercial value. The
exposure of the iron also tends to protect the remaining tin,
and the iron scrap is left imperfectly stripped, and therefore
of smaller value than if clean. The tin is deposited from
acid solutions in a spongy or pulverulent form, and its
fusion to form an ingot involves loss. Some market may,
however, be found for various salts of tin, notably stannous
chloride (made by dissolving the tin in hydrochloric acid),
which is used as a mordant. A more rational method
is to make the tin scrap the anode in a solution of caustic
soda, in which the metal is soluble, forming sodium stannate ;
the iron remains substantially unattacked. The stannates
are, however, somewhat unstable, and are easily decomposed
by carbonic acid, so that solutions exposed to the air are
apt to deposit tin as oxide. This tendency and their poor
conductivity have apparently prevented their successful
use. It may be noted that it is possible to strip tin plate
both by acid and alkaline solvents without the aid of a
current, and that, if the purely chemical method fails, there
seems to be no valid reason why an electrolytic method
should serve better. The difficulties of collection, cleaning
and handling mentioned above probably account for the
comparative failure of all methods of recovery, and the
remunerative utilisation of old " tins " and tin scrap is
likely long to be a pet problem for the professional inventor.
Attempts have been made to recover the tin from tinned
lead scrap. The lead sheet is often provided with a coating
of tin by covering thicker lead plate with tin and rolling
this down to the required gauge. Such tinned lead sheet
is used largely for bottle capsules. Recovery of tin from
126
TIN
these is easy, because, unlike iron, lead is electro-negative
to tin, and, on making the scrap tin the anode in an
electrolyte of sulphuric acid the tin dissolves, leaving the
lead unattacked. Both tin and lead are thus readily
separated and recovered. To the difficulty of collection
referred to above is to be ascribed the failure to base an.
industrial process on these principles.
127
ANTIMONY
THE chief ore of antimony is its sulphide, which is usually
reduced to metal by dry metallurgical processes. These
processes are relatively simple, not unduly expensive, and
produce a metal of sufficient purity for most purposes. It
is clear, therefore, that the need for an electrolytic process
is not great. The chief advantages that can be claimed for
a process of this kind are the possibility of treating ores too
poor to pay when smelted by the ordinary methods and the
feasibility of reducing the metal by water power in inacces-
sible districts where fuel is scarce. Such plain economical
considerations are too often overlooked when electrolytic
methods are invented or discussed.
At present only one process has succeeded in producing
metallic antimony on a commercial scale. It is worked
by Siemens & Halske, and the details of manufacture are
not publicly known. There is, however, a patent of the
same firm dealing with the same matter, and it is probable
that this patent describes and protects the process now being
worked. The leading principles of the patented process
are as follows : Antimony ore containing the- metal as its
sulphide (Sb 2 S 3 ) is leached with a solution of sodium sulphide.
The antimony sulphide dissolves, leaving the siliceous
gangue. The solution containing the antimony is then
passed through the cathode compartments of a series of
electrolytic cells, and is deposited on iron cathodes. The
anode compartments contain a solution of common salt in
which are carbon anodes ; chlorine is given off at these, and is
utilised for the manufacture of bleaching powder or chlorate.
The solution passing from the cathode compartments con-
sists chiefly of sodium sulphide containing little or no anti*
mony, and is used to leach a fresh portion of ore. The object
128
ANTIMONY
of thus working with a porous diaphragm and producing
a bye-product (chlorine) is to avoid the oxidation of the
leaching solution, viz. the sodium sulphide, which is inevit-
able if the electrolysis is conducted in an undivided cell and
the sulphide solution comes in contact with the anode.
The antimony prepared by the Siemens-Halske process
is in the form of plate about 2 mm. in thickness and having
a ridgy and warty surface, the appearance of which recalls
in some measure that of some samples of electrolytic copper.
The metal is nearly pure, and can, if necessary, be further
refined by the ordinary process of dry refining, which con-
sists in fusing the metal with a flux composed of crude
potash melted with antimony sulphide. This flux contains
potassium sulphide, which removes from the antimony any
residual antimony sulphide, forming a thioantimonite. For
most purposes, however, the antimony is pure enough in the
state in which it is deposited. The following analyses show
the quality of the unrefined electrolytic antimony, of the
same metal after refining, and of refined antimony prepared
by the ordinary dry process :
Unrefined
Refined
Refined antimony
electrolytic anti-
electrolytic anti-
made by dry
mony.
mony.
process.
Per cent.
Per cent.
Per cent.
As ...
Trace
S ......
0-288
0-0001
0-1000
Fe . . . .
0-008
0-0046
0-0100
Pb and Cu . .
Trace
0-0084
0-0303
Na . . . .
0-014
Sb ....
99-690
99-9869
99-8597
100-000
100-0000
100-0000
It will be seen that unrefined electrolytically prepared
antimony is almost pure, save for a little sulphur, doubtless
arising from the fact that the metal is deposited from a
solution rich in that element. Also, that refined electrolyti-
cally prepared antimony compares favourably with that
129 K
PRACTICAL ELECTRO-CHEMISTRY
made by the ordinary dry process. It is noteworthy that
antimony of good quality which has been cast in ingot
form shows its crystalline character by a well-marked
stellate appearance. This appearance is known as the
" star " of antimony, and is usually accepted as an index
of purity. It is evident that in the case of electrolytically
prepared antimony, stripped direct from the cathodes and
not melted and cast, this " star " is absent ; it is replaced
by the peculiar warty surface referred to above, which may
also be taken as an indication of the source of the metal and
as a guarantee of good quality. Although the Siemens-
Halske process is the only method by which antimony has
been successfully prepared on a commerical scale, other
methods have been devised and to some extent worked out.
Of these Borchers' process may be mentioned. Borchers
has studied the conditions of precipitation of antimony
from solutions of its sulphide in sodium sulphide, and has
designed a plant as the result of his experiments. It does
not seem, however, to have been tried on a manufacturing
scale. Using solutions of antimony sulphide (Sb 2 S 3 ) in
sodium sulphide (Na 2 S) with and without caustic soda, 1
and working without a diaphragm, he found that the whole
of the antimony could be deposited, but that at the anode
there was not merely a separation of sulphur (and consequent
formation of poly sulphides), but an oxidation of the sodium
sulphide to thiosulphate (hyposulphite). On account of
this action the sodium sulphide solution would decrease in
effectiveness as a solvent for fresh portions of antimony
sulphide, and the cyclical working of the process would be
impaired. A point would soon be reached at which the
sulphide solution, exhausted of its antimony, could no longer
dissolve a fresh quantity, and it would have to be replaced
by a new supply of sodium sulphide. Another difficulty
of the process is the fact that the antimony is deposited
in the form of powder, and has to be collected and fused
1 Two to three per cent, of common salt was added to improve
the conductivity of the electrolyte.
130
ANTIMONY
before it is marketable. Having regard to these funda-
mental defects inherent in the method, a discussion of the
merits of the plant proposed to work it is evidently super-
fluous.
Another method which claims attention is that of J. Izart.
In this a solution of sodium sulphide is used to leach anti-
mony sulphide ores, and the resulting solution containing
sodium thioantimonite is electrolysed in the cathode com-
partment of a cell in the anode compartment of which is
a solution of carbonic soda. The object of this arrange-
ment is to prevent the formation of polysulphide, which
substance would not be serviceable for extracting a fresh
portion of the ore, and would also tend to redissolve the
deposited antimony. By the use of a porous partition and
of a solution of caustic soda on the anode side these incon-
veniences are avoided and the antimony is deposited on
the cathode, sulphur (as sodium sulphide) appears in the
anode compartment, and the only waste lies in the consump-
tion of a quantity of caustic soda equivalent to the sulphur
originally present as antimony sulphide. In short, there
is a surplus of extracting liquor which has to be paid for by
the purchase of caustic soda. Unless some inventor can
devise a method of oxidising or removing the sulphur
originating from the antimony ore without forming a poly-
sulphide or consuming a fresh quantity of caustic soda, this
difficulty must be faced. The whole matter is a little academ-
ical, because the trade in antimony is not large and there is
no acute competition in supplying its requirements. If a
new thermo cell, of high efficiency, with antimony as one
element were devised, poor ores would be in demand and
more would be heard of electrolytic methods of winning this
metal.
ZINC
ZINC is a metal the winning of which by electrolysis pre-
sents peculiar advantages. Its refining, on the other hand,
can be best accomplished by non-electrolytic processes.
The commonest ore of zinc is blende (zinc sulphide), from
which zinc can be extracted by the usual metallurgical
methods only after the ore has been roasted and a crude zinc
oxide produced. This oxide, on heating with carbon, is
reduced, yielding metallic zinc. To effect the reduction a
temperature of about 1,300 C. = 2,372 F. is required ;
the boiling-point of the reduced metal is, however, only
930 C. = 1,706 F. From this it follows that, when a
mixture of zinc oxide and carbon is heated to a temperature
sufficiently high to reduce the oxide to metallic zinc, the
metal is generated as vapour, and cannot be directly run
down to a regulus, as can less volatile metals, e.g. copper
and iron. In consequence of this the winning of zinc by
ordinary metallurgical methods is always effected by dis-
tilling a mixture of the oxide and carbon (powdered coke
or non-caking coal) in retorts of refractory fireclay. (The
bearing of this disquisition on the electrolytic winning of
zinc will be seen immediately.) The reduction of zinc
oxide to zinc is represented by the equation
ZnO + C + Zn + Co,
and absorbs 56 Cal per gramme equivalent of zinc obtained.
This quantity of heat has to be supplied to the charge
through the walls of the retort, and even the best methods
of heating for this purpose are so wasteful that the quantity
of fuel used vastly exceeds the calculated minimum.
Another large cause of expense is the renewal of the
somewhat costly and fragile retorts in which the distilla-
132
ZINC
tion is conducted. From a consideration of these facts
it is evident that there is ample room for an economical
method of winning zinc from its ores, whether by electrical
or other means.
Recognition of this circumstance induced the Brothers
Cowles about the year 1882 to attempt to distil zinc in an
electric furnace, the form of which is shown in the figure.
The fireclay retort A is embedded in a refractory non-con-
ducting material B, and is closed by a graphite crucible D,
which serves as a stopper and as a receiver of the zinc dis-
tilled from the retort. The current is passed through the
FIG. 25.
charge in the retort between the plug D and a graphite plate
0, which forms the other end of the retort. Gas, e.g. CO,
generated during the reduction of the zinc escapes through
the pipe E. The principle underlying this endeavour is
perfectly sound, but the apparatus is not well adapted for
its purpose, and did not succeed in practice. Recently
fresh attempts have been made to realize this idea. Dorse-
magen has proposed to heat a charge of calcined siliceous
zinc ore and coal in a furnace of the crucible type with ver-
tical electrodes. He claims that zinc is reduced and vola-
tilized, and silicon carbide is left. Experiments have been
made at Crampagna, Ariege in France, on the reduction of
133
PRACTICAL ELECTRO-CHEMISTRY
zinc electrically. The furnace took 100 kilowatts, and using
an ore containing 40 per cent. Zn., yielded about 5 kilo-
grams of zinc per kilowatt day. It is stated that 90 per cent,
of the zinc in the ore can be obtained, and that raw blende
can be used.
Putting this aside as a matter for inquiry and not an
accomplished process, one finds that already there exists
a growing industry in the production of electrolytic zinc.
Zinc is a metal so electropositive, and needing so much
energy for its reduction, that when aqueous solutions of
its salts are electrolysed there is a tendency to produce
hydrogen instead of zinc at the cathode. Moreover, from
most zinc solutions the metal is deposited in a spongy and
incoherent condition, unless special conditions, e.g. as
regards acidity and current density, are fulfilled. These
circumstances have rendered the device of a workable
method for depositing zinc electrolytically peculiarly diffi-
cult.
As mentioned at the beginning of this chapter, zinc is
not refined electrolytically. In the event of a demand for
especially pure zinc arising, it could at once be met with
ease by the fractional distillation of ordinary commercial
zinc in vacuo a process which can be accomplished at a
temperature but little above the softening-point of glass,
i.e. at a barely visible red heat. 1 The description of electro-
lytic processes for zinc will, therefore, relate chiefly to those
concerned with the winning of the metal from its ores.
PRINCIPLES OF ELECTROLYTIC DEPOSITION OF
ZINC
Several conditions must be carefully observed in order
to obtain a coherent deposit of zinc. Many inventors and
1 It is a curious fact that lead, which by itself is not volatile at a
low temperature, has a strong tendency to pass over with the zinc.
Therefore if zinc is distilled indiscriminately lead will be carried
over, but fractional distillation would probably allow of the pre-
paration of a metal substantially free from lead.
134
ZINC
investigators have laid down precautions more or less
empirical, but their instructions need not be considered,
because the whole subject has been investigated in the
most thorough manner by Mylius and Fromm (Zeits. /.
anorganische Chemie, 1895, p. 144), and from the data
which they have established by small scale experiments
the working conditions in manufacture can be deduced.
It must not be supposed that such knowledge can be trans-
lated at once to the works with a certainty of immediate
success ; nothing but close study of the actual working
of a process on a commercial scale will suffice ; nevertheless
the guiding principles which must be regarded are estab-
lished, and each manufacturer must apply them for himself.
This may seem cold comfort to the technologist, but it is
all he can expect to get, for, as a matter of fact, the few
processes for the electrolytic reduction of zinc which are
working successfully are guarded as secrets in the details.
It is not to be supposed that in these there is any great
divergence from what is common knowledge, but it is fair
to conclude that by attention to numerous small points
of working the manufacturers using these processes have
been able to apply remuneratively the principles about to
be discussed, and it is manifestly unreasonable to expect
them to make public what has been acquired at the cost of
much time, money and labour.
The chief difficulties in the electrolytic deposition of
zinc from a solution of its sulphate are :
(1) The evolution of hydrogen at the cathode instead
of the deposition of zinc there.
(2) The precipitation of the zinc in a spongy condition.
As might be predicted, the evolution of hydrogen is
most apt to occur when the electrolyte is poor in zinc, for
in that case there are likely to be too few zinc ions at the
cathode at any given instant, and the current is thus occu-
pied in the liberation of hydrogen from the water or sulphuric
acid which is relatively abundant in the neighbourhood of
the cathode. The electrolyte should, therefore, be fairly
concentrated, e.g. should contain at least 10 per cent, of
135
PRACTICAL ELECTRO-CHEMISTRY
the crystallised salt ZnS0 4 7 H 2 O. Next, it should be
neutral or slightly acid. If unduly acid, hydrogen as well
as zinc will be liberated at the cathode. Thirdly, a high
current density should be used, e.g. 1-2 amperes per square
decimetre, i.e. about 9-18 amperes per square foot. With
a concentrated electrolyte the current density may be
considerably increased and good adherent deposits may be
obtained. Fourthly, the electrolyte must not be basic,
i.e. it must contain no zinc oxide over and above that
necessary to form a neutral salt. Neutral zinc salts dissolve
small quantities of zinc oxide, and from such solutions
spongy zinc is precipitated. It must also contain no oxidis-
ing impurity. These last two conclusions were arrived at
by Mylius and Fromm irom a systematic study of the
character of the spongy zinc which is often deposited. It
has been suggested that the formation of this spongy zinc
is caused by the presence of a hydride (ZnH 2 ). There is
no evidence of this, and against it is the fact that the spongy
deposit always contains zinc oxide or a basic salt of zinc,
which can be detected and isolated by dissolving the metallic
zinc in mercury. The quantity of oxide thus left is under
1 per cent., but is sufficient to produce sponginess. When
to a solution of a zinc salt a small quantity of an oxidant,
e.g. hydrogen peroxide or zinc nitrate, is added, such a
solution on electrolysis yields spongy zinc ; under identical
conditions of temperature, concentration, current density
and the like, a solution free from these oxidising impurities
gives a normal deposit of coherent reguline zinc. Curiously
enough the presence of a small quantity of arsenic or anti-
mony in the electrolyte will cause the formation of spongy
zinc ; the rationale of their action is obscure, but the obser-
vation is important in that it indicates that the electrolyte
must be carefully purified for the successful deposition of
zinc on a commercial scale. The fact that the presence
of zinc oxide induces the deposition of spongy zinc explains
why a basic electrolyte is peculiarly apt to produce an
unsatisfactory deposit ; also, seeing that a strong solution
of a neutral zinc salt, such as the sulphate, will dissolve
136
ZINC
more zinc oxide than will a weak solution, it may be expected
that in a strong solution a slight excess of base will be less
detrimental than in a weak solution. Experiment shows
that that is the case.
Foerster and Giinther have made a study of the con-
ditions necessary to be observed in order to obtain a good
coherent deposit of zinc from solutions of its chloride. This
study forms a useful supplement to the work of Mylius and
Fromm, cited above. The electrolysis of zinc chloride in
aqueous solution may prove applicable in metallurgical
practice, and a knowledge of its principles cannot be neg-
lected. As in the case of the sulphate, the chief difficulty is
in obtaining the metal in a reguline and coherent condition.
There is an inconvenient tendency to form spongy deposits.
In the experiments about to be described, Silesian zinc of
exceptional purity, containing not more than 0-03 per cent,
of lead and 0-05 per cent, of iron, was used as the anodes.
The cathode was a piece of polished sheet zinc. A solution
of zinc chloride was used as the electrolyte, and was tried
neutral, acid and basic in turn.
It being established that the production ot spongy zinc
is primarily caused by the presence of zinc oxide, it appears
probable that deposition of zinc of good quality is more
likely to be attained with a solution of zinc chloride than
with one of zinc sulphate, because zinc oxide is more soluble
in the former, and is therefore less likely to make its appear-
ance at the cathode and impair the quality of the zinc there
precipitated. This is borne out by experiment, for a
solution of zinc chloride, containing 54-6 grammes of Zn per
litre, when electrolysed with a current density of 1-4 amperes
per square decimetre, continued to give a good deposit
until the electrolyte became so basic as to form a precipitate
of zinc oxy chloride. This occurred when there was present
for every 14 molecules of ZnCl 2 1 molecule of ZnO in solu-
tion. An obvious advance on this is to use a slightly acid
solution of zinc chloride to hinder the formation of a basic
chloride. But when the electrolyte is acid, hydrogen as
well as zinc appears at the cathode, current is wasted, and
137
PRACTICAL ELECTRO-CHEMISTRY
the deposit becomes uneven because of the local irregu^
larities of current density, due to bubbles of hydrogen,
causing spots and patches on the surface to be inaccessible
to the electrolyte while they persist there. A device which
has been employed by Mylius and Fromm can be resorted
to for the suppression of this hydrogen. It consists in
placing a small independent anode near the cathode and
passing by its means into the electrolyte a current sufficient
to evolve enough chlorine to combine with the objectionable
hydrogen. By adopting this plan a good deposit of zinc
can be obtained in a slightly acid solution of zinc chloride.
This observation is specially worthy of remark, because it
probably explains the attempts that have frequently been
made, as in the Ashcroft process (q.v.), to obtain good
deposits of zinc by the use of an oxidising agent. These
attempts have occasionally succeeded, although usually
based on erroneous assumptions, e.g. that the sponginess
of the deposited zinc was due to the presence of a zinc
hydride. We now see the true reason, viz. that the use
of an oxidant in regulated amount allowed an acid electro-
lyte to be used (thus avoiding the deposition of a slightly
oxidised, and therefore spongy, zinc), while at the same time
suppressing the hydrogen, which is liable to cause local
irregularities of current density, and therefore rough, warty
deposits.
As might be premised from the work recorded above, a
basic solution of zinc chloride, if not containing so much
oxide as to make it turbid, may give good deposits at first ;,
as the process goes on it becomes more basic and spongy
zinc begins to be formed.
In these experiments it was noticed that before the
electrolyte became so basic as to be turbid the deposit
began to change in character, forming long growths (appar-
ently of compact reguline zinc) from the edges of the
cathodes.
It must not be supposed that zinc oxide * in the electro-
1 Later researches throw doubt on the belief that a spongy de-
posit is necessarily caused by zinc oxide, for in a solution containing
138
ZINC
lyte is the only material capable of causing the formation
of spongy zinc. Various foreign metals in the electrolyte
have the same effect, and on this account the industrial
electro-deposition of zinc, especially from solutions obtained
by leaching out complex ores, will always be a somewhat
delicate operation, requiring care and skilled supervision.
The energy required to reduce zinc sulphate electro-
lytically to metallic zinc can be readily computed. The
decomposition represented by the equation
ZnS0 4 Aq = Zn + H 2 S0 4 Aq +
requires for its realisation the expenditure of 106 Cal, i.e.
106 Cal must be provided for winning 65 grammes of zinc.
This corresponds with 2.564 H.P. hours per ton ; therefore
the theoretical output of zinc per H.P. year (365 days of
twenty-four hours each) is 3-42 tons.
Now the critical voltage for the decomposition of zinc
sulphate (calculated from its heat of formation in manner
similar to the example already given) is 2-25 volts. To
obtain the output per H.P. year given above it would be
necessary to work at the critical voltage. But in practice
a voltage of about twice this, viz. 4-5 volts, would pro-
bably be required. Further, having regard to the
tendency for the current to reduce hydrogen instead of
zinc the current efficiency is not likely to be more than 80
per cent. ; the voltage efficiency is 50 per cent., therefore
the energy efficiency is per cent. = 40 per cent.
It follows that the output per H.P. year is not likely to
exceed 1-368 tons. With cheap water power, costing say
2 10,s. per H.P. year, the cost of energy for reducing 1 ton
of zinc is 1 16s. 6d. With steam power at Id. per H.P. hour,
i.e. 9 165. per H.P. year, the cost for 1 ton of zinc would be
7 3-s. 3d. The selling price of zinc being about 20 per
ton, it is clear that the cost of electrolytic reduction by
excess of caustic alkali, and therefore capable of dissolving zinc
oxide, spongy zinc may form.
139
PRACTICAL ELECTRO-CHEMISTRY
steam power would be a large part of the whole value of
the product, and that the margin for such heavy expenses
as roasting, extracting, maintenance of plant to say
nothing of the cost of the zinc in the ore is inconveniently
small. It is only where very cheap water power is available
that the electrolytic winning of zinc from its aqueous solutions
may be practised with a fair prospect of success. The case
is somewhat different where the zinc is, as it were, a bye-
product. Processes falling under this head will be dealt
with below.
PROCESSES FOR THE PRODUCTION OF
ELECTROLYTIC ZINC
Usually these processes have been designed to produce
zinc as a bye-product of some other manufacture, and not
for the winning of zinc from its ores as the principal object,
One of the chief causes of the various attempts which have
been made to invent a workable electrolytic process for
zinc is the growing necessity of treating mixed sulphide
ores, consisting of blende and galena (zinc sulphide and
lead sulphide) so intimately associated that their separation
by any method of mechanical " dressing " is well-nigh
impracticable. Such ores are also difficult to smelt by the
ordinary processes, and many plans have been proposed
to treat them by wet extraction methods.
THE SIEMENS-HALSKE PROCESS
Ore consisting essentially of lead sulphide, zinc sulphide
and gangue, and containing about 20 per cent, of zinc, 30
per cent, of lead, and 20 ounces of silver per ton, is roasted
at a low red heat so as to oxidise the sulphides and convert
them into oxides and sulphates.
It is desirable that the temperature should be kept low,
in order that a large proportion of the sulphides should be
converted into sulphates instead of oxides.
This requires a long time and much stirring of the ore.
Altogether this stage of the process, which sounds simple
140
ZINC
enough, is rather difficult and expensive. The roasted
ore is extracted with dilute sulphuric acid (about 10 per
cent, strength), and the zinc is dissolved as sulphate, leaving
the lead (also as sulphate) as an insoluble residue. This
is smelted by the usual dry methods. The bulk of the
silver, which is always present in ores of this class, is also
left with the lead, though some may go into solution. Of
course the value of the silver is an important part of the
whole value of the ore, and its careful extraction and recovery
are necessary to make the process remunerative.
The solution of zinc sulphate needs to be purified from
iron, copper, and other foreign metals by ordinary chemical
methods, such as limited precipitation with lime and chloride
of lime. The preparation in this manner of a tolerably
pure solution of zinc sulphate is by no means an easy matter.
These non-electrolytic parts of the process are the cause of
quite as much difficulty as the electrolysis itself. When a
satisfactorily pure solution of zinc sulphate has been obtained
it is electrolysed, lead anodes being used and thin zinc
cathodes. The conditions, stated above, necessary for
obtaining a good coherent deposit of zinc must be carefully
observed. In this process zinc is not merely transferred,,
it is actually reduced from the solution of its sulphate, and
the electrolyte becomes more acid as the reduction proceeds.
When the acidity is so great as to cause the evolution of an
unduly large amount of hydrogen at the cathode, the solu-
tion is run off and used again for extracting roasted ore.
Thus there is in circulation a large quantity of a solution
of zinc sulphate, acid with sulphuric acid, which is alternately
robbed of a portion of its zinc and again supplied with an
equivalent amount. But in each cycle of operations the
solution acquires impurities from the roasted ore, and these
must be eliminated before it can be used again as an electro-
lyte. The process has been tried by the Smelting Company
of Australia, at Illawarra, in New South Wales, but no
information as to its success has been published. Its.
weak points are sufficiently indicated in the foregoing
description.
141
PRACTICAL ELECTRO-CHEMISTRY
THE ASHCROFT PROCESS
This process is designed to work up refractory sulphide
ores of the same grade as those intended to be treated by
the Siemens-Halske process (q.v.). The ore is finely ground
and is roasted to convert the sulphides of zinc and lead
into oxides and sulphates.
The difficulty and expense attending the thorough roast-
ing of this class of ore have already been spoken of. The
remarks then made apply equally here. The solvent used
is ferric chloride, which is used up and replaced (not re-
generated) in the manner about to be described. In the
first place the roasted ore is leached with a solution of
ferric sulphate and chloride ; zinc sulphate and chloride
are formed and ferric hydroxide is precipitated. The
-extracted residue of lead sulphate, gangue, and ferric hydrox-
ide is smelted in the usual manner, the oxide of iron aiding
as a flux. The solution containing zinc is first passed over
scrap zinc to precipitate any silver which may be in solution,
and is then circulated through the cathode compartments
of a series of electrolytic cells and there deposits a portion
of its zinc. On cathodes of sheet zinc the good and coherent
quality of the deposit of zinc is said to be promoted by
-allowing the solution to be slightly basic. Now, seeing
what has been said above (p. 135) on the bad influence of
basic salts of zinc on the quality of the metal deposited,
it is fairly evident that it is unlikely that this method of
working can be successful. Only about one-third of the
total quantity of zinc in solution is deposited during the
passage of the electrolyte through the cathode compart-
ments, and the liquor then passes to the anode compart-
ments, which are separated from the cathode compartments
by a porous partition of cloth. The level of the electrolyte
in the anode compartments is kept lower than that in the
cathode compartments in order to prevent the liquid from
passing from the anode to the cathode compartment through
the diaphragm. This is because the anode liquid contains
iron salts, which would interfere with the deposition of the
zinc if they found their way into the cathode compartment.
142
ZINC
In some of the anode compartments are iron anodes, which
dissolve in proportion as zinc is deposited in the cathode
compartments, forming ferrous sulphate (or chloride). In
the remaining anode compartments, viz. those through
which the liquor passes out of the group of electrolytic
cells, the anodes are of carbon instead of iron. At these
insoluble anodes the ferrous salts previously formed at the
iron anodes are oxidised to the ferric state, and the liquor
becomes capable of acting again as a leaching agent for a
fresh portion of the ore. It will be seen that the process
is comparatively complex. The solvent action of the
solution of ferric salts is by no means particularly vigorous,
and anything approaching complete extraction of the ore
is difficult to attain. The precipitation of the iron by
means of the zinc oxide in the roasted ore is difficult to
effect completely, and if iron be left in solution the deposition
of the zinc is interfered with. The plan of depositing zinc
from a slightly basic solution is (as has been shown above)
based on an erroneous view, and is likely to hinder rather
than help. The renewal of the leaching liquor by the dis-
solution of iron anodes and the subsequent oxidation of
the ferrous salts thus produced necessitates the use of a
diaphragm to prevent commingling of the anode and cathode
liquors. The theoretical advantage gained is that the com-
paratively cheap energy rendered available by the dissolu-
tion of the iron aids in the deposition of the zinc by reducing
the voltage required for this purpose.
Whether as much as is gained by this is not lost by the
increased resistance of the electrolyte and diaphragm is a
nice point. That these difficulties are not imaginary is
shown by the fact that the sulphide corporation which
worked the Ashcroft patents spent large sums of money
without bringing the method to a successful issue. Great
efforts were made to put the process on a working basis,
and the history of these attempts is contained in a paper
by Mr. Edgar A. Ashcroft, which was read before the Insti-
tute of Mining and Metallurgy in June, 1898. The gist of
this paper, as far as is necessary for a comprehension of the
143
PRACTICAL ELECTRO-CHEMISTRY
difficulties encountered in the treatment of mixed sulphide
ores, is given in the ensuing paragraphs.
THE ASHCROFT PROCESS AS WORKED AT
COCKLE CREEK
The ore treated proved to be poorer than was antici-
pated, containing about 20 per cent, of lead, 25 per cent,
of zinc, and 10 ounces of silver per ton, instead of 30 per
cent, of lead, 30 per cent, of zinc, and 45 ounces of silver
per ton, as was expected. Thus, ore containing metals
with a gross assay value of only 7 18s. per ton was available
instead of ore worth 13 85. Seeing that all calculations of
profit had been made on the latter, the ultimate failure of
the process is not surprising.
The ore is dried, ground in Krupp ball mills to a fineness
such that it will pass a 60 x 60 mesh sieve, and roasted in
a long reverberatory furnace with a terraced hearth, so
that the roasting can be done systematically and the ore
well turned and rabbled as it descends from the higher steps
at the far end of the hearth to the lower steps nearer the
bridge.
The roasting is conducted at as low a temperature as
possible, in order that the product may be sulphate rather
than oxide. The operation is carried out by hand labour,
but would probably be better effected in a mechanical
roasting furnace. The roasted ore is reground and its zinc
leached out by means of " sulphuric acid, with or without
ferric sulphate." In the paper from which these facts are
taken it is not specifically stated that the use of ferric salts
as leaching agents has been abandoned, but the general
tenor of the description conveys the impression that this is
the case. In fact it may fairly be assumed that, at least in
the later stages of the trial, the roasted ore was simply
extracted with a solution from the cathode compartments
of the cells, containing free sulphuric acid, and of course
zinc sulphate. The characteristic reaction on which the
claim of the process to be considered novel is based, thus
disappears.
144
ZINC
The leaching is done in large wooden vats with agitators,
and the solution is kept at about 80 C. = 176 F. When
the bulk of the free acid is neutralised and the greater part
of the zinc in the ore is extracted, the mixture is filter-
pressed, the residue sent to the smelting furnaces for reduc-
tion to argentiferous lead, and the solution of crude zinc
sulphate purified in order to make it fit for electrolysis. Iron
is always present, and is peroxidised in the anode compart-
ments and precipitated when the electrolyte, partially de-
pleted of zinc, is used to leach out a fresh portion of ore.
Manganese is also present, and is considered very objection-
able. It is stated that its removal by means of bleaching
powder and other oxidising agents is too costly, and there-
fore it is allowed to accumulate until its influence becomes
excessive, when a portion of the liquid is removed and re-
placed by dilute sulphuric acid or a fresh extract from the
ore. The fraction of the liquid thus taken out of the cycle
of operations may be worked up for zinc by evaporation
to dryness, decomposition of the zinc sulphate by heating
(the S0 2 + evolved being reconverted into sulphuric
acid), and reduction of the crude zinc oxide thus obtained
by the usual process of distillation with carbon. It is stated
that sufficient purification of the liquor to be electrolysed
can be effected by allowing it to fall in cascade over cast-
iron scrap or borings, and that this (apart from the periodi-
cal necessity for removing a fraction on account of the
accumulation of manganese) is the only operation necessary
between the leaching vats and the electrolytic cells. In
spite of these attempts at simplification, the electrolytic
separation of zinc by this process has not yet proved to be
successful.
Since the failure of the process tried at Cockle Creek,
Ashcroft, in collaboration with Swinburne, has devised a new
process for the treatment of mixed sulphide ores. This
" Phoenix " process is not in its essence electrolytic, for its
fundamental reaction consists in attacking the sulphides
with chlorine at a low red heat in a vessel resembling a
converter, tapping off the mixed chlorides, selectively
H5 L
PRACTICAL ELECTRO-CHEMISTRY
precipitating the chief metals other than zinc, and finally
obtaining a solution of zinc chloride approximately pure.
This solution is boiled down and the zinc chloride electrolysed
in the fused state. The anodes are carbon and the cathode
consists of fused zinc ; the electrolyte is kept fused by the
current. The zinc is drawn off at intervals and the chlorine
sent back to the converter to chlorinate a fresh portion of
ore. Quite apart from the merits of the process for treat-
ing complex sulphide ores, the mode of electrolysis is worthy
of attention. The inventors have recognised that zinc is
one of those metals best reduced from a fused electrolyte ;
they have also realised that the electrolyte is best kept fused
by heat internally generated. These principles may be applied
to the reduction of zinc from its salts irrespective of the
source of the zinc ; it need not be derived from complex
sulphide ores.
THE HOEPFNER PROCESS
Numerous processes for the electrolytic winning of zinc
have been devised and patented by Hoepfner. One of these
has been worked on a semi-manufacturing scale (about
100 H.P. being used) at Friifurt in Germany. The raw
material is an iron ore containing about 10 per cent, of
zinc. This is roasted and extracted with sulphuric acid ;
the solution is treated with common salt in order that the
zinc may ultimately be obtained as chloride. The liquor
is purified from manganese by the use of caustic soda and
bleaching powder, and from lead and copper by means of
zinc dust. The purified solution is cooled to 25 F., the
sodium sulphate crystallised out, and the resulting solution
of zinc chloride is electrolysed. The anodes are gas carbon,
and the cathodes are revolving zinc plates. They are
separated by diaphragms of nitrated cellulose. A pressure
of 3-7 volts per cell is required, and a high current density as
much as 36 amperes per square foot may be used. The
electrolyte is circulated independently through the anode
and cathode compartments. The products are zinc, which
146
ZINC
is obtained on the revolving cathodes in a coherent state,
and chlorine, which is used for making bleaching powder.
This process has recently been modified and developed
to some extent. The ore after roasting for sulphur is
mixed with about 20 per cent, of salt and is again roasted to
chloridise the zinc. The solution obtained by extracting
the roasted mass is cooled to 5 C. = 23 F. to separate
the sodium sulphate found in the process of roasting, and the
solution of zinc chloride is electrolysed in cells provided
with diaphragms of nitrated cotton. Carbon anodes are
used ; the cathodes are discs which rotate and the electrolyte
is kept fairly strong in order that a coherent deposit may be
obtained. The plant is stated to treat 18 tons of spent
ore (containing 10-16 per cent, of zinc) per day.
Another process devised by Hoepfner is being worked
by Brunner, Mond & Co. in this country. As far as the
electrolysis is concerned, it is generally similar to that
described above, the electrolyte being a solution of zinc
chloride. The adoption of a process of this kind by an
alkali works becomes intelligible when it is considered
that the electrolyte (zinc chloride) is obtained by acting
on zinc oxide (roasted zinc ore) with calcium chloride solu-
tion and C0 2 ; calcium carbonate is precipitated, and zinc
chloride goes into solution. It is said that this reaction
works smoothly. The zinc obtained may be regarded as a
bye-product, covering the cost of the ore and part of that
of the process, the real object of the alkali maker (using the
ammonia-soda process, and therefore not obtaining hydro-
chloric acid as a bye-product, as does the Leblanc maker)
being to recover chlorine from his waste calcium chloride
liquors. The plant at Brunner, Mond & Co.'s works is about
to be increased to 1,200 H.P. The output is estimated at 4
kilos of zinc per horse power per day.
MOND PROCESS
Mond has devised the following apparatus to overcome
the difficulties experienced in obtaining a good adherent
deposit of zinc.
147
PRACTICAL ELECTRO-CHEMISTRY
The cathode consists of not fewer than three long, rotating
mandrels, the bearings of which are arranged in such a way
that horizontal motion is permitted. These cylinders are
kept pressed together by means of springs and are slowly
rotated in the electrolyte. To prevent the same parts of
the surface of the cylinders from coming in contact too fre-
quently, the cylinders are all of different diameters, and to
give the deposit a good burnish a slight sliding motion is im-
parted to one or more of the mandrels whilst revolving.
A solution of zinc chloride is used and insoluble carbon
or lead anodes are said to be employed.
FIG. 26.
The deposited zinc is removed as tubes from the mandrels
and is cast into ingot form in the ordinary way ; it is stated
to be almost pure.
This corresponds with 1*46 tons per H.P. year as against
1-368 tons calculated on the data given on p. 139 which
relate to the decomposition of the sulphate.
THE DUISBERG PROCESS
A process which is successfully producing zinc on a com-
mercial scale is that devised by Dieffenbach and worked at
148
ZINC
Duisberg in Germany. The details of the process are kept
secret, but it appears that the electrolyte is a solution of
zinc chloride. It is probable that such an electrolyte will
prove better than one of zinc sulphate, partly because the
carbon anodes last better in a chloride solution, partly
because chlorine is a valuable bye-product. The success
of the process at Duisberg may be gathered from the state-
ment that the output is 90 tons of zinc per month and that
the plant is being increased.
ELECTROLYSIS OF FUSED ZINC CHLORIDE
The use of fused zinc chloride instead of aqueous solu-
tions of the chloride and other salts of zinc has attracted
the attention of inventors ; with one exception no work-
ing process has resulted from their efforts, but the appara-
tus, devised by Borchers (shown in fig. 27), will serve as an
example of the attack of the problem on rational lines.
A leaden vessel A having a grooved rim is used to contain
the fused zinc chloride. The rim is filled with zinc chloride
in the fused state and the cover D placed in position. Water
is turned into the trough c surrounding the grooved rim,
and the zinc chloride is thus caused to solidify, sealing the
cover. A sheet of zinc B bent to the shape of the vessel is
149
PRACTICAL ELECTRO-CHEMISTRY
used as the cathode, and the carbon rod E as the anode.
F is a pipe serving to carry off the chlorine, and G is a plug
closing a hole through which fresh zinc chloride can be intro-
duced from time to time. At the beginning of electrolysis
the main quantity of zinc chloride in the vessel A is fused
by the application of external heat ; afterwards it can be
kept fused by the heat generated by the passage of the
current, provided a sufficiently high current density be
used. The two weak points in the process are the diffi-
culty of preparing zinc chloride in quantity anhydrous and
sufficiently pure to serve for the preparation of electrolytic
zinc, and the fusibility of the leaden vessel. Lead melts
at 325 C. = 617 F. and zinc chloride fuses at 262 C. =
504 F., so that the margin of safety is not large.
A far better apparatus has been devised in connection
with the Phoenix process (p. 145). The purified ^zinc
chloride is electrolysed in the vessel shown in the figure,
which is a tank (A) built of firebrick, and having three car-
bon anodes of peculiar shape ; one (B) is shown in the
figure, which represents a section of the tank. The cathode
is a layer of fused zinc, C. D is the steel connection with
the cathode, and E is a plug closing the tapping tube
through which the fused zinc can be drawn. The electro-
lyte is kept fused by the current, and the whole arrangement
is comparable with a cell for the reduction of aluminium.
150
ZINC
WORKING UP " ZINC AMALGAM " FROM THE
PARKES PROCESS
One of the most effective methods of desilverising lead
consists in treating the molten metal with an immiscible
solvent, viz. zinc. 1 The zinc floats on the surface of the
lead and is periodically removed. During its contact
with the lead to be desilverised, the zinc absorbs not only
silver, but also a certain amount of lead. It also oxidises
to some extent. Thus it comes about that the " zinc
amalgam," as the crust of zinciferous matter floating on the
bath of lead is termed, is a loose, friable mass, varying con-
siderably in composition according to the conditions of
working. Its composition ranges from 55 to 77 per cent,
of lead, 12 to 40 per cent, of zinc, 2'5 to 5 per cent, of silver,
with various other metals and oxides. The ordinary
metallurgical method of working up this complex alloy
consists in liquating the excess of lead (which is returned to
the desilverising pots), and distilling the residual mixture
of zinc, silver and oxide of zinc with a little charcoal. The
zinc already present as metal and that reduced from the
oxide by the charcoal distils off, and crude silver remains,
which is purified in the usual way. It is proposed to im-
prove on this process by refining the alloy electrolytically,
the object being to dissolve out the zinc and to leave the
lead and silver. This is not altogether easy, because the
alloy is too brittle and contains too much oxide to be cast
into plates. Also the quantity of soluble material (zinc)
which is to be extracted is small compared with the quantity
of insoluble material (lead and silver). Thus, any form of
anode will become crusted with this insoluble material,
and its dissolution will be hindered thereby. A better grade
of " zinc amalgam " is said to be produced when the desilver-
ising of the lead is conducted with zinc containing a small
percentage of aluminium, in that it contains a smaller
1 A full description of this elegant process can be found in any
good metallurgical text-book.
PRACTICAL ELECTRO-CHEMISTRY
proportion of lead. But even in this case the product con-
tains much lead, dissolves slowly, and cannot be cast into
serviceable anodes. It must therefore be treated in frag-
ments, lying loose on a plate, or contained in a metallic
basket serving as the anode. Such a receptacle may be made
of lead. A solution of zinc sulphate should serve as the
electrolyte, and the conditions for the deposition of zinc in
a coherent form should be maintained as nearly as possible
like those which have been already laid down as suitable
for the winning of zinc (p. 134 et seq.). After a time the
outer parts of the fragments of zinc-lead-silver alloy will
become robbed of their zinc and converted into a spongy
mass of argentiferous lead. A kernel of unattacked zinc-
lead-silver alloy will remain. The dissolution of the zinc
from this will be slow, partly because its conductive con-
nection with the plate or basket serving as the anode is
impaired by the formation of a film of lead sulphate on the
spongy lead. A remedy for this state of things is the removal
of the partly spent fragments, the liquation of the coating of
argentiferous lead from the core of zinc-lead-silver alloy, and
the re treatment of the kernels thus isolated. This is a cum-
brous arid costly proceeding, and is not likely to conduce to
the success of the process.
Somewhat sketchy information is extant concerning the
treatment of " zinc amalgam " containing aluminium. It is
said that this can be successfully worked up in an electrolyte
consisting of a strong solution of the chlorides of zinc and
magnesium, and that the zinc is collected on revolving disc
cathodes, as in the Hoepfner process (see above). The zinc
obtained is substantially pure, and the lead and silver left
as anode sludge contain but little zinc, and can be cupelled
at once to recover the silver.
Summing up, one may say that the electrolytic treat-
ment of zinc is in a backward state. For that purpose
which promises most reward the winning of zinc from
its ores, especially from mixed sulphides no satisfactory
process depending primarily on electrolyis has yet been de-
vised. For the mere purification of zinc already won from
152
ZINC
its ores which is comparatively easy there is not, nor can
be, any demand which is not easily supplied by the simple
method of redistillation. In the case of the one crude pro-
duct from which zinc may be advantageously separated by
electrolytic means " zinc amalgam " there are many
difficulties in treatment. A moderate success must be
chronicled in "cold galvanising," i.e. coating iron and steel
with zinc electrolytically deposited. This is dealt with in
the section allotted to the art of electro-deposition.
153
SECTION III
Winning and Refining Metals in
Igneous Solution
ALUMINIUM
A LUMINIUM differs from all other metals used as such
AIL in the arts, in that at the present time it is produced
solely by electrolytic methods. Ordinary metals copper,
zinc, silver, etc. which are employed not for their chemical
peculiarities, as are sodium and magnesium, but on account
of their physical and mechanical properties, are obtained
partly or chiefly by other than electrolytic means. Alu-
minium alone is manufactured exclusively by electrolysis.
Thirty years ago this was not the case ; aluminium was then
made wholly by methods which were purely chemical.
Even twenty years ago no serious attempt had been made to
manufacture aluminium by an electrolytic process.
The reduction of aluminium from its oxide by smelting the
latter with carbon is impracticable at ordinary furnace
temperatures. 1 The cognate metal iron can be reduced with
ease, and this is done daily in the blast furnace. If we
substitute A1 2 3 for Fe 2 O s and heat it with carbon no metal
is obtained. It is only at the extremely high temperature
of the electric arc (about 3,500 C. = 6,332 F.) that reduc-
tion occurs! Even then, if alumina be heated in contact
with carbon, it is not Al but the carbide A1 4 C 3 which is
obtained. If a metal with a great affinity for oxygen is
used instead of carbon, e.g. manganese or magnesium,
reduction equally fails to take place. But if another com-
pound of aluminium, namely the chloride, be used instead
of the oxide, metals of this class will reduce aluminium there-
from. The original chemical method of Deville is based on
1 The equation A1 2 O 3 + 3 C A1 2 + 3 CO requires the addition of
305 Cal. for its realisation.
157
PRACTICAL ELECTRO-CHEMISTRY
this fact. Anhydrous aluminium chloride is prepared by
heating a mixture of alumina and carbon in a stream of
chlorine. By adding sodium chloride to the mixture, the
double chloride Al 2 Cl 6 6NaCl is obtained, and this is the
substance used in the old Deville process. When this
double chloride is heated with sodium it is reduced according
to the equation
A1 2 C1 6 6 NaCl-f- 6 Na = A1 2 + 12 NaCl.
Instead of the double chloride, the double fluoride
Al 2 F 6 6NaF (cryolite) may be treated with sodium for the pro-
duction of Al. Any process of this kind involves, in the
first place, the manufacture of sodium. Deville's process
remained costly until cheap sodium was produced by Cast-
ner ; who made the metal by reducing caustic soda by means
of an intimate mixture of carbon and iron. It will be ob-
served that the large amount of energy necessary to sever
aluminium from oxygen is provided in this chemical process
in two stages. In the first aluminium chloride is produced,
the heat of combination of which is 322 Cal. as against 392
Cal. for the oxide. In the second a metal (sodium) reducible
by carbon and having a high heat of combination with
chlorine is manufactured. This, being caused to react with
the aluminium chloride, accomplishes what it could not do
had it been applied to aluminium oxide. Thus ultimately
almost all the energy needed to reduce alumina has been
obtained from carbon in two stages, each being ineffective
alone. Now one of the great advantages of electrolytic
methods is that the energy needed for their execution can
be supplied at any desired pressure. The critical pressure
for the electrolytic decomposition of alumina according to
Gin is 2*82 volts, and this is. of course, well within
working limits. Seeing that this value is considerably
higher than the critical pressure corresponding with the
electrolytic decomposition of water, it is clear that the
reduction of aluminium cannot be accomplished in aqueous
solution ; it must be carried out in a fused electrolyte.
The realisation of these conditions in practice constitutes
ALUMINIUM
the modern electrolytic method of aluminium manu-
facture, which has completely ousted the Deville pro-
cess and its modifications depending on purely chemical
procedure.
ELECTROLYTIC REDUCTION OF ALUMINIUM
The process on which the world's supply of aluminium
now depends consists in the electrolysis of alumina. Alu-
mina, having a very high fusing-point, is conveniently dis-
solved in a fused salt of aluminium, e.g. the fluoride or the
double fluoride of aluminium and sodium. This is accom-
plished by several processes, which will be described in turn.
THE HEROULT PROCESS
This process as at present worked is the type of all success-
ful processes for the production of pure aluminium as dis-
tinct from aluminium alloys. There are several other
processes, known by the names of their devisers, which pro-
fess to be distinct from the Heroult, but the distinction if it
exists is in law rather than in fact.
The Heroult process is that worked by the Aluminium-
Industrie-Aktien-Gesellschaft at Neuhausen in Switzerland
and by the British Aluminium Company at Foyers in Scot-
land, the two largest manufacturers of aluminium in Europe.
It is noteworthy that in the original patents the preparation
of aluminium bronze rather than of aluminium was contem-
plated, and that all the accounts of the process apply to the
production of the alloy. The general arrangement, of the
Heroult furnace as originally devised may be understood
from the following description.
The claim in Heroult's German patent is for the continu-
ous electrolysis of aluminium compounds between a carbon
anode and a cathode consisting of a bath of a metal, e.g.
copper, in a state of fusion, the whole being contained in
a crucible provided with a tapping-hole. The process as
actually carried out embodies more than this. As stated
159
PKACTICAL ELECTRO-CHEMISTRY
above,the electrolyte consists of alumina dissolved in cryolite
or in an artificial mixture of aluminium fluoride with sodium
fluoride. This electrolyte is kept fused, not by heat exter-
nally applied, but by heat generated by the passage of the
current. The waste which thus occurs, in that costly
electrical energy is used for mere heating, is more than com-
pensated for by certain practical advantages. These are,
first, that whereas any method of external heating would
require the transmission of every unit of heat through the
walls of the containing vessel, the electrical method applies
FIG. 29.
thereat precisely where it is needed ; secondly, that whereas
in external heating the fused electrolyte would be in contact
with the walls of the containing vessel and would dissolve
and destroy any material but platinum, with electrical
heating the walls remain cool and may be thickly lined with
a congealed crust of the electrolyte itself ; thirdly, that the
temperature of the electrolyte is more readily controllable
1 60
ALUMINIUM
by altering the current and distance between the electrodes
than by regulating an external heating apparatus. It may
be safely said that one of the chief features of the Heroult
process is this method of maintaining the electrolyte in a
fused state. The original Heroult apparatus designed for the
electrolysis of alumina in contact with a copper cathode is
shown in the above figure.
A is an iron box. lined with carbon plates B. The central
cavity contains melted copper c and the electrolyte (alu-
mina dissolved in cryolite) D. The copper is made the
cathode of the cell, electrical connection being obtained
by the cable E clamped to the wall of the iron box. The
tapping hole P is closed by a rod arranged to act as a screw
valve, as shown in the figure. The cell is provided with
a cover of carbon, having two holes, G, G, through which
alumina may be fed and having a central hole large enough
to clear the anode H, which is built up of carbon plates suit-
ably clamped together.
The furnace is started by placing copper in the lower
part, bringing the anode in contact with the metal, thereby
fusing it, adding the electrolyte, and gradually withdrawing
the anode from contact with the copper both copper and
electrolyte are maintained in fusion by the current. Alu-
minium is separated at the cathode and alloys with the
copper, the product being tapped off at intervals. Fresh
alumina and copper are fed in at G, G, as may be required.
This furnace is apparently equally well adapted for the
production of pure aluminium, for if that metal be substi-
tuted for copper at the start and alumina alone be fed in,
the sole cathode product will be aluminium, which can be
tapped off as it accumulates. Another of the earlier designs
is shown diagrammatically below.
A is a wrought-iron box with hollow sides through which
water can be circulated by the pipes B and c. It is un-
lined save for the coating of solidified electrolyte (cryolite
or other double fluoride of aluminium formed and main-
tained by the coolness of its walls). D is a steel plug with a
mushroom head passing into the lower part of the cell. It
161 M
PRACTICAL ELECTRO-CHEMISTRY
makes a mechanical fit with the bottom of the iron box,
and its junction therewith is protected by a layer of solidi-
fied electrolyte. Its head projects into the bath of melted
aluminium. Above this is the fused electrolyte E, into
which dips the carbon anode F. The cell is covered by a
fireclay slab G. It is evident that if found preferable this
slab could be replaced by a hollow iron lid, cooled by cir-
lation of water so that it would protect itself against the
electrolyte splashed up on to it from the bath. Such de-
tails as feeding and tapping arrangements are intentionally
omitted from the illustration.
FIG. 30.
The actual apparatus in use in a large European manu-
factory is arranged as follows : The cells consist of a rectangu-
lar case or box made of cast iron plates clamped together,
about 4 x 2 x 1J ft. These cells are lined with carbon blocks,
and contain cryolite saturated with aluminium. A pool
of fused aluminium acts on the cathode ; a group of large
carbon blocks serves as the anode. The whole arrangement
is shown in the accompanying figure.
The temperature is very moderate, e.g. about 800 C =
1,472 F. The bath is open to the air, and emits no fumes.
162
ALUMINIUM
The aluminium in a fully molten state is tapped off at
intervals. Alumina is fed in from time to time. The
process works smoothly and easily, and appears to suffer
from none of the troubles and defects which have been
ingeniously provided for by many inventors.
In the Heroult process the source of the aluminium
is alumina. The cryolite or other double fluoride of alu-
minium and sodium serves only as a solvent for the alumina.
The case may be likened to that of the electrolysis of zinc
chloride dissolved in water, where the water acts simply
as a solvent, the products being zinc and chlorine. The
products are, therefore, aluminium at the cathode and
Clamp & cable
Clamp &. cable
Fused
aluminium
carbon anode
-iron,
casing
carbon
lining
tapping hole
FIG. 31.
oxygen at the anode. The anode, being of carbon, is attacked
by the oxygen there produced and yields carbon monoxide.
If this attack be considered as an integral part of the pro-
cess of electrolysis, the critical voltage for the Heroult
process will be that corresponding with the equation
which requires 306 Cal., corresponding with a pressure of
2-2 volts. This reduction of voltage from the 2- 82 requisite
163
PRACTICAL ELECTRO-CHEMISTRY
for the electrolysis of A1 2 3 with an unattackable anode is,
however, dearly bought by the consumption of expensive
carbon anodes. This corrosion of the anode is a serious item
of expense, as will be seen when the whole cost of the pro
cess is considered below. The reason is that although chemi-
cally carbon in any form would suffice for combination with
the oxygen, yet for working conditions it is necessary that
the carbon should be electrode carbon, mechanically fairly
strong, sound and homogeneous, of good conductivity and
nearly free from ash. The ash, consisting chiefly of silica,
alumina, and oxide of iron, will dissolve in the electrolyte
and eventually contaminate it to such an extent that the
aluminium produced will be no longer pure, but will
contain silicon and iron, both objectionable impurities ;
ultimately the collection of impurities in the electrolyte will
compel its renewal or purification, the former being probably
the more practicable proceeding. For the production of
pure aluminium it is necessary also to use a moderate cur-
rent density ; if a certain maximum be exceeded, sodium and
fluorine will appear at the cathode and anode respectively.
The extreme chemical activity of fluorine makes it highly
objectionable, because of its corrosive action on everything
with which it may come in contact, while the occurrence of
sodium in the aluminium causes the metal to be easily oxi-
dised, the oxidation taking place locally and leading to
serious deterioration.
Provided a proper supply of alumina be maintained there
should be no risk of decomposing NaF, for its heat of com-
bination is approximately 100 Cal., corresponding with a
critical voltage of 4-3 volts as against 2*82 for the decomposi-
tion of A1 2 3 , or 2-2 volts, if the oxidation of the carbon anode
is assumed to act as an auxiliary source of electrical energy.
This diminution of voltage (supposing it to occur) by no
means compensates for the cost of the carbon electrodes ;
it would be better to work with insoluble electrodes, e.g. of
platinum, were that feasible.
The Heroult process was first put to work at Neuhausen
in 1888, 300 H.P. being used. In the following year the right
164
ALUMINIUM
to use 4,000 H.P. was acquired. The plant put down con-
sisted of two turbines of 600 H.P. each and one of 300 H.P.
The turbines were arranged horizontally, their vertical
shafts carrying the dynamos at their upper end. This
simple and compact arrangement has since been adopted
at the great power house at Niagara, and in many situations
is the best that can be desired. A further increase has since
been made by putting down five more turbines, each of
610 H.P. driving dynamos, each of which gives 7,500 amperes
at 55 volts. The whole installation suffices for the produc-
tion of 2.500 kilos of aluminium per day of 24 hours, or for
a working year of 300 days, 750 tons. There has been a
further increase of plant lately, and the output has risen
to 1,800 tons per year. This rapid development has been
equalled by other installations, and at the present time the
world's output of aluminium cannot be far short of 7.000
tons per year. Considering the comparatively limited and
special uses of the metal, it is remarkable that this quan-
tity should find a market.
The Heroult process is in use at Foyers, in Scotland,
where 3,000 H.P. are available, corresponding with a capacity
for an output of 4.000 pounds per day, i.e. 535 tons per
year of 300 days. The raw material for this works is ob-
tained from bauxite imported from France and worked up
at Larne, in the north of Ireland. The preparation of pure
alumina is necessary as a preliminary stage in all modi-
fications of the Heroult process, and a description of the
method may be usefully given.
The bauxite has the following average composition :
Per cent.
Alumina ........ 56
Ferric Oxide 3
Silica , 12
Titanic Acid 3
Water 26
100
The material is crushed so as to pass a quarter-inch mesh
sieve, and is gently roasted in a revolving calcining furnace,
165
PRACTICAL ELECTRO-CHEMISTRY
the temperature being regulated so as to destroy any organic
matter and ensure that all iron shall be present as Fe^-Os,
and nevertheless not to render the alumina insoluble. The
roasted material is powdered so as to pass a sieve having
30 meshes per linear inch, and is digested with a solution
of caustic soda, of specific gravity 1-45 at a pressure of 70-
100 pounds per square inch. After digestion for two or three
hours the solution is diluted to a specific gravity of 1-23 and
is passed through filter presses ; the clear liquid is then ready
for precipitation. In former processes for the manufacture
of alumina, the alkaline aluminate was decomposed with
C0 2 and the alumina was thus precipitated. The disad-
vantages of this process, apart from the cost of the C0 2 ,
are that any silica present in solution is also throw r n down
and contaminates the alumina, and moreover the alkali
is converted into carbonate, and has to be recausticised
before it can be used again for extraction. By Bayer's
process, which is that now in use, the caustic solution of
alumina is treated with a small portion of alumina precipi-
tated in a previous operation ; it is thereby caused to de-
posit about 70 per cent, of its dissolved alumina if the solu-
tion is well agitated and the precipitation allowed to continue
for about 36 hours. The clear liquor is drawn off and the
alumina washed in a filter press and dried to some extent
by a blast of air, being then roasted at about 1,100C. =
2,012 F. in order to render it both anhydrous and non-
hygroscopic. The latter quality is necessary, as otherwise
the alumina would absorb water during storage and would
not be fit to feed into the electrolytic cell.
The caustic soda solution, diluted but retaining a portion
of its alumina, is concentrated in a triple-effect vacuum
evaporator to its original specific gravity of 1-45, and is then
ready for the extraction of another portion of bauxite. It
will be seen that the caustic soda serves merely to pick out
the alumina from its accompanying impurities, and to de-
posit it, as it were by the word of command, in a pure state.
Both silicon and iron are objectionable impurities in
aluminium, and great pains are therefore taken to exclude
1 66
ALUMINIUM
both from the raw material (alumina). This necessity for
the careful purification of ore (alumina) differentiates the
manufacture of aluminium from that of any metal prepared
by ordinary smelting processes, and adds considerably to
the cost of manufacture. Indeed, the cost of the alumina
necessary to produce 1 pound of aluminium may be J to J
the total manufacturing cost of the aluminium.
The need for using pure alumina has been one of the diffi-
culties of aluminium manufacture. It would be better to
electrolyse an impure alumina (even bauxite direct), to
obtain thereby an impure aluminium and to purify this pre-
ferably electrolytically. The operation has been attempted
by the Pittsburg Reduction Co. by making moderately
crude aluminium the anode in a bath of fused aluminium
fluoride and collecting on the cathode pure aluminium.
The aluminium being more readily attacked than its usual
impurities is dissolved and transferred much as copper is in
a sulphate solution, and the impurities are left behind undis-
solved or non- transferable, very much as is the anode
sludge or dissolved impurities in copper refining. The
idea seems feasible.
THE HALL PROCESS
This is a process presenting many similarities to the He-
roult. The raw material is purified alumina ; it is dissolved
in a fused bath of aluminium fluoride and sodium fluoride
mixed in about the proportions A1 2 F 6 2 NaF. The sodium
fluoride may be replaced by calcium, potassium or lithium
fluoride. In the various patents by which the process is
disclosed the electrolyte is to be kept fused by external
heating. If this is actually practised difficulties will cer-
tainly arise from the attack of the containing vessel by the
electrolyte. It has been pointed out above (p. 160) that
protection of the containing vessel may be best secured by
a congealed coating of the electrolyte itself, and this is only
possible when the heating is internal, i.e. produced by the
167
PRACTICAL ELECTRO-CHEMISTRY
passage of the current. If then the heating is internal,
the Hall process is practically identical with the Heroult.
An official description of the Hall process has been pub-
lished by Hunt, the President of the Pittsburg Reduction
Company, which uses the Hall process at New Kensington,
Pennsylvania, and at Niagara Falls. At each place it has
1,600 H.P., with an output of about 2,000 pounds per day of
Al. It is intended to increase the Niagara works consider-
ably.
The vessel (A, see Fig. 32) containing the electrolyte is
an iron trough lined with carbon plates B. It is made the
cathode by connection with the dynamo by the copper strip
E
E
E
E
I
~
-r
_- -
V-j
FT"
E_
~__;
.
and cable c. From a copper rod D the carbon anodes E, E
are hung, and dip into the electrolyte r. They can be lowered
as they are consumed. It is stated that the carbon lining
is not sensibly affected, and that both it and the iron pot last
a long time. This is very dubious if the heating is external.
Under such conditions the electrolyte would soak through
the carbon and be likely to attack the iron. It is highly
unlikely that the heating is actually external.
The following details have been published by the inventor
of the process. The chief raw material (alumina) costs
about 2|d .-3d. per pound, and can be obtained in the re-
quisite purity from various chemical manufacturers in the
States, who prepare it from cryolite. It is, of course, hy-
drated and substantially free from all impurities save soda.
1 68
ALUMINIUM
Cryolite, costing about 3d. per pound, can be obtained from
the same manufacturers. Hydrofluoric acid (necessary for
the preparation of the aluminium fluoride) can be bought
in quantity of the required quality at 2d.-2%d. per pound.
Carbon for lining the electrolytic cells is prepared from
good coke or retort carbon and tar baked in the usual way ;
the cost is about 1 %d.-2d. per pound. The electrolyte is
prepared by treating a mixture of alumina, cryolite and
fluorspar with hydrofluoric acid in a lead-lined vessel. The
mass is dried, fused in the carbon-lined steel troughs de-
scribed above, and electrolysed. After some hours, when the
mixture is thoroughly fluid, alumina is fed in, and then
acts as the electrolyte proper, as in the Heroult process.
The separated aluminium collects at the bottom of the car-
bon-lined cell, care being taken to keep the electrolyte
specifically lighter than the fused metal. This can be aided
by the addition from time to time of the double fluoride
A1 2 F 6 2 KF. The general rules to be observed in the manu-
facture of aluminium by this (or, indeed, by any cognate)
process have been laid down by Hunt, whose authority has
been cited above.
(1) The solvent, with its dissolved alumina, must.be
fusible at a moderate temperature.
(2) The solvent must dissolve alumina freely, e.g. must
take up at least 20 per cent, at the working tem-
perature.
(3) The critical voltage for the solvent must be higher
than that for the alumina.
(4) The specific gravity 1 of the solvent at its working
temperature must be lower than that of aluminium
at the same temperature, so that the metal may
collect at the bottom of the cell.
1 The specific gravity of solid aluminium, is not higher than 2- 7,
and that of cryolite is about 3. When these materials are fused,
however, the alteration of their specific gravities is very consider-
able, and the relation of the specific gravities is reversed. Richards
has published an interesting table which shows why it is possible
169
PRACTICAL ELECTRO-CHEMISTRY
The same writer supplements Hall's earlier account in
several respects. Thus the solvent may be formed of vari-
ous fluoride mixtures and yet comply with the conditions
laid down above. The solvent most commonly used is one
of 677 parts by weight of aluminium fluoride, 251 of sodium
fluoride, and 234 of calcium fluoride. The ingredients may
be fused in separate vessels or in the electrolytic cell by the
passage of the current. It should be noted that this is not
in accordance with the general tenor of Hall's patents, in
which fusion is effected by heat externally applied. When
the solvent is fused, alumina to act as the electrolyte is fed
in in the proportion of about 20 per cent, of the weight of
the solvent, and this proportion is maintained as electrolysis
proceeds. The bath is kept below 982C. The separated
aluminium is baled out from time to time.
The descriptions of the Heroult and the Hall processes
which have been given above show their principles and mode
of working to be substantially identical. It may be accepted
without much hesitation : (1) that the electrolyte in each
case is alumina dissolved in a fused mixture of aluminium
fluoride and the fluoride of the metal of an alkali or alkaline
earth ; (2) that the bath in each case is kept fused by the
heat generated by the current itself ; (3) that carbon
anodes are used ; (4) that the cathode in actual working is
or soon becomes a pool of liquid aluminium ; (5) that
the containing vessel is iron with a lining of carbon. A
typical design which, though taken from a Heroult plant,
probably represents fairly enough the Hall apparatus as
worked, is given on page 163.
in the Heroult and similar processes to keep the separated alu-
minium at the bottom of the bath.
Specific Gravities
Fused Solid
Commercial aluminium . . .2-54 2-66
Commercial Greenland cryolite . .2-08 2-92
Cryolite saturated with alumina . .2-35 2-90
Cryolite mixed with aluminium fluoride
in the proportion required by the
formula A1 2 F,, 2 NaF . . .1-97 2-96
This mixture saturated with alumina . . 2-14 2-98
1/0
ALUMINIUM
THE MINET-BERNARD PROCESS
This process is the only other method of preparing alumin-
ium which need be referred to. According to the patent
taken out jointly by Minet and Bernard, the electrolyte is
a mixture of aluminium fluoride and sodium chloride,
fused in a metal vessel by heat externally applied. The
vessel may act as cathode or a separate carbon cathode
may be used ; the anode is of carbon. When the vessel
is made the cathode a portion of its substance is dissolved
by the separated aluminium, and the metal obtained is only
fit for the production of alloys. For pure aluminium a
carbon cathode is requisite. It will be seen that the
whole arrangement is very crude, and meets none of the
difficulties which have been discussed above. The Minet-
Bernard process is said to be in use at one works ; if this be
true the process must have been considerably modified,
and, it may be fairly assumed, on the lines which have been
laid down in considering the Heroult process.
OTHER METHODS
The method of production of aluminium by the elec-
trolysis of alumina dissolved in a fluoride bath is not with-
out certain drawbacks. In the first place, the bath is highly
destructive of most materials that can be used as containing
vessels, and thus makes necessary the use of various devices,
which have been described above, to prevent it from acting
thereon. In practice this difficulty is met by making the
bath large in comparison with the active area, and thereby
protecting it with a layer of scarcely fused electrolyte.
Secondly, expensive carbon anodes are necessary, and these
are consumed by the oxygen of the electrolyte (alumina).
Thirdly, these same anodes inevitably contain ash, consisting
largely of silica and oxide of iron, impurities which dissolve
in the bath and eventually contaminate the alumina. It
is therefore not surprising that other methods should be
worth considering.
171
PRACTICAL ELECTRO-CHEMISTRY
Before the advent of a practicable electrolytic method,
many attempts were made to devise a chemical process
cheaper than the chemical method of Deville. The key
to all possible processes is this : alumina cannot be reduced
to aluminium at ordinary furnace temperatures by carbon
or any other practically available reducing agent. Its heat
of combination is too high, viz. 392 Cal. The reduction
must therefore be affected in two stages, as in Deville's
process, where the anhydrous chloride is first produced
(heat of combination 322 Cal.), and then this in turn is
reduced with sodium. In like manner the anhydrous
sulphide may serve as an intermediate step. A1 2 S 3 has a
heat of formation of 124-4 Cal. The heat of formation of
manganese sulphide is about 45 Cal., so that the equation
A1 2 S 3 + 3 Mn = A1 2 + 3 MnS
should be possible, as it would evolve 3 x 45 124-4 Cal.,
i.e. 10-6 Cal. Unfortunately pure manganese cannot be
produced by any ordinary smelting operation, the material
obtained always containing a good deal of iron, silicon and
carbon. Thus direct reduction of aluminium sulphide by
any ordinary chemical process is hardly to be looked for ;
the sulphide itself has, however, certain merits as a material
for electrolytic reduction.
In the first place, its critical voltage is 0-89 volt, as against
2-82 for AL0 3 . Secondly, its anode product is sulphur,
which does not combine with carbon until a high tempera-
ture has been reached ; the carbon anode should therefore
remain unattacked. These obvious merits have induced
sundry inventors to devise processes in which the sulphide,
instead of the oxide, of aluminium is to be employed. The
great obstacle in the way of this class of process is the
manufacture of aluminium sulphide. A1 2 S 3 is decomposed
by water yielding A1 2 3 and H 2 S, and consequently cannot
be prepared by any wet method. The reaction
A1 2 3 + 3 C + 3S = A1 2 S 3 + 3 CO
requires the addition of 211 Cal. in order to bring it about.
172
ALUMINIUM
This may be met in some measure by using CS 2 instead of
C and S independently (CS 2 being an endothermic substance),
and bringing into the reaction a store of energy previously
acquired. If the equation
2 A1 2 O 3 + 3 CS 2 = 2 A1 2 S 3 + 3 C0 2
is possible, it would still require 100-5 Cal. per gramme equi-
valent of alumina converted into aluminium sulphide ; if the
reaction took the form
A1 2 3 + 3 CS 2 = Al a S 3 + 3 CO + 3 S,
there would still be lacking 168 Cal. These facts show
clearly enough that the first necessity for processes proposing
the electrolysis of aluminium sulphide is an improved
method of manufacturing that substance. According to the
patents of Bucherer it can be obtained by the joint action
of sulphur and carbon on alumina in the presence of sulphides
of the alkali metals, double sulphides of the form
A1 2 S 3 3 Na 2 S
(thio-alummates, in fact) being produced. The only addi-
tional source of energy which makes this proceeding more
hopeful than that expressed by the equation
A1 2 3 + 3 C + 3 S = A1 2 S 3 + 3 CO
is the combination of Na 2 S with A1 2 S 3 , and this is likely, to
yield but little energy.
Assuming that aluminium sulphide is produced, it can,
according to Bucherer, be dissolved in fused sodium chloride
and electrolysed as in the Heroult process. The fusion of
the mixture may be effected either by the current or by
external heating, the former for choice, because fused sodium
chloride attacks any material available for a containing
vessel. It is said that the Aluminium-Indus trie- Aktien-
Gesellschaft at Neuhausen (the company which first ex-
ploited the Heroult process) is trying a sulphide method,
but no information as to its working has been made public,
173
PRACTICAL ELECTRO-CHEMISTRY
Experiments have been made by Tucker and Moody on
the production of aluminium by the action of calcium
carbide on alumina at a high temperature ; the method
is analogous to that for the production of chromium by the
action of aluminium on chromic oxide ; that is to say, in both
cases the ultimate source of energy is electrical, but the
application is indirect. In these experiments it was found
that whereas alumina is not reduced by carbon alone in
the electric furnace 1 it can be reduced to aluminium by
calcium carbide. A charge of 150 grm. A1 2 3 , 200 grm. CaC 2
and 60 grm. carbon (the latter to compensate for casual oxida-
tion) when heated in a furnace supplied with a current of 275
amperes at 50 volts proved satisfactory. The calcium
carbide may be regarded as a convenient and accessible
form of calcium, because the heat of formation of calcium
carbide is quite small ; the comparative ease with which
calcium carbide is produced in the electric furnace is due
not to any exothermic reaction between Ca and C, but to
the non-volatility of the CaC 2 .
As the cost of pure alumina is rather high, attempts
have been made to produce fairly pure alumina by fusing
crude alumina in the electric furnace with carbon as a re-
ducing agent. Iron and other impurities are reduced and
separated from the alumina, which then serves as a source
of aluminium in an electrolytic cell. The idea is rational ;
whether it is successful in practice is not yet common know-
ledge.
THE COST OF PRODUCTION OF ALUMINIUM
From the foregoing pages it will seem that under even
favourable conditions the amount of energy needed for the
reduction of a given weight of alumina to aluminium is very
large, viz. 272,222 joules per gramme equivalent. Therefore
1 This is literally true, but, as stated, might give rise to miscon-
ception ; alumina is reduced by carbon alone, not to metal, but to
the carbide A1 4 C 3 .
174
ALUMINIUM
an apparatus working with theoretical efficiency would
produce 88-8 grammes of aluminium per H.P. hour, i.e. 4-7
pounds per H.P. working for 24 hours. It is certain, how-
ever, that this theoretical efficiency is never approached.
The current efficiency is not likely to be higher than 50 per
cent., and the voltage required will be not less than double
the critical pressure, 2-8 volts. The energy efficiency will
therefore be 50 per cent, x 50 per cent. = 25 per cent., and
the output per H.P. hour not greater than 22-2 grammes, or
say | ounce. This agrees with estimates made by Borchers,
based on small manufacturing experiments, and with the
most reliable figures which have been published concerning
processes actually at work on a large scale. It may be
taken from these facts that a plant of 1,000 H.P. (net, de-
livered at the terminals of the electrolytic cell) will manu-
facture 194 tons of aluminium per year of 365 days of 24
hours each. This is not a large quantity of metal, and it is
easy to see that an aluminium factory to have a fair output
must be in a position to use several thousand H.P. In fact,
in this, as in many electro-chemical industries, 1,000 H.P.,
large as it seems from a mechanical point of view, is a con-
venient unit to think in. The capital cost of a water-power
plant must include the expenditure for dams, conduits,
turbine pit, turbines with buildings, and land necessary for
their accommodation. It will obviously vary greatly accord-
ing to the circumstances of each case. If much civil engineer-
ing is required, i.e. if the water has to be impounded and
a new and artificial outlet and course have to be provided
for it, the capital expenditure may be very large. But
if the prospect of the undertaking's success is to be good, the
total outlay for this work and for the power plant should
not exceed 50 per electrical H.P. That is, a capital outlay
of 250,000 will be necessary for a single plant of 5,000 H.P.,
which, though certainly a good size, is by no means colossal,
seeing that it is capable of producing no more than 1,000
tons of aluminium per year, the power being used continu-
ously. Such a plant will yield power at the rate of about
4 per electrical H.P. year, allowance being made for interest,
PRACTICAL ELECTRO-CHEMISTRY
depreciation and running charges. This corresponds with
4,000 for 194 tons of Al, i.e. 20 12s. per ton, or 2-2d. per
pound. The market price of aluminium is about Is. per
pound, whence it appears that the cost of power, though
a considerable item, is not so large as to make it certain
that a source of cheap power can be profitably used
for the manufacture of aluminium, irrespective of other
considerations, such as cost and accessibility of raw
material.
Probably the largest item of cost in the manufacture of
aluminium is the price of the alumina. About 2 pounds of
anhydrous alumina are needed to produce 1 pound of
aluminium, and the present price of alumina of good quality
made from bauxite is 2d. per pound. The next large item
is the cost of the carbon electrodes. The calculated con-
sumption of carbon is 66 f per cent, of the weight of alu-
minium produced, but in practice it usually amounts to 100
per cent. Taking the cost of carbon electrodes at 2d. per
pound, the expenditure on this score is as great as that
necessary for power. The approximate minimum cost of
manufacturing aluminium may be set forth as follows :
Per pound of
Al produced.
Power ........ 2-2d.
Alumina ........ 4*0
Carbon electrodes . . . . . .2-0
Labour superintendence, interest on and repairs to
furnaces 2*
I0-2d.
This estimate is so little below the market price of aluminium
(Is. per pound) that it is probable that some of these items
have lately decreased in cost. For example, alumina may
well come down to 10 per ton (say Id. per pound), and
carbon electrodes to a like figure. The minimum cost of
aluminium would then be 7-2d. per pound, and its selling
price may fall to 9d. per pound. At that price it is one and
a half times the price of copper weight for weight, and less
176
ALUMINIUM
than half its price bulk for bulk, so that it can be freely
used as an industrial metal of moderate price. 1
USES OF ALUMINIUM
There are four chief outlets for aluminium :
(1) As a reducing agent. Vast quantities are employed
as an addition to steel when it is about to be cast
to reduce and remove any entangled oxide, to cause
the metal to pour quickly and to produce sound
castings. It is sometimes used as an addition to
copper and copper alloys for a like purpose. In this
case the proportion added is small ounces per ton
and the aluminium, having done its work, passes
from the metal, and may leave no recognisable trace
in the finished material. The aggregate amount
thus used is very large, although individual doses
are minute.
It is also used to reduce refractory oxides, such
as that of chromium, and thus yield the metal in a
pure state ; this method has been much developed
of late, and is likely to oust purely electrical methods
in which the metal sought is reduced in an electric
furnace ; it may be noted that even here electrolysis
is necessary for the production of the reducing agent.
(2) As an industrial metal for small ware and structures
where lightness and resistance to corrosion are
required. The specific gravity of commercial
aluminium ranges from 2-67 to 2' 70. The metal can be
1 An estimate for the Hall process shows similar figures :
Per pound of
I Al produced.
Power 2- Id.
Alumina . . . . . .6-3
Carbon electrodes . . . . .1*6
Miscellaneous . . . . .0*6
10- Qd.
177 N
PEACTICAL ELECTRO-CHEMISTRY
worked as freely as brass, save that it is not readily
soldered. All kinds of small articles for daily use
boxes, travelling cups and flasks, cooking vessels
and such like are made in large quantities, and
are now cheap enough. Specially light boats may
have their fittings of aluminium, or even be built
of it entirely ; aluminium motor car bodies and
engine cases are also made ; bells have been prepared
from it.
(3) As a constituent of alloys. Aluminium added to
copper gives alloys of great strength and high
mechanical utility. Similar good alloys are ob-
tained when aluminium is added to brass. The
quantities used range from 1 to 10 per cent, of Al.
Another series of aluminium alloys is made by adding
1 to 10 per cent, of alloying metals, such as copper,
nickel and tungsten, to aluminium itself. Alloys
of this class are almost as light as aluminium, and
a good deal stronger. They may often be substituted
with advantage for unmixed aluminium, and used
for the purposes already mentioned under section 2.
(4) As a material for electrical conductors. Aluminium
is used for carrying large quantities of power to
considerable distances. The chief difficulty arises
from the joints, which are less easy to make than
are those in copper conductors. The following table
gives a comparison of the two metals :
Copper Wire.
Aluminium Wire.
Sp. G
Q.QQ
2.AK
Conductivity
100
Al
Section for equal conductivity .
Weight
Tensile strength .
1
1
1
1-64
0-485
0-460
for equal conductivity
1
0-75
There is a saving in weight for equal conductivity if
aluminium is used instead of copper.
178
ALUMINIUM
IMPURITIES OF COMMERCIAL ALUMINIUM
On account of its method of production by the electro-
lysis of alumina, which is never quite pure, and by reason
of the introduction of additional impurities from the carbon
anodes, commercial aluminium almost invariably contains
small quantities of iron and silicon. If too high a current
density has been used or the bath allowed to become poor
in alumina, sodium may also be present. Thus commercial
aluminium rarely contains more than 99 per cent. Al. A
great deal of that put on the market is no better than 98
per cent., and a crude metal of 96 per cent, or lower is
manufactured for reducing purposes . The following analyses
show the nature and amount of the usual impurities in the
better grades of aluminium :
I.
II.
III.
Al ....
Si ....
Fe ....
99-59
0-25
0-16
99-00
0-87
0-13
98-45
1-29
0-10
100-00
100-00
99-84
One of the impurities mentioned above, viz. sodium, is
particularly objectionable, because the sodium, which is
present in minute specks and spots segregated from the mass
of the metal, oxidises readily and causes injury and corrosion.
Even when oxidised it continues to cause corrosion, for the
caustic soda produced itself acts on the aluminium surround-
ing it.
During the last year or two the purity of commercial
aluminium has increased considerably, and defects due to
the comparative crudity of the metal previously manu-
factured have almost disappeared. Good grades of alu-
minium and serviceable alloys are now available for indus-
trial purposes.
179
MAGNESIUM
MAGNESIUM, like aluminium, is a difficultly reducible metal
which can be most economically manufactured by electro-
lysis. Formerly magnesium chloride was reduced by means
of sodium, but the metal thus obtained had to be purified
by distillation. As magnesium boils at about 1.000 C.
= 1,832 F., this operation is somewhat difficult and costly,
and its avoidance is accomplished by the use of the electro-
lytic process, which, if properly conducted, yields a metal
sensibly pure.
Magnesium chloride is the raw material. It is obtained
in the double salt carnallite'(KCl MgCl 2 6H 2 0) from the
saline deposits of Stassfurt. Whereas an aqueous solution
of magnesium chloride, when evaporated to dry ness, is
largely decomposed, yielding magnesia and hydrochloric acid,
one containing also the chloride of an alkali metal can be
dehydrated without decomposition. The anhydrous double
chloride is fused and electrolysed between a carbon anode and
an iron cathode. The process presents analogies to that for
the manufacture of aluminium, but differs in the fact that
the electrolyte is not the oxide of the metal dissolved in
its fused halogen salt, but is the halogen salt itself. The
essential parts of an apparatus for the electrolytic reduction
of magnesium are shown in the accompanying drawing
(Fig. 33) of one devised by Graetzel, which, in more or less
modified form, is a type of the plant now employed.
A is a cylindrical steel vessel, made a cathode by the
cable B. It is closed by an air-tight cover c, through which
passes an entrance pipe D, conveying a gas, e.g. nitrogen, or
furnace gases free from oxygen ; the surplus gas passes out
by the pipe D. E is a porcelain cylinder open at the bottom,
1 80
MAGNESIUM
and having slits in the sides. This contains the carbon
anode F, and carries a pipe G for the escape of the chlorine
generated at the anode. The vessel A is filled with carnallite,
which is kept fused by heat externally applied. The pro-
ducts of electrolysis are magnesium and chlorine. The first
floats on the fused carnallite, and is protected from oxidation
by the atmosphere of nitrogen or other inert gas supplied
through D. The chlorine liberated at the anode can pass
freely away, and is hindered from casual entrance into the
outer vessel by the porcelain cylinder E, which, nevertheless,
permits free flow of the current and of the electrolyte itself.
FIG.
As is usual in the electrolysis of the fused salts of difficultly
reducible metals, the design of an apparatus which will yield
the metal is comparatively simple ; the device of one which
will be fairly permanent in actual manufacture is less easy.
In that shown above no attempt is made to protect the walls
of the outer vessel from the action of the electrolyte ; should
this action be found severe, recourse must be had to the
method described under Aluminium, viz. the cooling of the
walls to form a protective crust. In this case the vessel itself
could not be the cathode, and an independent cathode, as
in the aluminium apparatus, would be necessary ; also the
181
PRACTICAL ELECTRO-CHEMISTRY
electrolyte would be kept fused by the current, and not by
heat externally applied.
It was noted above that the magnesium, as it was reduced,
collected on the surface of the electrolyte. This is far from
convenient, and necessitates, both in the present apparatus
and in one of the aluminium type, an envelope for the anode,
to hinder union of anode and cathode products. In the case
of aluminium, although the solid metal is specifically lighter
than the solid electrolyte, yet when both are fused
the metal is the heavier. This convenient relation
does not obtain with magnesium and carnallite. It is
possible that it might occur for some other feasible electro-
lyte, but exact data are lacking. The demand for mag-
nesium is too small to warrant much technical investigation
for the device of a perfect process. What is needed can be
made, and its cost of production is a secondary matter.
The heat of combination of MgCl 2 is 151,000 Cal. The
critical voltage for its electrolysis is therefore 3-26 volts.
The heat of combination of KC1 + MgCl 2 to form carnallite
is probably so small as not to effect this value appreciably.
In practice a high current density is used, e.g. 100 amperes
per square foot of cathode surface, and the voltage is cor-
respondingly high, in spite of the fact of fused carnallite
being a good electrolytic conductor. There are certain
details in the process which have been studied by Oettel,
and are of interest in that they indicate the sort of difficulty
not obvious from a consideration of the principles of a
process, but prominent enough when it is put to work. In
order to collect the magnesium which floats on the electro-
lyte, it is desirable that it should run together into large
buttons. Minute globules are difficult to collect and oxidise
in proportion to their surface, which is relatively great.
The cause of the failure of the metal to agglomerate in the
desired manner is usually the formation of a thin skin of
magnesia on the globules of metal, which prevents their
mutual contact, much as dirt and oxide on mercury will
prevent it from running together. This magnesia comes
from the electrolysis of MgS0 4 , present as an impurity in
182
MAGNESIUM
carnallite, the products of electrolysis being MgO and S0 2
and 0. This explanation is not completely convincing,
for it might well be supposed that MgO would dissolve in
considerable quantity in fused carnallite, and would thus
be harmless. Once dissolved it would be as readily reduced
as MgCl 2 , its heat of combination being nearly the same.
A more likely cause seems to be the presence of oxygen in
the gas used as a neutral atmosphere for the cathode com-
partment. This would act on the small globules of metal
as they rose from the cathode and swam on the surface of
the electrolyte ; it would coat them with a film of MgO, and
prevent their coalescing. Even if magnesia is fairly soluble
in fused carnallite, it would not be promptly removed
from these globules because they are not fully immersed
in the electrolyte. The removal of this film may be accom-
plished by adding fluorspar (calcium fluoride) to the melt ;
in like manner a mass of magnesium globules mixed with
carnallite, such as will be obtained by ladling out the con-
tents of the cathode compartment, may be caused to come
together by adding fluorspar and heating. A clear melt
and magnesium stripped from any coating of magnesia
will result. Melted magnesium in bulk, and not in globules,
can be handled without fear of its taking fire, or even oxidis-
ing largely, if it be kept not much above its melting-point,
viz. 750 C. = 1,382 F. ; if the temperature be allowed to
rise to a good red heat, combustion will occur.
The production of magnesium is more interesting as
illustrating many principles of electrolysis applied to fused
salts than important from a commercial point of view. The
latest statement from what is the chief and perhaps the only
factory now making the metal, viz. the Aluminium and
Magnesium Works at Hemelingen, is to the effect that the
demand for magnesium is decreasing. This may well be
due to the preferential use of aluminium as a reducing agent ;
even for flash-lighting, for which magnesium seems especially
suitable, aluminium has been proposed as a substitute.
Almost the only other purpose for which magnesium is
employed is as an addition to nickel to cause it to cast well.
183
PRACTICAL ELECTRO-CHEMISTRY
Here it doubtless acts as a reducing agent, and removes
entangled oxide. As it does not alloy with nickel, the
surplus magnesium does not appear in the finished casting.
An alloy of magnesium with aluminium 1 (termed magnalium)
has lately been prepared which is said to be not easily
corrodible ; no other useful alloy of this metal has yet
been obtained.
1 Mach, the inventor of these alloys, states that the alloy con-
taining 10 per cent, of magnesium can be worked like zinc, that
when the proportion rises to 15 per cent., the material resembles
brass; with 20-25 per cent, its behaviour is similar to that of
gunmetal when machined.
184
SODIUM
REFERENCE is made to the electrolysis of fused sodium
salts, and the production of sodium, in the chapter on
Alkali and Chlorine (q-v.). In the processes there dealt
with, however, the production of sodium is incidental, and
the metal itself is not isolated ; it serves only as an inter-
mediate stage in the formation of caustic soda or sodium
carbonate. When the metal sodium is the desired end-
product, other methods than those there described become
necessary.
Sodium was formerly manufactured by distilling sodium
carbonate with charcoal, the reaction being
Na 2 C0 3 +C 2 = Na 2 +3CO.
This process needed a very high temperature, was costly
in fuel and destructive of retorts, and was superseded by
the Castner process (a purely chemical method, not to be
confused with the Castner electrolytic process for sodium,
which is about to be described).
In this process caustic soda was used instead of sodium
carbonate, and the reducing agent was a mixture of iron
and finely divided carbon made by heating together oxide
of iron and tar. The function of the iron is to weigh down
the carbon and keep it immersed in the fused NaOH.
The reaction
4NaOH+C 2 = Na 2 C0 3 + Na 2 +2 H 2 +CO
requires 106 Cal., instead of 186 Cal. requisite for the reduc-
tion of sodium carbonate formerly practised, and moreover
takes place at about 800 C. = 1,472 F. instead of at about
1,500 C. = 2,732 P. These advantages more than com-
pensate for the use of the dearer raw material, caustic soda,
in place of sodium carbonate.
185
PRACTICAL ELECTRO-CHEMISTRY
At the present time all chemical methods for the manu-
facture of sodium are obsolete. The metal is produced
exclusively by electrolysis, the sole process employed
being one devised by Castner, the inventor of the chemical
method described above. It is noteworthy that the alkali
metals were first isolated by the electrolysis of caustic alka-
lies by Davy, and that the same process is now the only
method of commercial importance.
THE CASTNER PROCESS
As stated above, the raw material of the Castner electro-
lytic process for the manufacture of sodium is caustic soda.
This substance, in its commercial state, always contains
water (up to about 10 per cent.), and fuses more readily in
consequence. As the water is driven off, the melting-point
rises, but never exceeds a low red heat. It is to this fusibi-
lity of caustic soda that the success of the Castner process
is in a large measure due. The requisite temperature is
manageable, and the apparatus is not rapidly destroyed,
as it is when fused salt, for example, is used as the electro-
lyte. Further, the gas evolved at the anode is oxygen,
not chlorine, and it is therefore possible to use iron anodes,
which are little attacked by oxygen in alkaline liquids at
moderate temperatures. The conditions to be observed are
that the electrolyte should be kept but little above its fusing
point and that the products of electrolysis should be removed
as quickly as possible. An apparatus designed with these
ends in view is shown in Fig. 34. A is a cylindrical steel
crucible with an opening at the bottom through which the
iron cathode B passes. The crucible is set in a flue, so that
the body of it is heated while the neck c remains cool. The
caustic soda which fills the crucible consequently solidifies
in the neck c, and protects the joint made between the
cathode and the crucible. The anode D, which may be a cylin-
der with vertical slits to allow free flow of the electrolyte,
surrounds the upper part of the cathode. This upper part
is encircled by a cylinder of wire gauze E, depending from
the collecting pot F. As electrolysis proceeds, fused sodium
1 86
SODIUM
f floats from the cathode and collects on the surface of the
fused caustic soda in the pot F. It is hindered from straying
into the anode compartment by the wire gauze, through
which it cannot easily pass on account of its high surface
tension. The extreme fluidity of caustic soda and the ease
ivith which it wets all surfaces allow that body, on the other
land, to pass freely through the gauze.
From the collecting pot the sodium can be baled from
;ime to time. This pot is, of course, full of hydrogen, which
serves to protect the sodium from chance oxidation. In
actual work small quantities of hydrogen occasionally
ignite ; thus a succession of small and harmless explosions
usually accompanies the process of electrolysis.
It may be said that the world's supply of sodium is pro-
vided by this process, which is at work at Oldbury, at Weston
Point, at several works in Germany, and at Niagara. At
the last-named place the Electro-Chemical Company use
about 700 H.P., supplied from the main power house at the
Falls. The output possible for such a plant may be calcu-
187
PRACTICAL ELECTRO-CHEMISTRY
lated. The heat of combination of NaOH is 102 Cal.
The critical pressure necessary for its electrolytic decomposi-
tion is, therefore, 425 - OC ' volts = 4- 4 volts. Assuming that
96,540
this voltage could be used and that theoretical current
efficiency could be attained, the output of 700 H.P. would
be 102 kilos Na per hour, i.e. 732 tons of Na per year of 300
days of 24 hours. In practice the joint current and pressure
efficiency is not likely to be greater than 50 per cent., whence
it follows that a plant of this size would turn out about 360
tons of sodium per year. The quantity is small, but pro-
bably ample for the requirements of the market. Sodium
is used only for a few special purposes, such as the manufac-
ture of sodium peroxide, the production of cyanides, and for
" quickening " mercury in gold amalgamation ; in the
larger industries it has as yet found no place. In the future
it may possibly be used as a compact, amenable and port-
able form of energy.
Recently another compound has been used for the pro-
duction of sodium. Darling has devised a process for the
electrolysis of the nitrate. This salt is fused by external
heat and electrolysed between an iron cathode and an iron
anode. The containing vessel serves as the anode, and to
separate the sodium from the oxides of nitrogen there evolved
a septum is necessary, much as in the case of the electrolysis
of magnesium chloride described on p. 180. This partition
consists of magnesia packed between two perforated steel
cylinders ; evidently the function of the arrangement is to
secure a mechanical separation ; there is no electrolytical
necessity for its employment. The advantage of using the
nitrate is that, provided the recovery of the nitrous gases be
satisfactory, the material is cheaper than caustic soda as a
source of sodium. The plant which has been tried has an
output of 700-800 Ibs. of sodium. There are 12 cells, and
each cell takes about 400 amperes at 15 volts. 1
1 As often happens, these statements are incompatible ; the
current used could not, even with theoretical current efficiency, pro-
duce more than 220 Ibs. per day of 24 hours, 12 baths being employed.
188
SODIUM
An ingenious method for preparing sodium is du
Ashcroft. In this the electrolyte is sodium chloride, which
is kept fused by heat generated internally ; the separated
sodium is collected in lead, which is transferred to a second
compartment and there made the anode of a cell contain-
ing fused caustic soda. In this the sodium is dissolved from
the lead and precipitated on an iron cathode. The caustic
soda undergoes no permanent change, serving merely as a
means to transfer the metal from its solution in lead to
the final cathode.
Potassium could doubtless be manufactured in the same
manner as sodium, but as it has no industrial use it need not
be dealt with here. Small quantities are prepared for
scientific purposes, probably by the older chemical processes.
The third member of the alkali group, lithium, has no
industrial use as a metal.
189
SECTION IV
Winning and Refining Metals and their
Alloys in the Electric Furnace
Carbides, Borides and Silicides
THE ELECTRIC FURNACE
r I ^HE high temperature attainable in the electric furnace
A has not merely served to produce certain metals
and alloys less easily won by older means, but has allowed
of the preparation of many substances not previously
known at least in an industrial sense. When the for-
mation of a given product needs a temperature exceeding
2,000 C. = 3,632 F. there is no choice in the matter,
because ordinary processes of combustion cease at or below
that temperature. By pouring electrical energy through
refractory electrodes into a box made of a material which
conducts heat badly, the temperature in the interior of that
box can be raised to that of the arc (computed at 3,500 C.
= 6,332 F.), and reactions unknown at ordinary furnace
temperatures proceed freely. For the further discussion of
the principles of this method of heating, see Section I., p.
21. For the purpose of the present section it is sufficient to
realise that by the use of the electric furnace it is possible to
attain temperatures far above those which can be reached
in any other way, at the exact place where the heat is
required and this without contact with any foreign matter
other than the electrodes and the walls of the refractory box
forming the furnace.
Probably the earliest attempt to use this peculiar advan-
tage of electrical heating was in the manufacture of zinc by
the process devised by the Brothers Cowles, who heated a
mixture of zinc ore and carbon in an electric furnace, the
zinc being reduced, distilled and collected (see p. 133).
This process was not successful, because the temperature
necessary for the reduction of zinc is not high enough to
193 O
PRACTICAL ELECTRO-CHEMISTRY
make ordinary furnace heating impracticable, and at the
time of the Cowles experiments the best conditions for
electrical heating were not fully understood.
The same inventors adapted their furnace for the pro-
duction of aluminium bronze. As this furnace is the type
and forerunner of many modern electric furnaces a sketch
of it in its simplest form may be usefully given. A fire-
brick box A, fitted with a cover B having a hole for the escape
of gases, is pierced with two openings one at each end,
through which pass large carbon electrodes. These are
coupled by heavy copper clamps to cables of large section.
A large current may thus be passed into the box and a power-
FIG. 35.
ful arc formed. The substance to be heated in this case
a mixture of alumina and carbon is packed round the
electrodes and fills the box. This form of furnace has been
modified in various ways, but its type remains fixed. It is
merely a device for heating by an enclosed arc. 1 The Cowles
furnace has now only an historical interest, but it was in
many ways so well conceived and carried out that a short
account of its more developed form may be given.
In this furnace, which was one of the latest forms in use
shortly before the Cowles process for the manufacture of
1 There need be no actual arc ; passage of the current through
a high resistance, such as that of a thin carbon rod or of the heated
charge itself, will equally determine the production in the midst of
the furnace of as high a temperature as that of the arc proper.
194
THE ELECTRIC FURNACE
aluminium bronze was given up, the electrodes consist of
bundles of large carbon rods and are inclined. The rods
c, c are set in massive metal caps, which are of copper if a
copper aluminium alloy is to be produced and of iron if
f err o -aluminium is to be made. This is because the elec-
trodes and their holder get very hot and the latter towards
the end of the run may melt, mingling with the charge.
The caps are connected by rods with the cables D, D. The
holders slide in the tubes E, E, and are moved forward as
FIG. 36.
required by the screws F,F, which pass through nuts attached
to the rods and bear against the flanges of the guide tubes.
A heavy fireclay cover with vents for the escape of gas com-
pletes the apparatus, which is throughout very simple and
massive. The disposition of the charge is important. The
brickwork constituting the body of the furnace is, of course,
lined with firebrick, but this is by no means sufficiently refrac-
tory to resist the high temperature which prevails in the
furnace. It is, therefore, protected by a lining of broken
charcoal. Lest this should become graphitic and agglo-
merate at the high temperature of the furnace it is
previously dipped in milk of lime, so as to leave a film of
lime on each particle. Thus satisfactory isolation of the
heated charge from the walls of the furnace is secured.
The form of alumina usually employed in the Cowles
process is corundum (crystallised aluminium oxide) ; the
first charge consists of 15 kilos of corundum and 30 kilos
of granulated copper, with enough carbon to make the mix-
ture conductive. To subsequent charges the slag from
previous operations is added ; this material is well worth
working up, seeing that it contains about 30 per cent, of
195
PRACTICAL ELECTRO-CHEMISTRY
aluminium and 25 per cent, of copper, both present chiefly
as metal. The charge is covered with coarsely powdered
wood charcoal and a current of 3,000 amperes at a pressure
of 50 volts turned into the furnace. This pressure is main-
tained as nearly as possible throughout the run, the electrodes
being drawn back as the resistance of the charge decreases.
About ten minutes after the current has been switched on,
the air and moisture in the materials will be expelled, and
the reduction of the alumina begins according to the equa-
tion
AI 2 3 + 3 C = 3 Co + A1 2 .
The CO escapes at the vent holes and is burned under
a chimney. The burnt gases, which may contain many
mineral particles volatilized or carried away mechanically,
are passed through a depositing flue. After two hours the
electrodes are about 1*1 metres apart and the charge is
worked off. The run is stopped and the electrodes are
drawn back as far as possible into their protecting iron tubes
so as to hinder their useless oxidation. When the charge
is drawn it is found to consist partly of unused charcoal,
together with slag and unreduced alumina, and, as the desired
product, a mass of crude aluminium bronze containing 14
to 20 per cent, of Al. From this, after analysis, alloys of
determined composition, e.g. 10 per cent, aluminium bronze,
5 per cent, aluminium brass, and the like, can be prepared.
It is found that even with proper working up of the slag not
more than two-thirds of the aluminium originally present
in the corundum is recovered as metal. The output per
H.P. hour is poor, being in the case just cited about 7'5
grammes ; in later practice at Milton in England as much as
25 grammes per H.P. hour was obtained. The theoretical
output can be readily calculated. Thus the reaction
A1 2 3 + 3 C = 3 CO + Al a
needs the expenditure of 305 Cal. for its realization, that
is to say, 305 Cal. are required for the production of 54
grammes of Al. Now 1 H.P. hour = 646 Cal., whence it
196
THE ELECTRIC FURNACE
follows that it should produce 114 grammes of Al. An
actual output of 25 grammes per H.P. hour, therefore,
represents an efficiency of only 22 per cent. Apart from
this low efficiency, the expenditure necessary for wood
charcoal and electrodes is considerable, so that the process
is comparatively costly. Further, the product is not of
particularly good quality, for in the tumultuous sphere of
reaction all oxides are reduced, and such impurities as iron
and silicon tend to appear in the crude aluminium bronze.
Thus it came about that as soon as the Heroult process and
its congeners had been got to work successfully the Cowles
process for the production of aluminium bronze was super-
seded. At present it is generally preferable to prepare such
alloys from the pure metals, but of course the alloys them-
selves could, if desired, be made in the Heroult furnace
(p. 163) by using a cathode of copper or other metal to be
alloyed with the aluminium. Indeed, the Heroult process
was originally designed for the direct production of such
alloys.
The chief interest of the Cowles process lies in the fact
that on its account a highly practicable form of electric
furnace was devised ; also that it took advantage of the
tendency of aluminium to alloy with certain metals rather
than to form a carbide. If it is attempted to prepare un-
alloyed Al by the use of the electric furnace, the chief pro-
duct will be A1 4 C 3 . In addition to this tendency to form
carbide, there is another obstacle to the production of pure
aluminium in the electric furnace. Moissan has shown that
alumina even when liquid is not reduced by carbon, and
that both bodies must be vaporised and their vapours very
strongly heated before the alumina is reduced ; the product
then consists of aluminium mixed with aluminium carbide.
It is only when a metal is present capable of alloying freely
with Al and preferably, as in the case of copper, with the
evolution of heat that a carbonless product is obtained.
According to the Cowles patents the original intention of
the inventors was to form such an alloy and then remove
the alloying metal, recovering pure Al. But such removal
197
PKACTICAL ELECTRO-CHEMISTRY
is impracticable, and the process naturally evolved itself
into one for the production of alloys.
The systematic and scientific study of the capabilities
of the electric furnace is due almost entirely to Moissan.
His investigations are far in advance of any industrial appli-
cation which they have yet received, and afford accurate
data for the manufacture of such of the various carbides,
silicides^and borides as may from time to time be found
commercially important. In view of this it is desirable that
an outline of his work should be given here, in order that
the applications already made may be the better understood.
FIG. 37.
The starting-point of his researches was the study of the
crystallisation of carbon, with especial regard to the pro-
duction of the octahedral or diamond form of crystals.
For this purpose it was necessary to cause a metal containing
carbon in solution to solidify in such a manner as to exercise
great pressure on the carbon at the moment of its crystalli-
sation. In order to saturate the chosen metal with carbon
it was requisite to heat the metal far above ordinary furnace
temperatures . Thus various forms of furnace were devised, in
which the substance to be heated was kept apart as much as
possible from the electrodes and from all other foreign bodies.
The difficulty of finding a substance of which to construct
the body of the furnace was considerable ; eventually lime
was chosen. A typical furnace is shown in Fig. 37. The
body A is made of blocks of lime scooped out in the middle
198
MOISSAN'S RESEARCHES
to form a small cavity, into which the electrodes B, B project.
The cables c, c are attached at the bottom of the clamps, so
that they may not be burned by the torrent of flame which
may burst out from the holes into which the electrodes pass.
As will be seen, it consists essentially of the same parts as
those of the furnace diagrammatically represented on p. 194.
The chief difference is in the materials of the walls of the
furnace, which in the former case are of firebrick and in
the present instance of lime. The lime is not only enor-
mously more refractory than the firebrick, but is also a
vastly worse conductor. With the aid of this apparatus
Moissan was able to bring about novel reactions and to pre-
pare substances previously unknown industrially.
By the use of this furnace with a small hearth on which
the energy represented by an output of 100 H.P. can be
expended, every known oxide can be reduced or volatilised.
Lime, magnesia, alumina and zirconia melt and volatilise
freely. Carbon boils, and its vapour can be used to reduce
refractory oxides also in ebullition. The chief conclusions
to be drawn from Moissan's work having an industrial signi-
ficance are as follows :
The stable form into which carbon, wiiether amorphous
or crystallised as diamond, tends to pass is graphite. Under
ordinary conditions carbon does not melt, but passes directly
into the gaseous state ; if subjected to high pressure, as it
may be by suddenly cooling a liquid, e.g. iron, in which it is
dissolved, it may be liquefied and then may crystallise as
diamond.
Lime, magnesia, molybdenum, tungsten, vanadium and
zirconium may be fused. Silica, zirconia, lime, aluminium,
copper, gold, platinum, iron, uranium, silicon, boron and
carbon may be volatilised. The oxides among these sub-
stances may be deposited in a crystalline form. Oxides
usually regarded as irreducible, e.g. alumina, silica, baryta,
strontia and lime, uranium oxide, vanadium oxide and
zirconia, may be reduced by carbon in the electric furnace.
Many metals which are reduced with difficulty in ordinary
furnaces, such as manganese,chromium, tungsten and molyb-
199
PRACTICAL ELECTRO-CHEMISTRY
denum, may be prepared in quantity. Moreover, in the
electric furnace these metals can be obtained of approximate
purity in spite of their great tendency to unite with the
oxygen and nitrogen of the air. It often happens that,
when a metallic oxide is reduced with excess of carbon in
the electric furnace, a carbide of the metal is first formed.
From this the pure metal can usually be prepared by fusing
the carbide with the oxide of the metal. The carbon is
oxidised and an equivalent of the metal is reduced. The
behaviour of such metals in dissolving carbon at high tem-
peratures, in rejecting it on cooling, and in losing it when
subjected to selective oxidation in general resembles that
of iron, which is well known and forms the basis of the
metallurgy of that metal. One class of bodies is particu-
larly stable at the high temperatures attainable by the
electric furnace to wit, that comprising the carbides, borides
and silicides. These substances are usually of simple com-
position ; SiC (silicon carbide), CaC 2 (calcium carbide),
Mn 3 C (manganese carbide), Fe 2 Si (iron silicide), FeB (iron
boride), CB 6 (carbon boride) will serve as examples. Some
members of the group are extremely hard. Thus carbon
silicide (or silicon carbide) is harder than emery, while boron
carbide and titanium carbide may actually serve to cut a
diamond not merely to polish it, as does silicon carbide,
but to produce definite facets. Others of the carbides have
another claim to interest from an industrial as well as from
a scientific standpoint. Every one knows nowadays that
calcium carbide is decomposed by water and yields acetylene ;
but it is not always realized that the property of thus giving
rise to a hydrocarbon is general for a large number of similar
bodies, e.g. the carbides of lithium, aluminium, thorium
and cerium. Lithium carbide (Li 2 C 2 ) yields acetylene ;
aluminium carbide (A1 4 C 3 ) gives methane ; cerium carbide
CeC 2 , a mixture of the gases acetylene, ethylene and methane,
and a notable proportion of liquid hydrocarbons. This
brief catalogue of facts will show how large a field for in-
dustrial research exists, and how well mapped are the paths
by which it may be entered.
200
METALS PRODUCED OR REFINED
BY THE ELECTRIC FURNACE
As has been shown above, the production of aluminium
has been attempted by means of the electric furnace without
success. Aluminium alloys have been successfully prepared
in similar manner, but this mode of preparation is now super-
seded. Certain other metals of industrial importance can
be prepared in quantity in the electric furnace, and there is
reason to believe that it is the best and sometimes the only
way of preparing them. For an account of such prepara-
tions it is necessary again to refer to Moissan's work.
CHROMIUM
Chromium has scarcely been known as a metal in the
reguline state until the last few years. It can be prepared
in the electric furnace first as a carbide and then as
the pure metal. The production of " cast chromium "
corresponding with cast iron, containing about 10 per cent,
of carbon, can be effected by heating a mixture of Cr 2 3
and carbon in the electric furnace. There is evidence of
the existence of two definite carbides Cr 3 C 2 (containing
13-33 per cent, of C) and Cr 4 C (with 5-45 per cent, of C),
but the cast metal may contain from 1-2 per cent, up to the
limit set by the higher carbide. The preparation of chro-
mium containing only a small percentage of carbon is not
201
PRACTICAL ELECTRO-CHEMISTRY
easy. It is true that the carbon can be removed by selective
oxidation by fusing the crude cast metal with chromic
oxide in a crucible lined with chromic oxide, but the result-
ing metal is " burnt," i.e. it contains a certain amount of
oxygen. A better plan is to refine it by fusing it with lime.
The tendency of lime to form calcium carbide causes it to
remove carbon from the chromium, and by this method a
metal with 1-5-1-9 per cent. C is obtained. Complete re-
moval of C is not practicable however in this way, because
at this point oxidation of the chromium itself occurs, and the
metal is ultimately converted into a calcium chromite. The
object to be attained can be reached by the aid of this very
body. Its tendency to oxidise chromium is not so great as
that of lime per se, and, therefore, when cast chromium con-
taining carbon is refined in a furnace lined with this material,
the oxidation and removal of the carbon take place in regu-
lated manner. Pure chromium is obtained. It is a bril-
liant metal of a grey colour, rather lighter than that of iron,
and though hard can be filed and polished without difficulty.
The various statements as to the extreme hardness of chro-
mium which have been current in text books have probably
arisen from the fact that the carbide Cr 3 C 2 is extremely
hard, scratching quartz and topaz but not corundum. Pure
chromium has a specific gravity of 6-92 at 20 C. It is not
attracted by a magnet. Its melting point is higher than
that of platinum, and cannot be reached by the use of the
oxyhydrogen blowpipe ; the carbides are less infusible.
The metal keeps its polish in the air, is almost unattackable
by acids, even aqua regia, and by fused alkalies. Its
mechanical properties do not appear to have been systematic-
ally examined ; if they are found as excellent as is its chem-
ical behaviour the metal should find an industrial use as a
structural material.
Chromium can be produced with ease in quantity and of fair
purity. A cast metal of the composition given below can be
made in lots of 10 kilos at a time by the use of a current of
1,000 amperes at 70 volts, i.e. 94 E.H.P. The analysis of
the metal gave :
202
METALS PRODUCED BY ELECTRIC FURNACE
Per cent.
Cr . 97-14
C.
Fe
Si
Ca
1-69
0-60
0-39
Trace
99-82
Such a material is well suited for adding to steel to pro-
duce special alloys containing known quantities of chro-
mium. These alloys, having for example 3-4 per cent, of
Cr, are employed for making projectiles, and have been
suggested for use in railway tyres, as they are both hard and
tough.
There is another method of preparing chromium, which
is in some respects better than the use of the electric furnace.
Chromic oxide is mixed with aluminium in powder, and is
fired by a fuse composed of a mixture of aluminium powder
and barium peroxide, in which a strip of magnesium is
embedded so that it may be kindled.
The heat of combination of aluminium with oxygen is
so great that it causes not only the reduction of the Cr 2 3 ,
but fuses the resulting Cr into an ingot. Such metal from
its mode of preparation is free from carbon, and indeed can
be prepared of great purity. Even here, it is interesting to
note, the method depends ultimately on an electro-metal-
lurgical process, viz. the electrolytic reduction of aluminium
(q.v.).
This method of employing aluminium has been used with
success for reducing other oxides, notably oxide of iron.
In this case the object is twofold, namely, to reduce the oxide
to metal and to reduce it at so high a temperature that it
will fuse or raise to a welding heat any joint in iron to which
it may be applied . The method is known as the ' ' Thermite ' '
process, and has been sucessfully used for welding pipes and
jointing rails.
The study of the properties of pure chromium prepared
by reducing chromic oxide by means of aluminium has led
to remarkable results. W. Hittorf has found that although
203
PRACTICAL ELECTRO-CHEMISTRY
chromium is so powerfully electro-positive as to reduce zinc
from its fused salts, yet in an aqueous solution of hydro-
chloric acid or of the chloride of an easily reducible metal it
is inert. Solutions of the chlorides of zinc, cadmium, iron,
nickel, gold, palladium and platinum are not affected ; cupric
chloride and mercuric chloride are reduced to their respec-
tive lower chlorides, but only when the solution is boiling.
This greater activity in a solution at a high temperature is
characteristic of the behaviour of chromium when used as
an anode. At the ordinary temperature it is indeed dis-
solved, but not with the production of chromous chloride ;
it forms chromic anhydride. As an anode in solutions of
metallic chlorides at their boiling point, however, chromous
chloride is produced. Chromium which has been made an
anode under such conditions as to cause it to yield chromic
anhydride assumes a passive state, like that known to occur
in the case of iron, 1 and is incapable of reducing metals
certainly less oxidisable than itself. The whole series of
phenomena exhibits many anomalies, and has not yet re-
ceived full explanation. It is sufficient here to indicate that
a remarkable and interesting addition to our knowledge
of the chemical qualities of a fairly common element has
accrued from the happy facility for the preparation of re-
fractory metals relatively pure and in a compact state,
which has been afforded us by electrolytic methods.
MOLYBDENUM
This metal can be prepared in similar manner to chromium.
It may be obtained free from carbon by heating a mixture
of the dioxide Mo0 2 with defect of carbon in the electric
furnace. It is white, has a density of 9-01, is as malleable
as iron, and can be filed, and, when hot, forged. It is only
slightly oxidised in ordinary air. When heated in contact
with carbon it absorbs a small percentage of that substance,
and can then be hardened by quenching in the manner
characteristic of steel. It forms a definite carbide (Mo 2 C),
1 Analogous effects have been observed with nickel and cobalt.
204
METALS PRODUCED BY ELECTRIC FURNACE
which is hard and crystalline and has a specific gravity of
8-9. The pure metal is very infusible, the carbide somewhat
less infusible.
Molybdenum is used to- a small extent in making special
steels. Moissan proposes to employ it instead of manganese
or aluminium to deoxidise steel in the converter. The ad-
vantages of this substitution would be that the oxide which
would be produced (molybdic acid, Mo0 3 ) is volatile and
would escape from the bath, and that the molybdenum which
might be left in the metal would have similar properties
to the iron with which it was mixed, notably in respect of
its malleability and power of hardening when quenched.
TUNGSTEN
This is another metal of the same group as those already
described. It is infusible save at the highest temperature
of the electric furnace, in which it can be prepared by re-
ducing tungstic acid (W0 3 ) by carbon. When the carbon
is used in defect, and the mass is not completely fused, the
pure metal results ; but if an excess of carbon be employed,
or if the reduced metal is fused so that it comes freely into
contact with the walls of the crucible, it takes up carbon,
giving a cast metal more fusible than pure tungsten. A
definite carbide (W 2 C, containing 3-16 per cent, of carbon)
may be prepared. It has a specific gravity of 16-06 at
18 C., and is hard enough to scratch corundum. Tungsten
free from carbon is soft enough to be filed ; it can be forged ;
it absorbs carbon readily and is hardened thereby, in this
respect resembling generally molybdenum and iron. It
is not attracted by the magnet ; its specific gravity is 18-7.
Tungsten, like molybdenum, is a metal which is used to a
limited extent to produce special steels. The precise proper-
ties and merits of alloys of this description are not well
understood, chiefly because they have not yet been subjected
to the systematic study necessary to give us the precise
knowledge which (thanks largely to the researches of Had-
field) we already possess of steels containing as characteristic
205
PRACTICAL ELECTRO-CHEMISTRY
constituents silicon, aluminium, manganese and nickel.
The easy and relatively cheap manufacture of metals almost
unattainable previously in the pure state will lead to the
examination of their capabilities as constituents of indus-
trial alloys. 1 In the case of tungsten, however, it appears to
be well established that its alloys with iron (tungsten steel)
is capable of being heated to redness without becoming
soft. This property has been turned to account in preparing
steel for tools which in large lathes are run at so high a speed
and with so heavy a cut that the point of the tool is at a dull
red heat ; in spite of this it retains its temper and its cutting
edge. It has also been found that steels containing vanadium
are peculiarly resistant to shock and their utilisation is
already proceeding.
As in the case of chromium, there is a rival method for the
manufacture of tungsten, viz. the reduction of tungstic
acid by aluminium. It is perfectly possible that this method
may prove preferable to reduction in the electric furnace.
Titanium, although at present of small industrial importance,
may be mentioned, because it has proved to be the most
infusible metal which has been prepared by the electric
furnace, far exceeding chromium, tungsten and molybdenum
in this respect. It also has a strong tendency to form a
nitride (Ti 2 N 2 ) and a carbide (TiC). The formation of the
nitride can be prevented by using so powerful a current
that the temperature in the electric furnace is higher than
allows of the continued existence of the nitride ; the carbide
can be disposed of by re-fusing cast titanium containing car-
bon with excess of titanic acid (Ti0 2 ). It will be seen that
even the most refractory of bodies may be reduced, fused,
carburetted, refined and decarburetted in quantity, and with
complete ease and certainty, by means of the electric fur-
nace, which thus takes rank as a valuable instrument of
research and a powerful industrial apparatus.
1 Since this was written the inquiries referred to have been made
and the physical properties of tungsten, molybdenum and vanadium
steels have been studied in considerable detail.
206
CARBIDES
THE production of carbides by heating together certain
metals or non-metals and carbon, or by reducing the oxides of
these elements with excess of carbon in the electric furnace,
is quite general, and has been closely studied by Moissan.
He has arrived at the following conclusions :
At the high temperature of the electric furnace certain
metals, e.g. gold, bismuth and tin, do not dissolve carbon.
Copper will absorb only a small quantity, which suffices,
however, to modify its properties considerably.
Silver at its boiling point dissolves a small quantity of
carbon, and expels it on cooling in the form of graphite ;
the metal containing carbon expands on solidification, just
as does cast iron. Pure iron and pure silver contract on
solidifying.
Aluminium dissolves carbon and ejects it as graphite ;
it also forms a carbide (A1 4 C 3 ).
The platinum metals dissolve carbon, and on solidifying,
eject it as graphite.
Calcium, strontium and barium form carbides of the type
R"C 2 ; lithium yields Li 2 C 2 . All these give acetylene when
acted on by water.
Cerium, lanthanum and yttrium give carbides of the form
CeC 2 , which, however, do not yield pure acetylene, but a
mixture of that gas and ethane.
Manganese gives the carbide Mn 3 C, which decomposes
water with evolution of equal volumes of methane and
hydrogen.
Uranium carbide (Ur 2 C 3 ) gives methane, hydrogen, ethy-
lene, and, what is most interesting, a quantity of liquid and
207
PRACTICAL ELECTRO-CHEMISTRY
solid hydrocarbons representing about two-thirds of its
total carbon. On this and cognate facts Moissan has erected
a new and ingenious theory of the mode of formation of
petroleum.
Other metals form definite carbides sharply distinguished
from the foregoing by their remarkable stability. Examples
are Mo 2 C, W 2 C, Cr 4 C and Cr 3 C 2 . These are of metallic
appearance, very hard, and fusible only at a high tempera-
ture.
The carbides of the non-metals silicon and boron (SiC
and CB 6 ) and that of the pseudo-metal titanium (TiC)
are distinguished by their hardness, which exceeds that of
corundum.
Out of this long list, only two carbides are of industrial
importance : the one, calcium carbide, belongs to the group
of those carbides producing a gaseous hydrocarbon by
the action of water ; the other, silicon carbide, is an
example of the carbides which are useful because of their
great hardness.
CALCIUM CARBIDE
In 1862 Wohler prepared calcium carbide by heating
an alloy of zinc and calcium with an excess of carbon. The
body was not isolated, but the fact was recognised that it
evolved acetylene on treatment with water. Travers in
1893 heated a mixture of calcium chloride, carbon, and so-
dium, and obtained a grey mass containing calcium carbide.
On the 12th December, 1892, Moissan published the
following statement in a paper communicated to the Acad-
emic des Sciences : "If the temperature " (in the electric
furnace) " reaches 3,000, the lime forming the furnace melts
and runs like water. At this temperature carbon quickly
reduces calcium oxide, and the metal is separated in quantity ;
it unites easily with the carbon of the electrodes to form a
carbide of calcium, liquid at a red heat and easily collected."
This paper was supplemented by a note to the Academie
on 5th March, 1894, in which the facts were set forth that
208
CARBIDES
there is but a single carbide of calcium, that its formula is
CaC 2 , and that it yields pure acetylene when decomposed
by water.
Towards the end of 1894 Willson announced that he had
produced a substance giving acetylene when acted on by
water, by heating lime and carbon in the electric furnace.
His discovery appears to have been accidental and indepen-
dent of Moissan's work, with which he seems to have been
unacquainted. As soon as calcium carbide was recognised
as a valuable commodity, Willson and others endeavoured
to protect its production by patent. The state of knowledge
at the time was, however, too well advanced to warrant
the creation of a monopoly of this kind, and at the present
moment it is doubtful whether any patents for the production
of calcium carbide in the electric furnace, except such as
relate to some particular form of furnace, are valid.
Calcium carbide, though colourless when pure, is, as ordin-
arily prepared in the electric furnace, a dark, semi-metallic-
looking solid ; it can be broken easily, and its fracture is
crystalline. Isolated crystals are reddish-brown in colour ;
their sections under the microscope are seen to be transparent
and deep red in hue. Calcium carbide has a specific gravity
of 2-22. It is insoluble in all ordinary organic solvents.
It burns when heated in oxygen, forming calcium carbonate ;
when fused it dissolves carbon, and on cooling deposits
the carbon as graphite. This property is common to many
carbides ; those of iron and molybdenum may be cited.
The most noteworthy reaction of calcium carbide is that
which occurs when it is brought into contact with water.
Decomposition takes place smoothly according to the equa-
tion
CaC 2 + 2 H 2 = Ca(OH) 2 + C 2 H 2 .
Given that the carbide is pure, the yield of acetylene is
that required by theory, and the gas is pure. Even with the
industrial material these conditions are approached. As
might be expected, the carbides of barium and strontium
(BaC 2 and SrC 2 ) can be prepared from mixtures of the respec-
209 P
PRACTICAL ELECTRO-CHEMISTRY
tive oxides with carbon by the aid of the electric furnace.
Both furnish acetylene when treated with water.
The manufacture of calcium carbide is carried out in
a, very simple apparatus. All that is necessary is a fire-
brick box containing a charge of lime and coke, which can be
fused together by the passage of a powerful current. Seeing
that the production of calcium carbide is effected solely
by reason of the high temperature attained in the electric
furnace, and not by electrolysis, either an alternating or
unidirection current may be used. The former is generally
FIG. 38.
the more convenient, because it can be brought from a
distance at a high voltage and transformed on the spot
where it is to be used by a stationary transformer.
The simplest arrangement is that originally devised by
Willson. It is shown in Pig. 38. The brickwork casing
A is lined with carbon B, so as to leave a hollow which serves
as the crucible. The crucible itself acts as one electrode,
the other being a stout carbon rod c. A small charge is
placed in the crucible and an arc established. The electrode
is gradually raised, and fresh charges are fed in. A fused
210
CARBIDES
mass of carbide is formed at the bottom of the crucible, and
can be run off by the tapping hole E.
This apparatus represents one type of carbide furnace,
namely, that in which the carbide is completely fluid and is
tapped at intervals. This method has the advantage that
as the carbide is periodically removed from the sphere of
action it cannot be overheated and thereby decomposed
a not impossible contingency. Also, being fluid, it runs
free from the solid half-changed charge, and is nearly pure.
"Block" carbide (described below) may contain entangled
in it a quantity of partly converted material and conse-
quently be a good deal less pure.
It will be seen that in the Willson furnace the charge is
completely enclosed, and the walls of the crucible themselves
constitute one electrode. This is a disadvantage, as the
current is distributed from the walls through all portions
211
PRACTICAL ELECTRO-CHEMISTRY
of the charge ; the advantage of a protecting layer of
unfused charge lining the cavity is thus lost, and the walls
are likely to be overheated and, being of carbon, to take part
in the reaction and suffer corrosion. These inconveniences
are partly remedied in the furnace shown in Fig. 39. A
rectangular iron box A is lined with carbon blocks B, which
form a cavity in which is the charge c. The upper electrode
is a carbon rod, and an iron plate embedded in the base
block and insulated from the iron casing forms the other.
By this device the flow of the current is confined to
some extent, the greater part passing from the base block
direct to the charge. The charge itself forms the lining
and covering of the zone of highest temperature, so that
heating takes place by means of a sort of smothered arc.
In practice the raw materials are packed round the end of
the upper electrode as closely as possible and suffice to con-
fine the heat to some extent. The carbide is tapped at E
from time to time.
The most efficient form of furnace for the production of
tapped carbide would be one in which a crucible is used,
as in the Willson furnace (which is practically of the
original Siemens type), so as to conserve the heat- by en-
closing the arc completely, and in which the walls are of
some refractory material other than carbon, which shall not
be capable of taking part in any reaction with the charge.
I have endeavoured to embody these ideas in the furnace
shown in Fig. 40.
The body of the furnace A is of firebrick, and the lining, B
is magnesia, which is sufficiently refractory and indifferent.
The lower electrode is a carbon block c, and the upper a
carbon rod D ; there is a tapping hole E. The lower part
of the furnace is contracted so that the section of the column
of fused carbide may be smaller than the section of that
part of the raw materials which is actually undergoing
conversion. By this means compensation is provided for
the fact that the conductivity of the carbide is greater
than that of the raw materials, sufficient heat being
generated by the passage of the current to keep the carbide
212
CARBIDES
fused and fit for tapping. As the lining of the crucible
is non-conducting and refractory the charge can be piled
well up round the electrode D, and heating performed by
an arc which is effectively smothered.
Most furnaces used in manufacture are of the intermittent
type. Examples of these are as follows.
FIG. 40.
One devised by Willson, and used by him at Spray, North
Carolina, is shown in Fig. 41, which represents a pair built
together. The electrode c consists of a bundle of carbon
plates A, each 4 inches square and 30 inches long. They are
suspended from a thick copper rod, through which electrical
connection is made, and this hangs by a chain passing over
pulleys and controlled by a screw and nut D. The other
terminal is connected with the iron plate E, on which rests
213
PRACTICAL ELECTRO-CHEMISTRY
a layer of carbon F, composed of broken carbon pencils
or a baked mass of coke and tar.
The upper electrode is shown at its lowest point resting
on the lower electrode, but it will be understood that as the
charge is fused it is raised so that a conical pile of carbide is
gradually formed. The current is then cut off, and the mass
of carbide, after cooling, is withdrawn, broken up. and the
FIG. 41.
fully fused, nearly pure part picked out from the sintered
half -formed carbide, surplus coke, slag and similar debris.
By a natural improvement on the Spray furnace, the
King furnace has come into existence. It is shown in
two sections in Figs. 42a and 426.
The chief point of importance is the use of a small iron
truck to contain the carbide as it is produced. The truck A,
with its load of carbide, forms one electrode. It can be
run into place and removed as required. It is provided
with trunnions K, K, so that its contents may be tipped
out. It is given a small reciprocating motion by the
rod E, this being found useful in shaking down the charge
and preventing the formation of channels in it, and also in
214
CARBIDES
slightly altering the position of the arc so that all parts of
the charge are exposed to it in turn.
The other electrode consists of a bundle of carbon plates,
carried by a massive rod c, consisting of a conducting band
of copper strengthened with side bars of iron. In the figure
this electrode is shown resting on the floor of the truck,
but it will be understood the electrode is slowly raised as the
charge is fed in and fused, until it reaches the top of the truck,
which is then full of carbide and can be removed and re-
\
FIG. 42a.
FIG. 426.
placed by another ready for a fresh run. The raw materials
are fed into the furnace through the channels G, G, which
contain small rotating blades to control the descent of the
charge. The air flues shown are to keep the upper part of
the furnace fairly cool, the zone of fusion being confined to
the truck. This class of furnace is semi-continuous, the
only interruption to its working being that needed for re-
moving and replacing the trucks. Many attempts have
been made to construct furnaces strictly continuous in their
operation, that is -to say, having a continuous feeding-in of
215
PRACTICAL ELECTRO-CHEMISTRY
raw materials and a continuous discharge of fused carbide,
but they appear to be less manageable than furnaces of the
semi-continuous type.
A furnace of the semi-continuous type is the Horry furnace
used by the Union Carbide Co. at Niagara Falls. This
furnace is shown in the figure below. Two vertical elec-
trodes drop into an enclosure on the periphery of the drum,
into the mixture of raw materials ; current flows between
FIG. 43.
the ends of the two electrodes, and carbide is produced.
The periphery of the drum is closed by cross plates, a few
of which are removed at the point where the electrodes are
hung ; a block of carbide is formed here, and the drum is
revolved away from the electrode, fresh raw material being
supplied and more carbide formed. Ultimately a semi-
ring of carbide, held up by a series of cross plates, is
produced, and when this ring reaches the side opposite the
hanging electrode the carbide has become solid and can
216
CARBIDES
be removed. The electrodes may be covered by a hood, so
as to collect the carbon monoxide evolved by the reaction.
The utilisation of this gas is contemplated, but at the time of
my visit to the works had not been put into operation.
It may be noted that the fundamental reaction CaO +
3 C = CaC 2 + CO expends one-third of the carbon in produc-
ing carbon monoxide. A natural suggestion is to burn this gas
and use the heat for warming the charge before it descends
on to the hearth of the electric furnace, but this has not
yet been realised in practice. An extension of the idea
is to heat the charge non-electrically to as high a tempera-
ture as can be reached by ordinary furnace methods, leav-
ing the electric furnace to raise its temperature through the
remaining 1,000 C. or 1,500 C. necessary to cause the
reaction to occur. A considerable economy might be ex-
pected from this procedure, because calorie for calorie the
heat generated by the electric furnace is enormously more
costly than that generated direct from fuel, but up to the
present no practical realisation of the idea has been attained.
Borchers has suggested the enclosing of the electric
furnace with a water jacket, which shall serve as a boiler
to generate steam from heat that would otherwise escape
and from the heat of the block of fused carbide, which at the
end of the run has to be left in the furnace to cool. This
suggestion, even if carried out, would have but a trifling
effect in reducing the cost of the carbide. Many attempts
have been made to prepare carbide commercially without
the use of the electric furnace ; these have been uniformly
unsuccessful. It may be accepted that the lowest tempera-
ture at which carbide can be produced is 2,000 C= 3,632 F.,
and this is about the topmost limit of any non-electric
furnace. Borchers has experimented with a blast of air
enriched with oxygen, but the trials, though interesting,
have led to no commercial result.
Attempts have been made to obtain a more even distribu-
tion of temperature in a furnace using three-phase currents.
A plant of 800 H.P. has been erected at San Marcello
d'Aosta, in Italy, according to the patents of Ricardo
217
PRACTICAL ELECTRO-CHEMISTRY
Memmo. The simplest form of discontinuous furnace for
three-phase currents is shown in Fig. 44.
The carbons, although converging, cover a considerable
area, and fluctuations of current taken by the furnace,
due to a high resistance at any given point, are less severe
than with the ordinary single electrode. The furnace has
a capacity of about 70 cubic feet, and is made of brickwork,
lined with refractory bricks. The bottom on which the
fused carbide rests is made of magnesia bricks magnesia
being unattacked and not forming a carbide, as does lime.
As shown in the figure, the carbons c, c, c (which are 5
FIG. 44.
inches in diameter) are carried by stout iron rods A, A, A.
These pass through bronze collars, and can be screwed
up and down by the hand wheels B, B, B. The attachment
of the carbon rods to their sockets is apt to cause trouble
unless special precautions are taken. The carbon should
be inserted when both it and the iron are as hot as they are
likely to become in practice, and any crevices filled up with
a graphite cement. It is well to stop the run before the
carbons are quite consumed, lest an arc form between the
carbide and the holder, ruining the latter.
A semi-continuous furnace for three-phase currents is
shown in Fig. 45. The raw materials are fed in at the top,
218
CARBIDES
and fall on the cast-iron plate A, which is protected by a
layer of graphite. As the fusion proceeds this plate is
lowered by the screw B with its gear c. A column of carbide
is thus built up, the top of the column always forming one
electrode and the three carbons jointly the other electrode.
These are only moved slightly to compensate for their
gradual consumption. The carbide, when it has reached
the lowest part of the furnace, is sufficiently cool to enable
FIG. 45.
it to be withdrawn, and the running of the furnace can be
resumed.
Taking the question broadly, it may be said that modern
carbide furnaces are simple machines. If block carbide is
to be produced, a form of Siemens furnace with a smothered
arc, fed by hand and provided with any ordinary mechanical
device for raising and lowering the upper electrode and for
removing the pot containing the finished carbide will suffice.
It is probable that the bulk of carbide made in Europe is
219
PRACTICAL ELECTRO-CHEMISTRY
prepared in this manner. Tapping furnaces are generally
less simple and handy, and it is doubtful whether the better
quality of their product will outweigh the advantages of the
more elementary type.
The quality of the raw materials for the manufacture of
calcium carbide is of importance. Both lime and carbon
should be as nearly pure as possible. The lime should not
only be free from siliceous impurities, but should also be
free from magnesia. This base is unattacked by carbon
at the temperature of the electric furnace ; it is not re-
duced, nor does it yield a carbide. The most convenient
form of carbon is coke ; charcoal can be used and contains
a smaller percentage of mineral impurities, but it is incon-
veniently bulky. The coke should contain as little ash as
possible. Coke of good ordinary quality contains about
10 per cent, of ash ; for calcium carbide manufacture the
quantity should be considerably less 5 per cent., or better
if procurable. The lime may, of course, be used as carbon-
ate, but this alternative is not desirable, because the work
of decomposing the carbonate is thrown on the electric
furnace, the energy of which is costly. It is usually better
to prepare the lime in an ordinary kiln. The comminu-
tion and mingling of the raw materials have been the sub-
ject of much study. It was at first supposed that the raw
materials should be finely ground. Now, however, it is
found that pieces as much as 1 inch in diameter will serve
perfectly well, and the preparation of the raw materials
resolves itself into a sort of cracking process instead of
grinding. The machines most in vogue are of the coffee-
mill type, eminently adapted to produce coarse fragments
of uniform size almost free from dust.
From the nature of the case, seeing that in the electric
furnace the energy poured into it is effectively boxed in
and must be transformed on the spot where it is wanted
into heat of high temperature, it might be supposed that the
manufacture of calcium carbide is a fairly efficient process.
Enquiry shows that this is a true view. The energy strictly
necessary may be computed thus. Moissan has shown that
220
CARBIDES
the heat of formation of calcium oxide is 145 Cal., and
that the reaction CaO + C 3 = CaC 2 + CO takes place at
3,300 C. = 5,972 F. The specific heat of CaO may be
taken as approximately 0-12; that of carbon as 0-47. *
The energy necessary to raise 56 grammes of CaO and 36
grammes of C to this temperature is 79- 5 Cal. The formation
of calcium carbide from Ca and C is esteemed an endother-
mic reaction requiring 48 Cal. The total energy needed is,
therefore, 79-5 + 145 + 48 Cal. = 272-5 Cal. From this
must be deducted the energy evolved by the oxidation of
C to Co, i.e. 29 Cal. Therefore, the energy to be supplied
to form 64 grammes of CaC 2 is 243-5 Cal. In this calcula-
tion the energy absorbed or evolved by the formation of
CaC 2 from Ca and C 2 is a doubtful quantity. Later
computations make it considerably smaller, e.g. 0-65 Cal.,
and some authorities regard it as slightly exothermic,
evolving 3-9 Cal. The estimate given is likely to err on the
right side, the more so as no credit has been taken for
possible regeneration by utilising the sensible heat of one
charge for warming up the next. Thus it may be taken for
practical purposes that the formation of 1 ton of CaC 2
requires 5,889 H.P. hours, or conversely for each H.P. per
day of 24 hours 4- 1 kilos of CaC 2 may be formed ; if, how-
ever, the more favourable view be taken, this value becomes
4,320 H.P. hours per ton of carbide.
It must not be forgotten that this estimate includes the
whole of the heat needed to raise the raw materials to the
temperature of the reaction and supposes that this heat
is lost. In practice at least a portion of it will be used in
pre-heating the raw materials before they are exposed to
the full temperature of the furnace. In like manner no*
credit is given for the heat which can be obtained by the
combustion of the CO evolved in the production of the car-
bide. The output claimed by some works is as much as
5 kilos per H.P. per 24 hours, say 1'8 ton of carbide per H.P.
1 These values are confessedly approximate. That of carbom
increases greatly with the temperature, and the figure adopted applies;
to temperatures not lower than 900 C. -=1,652 F.
221
PRACTICAL ELECTRO-CHEMISTRY
year ; but it is probable that this is carbide containing only
90 per cent, of actual CaC 2 . It is usual to consider that
in practice 1 ton of carbide can be produced by 1 H.P. year.
Recent information shows that as much as 1'5 tons may
be obtained in actual work, which agrees closely with
the calculated output, given on the preceding page, viz.,
5889 H.P. hours per ton.
An early experiment on this question may be quoted. In
1896 an American paper, The Progressive Age, retained
Messrs. Houston, Kennelley, and Kinnicutt, electricians and
chemists of repute, to make experiments at Spray, North
Carolina, on the cost of production of calcium carbide. These
experiments were on a manufacturing scale, and appear to
have been well conceived and well executed ; their results
were published in full and without comment. The plant
used consisted of a turbine of about 300 H.P., coupled
to alternators which delivered current at 1,000 volts to
transformers, whereby the pressure was reduced to 100
volts. Two furnaces were used, each with a floor area of
3 feet x 2 feet 6 inches, and having an iron base plate
covered with carbon 8 inches thick. This served as the
lower electrode ; the upper was a built-up carbon block
3 feet x 12 inches x 8 inches. It could be raised gradually
from the base plate as the mass of calcium carbide formed
thereon ; its consumption was y 1 ^ inch per hour. The
charge consisted of coke and lime, containing 52 per cent, of
CaO and 37 per cent, of C, the balance being moisture and
impurities. At the start a few shovelfuls of this mixture
were placed on the lower electrode, and an arc established
between this and the upper electrode. Fresh portions of
the mixture were added as the reaction proceeded, until
the cavity of the furnaces was filled with a pyramidal mass
of crude carbide.
Two runs were made, each with a charge of 2,000 pounds ;
in each case an output of about 200 pounds of calcium car-
bide was obtained. The carbide gave 80 to 85 per cent,
of its calculated yield of acetylene. In the first run 193-1
H.P. for 3 hours was used, corresponding with 579-3 H.P.
222
CARBIDES
hours, i.e. 432 kilo-watt hours. In the second run the energy
consumed was equivalent to 195-3 H.P. for 2 hours 40 minutes,
corresponding with 520-8 H.P. hours, or 388-5 kilo-watt
hours. Taking the output of carbide as 200 pounds in
each run, the first gives 3-75 kilos of 80-85 per cent, carbide
per H.P. per 24 hours, and the second 4-15 kilos for the
same expenditure of energy. These values are well below
the 5 kilos provisionally fixed above, and have the advantage
over the various figures commonly quoted of having been
derived from actual experiment. The cost of the carbide
prepared in these experiments may be calculated thus :
The plant is one delivering 200 E.H.P., and turns out 4
kilos of 85 per cent, carbide per H.P. per 24 hours in all,
292 tons per year of 365 days, running day and night. This
may be conveniently stated as 327 short tons (of 2,000
pounds), because the remaining figures are taken from the
American source cited above and refer to this unit of weight.
The cost of power per h.p. year is 6 dollars. 1 The capital
expenditure for the plant (other than power plant) is 12,000
dollars. Taking labour for making the carbide at 11 dollars
per day ; lime at 6-3 dollars per ton ; coke at 4-5 dollars
per ton, and carbon for the electrodes at 6 cents per pound,
the cost of producing 292 tons of carbide is found to be
as follows :
Dollars.
Power . 1,200
Interest and depreciation ... . 1,200
Labour 4,015
Lime 1,260
Coke 1,134
Carbon electrodes 450 2
9,259
This works out at a little more than 28 dollars per ton of
1 This is very low; 10-20 dollars is a more ordinary figure.
2 This is so considerable an item that in well-equipped works the
carbon electrodes are made on the premises, not bought from an
electrode manufacturer.
223
PRACTICAL ELECTRO-CHEMISTRY
2,000 pounds, i.e. 6 65. per ton of 2,240 pounds. In this
estimate, which, though confessedly only approximate,
is based on actual prices and experimental data, the chief
points to be noted are that it is much below the present
selling price of carbide, which is about 12 per ton of
2,240 pounds ; that the cost of power is low, and that
of labour and material high. All these items would
vary largely according to the local conditions. Power
(even water power) may well cost 20-25 dollars per year,
and per contra the price of lime may be not more than 3
dollars per ton, and that of coke 2J dollars per ton. Thus,
though the cost of power is a large item, yet it is not of such
preponderating importance as to make a calcium carbide
factory necessarily a success because it can obtain the
energy it requires at a low rate ; the industry may be ham-
pered beyond hope by dear and bad coke and lime. These
considerations are of particular importance when consider-
ing the prospect of success possessed by a given scheme for
utilising water power in a manufacture of this kind.
The fact that the cost of power, though so considerable
a factor, is not overwhelming in its influence on the manu-
facturing cost of carbide, makes it possible to establish and
work successfully a carbide factory quite independently
of water power. For example, any works possessing modern
coke ovens from which bye-products are recovered produces
large quantities of combustible gas ; in . like manner the
quantity of blast furnace gases from an iron works is far
larger than can be profitably utilised for heating the blast
and raising steam for the ordinary requirements of power
for blowing and for handling the materials. The surplus
gas can be used with economy in large gas engines, e.g. of
500 or 1,000 H.P., and energy thus obtained almost as cheaply
as from a water power. For example, at an inclusive cost of
To^- P er H - p - hour, which is by no means unattainable, the
price per H.P. year is 3 13s., a figure which approaches that
of a moderately cheap water power. The real obstacle to the
general utilisation of such power is not its cost, but the some-
what restricted market for carbide, causing it to be readily
224
CARBIDES
swamped by any great increase of supply ; even with that
restriction, however, the manufacturer having cheap coke
and lime in an industrial centre will stand at least as good
a chance as his rival with slightly cheaper power but away
from such supplies.
As regards the conditions to be especially kept in view
by the manufacturer, it is sufficient to say that the raw
materials of each charge should be converted as nearly as
possible completely into calcium carbide, to avoid the neces-
sity of heating them over again as will be requisite if they
have to be worked up with the next charge ; but, however
carefully the operation is conducted, there is likely to be a
comparatively large part of the charge which has served
as a protection and envelope to that which has been fused,
and must be reworked or thrown away. Slag and similar
inert products must be picked out. The quality of the car-
bide should be measured by the volume of acetylene which
a given weight evolves when acted on by water, and the
material should be bought and sold on this assay.
One kilo of pure CaC 2 evolves 348-4 litres of acetylene,
the gas being measured at a pressure of 760 mm. and a
temperature of C. This quantity corresponds with
5-587 cubic feet for 1 pound. The commercial product
rarely gives more than 300 litres per kilo, and often only
280 or even less. Even the better of these is only 86 per
cent, of full strength. It is clear, therefore, that a good
deal may be done to improve the quality of calcium car-
bide as now manufactured.
SILICON CARBIDE
The other carbide of industrial importance 1 is silicon
1 There is at present no other carbide than calcium carbide and
silicon carbide which is used as such commercially ; barium carbide
has, however, been proposed as a source of cyanide ; in this case
the carbide is used to absorb nitrogen, the cyanogen converted into
alkali cyanides and the barium serving again for the production of
carbide.
225 Q
PRACTICAL ELECTRO-CHEMISTRY
carbide (SiC), which can be prepared synthetically by the
direct union of its elements at the temperature of the electric
furnace. Commercially, the oxide of silicon, i.e. silica
such as quartz, is used as the source of silicon, which is
reduced from silica by carbon and combined with a further
quantity of carbon at a single operation, according to the
equation Si0 2 + 3 C = SiC + 2 CO. The commercial name
for silicon carbide is carborundum, a word constructed to
convey the idea that the material is of the nature of corun-
dum (crystallised alumina), but contains carbon. Of
course there is no chemical similarity of carborundum to
corundum. Pure silicon carbide is colourless and crystal-
lises in hexagonal plates. It contains 70 per cent, of silicon
and 30 of carbon ; its specific gravity is 3-12 ; it is hard
enough to scratch ruby. It is extremely stable and does
not oxidise even when heated in air to whiteness. It is
insoluble in all acids, but is attacked by fused caustic potash.
This great refractoriness is in striking contrast to the ease
with which the other industrial carbide, calcium carbide,
is decomposed by water. Although pure SiC is colourless,
the crystals usually obtained from materials not perfectly
free from iron and similar impurities are slightly coloured,
and may be blue, yellow, or brown. The commercial pro-
duct is dark brown or black.
Silicon carbide was discovered and first manufactured
by Mr. E. G. Acheson. His process is in use under his
direction at the works of the Carborundum Co. at Niagara
Falls. There is stated to be a carborundum works in Austria
and another in Savoy, but probably the greater part of the
world's output still comes from the original works.
The furnace used is built of bricks put together without
mortar or cement, both because of the need to allow free
escape of gases and because the whole structure has to be
pulled down at the end of this run. The furnaces used
until lately at the works of the Carborundum Co. at Niagara
Falls were about 15 feet long, 7 feet high and 7 feet wide.
At each end is a heavy bronze casting to which the leads are
connected, and which on the inner side carries a bundle of
226
CARBIDES
sixty 3-inch carbon rods 2 feet long. These project into
the furnace cavity proper, and between them is a cylindrical
mass of coarsely powdered coke making electrical connec-
tion between the carbon electrodes ; this core of coke is
about 9 feet long and nearly 2 feet in diameter. Thus
it will be seen that the manufacture of silicon carbide, unlike
that of calcium carbide, is effected by heating a resistance
and not by an arc. The general arrangement of a carborun-
dum furnace is represented diagrammatically in Fig. 46.
A is the loosely-built brick box, carrying the heavy metal
holders B, B, to which the cables are attached. The carbon
rods c, c are set in these holders and project well into the
furnace. The conductive cylinder of broken coke is shown
between the ends of the carbon rods at D. The charge
FIG. 46.
which is packed round this heating core and fills up the
cavity of the furnace consists of 34-2 per cent, of coke,
54-2 per cent, of sand, 9 - 9 per cent, of sawdust, and 1*7
per cent, of common salt ;* it weighs about 10 tons, and the
yield of carborundum from this quantity is not more than
2 tons. The calculated yield of 10 tons of silica and carbon
mixed in equivalent proportions, i.e. 62- 5 per cent, of silica
and 37-5 per cent, of carbon, is 4J tons of silicon carbide,
whence it will be seen that the output is poor. The
1 The function of the sawdust and coke is probably to render
the charge sufficiently porous to allow of the escape of the carbon
monoxide, which is abundantly produced in the running of the fur-
nace.
227
PRACTICAL ELECTRO-CHEMISTRY
reason for this is that a great part of the charge serves as
a covering to the central part, and confines the heat thereto.
The outer layers are only partly converted into carborun-
dum ; they are worked in with the next charge. The later
type of furnace does not differ in principle from that de-
scribed, but the details of construction have been modified.
The furnace is 30 ft. x 9 x 9 over all, and the resistance
core which carries the current and round which the charge
is packed, is found of square sectional carbon rods laid
zigzag, with cross blocks at the angles, as shown in the
figure. For a furnace of the size given about 1,000 H.P. is
required. When the current is switched on, heating pro-
ceeds slowly until, after about 2 hours, carbon monoxide is
evolved at all openings in the rough brickwork and from the
FIG. 47.
upper surface of the charge, and there burns with a blue
flame. The current is passed for about 36 hours, at the
end of which time it is found that the reaction has proceeded
as far as it is feasible to push it, and the current may be
switched off and the furnace allowed to cool. The whole
operation of loading, heating and drawing occupies about
72 hours. On pulling down the walls of the furnace the
charge is found to be composed of several layers ; the outer
consists of about 11 per cent, of salt (volatilised from the
inner part of the charge), 56 per cent, of silica, and 33
per cent, of carbon, this representing the portion which has
not been hot enough to form silicon carbide. Within this is
a layer of harder material of a greenish colour and roughly
concentric with the core ; this consists of amorphous silicon
carbide mixed with unaltered raw materials. It is not hard
228
CARBIDES
enough to be used as carborundum, and has to be worked
up with the next charge. The next inner layer is crystallised
silicon carbide, carborundum proper. The crystals constitut-
ing this layer are small on the outside, and increase in size
towards the core. The total thickness of the useful layer
may be some 16 inches. Within this again is the core of
coke or carbon rods which has been converted into graphite
by the high temperature to which it has been subjected. The
layer of properly crystallised silicon carbide is broken up,
crushed in edge runners, washed with water and acid, dried
and graded by sieving. The following analyses illustrate
its composition :
T
II
Si
62-70
69-10
C
Fe 2 O 3 +A12Q 3 . . .
CaO . . .
36-26
0-93
30-20
0-49
0-15
MgO
0-11
100-00
99-94
Carborundum is used as an abrasive. Its extreme hard-
ness makes it preferable to emery for some purposes. Car-
borundum wheels are stated to cut so much more freely
than emery wheels that the article being ground is much
less heated than it would be by an emery wheel. The
guiding wheels are made by heating carborundum with 20
per cent, of ordinary porcelain mixture to such a temperature
that the porcelain sinters and binds the whole into a coherent
mass. Carborundum paper and cloth, similar to the pro-
ducts made from emery, are also prepared. Silicon carbide
is used as a means for adding silicon to steel in regulated
amount ; at a price of 80 dollars per ton about 1J million
pounds per year are thus employed. The electrical power
needed for producing carborundum has been reduced from
229
PRACTICAL ELECTRO-CHEMISTRY
15-5 to 8-6 kilo- watt hours per kilogramme. The output
for 1903 is given as about 3,000 tons.
SILOXICON
The study of the reactions concerned in the production
of silicon carbide has resulted in the production of another
material intermediate, as it were, in composition between
silica and silicon carbide. This body, termed siloxicon by
Acheson, is found by reducing silica with carbon, but not
carrying the reduction as far as to produce carborundum.
The greenish-yellow material found surrounding the core of
silicon carbide in the ordinary running of a carborundum
furnace, probably contains siloxicon as well as amorphous
silicon carbide ; in practice the same partial reduction is
secured more systematically. Mr. Acheson's description
of the method, in a letter to the author, maybe usefully
transcribed ; he says " siloxicon is an oxygen-carbon-silicon
compound which forms in the electric furnace from proper
mixtures of silica and carbon at about 2,500 C ( 4,532F.).
It is exceedingly refractory, neutral towards acid and
basic slags, infusible and insoluble in molten metals. At
fusion temperatures it is decomposed by pure alkalies, and
in the presence of free oxygen it oxidises at about 1,500 C.
In a neutral or reducing atmosphere, however, it is unaffected
until its temperature of decomposition is reached, which
is well over 3,000 C. Upon decomposition the oxygen is set
free, the carbon and silicon uniting to form carbide of silicon,
which is in itself an exceedingly refractory material." It
will be seen from this that the primary condition of pro-
duction is a moderated temperature, a condition easily
secured by regulation of the current. The composition of
siloxicon is given by the following analysis :
230
CARBIDES
Si 57-7
C 25-9
Al . . . 0-4
Fe . 21
Ca -."'.. . . Trace
Mg . ... Trace
O (by difference) 13-9
100-0
Corresponding approximately with the formula Si 5 C 5 O 2
Siloxicon is suitable for use as a furnace lining. If made
into bricks, it is mixed with 2 per cent, of alumina and baked
at a temperature but little below its oxidising point.
It may be applied as a lining by mixing it with coal tar, or
with a solution of silicate of soda, and painting it upon the
surfaces to be protected.
ARTIFICIAL GRAPHITE
Another characteristic product of the electric furnace is
artificial graphite. The commercial production of this sub-
stance is also due to Ache son. The furnaces for preparing
graphite are similar to those used for carborundum. Each
is about 30 feet long and takes 8,000-9,000 amperes at 80
volts, corresponding with about 1,000 H.P. ; the charge is
about 3J short tons. Carbon electrodes, each with a cross
section of 400 square inches, are used, and between them
is placed the carbon to be converted into graphite. Two
materials are manufactured. For the first, namely graphite
in mass, anthracite crushed to the size of a pea is employed.
This is packed round a core and converted bodily into
graphite which can be powdered and moulded precisely
as is natural graphite, and used for the same purposes.
Carbon electrodes are the second product. These are made
231
PRACTICAL ELECTRO-CHEMISTRY
by heating in the same kind of furnace ordinary carbon
electrodes moulded and baked in the usual manner ; they
retain their form, though their nature has been altered
fundamentally. Graphite electrodes have been found
particularly suitable for many electrolytic processes in which
ordinary carbon electrodes are disintegrated. The charge,
whether of anthracite or of carbon electrodes, is, of course,
protected while being heated ; for this purpose it is covered
with a mixture of sand and coke such as is used for making
carborundum. Graphite prepared by the Acheson process
is almost pure ; it is substantially free from ash, containing
less than that present in the raw materials. Fitzgerald
gives examples ; an anthracite containing 5-78 per cent, of ash
gave graphite with 0-03 per cent. ; a carbon electrode having
2 per cent, of ash yielded only 0-04 per cent, after being
graphitised. In 1903, 1,200 tons of graphite were produced
from amorphous carbon.
The mechanism of the formation of graphite in the
electric furnace is obscure. The most obvious and natural
explanation is that as graphite is the final and most stable
form of carbon at a high temperature, conversion takes
place simply by reason of that high temperature. But
Acheson, whose opinion must be received with respect,
maintains a different view. He considers that the graphite
is produced by the decomposition of carbides, instancing
the formation of graphite by the dissociation of silicon
carbide at a high temperature. The carbides which serve
by their decomposition for the production of graphite are
formed from silica and metallic oxides, e.g. oxide of iron, and
the amorphous carbon which is to be converted. The
quantity of these oxides is altogether insufficient to combine
with the whole of the amorphous carbon at one time ; hence it
must be assumed that carbides are formed and decomposed,
formed and decomposed until the whole of the amorphous
carbon has at one time or another existed as a carbide, has
been released from its combination, and has appeared as
graphite. Further, it seems that the elements (silicon, iron
or what not) which have served as carriers of carbon from
232
CARBIDES
the amorphous state to combination as a carbide, and finally
to the condition of graphite, are volatilised and disappear,
having done their work and leaving the carbon with which
they have been in transitory union as graphite substantially
free from other elements. These ideas are worthy of study
and consideration ; their establishment and acceptance
require further experimental work.
233
BORIDES
No borides are as yet prepared on an industrial scale. A
few words may, however, be added to those already set
down in an earlier part of this section concerning Moissan's
work on the synthesis of borides.
Boron, like silicon and carbon, combines with certain
metals and non-metals to form bodies which are stable and
of simple composition. Examples are the borides of iron,
nickel, and cobalt FeB, NiB, CoB and CB 6 . The first
three can be prepared at ordinary furnace temperatures,
but carbon boride is a typical product of the electric furnace ;
it is formed when the two elements are heated together
at a temperature of about 3,000 C. = 5,432 F. Good
crystals can be obtained when the union of the constituents
is brought about in a bath of copper or silver acting as a
solvent. Carbon boride (CB 6 ) crystallised from fused
copper is a black crystalline substance of specific gravity
2-51. It ignites when heated in oxygen to 1,000 C. =
1832 F., but burns with difficulty because the boric anhy-
dride produced forms a protective skin. It is insoluble in all
acids, but is attacked by fused alkalies. Its most note-
worthy property is its extreme hardness. Silicon carbide
is considerably harder than corundum, but nevertheless will
only polish diamond without actually cutting it ; carbon
boride, however, will cut diamond, not perhaps as well as
"diamond itself, but still definitely enough. In the scale
of hard materials diamond must still stand first, but next
to it is carbon boride, then titanium carbide, then silicon
carbide, and perhaps corundum as the fifth. The indus-
trial preparation of carbon boride as an abrasive may
prove useful and remunerative.
234
SILICON AND SILICIDES
SILICON and its compounds with metals can be produced
easily enough in the electric furnace. Their manufacture
is already practised, and its extension waits for the dis-
covery of new and useful applications. At present it appears
that the chief directions in which these are likely to be found
are the production of special alloys and the use of silicon,
alone or combined, as a fuel local in its effects. Silicon
used for the sake of the heat evolved by its oxidation has
the advantage that its product of oxidation is a solid, and
the loss of heat concomitant with the formation and dis-
sipation of a gas is avoided. On this account silicon may be
preferable to carbon as a material for heating ; the fact has
been recognised in what is picturesquely termed Klein-
bessemerei, clumsily expressed in English as the production
of steel on a small scale in a Bessemer converter.
Silicon itself is made by reducing silica with carbon
very much as in the preparation of silicon carbide, but using
a smaller proportion of carbon. It is now prepared not only
in powder but in lumps, and can be used as a reducing
agent for steel and as an addition to cast iron ; in this
latter case it replaces the silicon burnt in the cupola, and
allows a grey iron to be obtained when, but for the addition,
a hard brittle white iron would be obtained. As in all
similar cases the limit to its use is fixed by its price. Ferro-
silicon is made in large quantities by reducing a mixture
of silica (clean sand) and ferric oxide (good haematite ore)
with coke in a furnace similar to a carbide furnace with a
smothered arc, the product being tapped as the reduction
proceeds. Ferrosilicon containing a moderate percentage of
Si can be made in the blast furnace, but as grades which are
235
PRACTICAL ELECTRO-CHEMISTRY
rich in silicon are difficultly fusible, the use of the electric
furnace is essential for these. It is used for dosing steel
and cast iron just as is silicon, the substance needed being
silicon and the iron being so much lumber. Commercial
ferrosilicon is a mixture of various silicides. A definite
silicide Fe 2 Si has, however, been isolated, corresponding with
chromium silicide Cr 2 Si, which can be prepared in a similar
way, and has at present found no application.
Silicide of copper containing 10, 15 and 30-35 per cent, of
silicon is an article of commerce. The richest of these
compounds corresponds approximately with the formula
CuSi. Silicide of copper is employed instead of phosphide of
copper as a reducing agent useful in the production of
bronzes and other copper alloys, and especially for adding
to copper itself to form so-called silicon-bronze, used for
telephone and telegraph wire. In these materials silicon
is not necessarily present in more than minute quantity ;
having deoxidised the metal to which it has been added, it
may disappear in the slag.
Other silicides of interest, though not yet of commercial im-
portance, are those of the alkaline earth metals, CaSi 2 ,
BaSi 2 , SiSi 2 . These bodies are analogous in composition
to the carbides of the same metals, and are prepared by
heating the oxide of the metal, e.g. lime, with silica and
carbon, the latter being in sufficient quantity to reduce the
lime to calcium and the silica to silicon. They are white or
bluish white crystalline substances which oxidise slowly
in air at ordinary temperatures, and more quickly when
heated. They react with water, but do not yield a com-
pound of silicon and hydrogen corresponding with acetylene.
The reaction of barium silicide on water may be stated thus
BaSi 2 + 6 H 2 0=:Ba(OH) 2 + 2 SiO 2 + 5H 2 : Strontium
and calcium silicides give a similar reaction, but less
vigorously. On account of the ease with which the
barium compound reacts it has been proposed as a portable
source of hydrogen and as a reducing agent for indigo.
The behaviour of the silicides with acids is curious. BaSi 2
reacts thus :
236
SILICON AND SILICIDES
2BaSi 2 + 4HC1 + 4H 2 = 2 BaCl 2 + 2SiH 4 + 2Si0 2 + 2H 2 ,
whereas calcium silicide gives :
CaSi 2 + 2 HC1 = CaCl 2 + Si 2 H 2 .
Si 2 H 2 is analogous in composition to acetylene ; it is a
yellow crystalline substance easily oxidisable. Strontium
silicide gives a reaction such as might be expected from
a mixture of CaSi 2 and BaSi 2 . Calcium silicide may prove
useful as a reducing and desulphurising agent for steel.
There are other industries dependent on the use of the
electric furnace which are excellent illustrations of its
peculiar powers. At the beginning of this section it was
stated that the characteristic property of electric heating
was the application of heat at the precise point needed, and,
as a corollary, the ease of enclosure of that heat. The
production of carbon disulphide is a good example. Carbon
and sulphur unite, but the reaction is endothermic and the
necessary energy has to be supplied through the walls of a
retort when the heating is conducted in an ordinary furnace.
The obvious disadvantages of this procedure are overcome
when the heating is electrical. Taylor has erected a furnace
at Penn Yan, in the State of New York, which is of the shaft
type, 40 feet in height and 16 feet in diameter, generally
resembling a smelting furnace, but heated electrically. The
electrodes are of carbon, and are set in the hearth of the
furnace. The carbon is fed down the shaft through a bell
and the sulphur through an annular chamber. These,
arriving at the hearth, are heated to a temperature sufficient
to cause the formation of CS 2 , which escapes at a side opening
near the bell. The furnace takes 4,000 amperes at 40-60
volts say 300 electrical H.P. and it is stated not without
some ground, that the only drawback to the electrical manu-
facture of carbon disulphide is that the market for this
solvent is somewhat limited.
The preparation of phosphorus affords another case. It
is not too much to say that the whole manufacture is
now electrical. The materials, calcium phosphate, silica
and carbon, are heated in a furnace of the resistance type,
237
PRACTICAL ELECTRO-CHEMISTRY
and the phosphorus distils and is collected. The chemistry
of the operation is that of the older chemical manufacture ;
it is the mode of applying heat which is new and economical.
The fusion of such refractory materials as silica and
alumina evidently can best be accomplished in the electric
furnace. Alumina is fused to produce an artificial corundum
employed as an abradent. Silica given that an oxidising
atmosphere is maintained may be fused to form quartz
glass, and tubes may be formed from the material. It is the
necessity of maintaining this oxidising atmosphere and the
difficulty of securing such an atmosphere in the presence
of carbon electrodes which has hindered the prepara-
tion of large vessels of molten silica. Further, silica is
volatile at a temperature little above its melting point ;
hence the very reasonable hope that large vessels fit for
industrial purposes may be prepared by fusing sand in the
electric furnace must be preserved until experiments have
been made on a larger scale. Meanwhile small apparatus
made of silica fused electrically or by the oxyhydrogen
blowpipe has come into common laboratory use.
While this book has been going through the press in-
formation has been received showing that large vessels of
fused silica have been successfully prepared in the electric
furnace.
238
SECTION V
Iron and Steel
Iron and Steel
AT the time when the first edition of this book was
published so little practical success had been attained
in the electrometallurgical production and treatment of
iron and steel that no section was allotted to the subject.
A good deal of experimental work had been done, but the
outcome was, at the time, inconsiderable. But within the
last few years a great change has occurred. That earlier
period of experimental struggle which seemed all but hopeless
has as its legitimate successor the present epoch of moderately
fruitful toil ; there is good reason to believe that this will
merge into an era of remunerative industrial activity.
Probably the chief reason why success has been won
slowly is that the efforts of the pioneers were directed amiss.
Their ambition was to smelt iron electrically from its ores.
That can be and has been accomplished, but it is less easy
and less immediately useful than the production of steel and
alloys of iron by electrical means, using as raw material the
ordinary products of the blast furnace. The situation
may be summed up by saying that as the blast furnace is a
fairly efficient thermal device, and as it uses fuel direct
instead of in a round-about electrical manner, it is difficult
to displace except in those places where fuel is extravagantly
dear, whereas in the conversion of cheap pig iron into high-
priced steel or special iron alloys the cost of energy necessary
for the process is not so large a part of the total cost as to
make electrical methods impracticable.
Evidently, even when it is intended to make steel from
blast furnace pig, a country having cheap water power
possesses an advantage. Hence it is not surprising that
241 B
PKACTICAL ELECTRO-CHEMISTRY
the Canadian Government thought it advisable to appoint
a Commission to examine and report upon existing processes
of iron reduction and steel manufacture in Europe. The
report of this Commission has been recently issued, and is
valuable as containing descriptions of most of the processes
now in use or being tried. Doubtless as the true object and
scope of electrical methods for the manufacture of iron and
steel became better appreciated, improved processes will
be devised, and it may well be that our own ironmasters
will realize that the huge horsepower obtainable from the
gases from their blast furnaces is eminently suitable for
running electrical steel furnaces, using the product of their
blast furnaces as a raw material. Rapid development of
electrical steel making will be far more likely then than it is
at the moment when power is sought in out-of-the-way
places having nothing industrially in their favour except
a superfluity of falling water.
SMELTING PROCESSES FOR IRON
At Livet, Keller Leleux & Co. have furnaces competent
to reduce iron from the ore. The principle is practically
that of a carbide furnace with a smothered arc. The ore is
fed in at 'the top, the fused mass forms a sort of bath between
the electrodes, which are of carbon, and at intervals the
product is tapped. A diagrammatic figure is shown below.
The point of interest is that there are four hearths and a
central well. The four hearths are in two pairs, in series,
the members of each pair being in parallel ; an alternative
path is provided, when the hearths are in turn emptied, so
as to make the load only moderately irregular. The charge
is the ordinary burden of a blast furnace, and the cost of
production is at least as large as that of iron made in the
blast furnace. Other methods have been devised, notably
by Stassano ; for this the Report of the Canadian Com-
mission should be consulted. These applications of the
electric furnace are of less immediate practical importance
than is its utilisation in the production of steel.
242
FIG. 48. KELLER FURNACE WITH FOUR HEARTHS.
243
PRACTICAL ELECTRO-CHEMISTRY
ELECTRICAL MANUFACTURE OF STEEL
There are various types of furnace in use for the manu-
facture of steel electrically. In all a charge such as is
employed in an ordinary open hearth gas furnace,
consisting, that is, of pig, ore, scrap and appropriate
fluxes, may be worked up. In fact, the operations of
FIG. 49. HEROULT STEEL FURNACE.
steel making are carried out in the manner and by the
methods commonly in use, the sole difference being that the
heat is produced in the furnace electrically instead of being
obtained directly from fuel burning above the charge to be
heated.
244
245
PRACTICAL ELECTRO-CHEMISTRY
The Heroult process which is at work at Kortfors in Sweden,
and is also in operation at La Praz in France under the
direction of the inventor, is an excellent example of
electrical methods applied to a steel furnace of modern
design. The general appearance of the apparatus is shown
in Fig. 50, and its construction may be seen from Fig.
49. Essentially it is a tilting furnace, through the arched
roof of which two large water-jacketed carbon electrodes
depend. The current passes from one of these through
a short air gap, to the bath of metal, and from the
bath of metal through a second short air gap to the
other electrode, and can be regulated either by adjusting
the position of the electrodes by hand or automatically,
according to the variation of voltage between each electrode
and the bath. The furnace, which has a capacity of 4
tons, is provided with basic hearth, and takes an alter-
nating current of 4,000 amperes at 1 10 volts. The electrodes
are 1*7 metre long by 360 x 360 mm., and last about a week,
representing an average output of 40 tons of steel.
In actual trials made at La Praz steel of various grades
was produced ranging from the softest material suitable
for transformers to metal of the grade of tool steel contain-
ing 1 per cent, of carbon. Phosphorus can be eliminated
as in the ordinary basic process. The consumption of
energy is 0-1 to 0-153 electrical H.P. year per 2,000 Ibs. of
steel produced. Taking the latter figure, this amounts to
1-53 dollars with electrical energy at 10 dollars per H.P. year,
and to this should be added 0-2 dollar for electrodes. A
small furnace fired by coal will use about 1,200 Ibs. per 2,000
Ibs. of steel, and this at 5 dollars per 2,000 Ibs. (not an
extravagant figure in districts where coal is scarce) corre-
sponds with 3 dollars per 2,000 Ibs. of steel. It may be
taken that materials and labour will cost much the same
in both types of furnace, hence there is an estimated balance
in favour of the electrical method of 1-27 dollar per 2,000 Ibs.
of metal. Higher cost of repair will probably absorb this.
It must be remembered, too, that the figures given above
are for small furnaces, e.g. up to 10 tons ; with large tilting
246
Section A B
FIQ. 51. KJELLIN FUBNACB.
247
PRACTICAL ELECTRO-CHEMISTRY
gas fired furnaces holding 100 or 200 tons the balance of
advantage might well be against the electrical apparatus.
In fact, here is an example of what was said in the intro-
duction to this section : at present the true province of
the electric furnace is not the manufacture of vast quantities
of cheap material, but small quantities of steel of the very
highest grade at a price only slightly above that of common
structural steel, and enormously below that of the pure
high carbon steels hitherto prepared by ancient and costly
processes.
The Kjellin process differs from all other electric furnace
methods in that the furnace is destitute of electrodes.
Briefly, the furnace is a transformer. The current supplied
to the primary of many turns is converted into heat in
the secondary, which has a single turn, and consists
of the steel to be heated. Fig. 51 shows the arrange-
ment of the furnace in use at Gysiuge in Sweden. A A
is the primary wound round one leg of the magnetic
circuit c c c c. B B is the secondary of molten steel
contained in an annular groove. D D is the furnace
casing and F F a flue through which air passes to keep the
primary cool ; a water jacket may be substituted. A
tapping spout is shown at H.
The furnace takes about 200 H.P., and will turn out about
a ton of steel every 6 hours. The charge consists of high-
grade pig and scrap, and the product is steel of the class of
crucible steel in that such impurities as sulphur and phos-
phorus are absent, but evidently with any content of carbon
which may be desired. In fact, the whole apparatus may be
regarded as a device by which steel may be made without
contamination either from fuel gases or electrodes. It is
not adapted for making cheap structural steel or for refin-
ing impure raw materials. The electrical energy required
averages 0-14 electrical H.P. year per 2,000 Ibs. of steel
produced, corresponding with a cost of 1-4 dollars with energy
at 10 dollars per H.P. year. The total cost of manufacture
may be taken as 34 dollars per 2,000 Ibs. of steel, an ex-
penditure materially smaller than that necessary for steel
248
IRON AND STEEL
of similar grade 'make by the ordinary furnace processes.
Gustave Gin has patented a furnace for the production
FIG. 52. GIN ELECTRIC FURNACE.
of steel which,though at present worked only experimentally,
shows so much of interest that a description will not be out
of place. The charge is contained in a channel A in a
refractory lining, and constitutes a resistance through which
249
PRACTICAL ELECTRO-CHEMISTRY
the current flows from the terminals B B. (see Fig. 52). A
vertical section is shown in Fig. 53. The terminals are
water cooled, the connections being shown at E F, Fig 53.
Fused pig is run in at H, and scrap or ore may be added ;
tapping takes place at K. The whole arrangement is similar
to an open hearth gas fired furnace, except that the heating
is electrical and is applied direct to the charge and boxed in
by the roof instead of being produced in the vault of the
furnace and reflected from the roof on to the charge.
SPECIAL STEELS
Evidently the electric furnace is eminently suited for the
production of special steels or special iron alloys, whenever
their cost is high enough to warrant the use of a somewhat
costly mode of heating. The more refractory such an alloy
may be, the more advantageous is the electrcial method.
Ferro-tungsten, ferro- vanadium, ferro-chromium, even alloys
of iron with manganese and nickel may well be made electri-
cally. The actual smelting can be performed in furnaces of
the Heroult type, and the production of special steels by
dosing ordinary scrap with regulated quantities of the iron
alloys aforesaid can be carried out either in the Heroult or
the Kjellin furnace. The objection based on cost to the
smelting of common pig electrically disappears when special
iron alloys are to be smelted. Here the cost is of small
moment compared with the necessity of obtaining a pure
product of regular composition. In like manner the ad-
vantage which can be secured by electrical heating when
high grade carbon steel is to be prepared, is enhanced when
steels of the class of modern high speed steels or shock-
resisting steels alloyed with selected elements, e.g. tungsten
and vanadium, are required.
It may be added that should it be found feasible to
smelt iron ore remuneratively in the electric furnace, the
subsequent refining to steel can obviously be carried out in a
second furnace, into which the molten metal can be tapped,
the arrangement being strictly analogous to a blast furnace
worked in conjunction with open hearth furnaces.
250
SECTION VI
Electro-Deposition
Electro-Deposition
art of winning and refining metals on a commer-
cial scale by means of electrolysis has been practised
for but a short time, and in that time has undergone a very
rapid development. The electro-deposition of metals in
thin films to form replicas of embossed, incised or ornamented
surfaces, or to cover, protect or embellish some other metal,
is of older date, and at the present moment is somewhat
eclipsed by the growth and importance of its congener.
But, although electro-deposition (in this limited sense)
may be a smaller trade, it is absolutely large and of great
practical importance. Whilst it is true that such commodi-
ties as pure electrolytic copper and calcium carbide are
necessaries of modern industry, it is no less true that electro-
types and electroplate are conveniences of modern life
which could ill be dispensed with.
The earliest application of electrolysis to the deposition
of metals in thin films, exactly clothing and reproducing
the surfaces on which these films are deposited, was made
in the case of copper. The art of electrotyping, as it is now
called, seems to have been discovered in 1838 by at least
three persons Spencer, Jacobi and Jordan almost simul-
taneously, and its utility for the accurate reproduction
of engraved objects was so obvious that its development
was rapid. A year or two later an efficient solution (that
of the double cyanide) for the deposition of silver was dis-
covered, and electroplating was established as an industry.
Copper is the only metal which is used for producing
electrotypes, though doubtless others could be employed
253
PRACTICAL ELECTRO-CHEMISTRY
if it were necessary or desirable to do so. Electro typing
differs from electroplating, nickel plating, and similar
forms of electro-deposition in that the deposited metal is
afterwards stripped from the surface on which it has been
deposited. Forming as it does an independent object, it
must needs be of fair thickness, whereas a plating proper
may be (and often is) the merest film. Thus in electrotyp-
ing it is necessary that the surface to be reproduced should
not be so absolutely clean as to allow the deposited metal
to adhere firmly to it ; the faintest imaginable film of
grease or oxide will prevent such adhesion. In plating,
on the other hand, perfect adhesion is essential, and the
art of the plater is directed to cleansing the surface of the
metal to be coated so effectually that the deposited metal
is afterwards inseparable. It is failure to attain this end
which often causes plating to strip and expose the metal
which it is intended to embellish or protect.
254
ELECTROTYPING
BY use this term is confined to the formation of copper
replicas of articles in relief or intaglio. The principles on
which the art depends are simple, and may be gathered
from what has already been said on the winning and refining
of copper. In their early days electrotypes were produced
by making the article to be copied the cathode of a Daniell
cell. A rod of zinc in a porous pot filled with dilute sulphuric
acid or zinc sulphate was coupled to the mould to be covered,
which was immersed in a solution of copper sulphate sur-
rounding the porous pot. The arrangement was then equiv-
alent to a short-circuited Daniell cell, and as the zinc
dissolved an equivalent of copper was deposited on the mould.
In the ordinary Daniell cell designed for the production of
a current to be used outside the cell, copper is deposited on
the copper plate (which may be replaced by lead, carbon,
platinum and. the like), and is usually fully adherent. If,
however, the copper plate be not absolutely clean the copper
deposited may be detached, and its surface which has been
in contact with the plate will exhibit faithfully the irregulari-
ties, such as dints or file marks, which may have existed on
the original plate. Such detached deposited copper is in
the fullest sense an electrotype of the surface of the cathode
plate. The application of the ideas here embodied is
simple. A mould of some material which can be cast on
the object to be copied, so as to produce an exact copy,
is made sufficiently conductive to serve as the cathode of
any convenient source of current in an electrolyte consist-
ing of sulphate of copper. The copper deposited on this
mould is prevented from sticking too firmly to the mould
'by care in choice of the surface of the mould, which, though
conductive, should not be perfectly clean, untarnished metal ;
255
PRACTICAL ELECTRO-CHEMISTRY
otherwise the deposited metal adheres, and becomes a pro-
tective coat.
For a full description of the various technical details of
electrotyping, special works must be consulted. The more
important requirements of the art are set forth below.
In order to take a cast of the object to be copied, various
compositions are used. Gutta-percha and mixtures of that
substance with fatty materials, plaster of Paris, and fusible
metal are types of the various plastic or fusible substances
which may serve to take an impression. If gutta-percha is
used it is softened at a temperature of about 100 C. = 212 F.,
and when thus made plastic is pressed on to the surface to
be reproduced. After cooling the gutta-percha becomes
hard, and may be detached and used as a mould, from which
the original object to be copied may be reproduced with
exactitude. The set gutta-percha, though hard enough to
retain fine lines, is yet sufficiently elastic to allow of the re-
moval of the cast from an object which is slightly undercut,
whereas fusible metal, plaster, or sealing wax would obviously
fail under these conditions. Various prescriptions for mix-
tures containing gutta-percha are available. One consist-
ing of 66 per cent, of gutta-percha, 33 per cent, of lard, and
1 per cent, of Russian tallow is approved as suitable for
making a mixture so fluid that it may be poured over the
engraved plate and will copy the finest lines. In such
prescriptions, which pertain rather to cookery than chemistry,
there is usually some ingredient chiefly valuable as an aid
to faith, as, for example, the 1 per cent, of tallow in that
quoted. The mould when made from gutta-percha, or
from a mixture of gutta-percha and some fatty material,
is non-conductive, and is usually brushed over with plumbago
so as to give it a conductive coating. On this the first film of
deposited metal is formed evenly, and subsequent deposition
is simple. The adhesion of the metal to the film of plum-
bago is slight, and the electrotype can be readily detached.
Sometimes plumbago is incorporated with the gutta-
percha mixture itself, but the rationale of the procedure
is not obvious. It is not necessary to make the body of the
256
ELECTROTYPING
cast conductive if the surface is a sufficiently good conductor
to allow of the deposition of a film of metal ; when this is
accomplished, no further aid to conductivity is needed.
Plaster of Paris is not very well suited for making electro-
type moulds. The ordinary grades are too coarsely ground
to reproduce fine lines. Sharper impressions may be ob-
tained with Keene's cement (which is calcium sulphate al-
most pure and completely dehydrated), but its setting is
slow. But, however they may be obtained, plaster casts
are porous, and are slowly soluble in water, so that their
sharpness would be blurred if they were exposed directly
to the electrolyte. Accordingly they are protected by soak-
ing them in paraffin wax or some similar waterproof material ;
the surface is made conductive by plumbago, as in the case
of gutta-percha.
Fusible metal is a suitable substance of which to make
casts. One of the best of the ordinary fusible alloys is
Wood's metal, composed of four parts by weight of bis-
muth, two of lead, one of tin and one of cadmium ; it melts
at 141 F. = 60-5 C. The conditions to be fulfilled by such
an alloy are that it shall melt at a temperature conveniently
low low enough not to injure the object to be reproduced
and that it shall expand on solidification so as to force
itself fully into contact with the object to be copied. Fusi-
ble metal evidently needs no coating to make it conductive ;
rather, it requires an almost imperceptible film of oil or to
be slightly tarnished in order that deposited metal may not
adhere to it. It is not much used because of its relatively
high price ; the inevitable waste and the possible deteriora-
tion of the alloy in remelting limit the use of this material
in spite of certain obvious advantages.
Only one other moulding material need be mentioned.
For work undercut or in high relief a flexible material is
useful. This may be made from common glue, softened by
soaking in cold water, and melted together with about one
quarter its weight of treacle. The composition may be
made waterproof by adding to it 2 per cent, of tannin, which
combines with the gelatine of the glue to form an insoluble
257 s
PRACTICAL ELECTRO-CHEMISTRY
leather-like substance, or by soaking the finished cast in a
10 per cent, solution of potassium bichromate and then ex-
posing it to strong light. Bichromated gelatine when ex-
posed to light becomes insoluble in water, and the cast pre-
pared from it may be immersed in an aqueous electrolyte
without much risk.
All these materials, except fusible metal aforesaid, need
to be provided with a conductive film to enable the first
layer of metal to be deposited. When plumbago is used
it must be fine and perfectly free from grit, lest it scratch
the delicate surface of the cast. Graphite made in the
electric furnace (see p. 231), being almost free from mineral
matter, would probably serve better than natural graphite.
Although plumbago is most commonly employed, various
other substances will serve. Thus any finely-powdered
metal, such as gold, silver, aluminium or bronze powder,
may be painted or rubbed on to the mould. It is doubtful,
however, whether any metal can be prepared either by grind-
ing its leaf or in other mechanical manner of as great fine-
ness as that of plumbago ; an equally delicate coating is
hardly to be expected. Metal may be chemically deposited
on the mould in several ways. Thus by Parkes's method the
mould is coated with silver by dipping it in a solution of
phosphorus in carbon disulphide, and then in one of a silver
salt. The phosphorus reduces the silver and coats the cast.
Similarly, the cast may be coated by any of the ordinary
silvering mixtures, such as are used for coating glass surfaces
with actual silver ; such mixtures, consisting of a silver salt
with a reducing agent, e.g. Rochelle salt, aldehyde or for-
mic acid, are freely employed in silvering mirrors, the pro-
cess displacing " silvering " with mercury and tin. A film of
silver obtained in this manner may have too clean a surface
to be suitable for electrotyping, because the deposited metal
may adhere to it ; this inconvenience may be remedied
by slightly tarnishing the silver with sulphide. A metallic
coating may be provided by immersing the cast in a solution
of copper sulphate and sprinkling it with very fine filings
of iron, these depositing copper. All these methods are, how-
258
ELECTROTYPING
ever, relatively unimportant ; covering with plumbago is
the simplest device, and for most purposes the best.
The mould, however it may have been prepared, is
coated with copper by making it the cathode in an elec-
trolyte prepared by dissolving 1J pounds of crystallised
copper sulphate in 1 gallon of water and adding J pound of
sulphuric acid. A current density of 10 amperes per square
foot will generally be found suitable. The concentration of
the electrolyte is maintained by the use of copper anodes,
which should be of pure electrolytic copper.
In short, the conditions to be observed are substantially
identical with those necessary for refining copper electroly-
tically, save that, as the rate of deposition is usually not
important and as pure materials may be used, a perfect
coating may be more easily obtained. The process may be
continued until an adequate thickness of metal has been
deposited. Frequently this is small, as the plate can be
backed with a fusible alloy. Various devices are employed
to obtain a satisfactory coating on irregular objects. An
indented surface will receive on its depressed portions a
smaller quantity of copper than will be deposited on its more
prominent parts. The difficulty may be got over by using
a small movable anode, e.g. a thick wire, which may be
approached towards the depression and thus decrease the
resistance at that point, correspondingly raising the current
density on the cathode at that point to its normal value.
Although it is not quite easy to obtain smooth regular
deposits of such substantial thickness, e.g. % in. or more as
to allow the deposited metal to be used without backing
or support, yet with care and skill this thickness can be
attained. For example, seamless copper pots for laboratory
use are made by electro-deposition and are certainly pre-
ferable to brazed goods. The chief precautions necessary
are to keep the electrolyte clean, to circulate it well so that
there may always be ample copper at the cathode, to have
the anode at a considerable distance from the cathode, in
order that the resistance between it and all parts of the cath-
ode may be nearly identical, and finally to use a low current
259
PRACTICAL ELECTRO-CHEMISTRY
density, taking abundant time for the work of deposition.
The building up of copies of objects in the round and not in
the form of plates more or less indented or embossed is a
difficult and delicate art too remote from the subjects of
this book to be treated of here. All necessary principles
for the deposition of the metal when once the mould has
been prepared have been already laid down.
Copper is not commonly deposited to form a protective
coating, as distinct from a thick layer which is to be stripped
and to reproduce the surface on which it has been deposited.
In certain cases, however, it may be used thus. It may be
deposited on iron and steel either itself to serve as a protec-
tion or to act as the basis for a coating of nickel. The appli-
cation, of copper to protect steel has been used for plating
ships, but more as an experiment than in practice. There
is no metal other than iron which would benefit sufficiently
by a protective coating of copper to warrant the extensive
use of copper electroplating, and in the case of iron certain
difficulties arise. The coating must be perfect, as otherwise
corrosion of the iron will take place at the exposed spot, all
the more vigorously for the presence of the copper. De-
position from the ordinary coppering solution consisting of
copper sulphate dissolved in water and acid with sulphuric
acid is impracticable, because iron is capable per se of deposit-
ing copper from such a solution and the copper is apt to come
down in a non-adherent condition. It is possible to " flash "
iron with copper, i.e. to give it an extremely thin film
by rapid immersion in a solution of copper sulphate, and
possibly a good coating might be built up on this film if the
article were at once made the cathode in a coppering solution.
The general method, however, is to deposit the copper from
an alkaline bath, which will not attack iron. In electrotyp-
ing, as stated above, it is essential that the surface of a metal
mould to be copied, though conductive, should not be chemi-
cally clean. In electroplating with copper, where perfect
adhesion is essential, the metal to be coated must be cleaned
most scrupulously. The process of cleaning is similar in most
cases, whether copper or some other metal is to be deposited.
260
ELECTROTYPING
The object to be coated is freed from obvious impurities by
filing or scraping so as to present a smooth, bright surface.
If of iron which has been machined or finished bright it
may have been greased to protect it from rust. In this case
the grease is wiped off as completely as possible, and the
slight film remaining is removed by washing in a volatile
solvent, such as benzoline or coal-tar naphtha. Seeing that
the least trace of grease is objectionable in that it prevents
the formation of an adherent film, it is usual to dip the goods
in a hot 10 per cent, solution of caustic soda after the bulk
of the grease has been removed by the volatile solvent. The
cleaned surface may still be tarnished with a film of oxide :
this is removed by dipping in an acid bath containing 10 per
cent, of sulphuric acid or 25 per cent, of ordinary aqueous
hydrochloric acid. The acid is rinsed off with clean water
and the plating begun at once If delay occurs the metal
will begin to oxidise again and the acid dip must be repeated.
The perfectly clean iron goods are then coppered in an
alkaline bath. That most commonly employed contains
cuprous cyanide dissolved in an aqueous solution of potas-
sium cyanide, being therefore similar to the solution of silver
cyanide dissolved in potassium cyanide ordinarily used for
depositing silver (see below). A suitable copper bath of
this class consists of 4 parts of the double cyanide of copper
and potassium, 0-5 parts of ammonia, 0-5 parts of potassium
cyanide, and 94 parts of water. A current density of 3
amperes per square foot is used. Another type of alkaline
copper bath is prepared by adding caustic potash or soda to a
solution of a copper salt containing a tartrate. The pre-
sence of tartaric acid prevents the precipitation of cupric
hydroxide, and allows the formation of an electrolyte which
is strongly alkaline, but nevertheless contains copper in
solution. The well-known capability of ammonia to dis-
solve copper oxide, and thus to yield an electrolyte which
is alkaline and nevertheless rich in copper, does not seem
to have been used in the copper-plating industry. It is
possible, as Oettel has shown, to obtain adherent and co-
herent deposits of copper from an ammoniacal electrolyte,
261
PRACTICAL ELECTRO-CHEMISTRY
but it is probable that the necessary conditions must be
observed somewhat too closely for convenience in an indus-
trial process. Moreover there is always loss of ammonia going
on, whereby the composition of the bath is altered and the
air of the work room made unpleasant. Such inconveniences
occur to some extent with cyanide baths, but are absent from
those containing an alkaline tartrate. Electro typing, plat-
ing, and other arts depending on the deposition of metals
electrolytically in thin films are now well-established trades.
They have passed from the hands of the chemist and electri-
cian to those of the works manager and foreman. Natur-
ally, therefore, they have suffered an accretion of recipes.
Save possibly in the art of tempering steel, there is no branch
of metal-working so fruitful in nostrums as that now under
discussion. Some of the many complex baths which have
been proposed contain ingredients the use of which is
intelligible ; in others there are substances whose function
is obscure ; in some occur materials apparently chosen
by lot. A bath devised by Roseleur, which is suitable for
iron and can be used for other metals, is prepared by
grinding up 3J ounces of copper acetate with a little
water so as to make a smooth paste, adding to this
3J ounces of crystallised carbonate of soda and 1J
pints of water. Copper carbonate and sodium acetate
result from this reaction. The copper is then reduced
to the cuprous state by the addition of 3J ounces of
sodium bisulphite, dissolved in 1J pints of water. The
cuprous salt is then dissolved by potassium cyanide, of which
3 J ounces are used, dissolved in 5 pints of water. This is
probably an easy way of producing a cyanide solution of
cuprous cyanide, but there is no reason to suppose that an
equally good result could not be obtained by starting with
cupric chloride, precipitating it with sodium carbonate,
reducing this with sodium bisulphite, and forming a double
cyanide solution by adding excess of potassium cyanide. In
like manner, one might equally well precipitate copper sul-
phate with caustic soda, reduce the precipitated cupric
hydroxide with sulphurous acid, and add cyanide in excess.
262
ELECTROTYPING
An electrolyte of the tartrate class may be prepared by
dissolving 5J ounces of copper sulphate in a gallon of water,
adding 1 J pounds of Rochelle salt (double tartrate of potas-
sium and sodium) and then 13 ounces of caustic soda. In
these alkaline baths copper anodes dissolve less readily than
in the ordinary acid electrolyte, and it is sometimes neces-
sary to maintain the strength of the bath by adding a fresh
supply of a copper salt. When the iron goods have received
a fair coating of copper in an alkaline bath they may be
transferred to the usual acid electrolyte, and the required
thickness of copper obtained as in ordinary copper plating.
The double operation and the need for obtaining a particu-
larly perfect and somewhat thick covering of copper in order
to protect the iron effectually make the use of copper plating
on iron less common than would be expected from a consider-
ation of its obvious advantages. It has, however, a consider-
able application in the coppering of rollers for printing
designs on calico and other materials. Such rollers are of
iron or steel, and are coated with copper thick enough to be
engraved upon. The process is that already given, viz.
deposition first in an alkaline and then in an acid bath,
special care being taken to obtain a uniform thickness of
metal. The bath may be a vertical cylinder lined with a
pure copper plate serving as the anode, and having the rol-
ler placed concentrically with the cylinder and arranged
so that it can be rotated. The electrolyte is circulated
and the current density maintained as uniform as possible
over the surface of the cathode. Alternatively the de-
position may be carried out in a horizontal trough, with a
large anode of pure copper plate covering the bottom and
sides and with the roller rotating within this trough, the
whole arrangement resembling that used in the Elmore
process for making copper tubes.
Iron and steel are sometimes given a thin coating of cop-
per in an alkaline bath as a preliminary to the deposition
of nickel. Nickel can be deposited direct on iron, but it
usually adheres better if the metal is first given a film of cop-
per. The matter is further dealt with under Nickel Plating.
263
ELECTROPLATING
IN the trade this term usually means electroplating with
silver. For our purpose it may be conveniently extended
to include the covering by electrolytic methods of one
material with a thin and adherent layer of another. The
old term for silver vessels for domestic use is " plate." Goods
covered with silver by mechanical means (rolling on or
soldering) are termed plated goods ; when a method was
devised of covering an inferior metal with silver by electro-
lytic means, the process was called electroplating ; hence
the customary restriction of the term to silver.
264
SILVER PLATING
THIS is effected by making the objects to be coated act as
the cathode in an electrolyte containing silver, usually
in the form of silver cyanide dissolved in potassium cyanide.
Other electrolytes containing silver may be used, but this
is the most generally applicable. Before an article is plated
it must be carefully cleansed and made not merely mechani-
cally but chemically clean. The process of cleaning varies
to some extent according to the nature of the base metal to
be plated, but is usually effected in the following stages.
In the first place, all obvious impurities are removed by
scouring or similar mechanical means. Next, grease may
be got rid of by dipping the goods in a solvent, such as
benzoline or coal-tar naphtha. This process may be supple-
mented or replaced by immersion in a 10 per cent, solution of
caustic potash used hot. When once the removal of grease
has been effected, the goods to be plated must not be touched
with the fingers, lest a greasy film be again imparted to the
portions touched. A rinse in water follows, and then a dip
in acid, usually dilute nitric acid, to remove any film of
oxide or sulphide. Finally, a second rinse in water and the
goods are ready for the plating vat. All impurities have
been removed from the surface, and the clean metal (faintly
etched and roughened by the action of the acid) is ready to
receive a coating of silver. If there is delay between the
final dip and immersion in the bath, oxidation and tarnish-
ing may occur again and must be removed by dipping once
more in acid. Some discretion must be exercised accord-
ing to the nature of the metal composing the article to be
plated. The acid liquid is highly corrosive, and dipping
265
PRACTICAL ELECTRO-CHEMISTRY
must be done fairly quickly ; the alkali also will corrode
alloys containing much tin. For such reasons, as well as
to avoid wasting metal and acid, the process of cleansing
should not be continued longer than is strictly necessary.
For brass goods the acid dip may be replaced by one of
potassium cyanide, which will dissolve any slight film of
oxide, though more slowly than does the acid liquid. Iron
and steel are usually dipped in dilute hydrochloric acid or
sulphuric acid instead of nitric acid, the action of which is
somewhat too violent.
Soft metals and alloys, e.g. tin, pewter, lead and Brit-
annia metal, may be satisfactorily cleaned without an
acid dip. All these small differences depend on considera-
tions which are obvious to the chemist ; in the art of electro-
plating they are matters of workshop knowledge and tradi-
tion. An additional means for providing a faultless metal-
lic surface on which silver may be deposited consists in the
process known as " quicking." This consists in dipping
the carefully cleaned goods in a solution containing mercury
which is deposited by direct chemical action of the more
electro-positive metal on the mercury salt. Mercuric
nitrate in the proportion of 1-2 ounces per gallon of water
is commonly used ; another suitable quicking solution
consists of mercuric cyanide dissolved in potassium cyanide.
Momentary immersion is sufficient to give the goods a
complete film of mercury, to which the silver ultimately
deposited on them adheres well.
The goods thus carefully prepared are made the cathode
in a bath consisting of silver cyanide dissolved in excess of
potassium cyanide. A usual proportion is 10 grammes of
silver cyanide, 15 grammes of potassium cyanide, and 1
litre of water, but the precise strength is not important.
The bath may be prepared by precipitating silver nitrate
with its equivalent of potassium cyanide, filtering and
washing the silver cyanide, dissolving this in potassium
cyanide solution, and diluting with water to the requisite
extent. There are many variants of this prescription. Thus
silver nitrate may be treated direct with excess of potassium
266
SILVER PLATING
cyanide, or silver chloride may be dissolved in the same
mixture. Also a bath may be made up by dissolving silver
electrolytically in potassium cyanide, but there is no especial
advantage in the procedure.
Anodes of pure silver are used so that the strength of the
bath in silver may be maintained. Various devices are
adopted for obtaining a uniform coating of silver. If the
surface is much indented, small anodes may be brought
near to the concave or re-entrant portions so as to reduce
the resistance at that point and thus bring the current
density to an equality with that at the more prominent
parts.
When the part is very difficult of access or where the
article as a whole cannot be immersed so as to bring this
part into contact with the electrolyte, it may be silvered
by the use of the apparatus known as the " doctor," which
is merely a pad of rag moistened with the electrolyte and
having an anode embedded in it. This may be applied to
the part in question, the article itself serving as cathode,
and a deposit of silver can be, as it were, painted on to the
metal wherever necessary. Seeing that most of the metals
ordinarily silvered are electropositive to silver, there is
always a possibility that they may by direct chemical action
reduce silver from the bath and cover themselves with an
imperfect and irregular film of the metal. To avoid this
the use of the " striking bath," may be adopted. This is
merely a separate bath, containing as a rule less silver and
more cyanide than in the plating bath, e.g. 3 grammes of
silver and 30 grammes of potassium cyanide per litre. As
high a current density as possible is used in working, so as
to deposit almost instantaneously a film of silver all over
the object to be plated. The article can then be removed
to the plating bath proper and the process of coating it
with a fairly substantial layer of silver proceeded with.
For this latter purpose a current density of about 4 amperes
per square foot is generally suitable. Silver is deposited
from the ordinary cyanide solution as a dense coherent
coating, dull and lustreless. It can be brightened by any
267
PRACTICAL ELECTRO-CHEMISTRY
mechanical process of burnishing, and this is generally the
method adopted. But for certain goods, parts of which
are not easily accessible, it is convenient to deposit silver
as a bright film. This can be accomplished by taking
advantage of the curious fact that a cyanide bath con-
taminated with a small quantity of certain foreign sub-
stances will yield bright silver. The substance generally
used is carbon disulphide, but other materials of the most
varied nature, ranging from silver sulphide to gutta-percha,
have been recommended from time to time. The carbon
disulphide solution is made by shaking up a few ounces of
carbon disulphide with a pint or two of plating solution and
allowing the mixture to stand. There will then be obtained
a saturated solution of carbon disulphide (that body being
slightly soluble in aqueous liquids, although not miscible
therewith), which is added to the plating bath in the pro-
portion of 1 ounce to 10 gallons. The quantity of carbon
disulphide thus introduced is not more than 407)00 of the
total electrolyte, but nevertheless it suffices to cause the
deposition of the silver bright instead of matt. The cause
of this phenomenon is unknown ; as far as I am aware no
attempt has been made to study it systematically, to deter-
mine for example whether the silver deposited has the ordin-
ary properties of pure silver and whether it possesses an
identical micro-structure. Certain precautions are neces-
sary : the current density should be greater than that used
for ordinary silvering, agitation of the liquid should be
avoided, and the goods should be washed as soon as they
are removed from the bath lest tarnishing occur from the
formation of silver sulphide.
The greatest use of electroplating is to coat spoons and
forks and other domestic implements, and thus to provide
them with a surface equal to that of solid silver goods ; in
addition, it is used for embellishing all kinds of ornaments.
The deposition of an alloy of silver and cadmium is spoken
of on p. 286.
268
GOLD PLATING
(Electro gilding)
THE covering of baser metals with gold for their protection
and ornament involves the same idea as that which led to
the use of silver plating. It can be effected by the old
process of " water gilding," which consists in covering the
object to be gilded with an amalgam of mercury and gold
and driving off the mercury by heat. In modern practice,
however, the gold is deposited electrolytically. The process
is generally similar to silver plating, but there are certain
differences in detail. The goods to be gold plated must, as
usual, be cleaned with scrupulous care before being placed
in the electrolyte.
They are sometimes " quicked " by dipping in a mercury
solution, as in silver plating. The bath may be made by
adding potassium cyanide in excess to a solution of gold
chloride, the proportions being about 10 parts by weight of
gold and 100 of cyanide to 1,000 of water. The bath may
also be formed by making a large gold plate the anode in a
cyanide solution and passing a current until as much gold is
deposited at the cathode as is lost at the anode in a given
time. There will then be in solution a sufficient quantity
of gold, and the bath can be used forthwith. These double
cyanide solutions of gold are generally used hot, at about
100 F. to 150 F. ; the current density is about 0'8 ampere
per square foot.
There are many other prescriptions for gold plating baths,
an account of which belongs rather to a collection of recipes
than to the present book. It is sufficient to say that, unless
pure materials are used and the anodes are pure gold, there
269
PRACTICAL ELECTRO-CHEMISTRY
is a probability of baser metals, e.g. copper and silver, being
precipitated along with the gold and forming an alloy with
it. The thickness of gold usually deposited is so small that
it serves as an ornament rather than as a protection to the
metal beneath. This, if silver, may tarnish from the for-
mation of sulphide almost as readily as if the gold were not
there. Rapid washing in weak cyanide solution will remove
this tarnish, while not attacking the gold appreciably.
Metals, such as zinc, which are apt to deposit gold from its
cyanide solutions without electrolytic aid are usually pro-
tected before gilding by a coating of copper.
It is possible with gold, as with silver and copper, to
deposit a second metal which shall modify the colour proper
to the gold itself. Such deposition belongs to the art of the
jeweller rather than to that of the electro-metallurgist, and
can be but briefly dealt with here. From a mixed solution
of gold and silver or gold and copper, gold may be thrown
down containing a small proportion of silver which will
lighten its. colour or of copper which will deepen it. The
proportions of the two metals can be controlled by adjusting
the relation of their salts in the electrolyte and the current
density at the cathode.
The process is precisely similar to the electro-deposi-
tion of brass from mixed solutions of copper and zinc,
or of silver alloys from silver and copper or silver and cad-
mium. The use of the last-named metal was proposed
a few years ago for silver plating. Plating with an
alloy of silver and cadmium instead of with pure silver
is said to have the advantage that the coating does not
easily become tarnished by sulphureous gases in the atmo-
sphere, and therefore keeps its colour better than does pure
silver. The method, however, has not been generally
adopted.
270
NICKEL PLATING
WHEREAS silver is the most generally useful plating metal
for domestic implements to be used in eating and drinking,
nickel forms the best coating material for larger, more sub-
stantial and more exposed objects, such as the fittings of
railway carriages, the bright parts of motor cars, bicycles,
firearms and water-taps. The process of nickel plating is
wholly modern, for it is only within the last thirty years
that nickel has been produced in quantity at a reasonable
price. Its present price is about Is. Sd. per pound.
Nickel, although less agreeable in colour than silver,
has the advantages of being considerably cheaper and of
tarnishing but little in ordinary air. It becomes somewhat
dull and acquires a sort of bloom which is easily removed
by gentle rubbing, but it does not become covered with a
film of sulphide, such as disfigures silver after a short ex-
posure, and moreover it is much harder than silver. It would
be an ideal metal for plating many kinds of goods were it not
for its tendency to flake and scale if deposited in any thick-
ness. A good deal of the complaint which is made against
nickel plating would be more reasonably made against
the plater, who does not take sufficient care to obtain
a perfect, continuous and adherent coating, but some of
the trouble arises from inherent qualities of the metal.
When the coating is imperfect the metal beneath the nickel,
if it is electro-positive to nickel, is attacked at the exposed
points with greater rapidity because of the adjacent nickel,
and the nickel which should protect it is peeled off by corro-
sion proceeding beneath the coating. Nickel plating is
harder and more brittle than the metal in massive form,
271
PRACTICAL ELECTRO-CHEMISTRY
somewhat as electro-deposited iron (q.v.) is harder than pure
iron in mass, but the reason for this has not been examined.
Electrolytic iron is generally considered to owe its hardness
to the fact that it contains hydrogen, which modifies its
properties. In the chapter on the electrolytic refining of
nickel will be found an account of certain experiments
on the conditions necessary for depositing nickel in a cohe-
rent state, which go to show that the metal is substantially
free from impurities ; but no special search seems to have
been made for hydrogen. It is possible that with nickel,
4 as with iron, the presence of hydrogen may increase the
, hardness of the metal.
The process of nickel plating involves the preparation
'of the article to be plated with even more care than is
requisite for silver plating. Not only must the surface be
clean, but it must be smooth and indeed bright, because a
film of metal electrolytically deposited reproduces accurately
the imperfections of the surface on which it is deposited,
' and in the case of nickel it is impracticable to smooth these
out by burnishing because of the hardness of the electro-
deposited nickel.
The preparation of a highly polished surface on the metal
to be covered necessitates burnishing, that is the rubbing
down of all projecting parts and the drawing of them over
the depressed portions so as to form a continuous reflecting
surface. All the small inequalities due to the actual micro-
scopic structure of the metal of the plate disappear, and the
hold available for the deposited metal is correspondingly
diminished. It follows that the not infrequent failure of
nickel plating to adhere may be due in some degree to the
excessive smoothness of the surfaces which it is intended to
cover.
But this must not be taken as the chief cause ; nickel,
even when deposited on a matte surface, will peel from it
spontaneously and without assignable cause as soon as it
becomes more than a mere film. In general the layer of
nickel required for plating is so thin that this tendency is not
of much practical significance.
272
NICKEL PLATING
If by any chance a stout layer is required it can be ob-
tained by keeping the electrolyte warm, e.g. between 50 C.
and 90 C. (seep. 115). That this method has not attracted
the attention of nickel platers is no slur on their sagacity,
which perceives small merit in a thick coating.
In the ordinary process of nickel plating the electrolyte
used is a double sulphate of nickel and ammonium. The
normal double sulphate corresponds with the formula
NiS04(NH 4 ) 2 S0 4 6H 2 0, and as a rule a further quantity of
ammonium sulphate is added. The customary proportions
are about 50 parts by weight of the double sulphate and
25 parts of ammonium sulphate in 1,000 of water. The
bath tends to become alkaline in working, because of the
ammonium sulphate as well as the nickel sulphate being de-
composed and yielding ammonia at the cathode, while its
equivalent of sulphuric acid is neutralized at the anode by
the nickel thence dissolved. The alkalinity is neutralised
from time to time with sulphuric acid so as to maintain the
bath as nearly neutral as possible ; it is commonly considered
that the solution should be slightly acid rather than alkaline.
This is probably because a slightly alkaline bath tends to
deposit basic salts, which may interfere with the coating.
When a nickel solution is made strongly alkaline with am-
monia so as to precipitate and to redissolve the nickel
hydroxide first thrown down, there is no difficulty of this
kind, and good nickel deposits are obtained.
The conditions are similar to those obtaining with copper.
There a perfectly neutral solution or one faintly alkaline
is apt to give bad deposits from the presence of basic salts ;
this trouble is overcome by making the solution acid, and in
the case of copper, unlike that of nickel, the amount of acid
may be considerable ; but good deposits may also be ob-
tained in an alkaline solution if the alkalinity be consider-
able, e.g. the copper salt be treated with sufficient excess of
ammonia to redissolve the cupric hydroxide precipitated
by the addition of a small quantity of the alkali. Ammoni-
acal copper and nickel baths are used in analytical separa-
tions but not in industry.
273 T
PRACTICAL ELECTRO-CHEMISTRY
As is usual in electro-plating, there are many recipes
for nickelling solutions, in some of which weak acids, e.g.
boric, citric and tartaric acids, or their salts, figure largely.
It does not appear that such additions give any better results
than the ordinary sulphate solution worked with intelligence
and care.
From a double sulphate solution nickel may be deposited
on most metals. On iron and steel the deposit is sometimes
not satisfactory in that it shows a tendency to strip. This
is probably due to want of care in preparing the goods, which
may not be perfectly clean when immersed in the electrolyte.
Occasionally steel goods are coppered in an alkaline bath
before being nickelled, with the view of obtaining a better
and more adherent coating.
The nickel anodes used in nickel plating should be as
pure as possible. It is only of late years that the commer-
cial metal has attained a reasonable standard of purity,
but it can now be procured fairly free from grosser con-
taminations. Electrolytic nickel or nickel prepared by
the Mond process (volatilisation as nickel carbonyl and de-
composition of this body by heat) is usually of fair purity,
but the supply of either variety is small ; metal made by
older processes often leaves much to be desired.
A current density of 10-15 amperes per square foot is
used for " striking," i.e. rapidly covering the whole surface
with a film of nickel, and when this is accomplished the
density may be lowered to 3 amperes per square foot. This
is the conventional procedure, but it is probable that much
improvement might be effected if the studies in the electro-
deposition of nickel detailed in the chapter on nickel winning
and refining were perpended by the nickel plater. It is
curious to note that, old as is the art of electro-plating,
there -has been scarcely any attempt to study systematically
the conditions necessary to effect a satisfactory deposition.
The whole art is empirical witness the number of quaint
recipes.
Small goods which would be troublesome to attach indi-
vidually to the cathode are often plated in a metal cage.
274
NICKEL PLATING
This in the ordinary course of work becomes plated itself,
and must be stripped or replaced from time to time. The
inconvenience is remedied by Delval and Pascalis, who
make the cage of wood with separate cathode plates on which
the goods to be plated rest. The cage is a cylinder set
horizontally, and can be rotated. It is not completely
immersed in the electrolyte. Its various cathode plates
are connected independently to a commutator. When it is
rotated only those cathode plates which are immersed and
on which the goods rest are supplied with current ; the
others are cut out. Hence no current is uselessly employed
in depositing nickel on the cathode plates themselves, which
are at a given moment bare of goods. Those cathode plates
which are actually bearing goods of course receive a small
deposit, but the bulk is thrown down on the goods to be
plated.
It is scarcely requisite to provide a separate section for
cobalt plating. The metal is scarcer and dearer than nickel,
and there is no great weight of evidence to show that it
forms a better protective coating. It is claimed that cobalt
is harder than nickel and does not tarnish so easily, but the
statement rests on slender ground. Should cobalt plating
be shown to be better or more permanent than nickel, it
can be obtained in much the same way, viz. by deposition
from the solution of a double sulphate of cobalt and ammon-
ium.
The greater rarity and cost of cobalt forbid its general
employment unless it can be shown to be sensibly better
than nickel as a coating.
275
ELECTRO-ZINCING
Zmc forms a cheap and excellent protective coating for
iron and steel. It has the great advantage over tin and
lead that it is electro-positive to iron, and is attacked in
preference to the iron when the two metals in contact
with each other are exposed to corrosion. In consequence
of this property, even when the zinc coating of an iron arti-
cle, e.g. a tank, is imperfect and a part of the metal is ex-
posed, the iron will be to a great extent protected from
corrosion while the zinc remains in sufficient quantity to
make an effective couple. Evidently this protective action
will not take place in the case of a plate on which is a bare
spot of considerable area, so that moisture may lie thereon
FIG. 53o.
without reaching the surrounding zinc. The difference in
the conditions, which is of some practical importance, is
shown in the accompanying diagrams. A tray (Fig. 53a)
of galvanised iron has a part of the coating stripped at c,
and in the middle of this bare space is a patch of moisture D.
Clearly corrosion will occur here, unaffected by the neigh-
bourhood of the zinc. A similar tray (Fig. 536) with a similar
bare patch E is filled with water so as to cover the bare patch
entirely with the water. Both iron and zinc are in electro-
lytic connection with the water, and the zinc, being the
positive metal, is corroded in preference to the iron. Thus it
comes about that a zinc coating is generally more protection
276
ELECTRO-ZINCING
in the case of a tank than in that of a roof. In like manner
one may protect an iron boiler or ship by attaching pieces
of zinc to the plates where they are immersed in water, but
one would hardly meet with success in attempting to protect
a bridge by like means. All this is obvious enough, but is
nevertheless constantly overlooked, with the result that
zincing is sometimes condemned because it does not perform
electro-chemical impossibilities .
For most goods zincing or galvanising, as it is errone-
ously termed is most cheaply and conveniently applied
by dipping the iron or steel articles (after they have been
carefully cleaned and pickled in acid) in a bath of melted
zinc. The zinc alloys superficially with the iron and forms
a complete and adherent coat. For certain classes of goods
FIG. 536.
this method presents disadvantages. The bath must be
at a temperature somewhat above the melting-point of
zinc, 412 C. = 774 F. At this temperature the harder
grades of steel, such as are used for the stronger kinds of
wire, are annealed considerably and thus lose a part of their
high tensile strength from this cause. Again, the alloy of
zinc and iron formed on the surface of the article coated is
of small mechanical strength compared with the iron from
which it has been formed. With articles of heavy section
this is not important, but with goods of relatively small
section, which have to carry heavy strains, e.g. wires, cables,
chains, bolts, hooks and the like, the diminution in strength
is often of serious moment. Thus it comes about that for
certain classes of work there is a demand for a coating of
zinc which shall be applied cold and shall not alloy appreci-
ably with the surface of the iron to be protected. These
conditions are fulfilled perfectly by zinc electro-deposited.
277
PRACTICAL ELECTRO-CHEMISTRY
There is another advantage small and incidental, but real
enough in electro-zincing. As it is taken from an aqueous
bath all salts are easily washed from its surface ; goods
taken from a bath of fused zinc may retain spots and
crusts of the flux (sal ammonia) with which the surface of the
molten metal is covered. Good washing and scrubbing will
remove these, but mere rinsing will hardly suffice ; hence any
carelessness in finally cleaning the zinced goods may leave
sufficient sal ammonia adhering to cause serious corrosion,
and, in fact, to destroy the coating locally.
There are various difficulties in depositing zinc electro-
lytically so as to obtain a good adherent coating. These
have hindered the general employment of electro-zincing,
but they have now been overcome in great measure,
thanks to the perseverance of one or two inventors, and the
process is already fairly freely used, and its use is likely to
extend.
The conditions necessary to be observed in order to obtain
a good deposit of zinc electrolytically have already been
described in the chapter on the winning and refining of zinc.
The application of the principles there laid down will suffice
to allow of the deposition of a satisfactory coating of zinc
to metal to be protected.
The metal most commonly zinced or " cold galvanised "
is iron (or steel) ; it must be cleaned before being coated
by the usual pickling methods. The objects to be coated
are made the cathode in a solution of zinc sulphate contain-
ing about 10 per cent, of this crystallised salt (ZnS0 4 7H 2 0).
The electrolyte should be free from foreign metals. As it
should be kept neutral or slightly acid, basic solutions tend-
ing to deposit spongy zinc (see p. 136), some difficulty will
be experienced if it be attempted to maintain the strength
of the bath by using zinc anodes. It is preferable to use
an insoluble anode, and to add zinc oxide or metallic zinc in
regulated quantity so as to neutralise the sulphuric acid set
free at the anode. By this means the electrolyte can be
maintained in a neutral or faintly acid condition, and, more-
over, can be purified at the same time . The latter advantage
278
ELECTRO-ZINCING
is secured by reason of the fact that zinc, being a strongly
electro-positive metal, is capable, whether as oxide or as
metal, of precipitating less electropositive impurities, such
as iron or its oxide. The purity of the electrolyte
(which is of much importance) can, therefore, be maintained
by the means used to regulate its acidity. In order to
obtain a good coating of zinc a fairly high current density
should be employed, e.g. 10 to 20 amperes per square
foot. Other precautions, such as circulation of the elec-
trolyte, and maintenance of a uniform current density by
specially shaped and placed anodes when objects of irregu-
lar surface are to be coated, are similar to those which must
be observed in plating generally. Perfection of coating,
provided the coating as a whole adheres well, is of smaller
importance than in the case of less electro-positive metals.
A small exposure of the underlying metal may occur without
causing corrosion as long as there is abundance of surround-
ing zinc, at the expense of which the underlying metal may
be protected.
A highly polished surface is rarely necessary for electro-
zinced goods. Such articles are commonly for outdoor use,
and a high finish is not required. It would be absurd to
confer on a roof, a boat-hook, a crane chain, or a wire rope
the lustre proper to an ornament. But even here aesthetic
considerations have a certain force. Hot zinced goods have
a bright metallic appearance, and their coating is sometimes
made to exhibit brilliant crystalline markings by adding a
little tin to the zinc bath ; electro-zinced goods have usually
a somewhat dull and leaden appearance. Irrational though
it be, a prejudice exists in favour of the former. In spite
of this, the substantial advantages of a method for deposit-
ing zinc in the cold, especially for hard steel (which if heated
would be softened) and objects of small section (which are
weakened by hot galvanising), will cause the process of elec-
tro-zincing to come widely into use for a variety of purposes.
Cowper Coles, who has worked out a process for electro-
zincing which has been put successfully into use, has given
an estimate of the cost of the operation. He reckons that to
279
PRACTICAL ELECTRO-CHEMISTRY
cover steel plates of an average thickness of y\ ; inch with
zinc at the rate of 1 ounce per square foot (a sufficient coat-
ing) will cost 2 85. 6d. per ton of plate coated. This is
probably somewhat greater than the cost of hot galvanising,
but the extra cost is more than compensated for by the
advantages which have been set forth above. Further, a
tank of fused zinc for big objects such as large plates or for
things which are galvanised after they have been riveted up,
e.g. tanks, is troublesome to heat evenly, and contains a
good many tons of zinc, which represent so much capital
locked up. The quantity of zinc in an electrolytic bath
capable of coating objects of the same size is relatively insig-
nificant. The iron tank is also somewhat perishable, in
that it is attacked by the melted zinc and eventually eaten
through. The alloy of zinc and iron resulting from this
attack not only represents destruction of the tank, but
useless consumption of zinc, which would otherwise go
to coat the goods to be galvanised.
A recent application of electro-zincing which has proved
successful is the coating of tubes for water tube boilers.
Such tubes are particularly liable to corrosion, and, being
smalland narrow, are not sufficiently protected by zinc blocks
attached to the body of the boiler ; a lining of zinc throughout
their length must certainly prolong their life.
Except for iron, zinc is not much used as a coating. One
other and smaller application may be mentioned. Rollers
for printing designs are made of copper and coated with zinc,
on which the design is engraved. When the design is
obsolete or worn out the zinc can be stripped and a fresh
surface deposited. In the stripping it is difficult to avoid
attacking the copper to some extent. On this account it
has been proposed to use aluminium rollers and to deposit
zinc on these ; stripping can then be done by nitric acid,
which dissolves zinc freely and has only a trifling action on
aluminium. It may be noted, however, that aluminium is
not an easy metal to plate, because of the ease and rapidity
with which it acquires a film of oxide almost imperceptible,
but sufficient to prevent adhesion.
280
ELECTRO-DEPOSITION OF IRON
(Aciertype)
IRON is not used as an ornamental plating material, and
(naturally) not as a coating to protect the metal beneath
from corrosion. But the hardness and toughness of the
metal make it suitable as a protective coating against
abrasion or attrition. Thus it comes about that the use
of iron as a plating substance is confined to facing electro-
types in copper or similar soft metal which have to be
exposed to considerable mechanical wear. The only case
in which iron is used for its chemical, as distinct from it
mechanical, properties is that in which it is employed to
face electrotypes which come into contact with vermilion
or other pigments containing mercury. Copper electro-
types would reduce mercury from such pigments and be
destroyed by the layer of amalgam which would be produced
thereby ; with iron no such action occurs.
Apart from this minor use, the main merit of a coating of
electro-deposited iron arises from its hardness, which is
much greater than that of pure iron prepared by other
means. Hence the term " aciertype," implying that the
plating is not iron, but steel. The cause of the hardness
of electro-deposited iron is generally asserted to be the
presence of hydrogen, which is co-deposited with the metal
and influences its condition much as does a small percentage
of carbon. The quantity of hydrogen present in electro-
deposited iron may be considerable, e.g. 240 times the volume
of the metal, corresponding with 0-27 per cent, by weight.
This hydrogen is driven off when the metal is heated to red-
ness, and the characteristic hardness of electro-deposited
281
PRACTICAL ELECTRO-CHEMISTRY
iron T disappears at the same time. Nickel, like iron, is
deposited from an electrolytic bath in an extremely hard
state ; it is not known whether this is due to the presence
of hydrogen. The hardness of electro-deposited nickel is
sufficient to enable it to be used in the same way as iron
for facing electrotypes. Its greater resistance to corrosion
makes it preferable to iron ; therefore, the replacement of
aciertype by plating with nickel appears probable. At
the present time, however, there is sufficient use of electrolytic
iron to warrant a description of the means by which it may
be deposited.
It is a mistake to suppose that electrolytic iron is neces-
sarily pure. Not only is hydrogen deposited along with
the metal, but several other impurities may appear. In the
first place, it is clear on general principles that as iron is
a highly electro-positive metal its deposition will require
the use of a current of relatively high voltage ; this will
tend to deposit all metals present in the electrolyte as im-
purities which are electro-negative to iron. Further, iron
deposited from solutions containing organic salts, e.g.
oxalates, tartrates and citrates, usually contains carbon ;
as much as 0-08 per cent, may be present a quantity cap-
able of modifying the properties of the metal materially.
From solutions containing sulphates iron is thrown down
contaminated with a small amount of sulphur. In fact,
the preparation of pure Fe electrolytically is as difficult as
it is by purely chemical means, and this, as every chemist
knows, is one of the most exacting tasks which he can
set himself.
But to obtain a coating of iron which is satisfactory phy-
sically and mechanically, although, or rather because, it is im-
pure, is perfectly practicable. The usual electrolyte is a solu-
tion of ferrous ammonium sulphate (FeS0 4 (NH 4 ) 2 S0 4 6H 2 0)
in the proportion of 150 grammes per litre. A double chlo-
ride of (ferrous) iron and ammonium is also suitable. The
bath should be nearly neutral, and the whole of the iron in
the ferrous state. Pure wrought iron anodes should be
used, so that the supply of ferrous ions may be maintained ;
ELECTRO-DEPOSITION OF IRON
otherwise oxidation will occur at the anode and the elec-
trolyte will become partly ferric. This ferric salt will have
to be reduced at the cathode before it will again yield its
iron.
Recently Burgess and Hambuechen have prepared elec-
trolytic iron in quantity, and have proposed to use their
method for the production of pure iron by refining com-
mercial iron electrolytically, precisely as copper and other
metals are refined. The electrolyte used is a solution of
ferrous ammonium sulphate ; the current density is 6-10
amperes per square foot ; anodes of wrought iron or mild
steel are employed. The deposit has a tendency to curl off the
cathode in the way characteristic of nickel, but by adopting
certain precautions (the nature of which is not stated) a
thickness of J in. has been obtained in 4 weeks. The iron is
stated to be almost pure 99-9 per cent, or better but
nothing is said as to the presence in it of sulphur. I see
no reason why iron should not be refined electrolytically if a
sufficient use can be found for it.
Two elegant minor applications of electrolytes to the
treatment of metals may be mentioned. Large reflectors
suitable for search lights are now made by a process due
to Cowper Coles.
A glass disc is prepared and optically worked to a parabolic
surface. This if silvered would itself serve perfectly as a
mirror. But it is costly and fragile. Accordingly replicas of
it in metal are formed in the following way. The surface of
the glass is made conductive by depositing on it a thin film
of silver by any ordinary chemical silvering process. Copper
is then deposited electrolytically on this surface until a sub-
stantial coating is obtained stiff enough to be handled without
deformation. It is stripped from the silvered glass matrix,
and accurately reproduces its optical surface. To make
it reflect well it is plated not with silver but with palladium,
which is less apt to tarnish. By this neat device accurate
metallic mirrors can be prepared relatively cheaply. If hit
in action by a rifle bullet there is no general smash as
there would be with glass, but merely a hole which
283
PRACTICAL ELECTRO-CHEMISTRY
scarcely impairs the efficiency of the mirror for its pur-
pose.
Another neat little process is what is known as electro-
gravure. A cast of any object to be copied, e.g. a medal,
is prepared in some porous material. Plaster has been
tried, but now other substances are used which are as porous
as plaster, but less soluble. The cast is placed so that it
is not immersed in the electrolyte but is saturated with it,
and its surface is constantly kept wet. On this surface
a metal disc is put ; it is made the anode. The cathode
may be in any convenient position, provided the only path
between it and the anode lies through the shaped surface
of the cast. Now as the metal disc lies on the porous cast,
naturally it touches only those parts which are highest, and
is eaten away there ; it gradually settles down on the
cast, touching at more and more points, and being
correspondingly corroded until it touches all over ; it is
then an accurate reproduction of the original from which
the cast has been prepared.
284
THE ELECTRO-DEPOSITION OF
ALLOYS
WHEN a single metal is to be deposited in a state as nearly
pure as possible from a solution containing a second metal,
the heat of formation of whose salts is greater than that of
its own, the object can be attained by working with a vol-
tage below the critical voltage of the second metal. Con-
versely, when an alloy of the two metals is desired the vol-
tage used must be above this critical point. The two
metals will be simultaneously deposited, their proportions
varying with the proportions of their salts in the electrolyte.
The formation of alloys in this manner is more curious than
important, having a somewhat limited field of application.
The alloy most commonly deposited is brass. It can be
obtained by electrolysing a solution of zinc cyanide and
copper cyanide dissolved in potassium cyanide, the propor-
tions being about 15 grammes of copper cyanide and 8
grammes of zinc cyanide to 100 grammes of potassium
cyanide in a litre of water. The number of prescriptions
which have been published is very large, and many of the
recipes are frankly obscurantist. Acetates, chlorides and
sulphates of the metals may be employed ; ammonia and
its salts are freely used, and such unlooked-for ingredients
as bisulphites and arsenious acid are not unknown. The
double cyanide solution is used hot with brass anodes. The
reason why cyanide solutions are commonly employed is
probably because electro-brassing is generally applied to
zinc or iron, and these metals would spontaneously deposit
copper from most of its other salts. Should it be desired
285
PKACTICAL ELECTRO-CHEMISTRY
to deposit brass on any less electro-positive metal than cop-
per there is no reason why it should not be effected from a
mixed solution of the sulphates of copper and zinc, approxi-
mately neutral and mixed in such proportions as would
ensure a sufficiency of zinc ions being always present at
the cathode.
In similar manner alloys of copper and tin (bronzes)
may be deposited from mixed solutions of salts of the two
metals. Silver may be deposited alloyed with tin or cad-
mium, the advantage claimed being that plating of this
description is not only cheaper than silver, but also better
resists the discolouring action of air containing sulphureous
Most other metals (save those, like aluminium, which
are too highly electro-positive) may be deposited in thin
films by electrolytic means. Their applications are, how-
ever, too limited to warrant separate mention. For details,
special works on the electroplater's art must be consulted.
286
SECTION VII
Alkali, Chlorine and their Products
Alkali, Chlorine and their Products
MANY attempts, extending over a number of years,
have been made to manufacture alkali and chlorine
by the electrolysis of salt. The fundamental reaction NaCl
+ H 2 = NaOH + H + Clis simple, and is easily realised
experimentally. Its accomplishment on a large scale at
a remunerative rate is, however, more difficult. By much
costly experiment and experience, bought by many disas-
trous failures, it has been found that the following conditions
are essential for success : (1) the cost of power must be very
low, certainly not more than 10 per E.H.P. year ; (2) the
process should be continuous ; (3) the electrodes should be as
nearly permanent as possible ; (4) the products of electroly-
sis should be removed from the electrolyte continuously
as the process proceeds ; (5) the units of plant should be as
large as is practicable ; (6) the output per unit of plant
should be great, as otherwise the process is burdened by
an unduly heavy charge for interest on the necessary capital.
It is only lately that a few processes have succeeded in fulfil-
ling most of these conditions.
GENERAL CHEMICAL CONSIDERATIONS
It is convenient to regard the electrolytic decomposition
of sodium chloride as being primarily represented by the
equation NaCl = Na + Cl. This can actually be realised
when fused salt is electrolysed. The number of calories
required for the decomposition of 1 gramme equivalent (58-5
grammes) of salt is 97-7 Cal., and the critical voltage cor-
289 u
PRACTICAL ELECTRO-CHEMISTRY
responding with this heat of combination is 4-22 volts. 1
Various processes to obtain caustic soda (by the action of
the liberated sodium on water) and chlorine in this man-
ner have been devised. They will be described in due
course. The great obstacle to their use is the corrosive
action of fused salt on most materials that can be used for
making the vessels in which the electrolysis can be conducted.
Apart from this the process is attractive, because both
chlorine and sodium can be removed continuously from the
electrolyte, the resistance of the electrolyte is low, no dia-
phragm is required, and a large output can be obtained
from a small apparatus. Nevertheless, at the present time,
those processes which have attained a fair measure of suc-
cess are methods for the electrolysis of aqueous solutions of
common salt. When the electrolysis is conducted in the pre-
sence of excess of water, it may, for the sake of simplicity,
be supposed that the reaction takes place in two stages,,
thus :
(1) NaCl = Na + Q;
(2) H 2 + Na = NaOH + H.
In computing the energy required it is unnecessary to
consider the stages of the reaction ; the original materials,
salt and water, and the end products, viz. chlorine, hydrogen,
and a solution of caustic soda, may alone be regarded. The
number of calories required for the decomposition of 1
gramme equivalent (i.e. 58-5 grammes) of sodium chloride
according to the pair of equations given above is 53 Cal.,
and the critical voltage is 2-29 volts. Caustic soda and hy-
drogen, instead of metallic sodium produced by the electro-
lysis of fused salt, being the end products, the energy and
critical voltage required are naturally lower than those
requisite for fused salt. But against this must be set the.
higher resistance of the electrolyte, the need (usually)
1 This is the critical voltage corresponding with the heat of for-
mation of salt at the ordinary temperature ; but as salt is solid at
the ordinary temperature and is not an electrolyte, this critical
voltage is of only theoretical interest. The critical voltage of salt
at its fusing-point (772 C. 1,422 F.) is approximately 3-81 volts.
290
ALKALI, CHLORINE AND THEIR PRODUCTS
of a diaphragm, and the difficulty (overcome in the best
processes) of continuously separating the products of electro-
lysis from the electrolyte. As a standard by which the
various processes about to be described may be judged, the
calculated output for a process of theoretical efficiency may
usefully be computed. The decomposition of 58-5 grammes
of NaCl into caustic soda, hydrogen and chlorine requires 53
Cal. Therefore the quantity of salt decomposed by 1 E.H.P.
year ( = 5,646,205 Cal.) is 6-13 tons. Taking the cost
of an E.H.P. year at 9 165. for steam power and at 2 10s.
for water power, the cost of electrical power for decompos-
ing 1 ton of salt is 1 12s. with steam power and 8$.
with water power. These figures correspond with 2 6s. Sd.
-and 11s. Sd. for a yield of 1 ton of pure caustic soda, i.e. a
little better than the trade grade known as 77 per cent,
(which is calculated on the percentage of Na 2 and on an
erroneous atomic weight for sodium), together with 2J
tons of chloride of lime containing 35 per cent, of chlorine
available for bleaching purposes. This last figure is slightly
inexact, because commercial chloride of lime contains a
certain small percentage of chlorine which is not available
for bleaching purposes, and this represents so much of the
total chlorine won by electrolysis wasted. Nevertheless,
the approximation is sufficient for practical purposes, and
enables one to see that, having regard to the present selling
price of caustic soda and bleaching powder per ton, the cost
of the power required for electrolysis is not excessive. Even
when allowance is made for the facts that the current effi-
ciency of the best processes does not exceed 90 per cent, and
the pressure efficiency does not exceed 50 per cent., making
an energy efficiency of 45 per cent., it remains clear that the
cost of electrical energy is moderate enough.
That large profits have not been realised hitherto in the
electrolytic manufacture of alkali and bleach arises from the
heavy cost of the plant (including, in many cases, interest
on large sums sunk in experiments or expended in the
purchase of patent rights) and costly up-keep, management
and supervision charges.
291
PRACTICAL ELECTRO-CHEMISTRY
PROCESSES USING A FUSED ELECTROLYTE
A large number of these have been devised, patented,
tried and abandoned. One or two are at present being
exploited on a considerable scale. The chief obstacles
which inventors have encountered may be understood by
a consideration of the defects of the simplest possible ap-
paratus for the electrolysis of fused salt.
A fireclay crucible A (Fig. 54) is set in a furnace and filled
with salt, which is thus kept fused. A rod of iron serves
as a cathode c, and one of carbon functions as the anode D.
When a current is passed between these electrodes, sodium
FIG. 54.
is liberated at the cathode and chlorine at the anode. But
the sodium, which is liquid at a temperature far below the
fusing point of salt, is also lighter than liquid salt, and rises
to the surface and there takes fire and burns. The first
difficulty is here encountered, and it is clear that in a work-
able process means must be taken to protect the sodium
from the action of the air, and to draw it off without giving
it a chance to inflame. Next it is found that the carbon
anode D suffers severely from the action of the fused salt
and possibly from that of the chlorine. An anode thus
used gradually disintegrates, and its fragments float in the
292
ALKALI, CHLORINE AND THEIR PRODUCTS
electrolyte, contaminating it and causing many incon-
veniences. Lastly the fused salt creeps over the edge of
the crucible, runs down outside, and soaks into the ware.
The bulk of the salt acts similarly on the inside of the cru-
cible. Both from within and without the crucible is satu-
rated with fused salt, which at the temperature prevailing
may act chemically on the ware, and in any case causes
mechanical disintegration. The destructive effects pro-
duced by fused salt on the most refractory materials are very
remarkable ; they are due to a variety of causes, chemical
and mechanical, and for our present purpose it is sufficient
to accept their existence as a fact.
It is not altogether convenient to obtain metallic sodium
as the cathode product. The substance to be prepared is
caustic soda, and when sodium is obtained instead it has to
be oxidised and hydrated to caustic soda, thus involving a
violent reaction with water. Not only is this reaction super-
fluous and objectionable, but it also connotes a considerable
waste of energy, because more than the amount of energy
necessary to prepare caustic soda from salt has been expended
in the production of sodium, and then this surplus has to be
run to waste as heat in the aforesaid violent reaction with
water. These drawbacks, as well as that caused by the
sodium being considerably lighter than the fused salt, are
avoided to some extent in the following way :
THE VAUTIN PROCESS
Instead of a cathode of solid metal one of fused lead is
used, as shown in Fig. 55, which represents a form of ap-
paratus devised by Vautin. A, lead cathode ; B, decomposing
vessel in which the lead-sodium alloy is acted on by steam ;
c, carbon anode ; E, pipe for escape of chlorine ; D, pipe for
admission of steam ; F, pipe for escape of hydrogen ; G,
refractory lining.
The sodium, as it is liberated, dissolves in the lead and
is transferred to the vessel at the side of the electrolytic
cell, where the lead sodium alloy comes into contact with
293
PRACTICAL ELECTRO-CHEMISTRY
water or steam and reacts, the sodium yielding caustic
soda and the lead being fit for use again in the cell. As
lead and sodium unite with considerable energy to form an
alloy, the total expenditure of energy necessary to produce
a lead sodium alloy by the electrolysis of sodium chloride
with a cathode of fused lead is smaller than would be re-
quisite were sodium itself prepared. In like manner the
energy liberated by the action of water on the lead sodium
alloy is also smaller, and the reaction is thus more moderate.
FIG. 55.
Unfortunately a comparatively small proportion of sodium
makes an alloy with lead which is not very mobile, and the
sodium thus fails to diffuse freely to the steam space. In
consequence of this the surface of the lead becomes crusted
with sodium, which eventually floats up through the fused
salt and is reoxidised at the anode. If oxygen as well
as chlorine be present in the space above the level of the
electrolyte, this sodium will form oxide and enhance the
attack of the materials of which the cell is constructed.
The cell was designed for external heating, and the usual
troubles which have been discussed in the section on
aluminium and sodium naturally occurred. Even the
most refractory lining materials, such as magnesia, suffered
attack by the fused salt and its products. These difficulties
led eventually to the abandonment of the Vautin process.
294
ALKALI, CHLORINE AND THEIR PRODUCTS
THE HULIN PROCESS
This process was adopted by the Societe des Soudieres
Electrolytiques, which erected works at Clavaux Isere, where
energy is obtained from the water of the river Romanche.
A steel pipe, 900 metres long and 2-5 metres in diameter,
brings the water to the turbine house, where a head of 42
metres is available. The power obtainable is 5,000 H.P.
The turbines are coupled direct to the dynamos, which yield
375 kilowatts apiece. The works have been designed for an
output of 4 tons of caustic soda and its equivalent (about
10 tons) of bleaching power. According to recent infor-
mation the process has not proved to be successful in practice ;
its ingenuity, however, justifies a description.
The principle of the Hulin process is identical with that
of the Vautin process described above, save that the electro-
lyte consists of a mixture of lead chloride and sodium
chloride instead of sodium chloride alone. By this alteration
the cathode product is a mixture of lead and sodium, and
the continual supply of a proportion of lead together with the
sodium prevents the crusting over of the surface of the lead
with sodium, which, as mentioned above, is apt to occur
when a cathode of fused lead alone is used. In order to main-
tain a proper proportion of lead chloride in the electrolyte,
part of the current is sent through lead anodes instead of
carbon anodes, and these, being attacked by the chlorine
liberated at their surface, dissolve in regulated degree. The
plant may be represented diagrammatically by Fig. 56.
A vessel A contains the fused salt mixed with lead chloride,
and at the bottom a layer of lead sodium alloy B. The
carbon anode c dips into a suspended vessel D, containing
melted lead. This is thus made an anode and is attacked,
producing lead chloride. In practice it is probable that
separate lead anodes would be used, so that the current pas-
sing through them may be more easily regulated and the
proportion of lead chloride in the electrolyte readily con-
trolled. It is evident that, as lead is dissolved and repre-
295
PRACTICAL ELECTRO-CHEMISTRY
cipitated, no consumption of energy is theoretically neces-
sary for its transference from the lead anode to the lead
sodium cathode. But, as in practice a considerable current
has to be caused to pass between these electrodes through
an electrolyte of considerable resistance, it is evident that
there will be a noteworthy expenditure of energy. (The
principles governing such an operation are fully expounded
in the section on Copper Refining, p. 31). All this must be
reckoned as a disadvantage of the process, but in practice
may be more than compensated for by the convenience of
FIG. 56.
obtaining continuously an alloy of regular and suitable com-
position.
Certain figures have been published giving the results
of a trial of the Hulin process on a small manufactur-
ing scale. They may usefully be transcribed here. The
power available was about 120 H.P. ; a current of 2,000
amperes at 32 volts was obtained therefrom and sent through
four electrolytic cells of the type described above, arranged
in series. Each cell when working normally required a
voltage of 7 volts and had a current density of 700 amperes
per square foot at the cathode. This current density is
296
ALKALI, CHLORINE AND THEIR PRODUCTS
enormously greater than the highest current density hitherto
found practicable with electrolytic cells using solutions of
salt ; in these 10 to 20 amperes per square foot is a common
current density. The large output thus made possible for
a given cell will go far to compensate for the low energy
efficiency of the process, of which more anon.
The lead-sodium alloy is drawn off periodically and the
sodium is converted into caustic soda in one of two ways.
If water be allowed to act on the alloy in its cold solid state,
the reaction proceeds quietly and is not dangerous. A solu-
tion of pure caustic soda is obtained, which may be made
fairly strong by using the same liquid to act repeatedly on
fresh portions of the alloy. The liquor thus obtained, having
a specific gravity of 1-54 and containing 750 to 800 grammes
of NaOH per litre, may be boiled down to solid caustic soda
with a moderate expenditure for fuel. If steam could be
used to act directly on the fused alloy, a stronger solution
of caustic soda could be obtained, and moreover the lead,
freed from sodium and still liquid, could be returned at
once to the electrolytic cell. It is stated, however, that the
action of steam on fused lead-sodium alloy is dangerously
violent, and the method is, therefore, not employed. When
the solidified alloy is acted on with water, spongy lead is
left, which may be used for the plates of storage cells.
The alternative method is to roast the lead-sodium alloy
in air. Sodium oxide or peroxide, and lead oxide are
obtained, the latter apt to oxidise to peroxide and com-
bine with the soda, forming sodium plumbate. This salt
would be decomposed on treatment with water, yielding
s, solution containing caustic soda (and probably some
sodium peroxide) and leaving a residue of lead peroxide,
useful, like the spongy lead, for the plates of storage cells.
The solution containing caustic soda and sodium peroxide
would be boiled down for solid caustic soda, and in the pro-
cess the sodium peroxide would be decomposed, producing
an equivalent of caustic soda. Thus, save for the possible
presence of traces of lead, the solution of caustic soda ulti-
mately obtained should be pure.
297
PRACTICAL ELECTRO-CHEMISTRY
The following tabular statement indicates the degree*
of efficiency of the Hulin process, both as regards current
and energy :
Hulin process.
Theory.
Voltage required ....
Cl per ampere hour .
NaOH per ampere hour .
Cl per H.P. hour ....
NaOH per H.P. hour .
Current efficiency
Pressure efficiency
Energy efficiency
7 volts
0-907 gramme
1 -052 grammes
97 grammes
112 grammes
68 -6 per cent.
60 per cent.
41-1 per cent.
4-2 volts *
1-322 grammes
1-490 grammes
235 grammes
265 grammes
100 per cent.
100 per cent.
100 per cent.
. THE ACKER PROCESS
The Acker process as originally designed resembled the
FIG. 57.
Vautin ; salt was fused by heat externally applied and
1 This value is calculated from the heat of combination of Na
and Cl at the ordinary temperature. At the temperature of fusing
salt the critical pressure is probably lower, viz. 3' 81 volts (see p. 290).
In this case the efficiency of the Hulin process is even smaller than
appears from the figures given above.
298
ALKALI, CHLORINE AND THEIR PRODUCTS
electrolysed between a carbon anode and a lead cathode.
The lead sodium alloy was caused to flow by means of a-
steam jet from the cell to an outer compartment, where
its conversion into caustic soda was accomplished. The
usual disadvantages of external heating having made them-
selves felt, the process was improved by depending on
the current itself for the fusion of the salt as well as for its
electrolysis. The general arrangement is shown in the
figure. The vessel A of cast iron contains the lead B,
which serves as cathode, and the fused salt c, which is the
electrolyte. The carbon anodes are marked D. The cir-
culation of the lead sodium alloy is effected by the steam
jet E. With a cell of this kind it may be necessary to
start the operation by the aid of external heat, but when
fusion has occurred, it can be maintained by heat from
the current. In practice this would be economised by
surrounding the cell with a stout brickwork casing.
It is stated that the plant at Niagara Falls consists of 45
cells, each taking about 8,000 amperes at 7 volts, the total
power utilised being 3,250 h.p., corresponding with a yearly
output of 3,894 tons of caustic soda and 8,580 tons of bleach-
ing powder.
THE BORCHERS PROCESS
Borchers has designed an apparatus for the production
of alloys of sodium and lead or other fusible metals, and
incidentally for the preparation of chlorine . This apparatus ,
like many of those devised by that experimenter, presents
several apparent merits and is worth description. Like
most of the same inventor's designs, it appears not to have
been put to practical use.
A is a conical vessel of iron which serves to contain the
electrolyte (fused salt). It is set in a furnace so that its
contents may be kept fused. The lower part of the vessel
is grooved on the inside, the grooves serving to contain
molten lead, a supply of which is delivered from the vessel
E at the side of the electrolytic cell. This lead is made
the cathode by connection with a dynamo through the
299
PRACTICAL ELECTRO-CHEMISTRY
terminal at F. The anode c is a carbon rod, while D is a pipe
to carry off the chlorine. The lower part of the electro-
lytic cell is protected from the action of the electrolyte
by the lead contained in the stepped grooves shown in the
figure ; the upper part is protected by congealed salt, which
is caused to solidify and form a crust on the inside of the
vessel by the cooling action of a water-ring R.
This plan of protecting a vessel serving as an electro-
lytic cell by a crust of the solidified electrolyte is undoubt-
edly based on a sound principle. In the manufacture of
aluminium (q.v.) it is easily adopted, because the heat
necessary to maintain the electrolyte in a fused condition
is obtained from the current itself, and is therefore
FIG. 58.
internal ; thus it is simple to keep the walls of the con-
taining vessel at a temperature below the fusing-point of the
electrolyte. The local solidification of the electrolyte by
water-jackets and similar devices is less easy of accom-
plishment, but is practicable in certain cases, of which the
present appears to be one. The lead charged with sodium
flows away into the collecting pot B, whence it can be re-
moved for the extraction of its sodium ; the recovered lead
is returned to the vessel E and passes again through the
apparatus. Borchers states, " A plant of this kind, twenty
times the actual size of the foregoing illustration, is adapted
to a current of 300 amperes, which corresponds to a current
aty of about 5,000 amperes per square metre [3-2 amperes
300
ALKALI, CHLORINE AND THEIR PRODUCTS
per square inch of cathode surface]. The electro-motive-
force required may be only 6 or 8 volts, which is consider-
ably less than that needed for the reduction of sodium in
the unalloyed condition." The high current density would
tend to keep the electrolyte fused independently of exter-
nal heating ; any such internal heating by means of the
current secures convenience of working and prolongation
of the life of the plant at the cost of an extra consumption
of energy in a somewhat expensive form. The statement
of the pressure required is misleading in that it implies that
the critical pressure necessary to produce a lead-sodium
alloy is 6 to 8 volts. As shown above, it is not higher than
3-8 volts. The extra pressure is needed for forcing through
a current of a density as high as that employed in this case.
As has been already stated, ceteris paribus, the voltage which
will suffice for the production of a lead-sodium alloy is
lower than that which is necessary for the production of
metallic sodium, and therefore, given a certain current
density, the " 6 to 8 volts " is lower than the pressure which
would be needed for making sodium unalloyed ; neverthe-
less the method of stating this fact adopted by Borchers
is elliptical and consequently obscure and likely to cause
error.
PROCESSES USING DISSOLVED SALT AS AN
ELECTROLYTE
A great number of these might be described if this were
a history of electro-chemical invention. All but a very
few have, however, proved failures and may be dismissed
at once. Of the remainder which will be dealt with it may
be said that their use has been seriously attempted on a
large scale. It must not be thought from this that they are
all commercial successes. It must also be remembered
that there are probably in existence other processes which
are working remuneratively and are kept as secrets. This
is the natural and inevitable condition of things in a novel
301
PRACTICAL ELECTRO-CHEMISTRY
and difficult industry, and the consequent lack of com-
pleteness of information in a book treating of the industry
cannot well be avoided.
THE ELECTRO-CHEMICAL COMPANY'S PROCESS
This process, known in its original form as the Holland
,& Richardson process, has been employed on a large scale
FIG. 59a.
by the Electro-chemical Company of St. Helen's, Lanca-
shire. On account of difficulties in working, the process,
which is simple and in many ways well conceived, has been
given up. As an illustration of one type of electrolytic
methods for the manufacture of alkali and bleach, it may be
FIG. 596.
usefully described. The generating plant consisted of three
vertical compound engines of the marine type, each driving
two dynamos giving jointly 2,500 amperes at a pressure of
180 volts. The electrolytic cells are of the form shown in
the figures.
In Fig. 59a, A is a rectangular slate tank in which dips
302
ALKALI, CHLORINE AND THEIR PRODUCTS
an inverted stoneware trough B, containing the anode c,
composed of blocks of retort carbon cast into a lead cap D.
E is iron wire netting, serving as the cathode. The shape
of this netting may be gathered from the section, plan, and
perspective sketch given (Figs. 59a, 59c, and 596), and from
the diagram (Fig. 59d), where A is a longitudinal section of
FIG. 59c.
the cell and E is the profile of the piece of netting. In like
manner a section of the " bell " or inverted stoneware
trough is shown in Fig. 59e. Here the lead cap D has cast
into it numerous rough lumps of retort carbon, forming a
cheap and effective anode, connection with which is made by
lugs passing through the stoneware " bell." The ends
y////
FIG.
of the wire netting serving as cathode project above the
surface of the electrolyte and allow of electrical connection
being made. The apparatus is, therefore, cheap and
simple.
The method of working is as follows : Brine (nearly
saturated) is fed into the anode compartment through a
303
PRACTICAL ELECTRO-CHEMISTRY
trapped pipe to prevent escape of chlorine. At the same
time chlorine is drawn off through tubes from the top of the
stoneware troughs by a rotatory exhauster. The slight suc-
tion (less than 1 inch) maintained in the anode compartment
helps to remove the chlorine as fast as it is generated, and
tends to prevent it from diffusing into the cathode division.
In like manner the influx of brine into the anode compart-
ment tends to keep the liquid therein f airly free from caustic
soda, which would otherwise be gradually transferred from the
cathode compartment. The process of electrolysis is con-
tinued until the cathode liquid contains about 8 per cent,
of caustic soda. It is then drawn off and boiled down,
the salt being fished out and used to make a fresh batch of
brine.
The voltage required is stated to be 5 volts for each tank,
and the current efficiency when the cells are working nor-
mally, 66 per cent. The energy efficiency is, therefore,
30 per cent. The current density at the cathode is 10
amperes per square foot, and at the anode 14 to 15 amperes
per square foot. On account of the necessity for drawing
off the chlorine under slight suction, a certain amount of
air is inevitably drawn in through the numerous joints needed
to connect the large number of single cells with the main
chlorine trunks. Thus the gas (about 30 per cent. Cl) is
used in Deacon chambers l for making chloride of lime,
and is still more conveniently employed in the manufacture
of chlorate.
Undoubtedly one of the merits of the process is the sim-
plicity of the plant and the absence of a porous partition.
This latter feature has, however, a certain disadvantage.
In spite of the efforts, described above, to keep the anode
1 In the Deacon process (a purely chemical method) for making
chlorine, a somewhat dilute chlorine is prepared by the action of air
on hydrochloric acid in the presence of an active material composed
of burnt clay saturated with a solution of cupric chloride. The
chlorine always contains a large excess of air, and is not adapted by
conversion into chloride of lime in the ordinary bleaching powder
chambers. Larger chambers worked systematically are, therefore,
necessary to obtain a satisfactory absorption.
304
ALKALI, CHLORINE AND THEIR PRODUCTS
and cathode products apart, a good deal of mingling is apt
to occur, lowering the current efficiency and contaminating
the caustic liquor drawn off to be boiled down for solid caus-
tic soda. Moreover, in drawing off the contents of the cell
no means exists of alldwing only the cathode liquor to be
taken and retaining the anode liquor. Thus the whole con-
tents of the cell have to be boiled down in order to obtain the
caustic soda in the cathode compartment. These drawbacks
ultimately proved fatal to the success of the process, and,
in fact, it may be said that with the possible exception of
the Bell gravity cell there is no process at work in which
the anode and cathode compartments are not separated,
either by a porous diaphragm or by an intermediate electrode
FIG. 59e.
of mercury. The Bell gravity cell is essentially similar to
that described above. There is a stoneware bell containing
the anode, and the cathode is in the vessel into which the
bell dips. Fresh electrolyte is fed into the anode division
and caustic soda (containing, of course, much common salt)
is drawn off from the cathode division. What merit the
arrangement may have depends on the regularity of the
feed and the disposition of the electrodes.
*THE HARGREAVES-BIRD PROCESS
This process is one of those in which the cathode product
is removed as fast as it is formed, this being one of the
objects set down on p. 289 as desirable of attainment. The
alkali is obtained as sodium carbonate, instead of caustic
soda, and in this respect the process is inferior to those
305 x
PRACTICAL ELECTRO-CHEMISTRY
FIG. 60a.
ALKALI, CHLORINE AND THEIR PRODUCTS
methods which prepare caustic soda at a single operation.
An experimental plant which, in the early history of the
process, was set up at Farnworth, in Lancashire, may be
described as illustrating its chief features.
A gas engine of 20 H.P. nominal drives a dynamo deliver-
ing 2,100 amperes at 4-3 volts. The leads in this experi-
mental plant are somewhat too small in section, and thus
it happens that the pressure drops on its way to the elec-
trolytic cell, and at the terminals thereof has a value of 3-3
volts. Thus the single cell absorbs 9-3 H.P. The electro-
E|
Ut/2
i\V
...,,.-.-.,-,,::!
|j|y|2
~-
------ i.- -_r. i
!fff.-:c
----- r - -_-_-.:!
FIG. 61.
lytic cell is a cast-iron box, 10 feet long, 5 feet high, and 2
feet wide. Its general appearance is shown in Fig. 60a.
It is lined with firebrick set in Portland cement. The inter-
nal arrangement is shown in the vertical section (Fig. 60&).
A is the anode, consisting of a leaden rod passing through
holes drilled in a number of blocks of gas carbon. The rod
itself is protected by a special cement. 1 The cathodes are
sheets of copper gauze, which with their leads are shown
at B, B in the diagram. They support the diaphragms c, c,
and are themselves supported by distance pieces, which
1 According to a recent patent, connection between the carbon
"blocks and the metallic conductor on which they are strung is made
in a hollow vessel filled with oil, the object of which is to prevent
the electrolyte soaking into and attacking the carbon blocks and
their connections.
307
PRACTICAL ELECTRO-CHEMISTRY
keep them from contact with the walls of the cell, so that a
clear space is left on the outer side of the cathode diaphragm.
Brine is circulated through the anode compartment, passing
in its course a box where the chlorine is trapped and led off,
and where salt is added so as to compensate for that con-
sumed. A stoneware pump forces the brine back to the
anode compartment.
This arrangement will be understood from Fig. 61. A is
a box with a hopper B, through which salt can be intro-
duced. The salt enters a compartment cut off by the curtain
c, which does not quite reach to the bottom. Into the other
compartment the brine + chlorine from the anode division
of the electrolytic cell enters by the pipe D ; the chlorine
is trapped and delivered by the pipe G, while the brine is
pumped out through the pipe E and is returned to the anode
compartment.
The space between the cathode and the outer wall of the
cell is not filled with salt solution. Into it steam and carbon
dioxide are blown through the pipes D, D, and a solution of
sodium carbonate trickles away through the pipes E, E
(Fig. 606). The practicability of this procedure depends
on the character of the diaphragm. It is claimed that the
diaphragm is not pervious to ordinary solutions, but never-
theless allows electrolysis to proceed through it. Thus
the liquid in the anode cell cannot ooze through en masse,
but the cathode products of its electrolysis can pass through
the diaphragm and make their appearance at the exterior
surface of the cathode. According to the inventor's views
sodium is first liberated at the cathode, and there acts on
water and carbon dioxide, yielding hydrogen and sodium
carbonate.
It is evident from this that the diaphragm is a highly
important part of the apparatus. It is made by spreading
a mixture of asbestos, silicate of soda and Portland cement
on a paper-making sieve, which is stretched over a chamber
that can be evacuated. The asbestos mixture is thus
sucked together to form a compact felt. The sheet is dried,
and then soaked for some days in a hot bath of silicate of soda.
308
ALKALI, CHLORINE AND THEIR PRODUCTS
The finished diaphragm is about J inch thick, and is of good
mechanical strength. At the time of the author's visit to
the works the diaphragm then in use had been running day
and night for thirty-four days, and seemed to be still working
well. The C0 2 necessary for the carbonation of the soda
is (in the experimental plant) obtained from the exhaust of
the gas-engine, first scrubbed free from sulphur dioxide.
The chlorine is obtained of full strength, the joints of the
apparatus being few considering its output, and serious
leakage of air being thus avoided. At Farnworth the
chlorine is being used for making both bleaching powder
and sodium chlorate. With regard to the working of the
apparatus, it is stated that the current efficiency is 97 per
cent, and that the pressure required is 3' 3 volts. The energy
efficiency will be discussed in a succeeding paragraph. The
current density used is about 20 amperes per square foot
of cathode surface, and rather less on the anode, the exposed
area of which is somewhat greater than that of the cathode
on account of its irregularity. The results obtained by
the experimental apparatus have been so good that a large
installation is about to be started to work the process on an
industrial scale.
The plant of this large installation on the occasion of
my visit at the end of 1901 consisted of 56 cells practi-
cally identical with that which has been described above,
arranged in groups of 14. Power was supplied by two
engines each of 450 H.P. The whole 900 H.P. was not
in use, about 640 E.H.P. being actually absorbed by the
cells then in operation. The chlorine is collected from
the anode compartments by a stoneware rotatory pump ;
being of full strength it can be used in ordinary bleach-
ing powder chambers. The raw material is brine satur-
ated as pumped from the well, i.e. containing about 28
per cent, of sodium chlorine. After passing through the
cells there remains about 20 per cent, of sodium chloride,
and it is found preferable to run this depleted brine away
rather than bring it up to the saturated state by adding
solid salt. So little attention does the process require
309
PRACTICAL ELECTRO-CHEMISTRY
that a staff of nine men suffices for the power and electrolytic
plant. The C0 2 necessary for the cathode compartments is
provided by the waste gases of a smaU vertical boiler carefully
fired so as to keep the percentage of C0 2 as high as possible,
e.g. 12 per cent. The boiler gases are scrubbed in a lime-
stone tower and thus freed from S0 2 . This arrangement
was designed only for temporary needs, the ultimate inten-
tion being to use the gases from the main boilers, keeping
up the proportion of C0 2 by well-arranged mechanical firing.
The Hargreaves-Bird process presents several features
of merit and interest which may be usefully discussed. In
the first place, as stated on p. 305, the process is designed
to produce chlorine and sodium carbonate, and does not
attempt to manufacture caustic soda. The end products
being sodium carbonate, hydrogen and chlorine, instead
of caustic soda, hydrogen and chlorine, the amount of energy
which has to be supplied to bring about the decomposition
of the salt is smaller than that necessary when caustic soda
is produced, being indeed 42-96 Cal., instead of 53-06 Cal.
This corresponds with a critical voltage of 1-85 volts instead
of 2-29. Now the Hargreaves-Bird process is stated to have
a current efficiency of 97 per cent., and to require a working
voltage of 3-3 volts. Therefore its energy efficiency is
1-85
97 + per cent., i.e. 54-4 per cent. This is appreciably
3-3
better than that of most processes making caustic soda
instead of sodium carbonate. It must be noted, however,
that this calculated energy efficiency is somewhat higher
than the truth, because a certain amount of heat energy is
supplied to the apparatus by the steam which is blown in to-
gether with the C0 2 . Making all reasonable allowance
for this, the result remains satisfactory, though it is instruc-
tive to observe how large a waste of energy occurs even in
a well-devised process, distinguished from many of its
rivals by its economy of working.
The next point of interest in the Hargreaves-Bird process
is the comparatively large size of the apparatus. The
single ten-foot cell which was run continuously for consider-
310
ALKALI, CHLORINE AND THEIR PRODUCTS
ably more than a month is capable of producing 1 ton of
bleaching powder per week of seven days of twenty-four
hours each, and -53 ton of sodium carbonate. These figures
correspond with 13-3 pounds of bleach and 7-1 pounds of
sodium carbonate per hour for a single cell. These results,
obtained experimentally, have been equalled or exceeded
in the manufacturing plant. Most other processes which
have been tried can only be worked with cells which
are comparatively small, e.g. giving one-fiftieth of this
output per cell. The multiplication of parts thus needed
is a serious disadvantage, and therefore the large cell with
large output must be reckoned as a substantial merit
of the Hargreaves-Bird process. The next point is the
nature of the diaphragm. A good deal of mystery
attaches to this part of the apparatus. Whether the
inventor is right in stating that a diaphragm made as
described above is impervious to water, but will allow elec-
trolysis to proceed through its pores, is a question difficult
to answer. The fact remains that the diaphragm is effi-
cient for its purpose, which is to keep the salt solution in
the anode compartment, and to allow the cations of sodium
to make their way to the copper gauze cathode, and there in
the presence of C0 2 and steam to yield the cathode products
hydrogen and sodium carbonate. These continuously
escape from the cathode, and thus the sodium carbonate
is not liable to be in its turn electrolysed, as is the caustic
soda produced in a cell of the ordinary type, in which the
cathode product accumulates in the neighbourhood of the
cathode until it reaches a sufficient concentration to warrant
its removal and recovery by boiling down the cathode
liquor. It is by no means clear why the Hargreaves-Bird
process should not be used for the manufacture of caustic
soda, by blowing steam without carbon dioxide into the cath-
ode compartment. No doubt the plan has been tried ;
some working difficulty may have prevented its adoption.
Precise information on the subject is lacking. The fact
that caustic soda is not made is a drawback of the Har-
greaves-Bird process, the price of sodium carbonate being
PRACTICAL ELECTRO-CHEMISTRY
considerably lower than that of a chemically equivalent
amount of caustic soda. A process producing sodium
carbonate can, however, always turn out caustic by adding
to the process proper the simple chemical operation of
causticising the soda ash by means of lime. Thus caustic
soda or soda ash (sodium carbonate) can be made and sold
according to the state of the market. An ideal electrolytic
process would turn out either at need, by a slight alteration
in mode of working, but the time for this is not yet.
The remaining point of interest in the Hargreaves-Bird
process is that it makes chlorine of full strength. This is
chiefly due to the large size of the apparatus and the absence
of those innumerable tubes, all with joints, most of them
leaky, which are necessary in a process having numerous
small units of plant. Thus bleaching powder is made as
readily as by any purely chemical process, and no special
apparatus or process for utilising chlorine largely diluted
with air is necessary. Taking all these things into considera-
tion, it is clear that the Hargreaves-Bird process presents
much that is worth study ; it has been tried systematically,
first on a small and then on a manufacturing scale ; techni-
cally it is a success. 1
THE CASTNER-KELLNER PROCESS
This process is of the type in which an intermediate elec-
trode of mercury is used. With the exception of the
Hargreaves-Bird method, the Castner-Kellner process is
the only mode of manufacturing alkali and bleach elec-
trolytically which has been put into successful operation
on a large scale in this country. The principles on which the
process depends are well known. The details of construc-
tion of the cell and the mode of working are kept secret.
In consequence of this only a diagrammatic sketch (Fig. 62)
of the apparatus can be given.
1 A cell patented by Moore, Allen, Ridlon & Quincy is of the
Hargreaves-Bird type ; it does not appear yet to have been put to
industrial use.
312
ALKALI, CHLORINE AND THEIR PRODUCTS
The cell shown in the figure is divided into three com-
partments A, B, c, by two vertical partitions reaching almost
to the bottom of the cell, but not making a water-tight joint
therewith. Each partition reaches down into a shallow
groove, so that when the bottom of the cell is covered with
liquid each compartment is completely trapped. The liquid
used to cover the bottom of the cell is mercury, a layer of
which is indicated by the shaded portion in the figure. On
this mercury a layer of salt solution rests in two compart-
ments, and a layer of water in the centre compartment.
In the outer are carbon anodes, in the centre an iron grid
acting as the cathode. The cell is supported on a knife
edge at one end and on an eccentric at the other. On rotating
the latter the cell is given a slight vertical motion at that
end and rocks on its knife edge at the other. The layer of
mercury at the bottom of the cell is thus gently oscillated.
FIG. 62.
The cell is completely closed, and there are pipes (not shown
in the figure) for leading off chlorine from the anode com-
partments and hydrogen from the cathode compartment.
Means are also provided for supplying fresh salt solution
to the anode compartments and for drawing off the solution
of caustic soda which forms in the cathode compartment.
The action of the cell is as follows :
The mercury acts as an intermediate electrode between
the anodes and the cathode. At the anodes chlorine is
evolved and sodium is produced at the surface of the mercury
facing each anode. The sodium dissolves in the mercury,
and, on account of the oscillating movement of that liquid
passes into the cathode compartment. Arrived there the
313
PRACTICAL ELECTRO-CHEMISTRY
mercury acts as anode towards the iron cathode. The
sodium which it contains reacts with water, and caustic
soda and hydrogen appear at the iron cathode. The mer-
cury, therefore, acting as a true intermediate electrode,
functions first as a cathode towards the anode of the cell,
and then as an anode towards the cathode of the cell. But
besides this it serves effectively as a diaphragm to keep the
aqueous liquids in the anode and cathode compartments
separate. It also serves as a solvent for the sodium and a
means of transferring it from the anode to the cathode com-
partment. It is, therefore, at once an anode, a cathode, a
diaphragm, a carrier and a liquid seal.
The critical voltage necessary for the electrolytical decom-
position of salt by the Castner-Kellner^ process is precisely
that necessary for any other process having chlorine, caustic
soda, and hydrogen as its end products. Although sodium
is liberated at the cathode surface of the mercury facing the
anode of the cell, yet it is oxidised in due course at the anode
surface of the mercury facing the cathode of the cell. Thus
the extra energy needed for the liberation of sodium instead
of caustic soda in compartments A and c is precisely balanced
by the energy provided by the oxidation and hydration of
the same amount of sodium in compartment B. Looking
at it in another way, one may say that the critical voltage
between the anode and the cathode of the cell is the alge-
braical sum of the voltage between the anode and the mer-
cury and between the mercury and the cathode. The case
may be argued step by step thus :
(1) An aqueous solution of sodium chloride decomposed
so as to yield sodium and chlorine requires the
expenditure of 96-51 Cal. per gramme equivalent.
(2) The combination of sodium and mercury to form
sodium amalgam liberates 21-60 Cal. per gramme
equivalent.
Therefore on the anode side the energy required is 96-51
- 21-60 = 74-91 Cal., i.e. 312,125 joules, corresponding
with a critical voltage of 3-23 volts.
But on the cathode side we have :
3H
ALKALI, CHLORINE AND THEIR PRODUCTS
(1) Sodium amalgam being decomposed and requiring
for its decomposition 21-60 Cal. per gramme equi-
valent.
(2) The reaction of sodium with water, displacing hydro-
gen and forming caustic soda with an evolution
of 43-31 Cal.
Therefore on the cathode side we have a source of energy
amounting to 43-31 -- 21-60 Cal. = 21-71 Cal., i.e. 90,458
joules, corresponding with a maximum available voltage
of 0-94 volt.
This voltage is in a direction opposed to that previously
calculated for the anode compartment, wherefore the actual
critical voltage of the cell is their algebraic sum, viz. 3-23
0-94 = 2-29 volts, which is the value previously calculated
as the critical voltage of a cell electrolysing a solution of
sodium chloride without the use of mercury as an inter-
mediary.
It is clear from this that the use of mercury as an inter-
mediate electrode does not give rise to any increased con-
sumption of energy in the cell. Such advantages as it
presents are, therefore, free from a drawback which might
be feared on casual inspection. These advantages are
sensible enough. There is a complete separation of anode
and cathode products. Formation of such substances as
sodium hypochlorite and sodium chlorate by interaction
of caustic soda and chlorine is impossible under normal
conditions of working. From the cathode compartment
sodium chloride is completely absent and the caustic soda
obtained is pure. The ordinary porous diaphragm, which
has usually either a high resistance or a short life (and fre-
quently both), is abolished altogether. Against these advan-
tages must be set the large quantity of mercury required,
which represents a considerable amount of capital locked up.
The loss of mercury, given careful handling, is in no way
serious. Neither does there appear to be any ground for
the outcry against the process made in its early days on
the ground that sufficient mercury vapour escapes to endan-
ger the health of the workpeople.
315
PRACTICAL ELECTRO-CHEMISTRY
The current efficiency of the process is said to be high
(90 per cent.). The voltage usually required is 4 volts,
2-29
wherefore the energy efficiency is + 90 per cent. = 51-5
per cent., a value similar to that calculated for the Har-
greaves-Bird process (p. 310), viz. 54-4 per cent. But it
must be remembered that the Hargreaves-Bird process
yields sodium carbonate ; the Castner-Kellner gives caustic
soda. The smaller efficiency is more than compensated
for by the greater value of the product.
Few details of the practical working of the Castner-Kellner
process have been allowed to become public. The only
point of special interest which is generally known is that
it is advisable to purify the brine from calcium and mag-
nesium salts. These impurities are removed by recrystalli-
sation ; precipitation with caustic soda or carbonate of soda
may also be practised. A process has also been suggested
which consists in submitting the brine to a preliminary
electrolysis ; the alkali formed suffices to precipitate mag-
nesia from the magnesium salts present as impurities.
The process is at work in England at Weston Point,
near Runcorn. At the time of my visit to the works at the
close of 1901, the power available was about 7,000 H.P., and
of this about 5,000 H.P. was in use for making caustic soda
and bleach. Another 1,000 H.P. was employed for manu-
facturing sodium by the Castner process (p. 186). The
engines are of the vertical marine type, and were at that date
supplied with steam from boilers fired with coal ; since
that time a producer gas plant has been erected with the
ultimate intention of supplying gas engines, replacing the
present steam engines. Pending the installation of the gas
engines the producer gas is to be burned under the existing
boilers. This proposed transition from the steam to the gas
engine plant is certainly a sign of the times. Where cost of
power is an important fraction of the whole cost of a process,
and where a tolerably constant load can be reckoned on,
the greater fuel economy of the explosion engine gives it
substantial advantages. The points still remaining uncer-
ALKALI, CHLORINE AND THEIR PRODUCTS
tain are the cost of upkeep of a plant of this class, and, as
a subsidiary issue, what size shall the unit of power be.
Shall we be content with a gas engine of 250 H.P., or go at
once to a machine developing 2,000 H.P. ?
About 1,100 cells are available for making alkali and
chlorine at Weston Point ; of these about 1,000 will be in
use at any given time, so that something like 5 H.P. will
be absorbed by each cell and its yearly output (assuming an
energy efficiency of 50 per cent.) will be about 10-5 tons of
caustic soda and 22-6 tons of bleaching power per year.
The solution of caustic soda obtained is of fair strength (e.g.
20 per cent.) and is concentrated in double effect evapora-
tors, being finished in ordinary boiling-down pans. On
account of the large number of cells for a given output,
and the correspondingly large number of joints, the chlorine
is diluted with air ; it contains not more than 50 per cent,
of actual chlorine, but nevertheless is strong enough to be
satisfactorily absorbed in ordinary bleaching powder cham-
bers.
Another plant of 2,000 H.P., belonging to the Mathieson
Alkali Company, is running at Niagara,using current supplied
by the Niagara Falls Power Company, and of this, too, a
few details may be given. The output is stated to be 10
tons of caustic soda and 24 tons of bleaching powder per
day of 24 hours ; the current efficiency 85 to 90 per cent. ;
the pressure required 3-5 volts, i.e. the energy efficiency
is 55-6 to 58-9 per cent. These statements are found to be
concordant if we assume that the joint efficiency of the
transformers and dynamos is 80 per cent.
This is not an unreasonable loss, inasmuch as the current
has not only to be let down in voltage, but has to be trans-
formed from an alternating to a direct current. The current
comes from the power house at a pressure of 2,200 volts ;
it is transformed down in stationary transformers to a
pressure of 120 volts. At this pressure the current (which
is, of course, still alternating) passes to motor-transformers,
which transform it to a direct current delivered at a pressure
of 200 volts, this being a convenient voltage for working a
317
PRACTICAL ELECTRO-CHEMISTRY
group of electrolytic cells. The plant has lately been
increased, and it is said that 6,000 H.P. are now in use.
The anodes used are ordinary " squirted " carbons ;
they are subjected to a " special treatment," designed to
render them more refractory, and are said to last a year.
Connection is made with them by means of a lead cap cast
on one end. Recently in many processes of this kind gra-
phite electrodes made by the Acheson process (p. 232) have
been successfully adopted. The caustic soda solution
obtained is fairly concentrated, e.g. about 20 per cent,
strength. Much is sent in liquid form in tank-wagons to
soap-makers in Buffalo, which is about twenty miles from
Niagara. Some is boiled down and sold in the solid state
to the Electro-chemical Company, whose works are close
to those of the Mathieson Alkali Company. This company
(not to be confused with the English company of the same
name) uses it for making sodium by the Castner process
(q.v.). The Solvay process uses an intermediate electrode
of mercury, which is arranged so as to flow continuously over
a weir, its surface containing sodium going to the cathode
compartment, and a new surface being thus exposed in the
anode compartment. In this process, the salt solution
standing immediately over the mercury is kept of a higher
sp. gr. than that surrounding the anode, whereby access of
chlorine to the mercury-sodium surface is hindered. The
advantage of some such device will be understood in con-
sidering the cell shown in the figure below. This cell is one
of the many forms in which the principle of a moving inter-
mediate electrode is used and is generally similar to the
Castner-Kellner cell described above. Its construction
is clear from the diagram. Mercury flows over a weir at A
and across the floor of the cell past the division wall B and
out at the sill c. Sodium liberated at the surface of the
mercury in the left-hand compartment is redissolved to
some small extent by the chlorine dissolved in the brine of
that compartment. Hence the quantity of sodium flowing
into the right-hand compartment is not strictly equivalent
to the quantity of oxygen corresponding with the hydrogen
ALKALI, CHLORINE AND THEIR PRODUCTS
liberated in the right-hand compartment. Hence rather
more oxygen appears at the surface of the mercury in the
right-hand compartment than can be taken up by the
sodium here. As a result the surface of the mercury in
the right-hand compartment suffers oxidation to some
extent. Stated briefly, for a given current passing through
the cell there is some loss of the ultimate anode product
(the chlorine), and an equivalent loss of the intermediate
cathode product (the sodium), but no loss of the ultimate
cathode product (the hydrogen) ; therefore there must be
an excess of the intermediate anode product, oxygen. To
avoid this difficulty a part of the current in the right-hand
vtvfbpT/
FIG. 63.
compartment is short circuited through a resistance which
is shown diagrammatically in the cell itself, but evidently
may be outside. By suitably adjusting this resistance a
wastage of current in the right-hand compartment may be
secured, such that no more hydrogen is there generated than
is strictly equivalent to the chlorine liberated in the left-
hand compartment, and in consequence no more oxygen
is available at the surface of the mercury in the right-hand
compartment than is actually needed by the sodium dis-
solved in that mercury. The Rhodin process is one having
a mercury electrode. Its general principles are so similar
to those of the Castner-Kellner apparatus that prolonged
litigation has taken place between the companies owning
319
PRACTICAL ELECTRO-CHEMISTRY
the respective patents. The Bell mercury cell (not to be
confused with the Bell gravity method referred to on p. 305)
embodies similar principles, and separation of the anode and
cathode divisions is secured by a flowing intermediate
electrode of mercury. The Le Sueur apparatus, as first
FIG. 64.
devised, resembled that of Holland & Richardson in respect
of the fact that the anodes were blocks of carbon contained
in a stoneware bell dipping in a trough of salt solution. The
process, which is at work at Rumford Palls, Maine, has
lately been modified and very thin sheets of platino-iridium
are now used instead of carbon as anodes. These, though
320
ALKALI, CHLORINE AND THEIR PRODUCTS
high in first cost, are permanent, and their use is found to
be economical. At the bottom of the anode bell is an
asbestos diaphragm ; on this is stretched a sheet of wire
gauze, serving as the cathode. Each cell is 9 feet x 5 x
1J feet, therefore the units of plant are conveniently large.
A pressure of 4 volts is needed ; the current efficiency is
stated to be 70 per cent. These figures correspond with an
energy efficiency of 40 per cent. The liquid in the anode
c.ell is kept slightly acid with hydrochloric acid. By this
means any sodium hypochlorite which may be momentarily
formed by the incursion of caustic soda from the cathode
side of the diaphragm is at once decomposed, and caused
to yield its equivalent of chlorine instead of oxygen. The
construction of the Le Sueur cell with its original carbon
anodes may be understood from the accompanying figures.
A is the bell containing the anode B, shown separately and
in greater detail in the smaller diagrams . c is the diaphragm
and D the tank containing the whole apparatus. The bell
is canted so as to favour the escape of hydrogen from the
gauze cathode beneath the diaphragm.
Another diaphragm cell is that known as the Outhenin-
Chalandre which has been put into use at Chevres in Swit-
zerland. The chief points in the construction are shown
in the figure. The outer tank B contains an inner vessel A,
which constitutes the anode compartment. The anodes
i i are rods of carbon cast into a lead cap u. They hang
down between the porous cells o o, which are arranged in
tiers of six, slanting a little as shown. It will be understood
that there are alternate rows of anodes and porous tubes
from end to end of the tank. In the porous tubes are iron
cathodes c c, which form, as it were, the teeth of a comb,
of which M is the back ; the arrangement is shown in the
figure. The porous tubes at both ends are set into the
walls of the vessel A so as to make a tight joint. The
upper ends (on the right of the figure) of the porous tube
are not closed, 1 but communicate freely with the space be-
1 The appearance of closure is due to the fact that the caps used
to make a joint with the wall of the vessel A are seen in section.
321 Y
PRACTICAL ELECTRO-CHEMISTRY
tween the inner vessel A and the tank B. In like manner
the lower ends of the tubes are open to the corresponding
space at the left of the figure. The working of the cell is
quite simple. Brine is fed into the anode compartment, and
the chlorine there generated escapes by the pipe H. The
brine then passes through the porous cells ; the hydrogen
FIG. 65.
given off is trapped by the hood v and led away. The caustic
soda flows down the slanting cells as the hydrogen flows up,
sinks to the bottom of the outer vessel and syphons over by
the pipe x. The chief interest of this cell, which presents
no novelty in principle and is somewhat complicated in
structure, depends on the fact that there is an attempt by
322
ALKALI, CHLORINE AND THEIR PRODUCTS
p
the number and by the sloping arrangement of the diaphragm
to work in some degree systematically ; the soda-solution
can get away from the cathodes because they are numerous
and independent of each other, and at the same time the
anode compartment is single ; hence the number of joints
for a given output is moderate. The complex structure
may be a serious drawback in working. In short, the cell,
like all those which are mere modifications in detail and
involve no fundamental change of idea, can only be judged
by comparing its behaviour in practice with that of others
of its class.
A good many other processes are at work in different
parts of the world, but the details of their working have not
been disclosed.
The present position of the electrolytic manufacture may
be summarised thus :
The original simple idea of electrolysing a solution of
common salt, until a good deal has been converted into
caustic soda and chlorine, and trusting for a separation of
the products to the fact that chlorine being a gas will escape,
fails completely in practice ; at a very early stage there is
enough caustic soda at the anode both to combine with the
chlorine and to convey current on its own account. The next
step, namely, to provide some form of porous diaphragm,
has not proved so successful as might reasonably have been
expected. The Greenwood cell, now defunct, was a good
example of a diaphragm cell ; it certainly was cleverly
designed and had considerable merits ; its failure is not to
be attributed to its principle. One may even go further and
say that there is a good deal to be said for a simple diaphragm
cell if its unit can be made large enough. As far as I know,
however, there is only one simple diaphragm cell in practical
use at the present time, the Le Sueur.
Any simple diaphragm cell will produce caustic soda
solution of only a moderate strength and mixed with sodium
chloride. To obtain a pure solution of soda of fair strength
one of two devices must be employed. The first is that of
a cathode and diaphragm all in one, as in the Hargreaves-
323
PRACTICAL ELECTRO-CHEMISTRY
Bird and the Moore, Allen, Ridlon and Quincy cell (p. 312).
The other is the use of mercury as an intermediate electrode.
Both methods have considerable merits ; both have been
worked on a large scale. The balance of advantage seemed,
when trials were first made, to be on the side of the Har-
greaves-Bird type with its large units and few gas joints ;
but at the time of writing commercial success inclines the
other way. Probably the greater part of the world's output
of caustic soda and bleaching powder by electrolysis is now
made by some form of mercury cell of the Castner-Kellner
or Solvay type.
It is perfectly possible that the class of cell represented
by the Acker, in which fused salt is electrolysed, may prove
ultimately the best for the production of caustic soda. At
first sight it seems wasteful to make sodium when only
caustic soda is wanted, but the waste is one of energy, and
that is fairly cheap. Evidently a process of this kind is at
an advantage at a spot like Niagara Falls, where salt has to
be obtained from a distance in the solid state and not as
brine, and where the cost of power is low. At Middlewich,
where brine is pumped and waterfalls are absent, its advan-
tage is reduced.
The pioneers of the manufacture of alkali and bleach elec-
trolytically have done good service in stirring up their
chemical rivals . There is no prospect of any existing electro-
lytic process extinguishing the older method, but there are
plenty which are quite able to engage in lively and effective
competition, wholesome for all concerned.
PRODUCTS OTHER THAN CAUSTIC SODA AND
CHLORINE
Cognate with the industries dealt with above are those
concerned with the manufacture of caustic potash, chlorates
and hypochlorites. Substituting potassium chloride for
sodium chloride in a practicable apparatus such as the Cast-
ner-Kellner, one would obtain chlorine and caustic potash
324
ALKALI, CHLORINE AND THEIR PRODUCTS
instead of caustic soda. The trade in caustic potash, al-
though smaller than that in caustic soda, is nevertheless-
very considerable. For certain purposes, e.g. in making
soft soap and in preparing oxalic acid from sawdust, caustic
soda cannot be used in place of caustic potash. The dearer
alkali must be employed, and the demand for it is not likely
to decrease. The raw material, potassium chloride, is much
dearer than sodium chloride, and thus it is of more import-
ance to economise raw material than to decrease to its utmost
limit the cost of manufacture. Therefore an electrolytic
process, even if as costly as, or somewhat more costly than,
one which is purely chemical, has a greater chance of success
when working on potassium chloride than on sodium chloride
by reason of its economy of raw material. The cost of raw
materials, of power, and the selling price of products when a
potassium salt is used may be compared with similar figures
for a sodium salt in the following table. The calculation is
made for a consumption of energy of 1 H.P. (at the terminals
of the electrolytic cell) acting for a year. The cell is assumed
to work with an energy efficiency of 57 per cent.
POTASSIUM CHLORIDE
Weight of
electrolyte
decomposed.
Products.
Value of raw
materials.
Value of
products.
Caustic
potash.
Chloride
of lime.
4-4 tons
3-3 tons
5-2 tons
33 85.
121
SODIUM CHLORIDE
Weight of
electrolyte
decomposed.
Products.
Value of raw
materials.
Value of
products.
Caustic
soda.
Chloride
of lime.
3-5 tons
2-4 tons
5-2 tons
5 35.
54
325
PRACTICAL ELECTRO-CHEMISTRY
It is assumed that steam power is used in each case,
and that a H.P. year costs 9. Comparison of this with
the value of the raw materials dealt with by that power,
viz. 33 85. for potassium chloride and 5 35. for sodium
chloride, shows at a glance the much smaller proportion
which the cost of the energy bears to the cost of the raw
materials in the manufacture of caustic potash than that
which it does in the manufacture of caustic soda. The
difference is still more marked when the selling price of
the products is used as the basis of comparison. It is
easy to see that the electrolytic manufacture of caustic
potash by a process not wasteful of raw materials and
turning out a product of high grade should be remunera-
tive, even if the cost of energy be somewhat larger than that
given. There is not, as far as present information goes,
any electrolytic process specially devised for the production
of caustic potash as distinct from caustic soda.
ELECTROLYTIC MANUFACTURE OF CHLORATES
If the products of the electrolysis of sodium chloride
(hydrogen, caustic soda and chlorine) are brought together
and caused to combine, they reproduce the common salt
and water from which they have been derived. If one of
these products, viz. hydrogen, be eliminated, the caustic
soda and chlorine interacting will produce either a mixture
of sodium hypochlorite and chloride or one of sodium chlor-
ate and chloride, according to the temperature at which the
reaction is caused to occur. Thus :
(a) 2NaOH + 2C1 = NaCl + NaOCl + H 2 ; or
(6)6NaOH + 6C1 = 5NaCl + NaC10 3 + 3 H 2 0.
It must not be supposed, because a portion of the sodium
chloride used in preparing the caustic soda and chlorine
is regenerated, and thus chlorine appears to be uselessly
consumed, that there is any waste of the oxidising or chlori-
nating power of the chlorine. For 1 molecule of sodium
326
ALKALI, CHLORINE AND THEIR PRODUCTS
hypochlorite (NaOCl) is equivalent in oxidising power to
2 atoms of chlorine, and similarly, 1 molecule of sodium
chlorate is equivalent to 6 atoms of chlorine. It may,
therefore, be accepted that the oxidising and bleaching
products formed when the anode and cathode products
(excluding H) of the electrolysis of sodium chloride are
brought together are precisely equivalent in oxidising or
bleaching value to the chlorine normally evolved in the
anode compartment. It might be assumed from this that
the simplest manner in which a bleaching solution could be
prepared would be by electrolysing a solution of common
salt or other suitable chloride in a cell without a diaphragm.
But such electrolysis could be conducted only up to a certain
point. The hypochlorite (or chlorate) formed by the union
of the caustic soda from the anode and the chlorine from the
cathode would not be confined to the neighbourhood of the
anode. It would be free to diffuse to the cathode, and would
there be reduced to chloride. Thus the energy impressed on
the electrolyte would be consumed in oxidising chloride to
hypochlorite (or chlorate) and subsequently reducing it
again to chloride. The net result is merely the conversion
of electrical energy into heat an outcome unintended,
costly and useless. Therefore the simple plan whereby
sodium chloride can be directly oxidised by hypochlorite
(or chlorate) in an undivided electrolytic cell can be utilised
only under particular conditions ; in general a more complex
arrangement is necessary. The methods which promise
greatest prospect of success may be usefully discussed.
PRODUCTION OF HYPOCHLORITES
Sodium hypochlorite may be made by the electrolysis
of a solution of sodium chloride, using carbon electrodes,
employing no diaphragm, and mixing the anode and
cathode products by agitation. The temperature of the
electrolyte should be kept low, e.g. below 60 F. = 15 C.
The concentration of the sodium chloride solution may be
high, but that of the hypochlorite should be low, e.g. 10
327
PRACTICAL ELECTRO-CHEMISTRY
grammes per litre ; by special care in mixing and
cooling the electrolyte, it is claimed that as high a con-
centration as 20 grammes per litre may be reached, but
under ordinary conditions the lower value is high enough.
It is impracticable to convert more than a small fraction of
the sodium chloride into hypochlorite, because, as the con-
centration of the latter rises, it is itself acted on and Teduced
at the cathode. Therefore the commercial production of a
hypochlorite in this manner is confined to cases where the
electrolysed liquor can be used for bleaching purposes and
returned to be again oxidised and made again effective as a
bleaching agent. Should the use of the bleaching liquor
contaminate it seriously (as in the bleaching of paper), it
may not be feasible to return it to the electrolysing cell.
In this case the process described can only be used when
the raw material, e.g. sodium chloride, is so cheap and abun-
dant that it can be used wastefully. Similar bleaching
liquids suitable for circulation through a bleaching process
and return to the electrolytic cell can be prepared from cal-
cium chloride and magnesium chloride. In the case of the
latter, the liquid is particularly active, because magnesium
hypochlorite is an unstable salt, and is readily hydrolysed,
yielding free hypochlorous acid. It is this property which
has led to extravagant statements concerning the remarkable
bleaching and oxidising effects of an electrolysed solution of
magnesium chloride ; these are due to the presence of free
hypochlorous acid. Where it is desirable to obtain a parti-
cularly active bleaching agent, a solution of hypochlorous
acid formed by treating a solution of common bleaching
powder with carbonic acid can be adopted. Choice between
such a solution and one prepared by electrolysis is governed
wholly by their cost.
A method for electrolysing sea-water, known as the
Hermite process, and intended for the production of an
oxidising, deodorising and bleaching liquor, chiefly for
the treatment of sewage, has been tried in this country at
various seaside places without achieving any great success.
It merits no detailed description, being merely an arrange-
328
ALKALI, CHLORINE AND THEIR PRODUCTS
ment for producing a weak solution of hypochlorites by the
electrolysis of the chlorides naturally present in the sea- water.
In cost it compares unfavourably with that of bleaching
powder and similar chlorinating agents.
Should a demand arise for pure hypochlorites, i.e. for
solutions free from the large excess of chlorides inevitably
present in any chlorinating solution produced by the direct
electrolysis of a chloride without separating cathode and
anode products, it can be met by any successful process for
the manufacture of alkali and bleach, e.g. the Castner-Kellner
process. It will then be worked as an adjunct to the main
manufacture of caustic soda and bleaching powder ; the
cost of such a bleaching liquor will depend primarily on that
of the chlorine produced electrolytically, and if that is
smaller than the price of chlorine made by chemical pro-
cesses, there will be a corresponding saving in the cost of
production of the bleaching liquor.
A large number of apparatus for the preparation of bleach-
ing liquids have been devised. They differ in details of
construction, but if serviceable for their purpose all involve
the same principles of design. These are that there should
be numerous electrodes with small spaces between them
through which salt solution can be pumped at a regulated
speed ; that these electrodes should be unattackable ; very
thin platinum or platinum-iridium foil has proved useful and
not unduly costly. The next essential point is that the
bleaching liquor, if it is used again, must be well cooled before
it is returned to the electrolyser. The embodiment of
these ideas is shown in the following figure, which represents
an apparatus built by Siemens and Halske.
The electrolyser itself is a stoneware vessel A B, containing
some 10 or 20 electrodes, which are in series, so that a single
connection at each end suffices, the intermediate electrodes
acting as both anode and cathode in the usual way.
The electrodes are arranged to form a number of
separate narrow cells ; through these the solution to be
electrolysed flows in by the pipes E r at the bottom of the
vessel and overflows through the troughs c D at the top
329
PRACTICAL ELECTRO-CHEMISTRY
down into the collecting reservoir H, in which is a cooling
coil. From the reservoir the electrolyte, which is now a
bleaching liquor, is driven by the centrifugal pump G to the
tank in which the bleaching is to be conducted, or back into
the electrolyser. It will be understood that this process of
circulation can be varied according to the needs of the case.
The electrolyte, if not strong enough after a single treatment,
may be pumped back into the electrolyser ; if ready for
use it may be pumped to its work of bleaching and pumped
back again when exhausted ; or it may be rejected and
fresh salt solution pumped through the cells. Evidently
the direction of the salt solution is indifferent, provided
that a continuous and sufficient stream be sent through
the cell and the returning liquor be adequately cooled.
Another apparatus, made by the Elektricitats-Aktiengesell-
330
ALKALI, CHLORINE AND THEIR PRODUCTS
schaft vormals Schiickert & Co., illustrates the same prin-
ciples. A group of cells is constructed by dividing up a
vessel m by partitions s, made of slate or glass. The elec-
trodes k are of carbon, coupled as shown. The electrolyte
after passing between them flows into a cooling cell provided
with a zigzag pipe of lead or glass through which water is
circulated. Hence as the electrolyte passes from cell to
cell it is cooled on its way and its temperature maintained
low enough for efficiency. The manufacturers of this
apparatus consider that a temperature of 30 C. is as low
as is necessary, holding that the better output obtained with
FIG. 67.
better cooling does not compensate for the elaboration of
the cooling apparatus. The electrolyte used is a 10 per cent,
solution of salt to which a few grammes per litre of calcium
chloride, lime and sodium resinate have been added. It
is stated that a film of calcium resinate is formed at the
cathode, hindering the reduction of the hypochlorite at that
surface. The probable action of such a film is discussed in
the section on Chlorates (p. 336). With this apparatus
and its special electrolyte, hypochlorite solution containing
33 grammes of available chlorine per litre is said to be ob-
tained.
PRACTICAL ELECTRO-CHEMISTRY
PRODUCTION OF CHLORATES
What has been said with regard to hypochlorites applies
generally, mutatis mutandis, to chlorates. The obvious-
method of preparation is to manufacture caustic soda and
chlorine in any good electrolytic apparatus, and to use the
chlorine for the production of chlorates precisely as it is used
when its mode of preparation is purely chemical. Seeing
that the chlorine may happen to be diluted with air, drawn in
through the many joints usually requisite in an electrolytic
chlorine plant, its utilisation for making chlorate is, on the
whole, preferable to its employment for the production of
bleaching powder, which is best made with chlorine of full
strength. This view commended itself to the Electro-
chemical Co. (whose process is described on p. 302), who-
used a good portion of their output of chlorine for making
chlorate. Granting that chlorate is to be made with electro-
lytic chlorine, it becomes sufficient to indicate the usual
chemical process for chlorate manufacture.
Potassium chlorate is that which is manufactured in.
the largest quantity. It is not made directly by the action
of chlorine on caustic potash according to the equation
6 KOH + 6 Cl = 5 KC1 + KC10 3 + 3 H 2 0/
because five-sixths of the necessary caustic potash would
be converted into potassium chloride, a comparatively
low-priced salt. The plan used to get over this difficulty
is first to prepare calcium chlorate thus :
6 Ca(OH) 2 + 12 Cl = 5 CaCl 2 + Ca(C10 3 ) 2 + 6 H 2 0,
and then to act on this with potassium chloride thus :
Ca(C10 3 ) 2 + 2 KC1 = 2 KC10 3 + CaCl 2 ,
1 The action of chlorine on a caustic alkali gives a hypochlorite
as the main product when the solution is cold, and a chlorate when,
the solution is hot. The two reactions are shown on page 326.
332
ALKALI, CHLORINE AND THEIR PRODUCTS
giving potassium chlorate and calcium chloride. There
is, therefore, no waste of any potassium salt, and the use
of caustic potash, which is comparatively costly, is dispensed
with. The manufacturing operation consists in passing
the chlorine into hot milk of lime, contained in a series of
cylindrical vessels. The contents of the vessels are kept
agitated and the absorption of the chlorine is conducted
systematically, i.e. the chlorine as it enters is passed into a
vessel already nearly saturated, and as it leaves passes out
through a vessel containing fresh milk of lime. The liquor
containing calcium chlorate is run into settling tanks and
is there treated with potassium chloride. The solution,
which may be regarded as containing, potentially at least,
potassium chlorate and calcium chloride, is evaporated until
it attains a specific gravity of 1-35, when potassium chlorate
crystallises out. The calcium chloride liquor, retaining a
portion of potassium chlorate, is run off and cooled strongly
to induce a further fraction of the potassium chlorate to
crystallise. The crude potassium chlorate is recrystallised
to free it from adhering calcium chloride, and is then pure
enough for ordinary commercial purposes.
When, however, chlorate is made by some special pro-
cess of electrolysis, distinct from those designed for the
manufacture of alkali and bleach, certain difficulties arise.
The direct method of electrolysing a hot solution (e.g.
one at a temperature approaching that of the boiling-point
of water) of potassium chloride in a vessel without a dia-
phragm, and causing free mixture of the caustic potash and
chlorine produced, is feasible only up to a small concentration.
The recovery of the chlorate from a solution rich in chloride
by means of any process of crystallising out the chlorate is
somewhat expensive. Thus some means must be sought to
permit the production of a more concentrated solution.
Where no diaphragm or other means of separation exists,
the anode product, i.e. the chlorate, will reach the cathode
and be there reduced. At the same time the caustic alkali
formed at the cathode may itself serve to convey the cur-
rent and yield as ultimative products oxygen and hydrogen.
333
PRACTICAL ELECTRO-CHEMISTRY
In either case electrical energy is expended uselessly, m
the first instance appearing as heat in the solution, and in the
second being represented by the chemical energy of products
which are not required and are useless to the chlorate manu-
facturer. Several suggestions have been made to remedy
these disadvantages. Thus Kellner proposes to add to the
solution of potassium chloride a small quantity of a spar-
ingly soluble hydroxide, such as slaked lime or magnesia.
He takes a saturated solution of potassium chloride and
adds to it about 3 per cent, of slaked lime ; a portion of this
dissolves, but the greater part remains in suspension. The
electrolyte may, therefore, be regarded as saturated with
calcium hydroxide, and containing a store of undissolved
calcium hydroxide ready to dissolve should that already
in solution be used up from any cause. In order to provide
a supply of lime to all parts of the electrolyte, the liquid
is agitated so as to prevent the slaked lime from settling
out. On electrolysing this solution electrolysis is con-
fined practically to the potassium chloride ; the quantity
of calcium hydroxide in solution is so small that no appre-
ciable proportion of the current is conveyed thereby. The
chlorine evolved at the anode comes in contact with the
dissolved calcium hydroxide, and at the temperature proper
to the reaction forms calcium chlorate and calcium chloride.
The former reacts with the potassium chloride, yielding
calcium chloride and potassium chlorate. The latter, to-
gether with the calcium chloride produced by the reaction
of the calcium chlorate with the potassium chloride, is
decomposed by the caustic potash liberated at the cathode,
giving calcium hydroxide and potassium chloride. Thus
all the materials return to the status quo ante, except a
portion of the potassium chloride which has been conver-
ted into potassium chlorate. The function of the calcium
hydroxide is merely to provide a medium for the absorp-
tion and utilisation of the chlorine, which is then passed
on to the caustic potash at the cathode. It may be said
that the same effect could be produced by adding caustic
potash to the electrolyte, so as always to maintain a slight
334
ALKALI, CHLORINE AND THEIR PRODUCTS
preponderance of alkali to combine with the chloride before
it can reach the cathode. This is true, but the plan has the
disadvantage that on account of the solubility of caustic
potash the whole of that added would be in solution, and
not chiefly undissolved and in suspension as a reserve to be
drawn upon as occasion required. To have the whole of
the caustic alkali in solution would lead to the inconvenience
(dealt with above) of a part of the electrolysis proceeding
with the caustic potash as an electrolyte instead of the potas-
sium chloride exclusively acting thus. It would therefore
be necessary to add the caustic potash little by little as it
was required, whereas the slaked lime, on account of its
sparing solubility, regulates the supply of alkaline hydroxide
automatically. It will be seen that it is tacitly assumed that,
provided the chlorine be converted into chlorate, it will
not readily be reduced at the cathode, for whatever devices
are adopted the chlorate must ultimately come into contact
with the cathode. This assumption is probably true. It
is certain that if chlorine and hydrogen were liberated
in juxtaposition they would combine. It is probable that
hypochlorite brought into the immediate neighbourhood of
the cathode would be reduced ; it is by no means so likely that
a chlorate in the immediate neighbourhood of the cathode
will suffer a corresponding reduction.
This idea of Kellner is ingenious and appears sound in
principle. No information, however, is forthcoming as to-
rts having been used on a manufacturing scale. This lack
of specific information is characteristic of the chlorate manu-
facture, which is being quietly pursued by various firms
who guard their particular methods with much care. Never-
theless, it may be taken that all essential principles have been
treated of in the foregoing paragraphs, and that novelties
and secrets of manufacture relate rather to the form of ap-
paratus and small details of working than to any great or
fundamental difference from what is generally accepted and
understood.
The idea underlying the method of Kellner, which is de-
scribed above, receives fresh illustration from the researches-
335
PRACTICAL ELECTRO-CHEMISTRY
of Bischoff and Forster on the electrolysis of a solution of
calcium chloride. When calcium chloride is used instead
of potassium chloride, the calcium hydroxide liberated at
the cathode forms a coating thereon, which confines the
reducing action of the hydrogen simultaneously formed to
very narrow limits, acting, in fact, as a sort of diaphragm,
preventing access of the chlorate (or hypochlorite) to the
cathode. It is evident also that the solution of calcium
chloride must contain calcium hydroxide in solution,
and indeed in suspension, as portions of the film on the
cathode become detached. Thus the electrolyte is in much
the same condition as Kellner's, in which there is an auto-
matically regulated supply of alkaline hydroxide capable
of absorbing and utilising the chlorine evolved at the cathode.
The resistance at the cathode is considerably increased by
the film of calcium hydroxide adhering there, and in this
respect the arrangement is inferior to Kellner's. Another
difference is the greater solubility of calcium hydroxide in
calcium chloride solution than in water (or a solution of
potassium chloride). This is of doubtful advantage, inas-
much as the presence of any considerable quantity of alkaline
hydroxide in solution and acting as an electrolyte will tend
to waste current by allowing the formation of oxygen and
hydrogen as end products instead of the chlorate, which
is the object of manufacture.
The idea of screening the anode products from the re-
ducing action at the cathode by means of a diaphragm manu-
factured from the electrolyte itself has been applied in the
Schuckert apparatus described on p. 331. The primary object
of the apparatus is the manufacture of hypochlorite, but
the principle is the same. According to the English patent,
the electrolyte is made by adding to every 14 litres of a 10
per cent, solution of common salt 40 grms.of calcium chloride,
30 grms. of lime, and 50 c.c. of a strong solution of resin
in caustic soda. In this way a film, probably of calcium
resinate, is found on the cathode and hinders the hydrogen
from acting on the hypochlorite which is the product
sought. This arrangement is said to be effective, and may
336
ALKALI, CHLORINE AND THEIR PRODUCTS
well be suitable for chlorate as well as hypochlorite
manufacture.
A cell for the manufacture of chlorate, in use by the Na-
tional Electrolytic Co. at Niagara Falls, shows certain points
of interest, and may be illustrated. In its early form the cell
had cathodes of copper oxide which were designed to sup-
press the hydrogen and prevent reduction of chlorate. This
FIG.
device has been abandoned, and reduction is now avoided
as far as possible by providing a continuous flow of potassium
chloride solution and keeping the concentration of the
electrolyte in chlorate as low as 3 per cent. ; the chlorate
is recovered by refrigeration, and the electrolyte, after the
necessary make up with potassium chloride, is returned to the
cell. A group of cells is shown in the figure. A wooden
337 z
PRACTICAL ELECTRO-CHEMISTRY
frame A is lined with lead B, arranged to form a series of
compartments. The cathodes consist of a grid of copper
wires carried on insulating cross bars ; a single wire c and
its cross bars in section o are shown in the figure. The
anode is a sheet of platinum foil E closely applied to the lead
wall D of the compartment. The chloride solution is fed
in by the pipes G, and the chlorate solution withdrawn by the
pipes H. The distance between anode and cathode is small,
e.g. about J in., and free intermingling of their products takes
place ; the temperature in the cell is maintained at about
50 C. The plant at Niagara Falls takes about 2,000 H.P.
It may be noted, in closing this section, that a still higher
state of oxidation than that represented by the chlorates
may be attained electrolytically. When a solution of potas-
sium chlorate is electrolysed with platinum electrodes, and
with the observation of certain conditions about to be de-
scribed, potassium perchlorate (KC10 4 ) is formed. In
order to get a good yield, e.g. 70 to 90 per cent, of the total
oxygen in the form of perchlorate, the electrolyte should be
kept cool, certainly not above 10 C. ; the current density
at the anode should be high, e.g. 4 to 12 amperes per square
decimetre ; and the electrolyte should be a saturated solu-
tion of the chlorate, preferably the sodium salt, because of
its solubility being greater than that of the potassium salt.
It is noteworthy that a good deal of ozone is given off
during the electrolysis, and it has even been suggested
to utilise this fact for the manufacture of that gas.
At present, however, there is no great commercial
demand for either ozone or perchlorate. In the manu-
facture of both chlorate and perchlorate the addition
of a chromate to the electrolyte is sometimes practised.
The underlying idea is to provide some substance
which is alternately reduced and oxidised, transferring its
oxygen to the chloride to be oxidised to chlorate, or the
chlorate to be oxidised to perchlorate. How far such an
addition is useful depends on whether the action of this
carrier avoids the formation or curtails the existence of
transition products like hypochlorite. If it is successful
338
ALKALI, CHLORINE AND THEIR PRODUCTS
in this function it may serve much the same purpose as the
rapid circulation, careful control of temperature and re-
striction of the concentration of the electrolyte in the product
sought to be obtained, which are the ordinary precautions
of manufacture of oxidised chlorine compounds in cells
without a diaphragm.
339
SECTION VIII
The Electrolytic Manufacture of
Organic Compounds and Fine Chemicals
The Electrolytic Manufacture of
Organic Compounds and Fine Chemicals
SEEING that by means of electrolysis a reducing action
can be exerted on an electrolyte at the cathode and
an oxidising action at the anode simply by the impress of
energy without the introduction of any foreign matter, it
is evident that electrolytic methods for the preparation of
many chemical substances have a prima facie advantage
over purely chemical methods, which, from the nature of
the case, frequently involve the use of some substance which
ultimately, having done its work, forms no part of the pro-
duct which it is sought to obtain, but is rather an encum-
brance and impurity to be eliminated. To take a simple
case : If copper is to be precipitated as metal from its
sulphate, the work can be done by metallic zinc ; it can also
be done by passing a suitable current (using an insoluble
anode). In the one case the solution at the end of the opera-
tion is encumbered by zinc sulphate ; in the other there
remains no foreign substance, but there are present simply
the products of resolution copper and sulphuric acid.
Silver may be precipitated from its solution by a variety
of reducing agents, e.g. tartrates, formaldehyde and milk
sugar ; the products of their oxidation remain in solution.
It may be precipitated electrolytically without the addi-
tion of any foreign material. A solution of cupric chloride
may be reduced to cuprous chloride by means of sulphurous
acid, but the resulting solution is contaminated with sul-
phuric acid ; it may be reduced electrolytically and remain
343
PRACTICAL ELECTRO-CHEMISTRY
free from such contamination. A lead salt may be oxidised
so as to yield lead peroxide by the action of caustic potash
and chlorine ; it may be obtained pure and directly by elec-
trolysis. Nitrobenzene may be reduced to aniline by iron
and hydrochloric acid ; at the cathode the same product
may be obtained per se. Examples might be multiplied.
It must not be concluded that an electrolytic method of
preparing a given substance is necessarily preferable to a
strictly chemical method. Considerations of cost, conveni-
ence, speed of output, obtainment of a high yield or of use-
ful by-products, must all be taken into account, and these
sometimes turn the balance of advantage against the elec-
trolytic method.
The manufacture of organic compounds, such as dye-
stuffs, and of fine chemicals, is an industry relatively in-
significant, although absolutely considerable. The processes
used are simply laboratory processes writ large, and their
practice and control are in the hands of a few highly trained
chemists. It follows that the methods employed are in es-
sence laboratory methods, and that any advance which
may be made, being in few hands, is carefully guarded from
public scrutiny. Such published processes as are rational
and promising are here recorded.
The typical electrolysis of common organic substances
recorded in the text-books is that of sodium acetate. It may
be supposed to take place in two stages :
(1) 2 (CH 3 CO(ONa) ) = C 2 H 6 + 2 CO 2 + 2 Na ;
(2) 2 Na + H 2 = 2 NaOH + 2 H.
The salt is resolved, yielding ethane and carbonic anhydride
at the anode and caustic soda and hydrogen at the cathode.
The reaction is general, though not necessarily quantita-
tive, with alkali metal salts of the acetic series. The acids
themselves should split up thus :
2 (CH 3 CO(OH) ) = C 2 H 6 + 2 C0 2 + 2 H,
but in dilute solution act simply as aids to the electrolysis
of water, much as does sulphuric acid. This decomposition
344
ELECTROLYSIS OF ORGANIC COMPOUNDS
of the salts of organic acids may be correlated with that of
certain of the salts of inorganic acids, where the salt is re-
solved primarily into the metal and the acid radicle, both
undergoing consequent changes. Thus the resolution of
sodium sulphate in the presence of water may be represented
by the equations :
(1) Na 2 SO 4 = 2 Na + S0 4 ;
(2) 2 Na + 2 H 2 = 2 NaOH + 2 H ;
(3) S0 4 + H 2 O = H 2 S0 4 + O ;
the final products being caustic soda and hydrogen at the
cathode and sulphuric acid and oxygen at the anode. Simi-
larly the resolution of sodium benzoate maybe regarded as
passing through corresponding steps, and its ultimate result
may be represented thus :
C 6 H 5 CO(ONa) + H 2 = C 6 H 5 CO(OH) + NaOH,
Benzole acid.
and of the sodium salt of phthalic acid :
C 6 H 4 (COONa) 2 + 2 H 2 = C 6 H 4 (COOH) 2 + 2 NaOH.
Phthalic acid.
This is the simplest case. 1 If sodium acetate were decom-
posed in this manner it would yield CH 3 COOH and NaOH.
It is the splitting up of the acid radicle which gives the pro-
ducts set forth above. The salts of certain acids, such as
hydroxy acids like lactic and tartaric acids, give products
which suffer a further oxidation, which may extend to the
complete destruction of the radicle and the production of
so typical a product of limited oxidation as CO . The large
number of possible changes, which are controlled not only
by the materials electrolysed, but by the conditions of
1 Although this reaction proceeds smoothly in acid or neutral
solution, yet in alkaline solution decomposition goes farther, the
anode products being carbon dioxide, carbon monoxide and some-
times acetylene ; a smell of oil of bitter almonds is frequently
observed.
345
PRACTICAL ELECTRO-CHEMISTRY
electrolysis, make prediction of the course of a given reaction
dubious and compel constant experiment. This is in process-
of being carried out by several investigators.
The very obvious idea of reducing nitro-compounds by
exposing them to the action of the current at the cathode
appears to have been patented by Die Farbewerke vormals
Friedrich Bayer in 1893. According to this patent, the
nitro-compound is dissolved in sulphuric acid either con-
centrated or only slightly diluted and placed in a cell
surrounding the cathode ; the anode is immersed in
sulphuric acid of 70 to 90 per cent, strength. Examples
of this method of reduction are furnished. Thus nitro-
benzene, C 6 H 5 (N0 2 ), is dissolved in sulphuric acid in the
proportion of 20 kilos in 150 kilos of sulphuric acid and
electrolysed. The product is para-amido-phenol sulphonic
acid, C 6 H 4 (NH 2 )(OH) ; the reaction is supposed to take place
in two stages, with the intermediate formation of phenyl-
hydroxylamine, thus :
C 6 H 5 (N0 2 ) + 2 H 2 = C 6 H 5 (NH)(OH) + H 2 ;
Phenylhydroxylamine.
C 6 H 5 (NH)OH = C 6 H 4 (NH 2 )(OH).
Amido -phenol.
The ultimate product, para-amido-phenol sulphonic acid,
separates in crystals from the electrolyte and can be filtered
off through asbestos. In like manner, from ortho-nitrotol-
uene can be obtained ortho-amido-metacresol, and from
me ta-nitro toluene, meta-amido-ortho-cresol. Correspond-
ing amido derivatives can be prepared from dinitro-
benzene and dinitrotoluene. Such transformations are the
alphabet of industrial organic chemistry, and the only inter-
est or importance attaching to their execution by electrolysis
turns on questions of cost and yield. On these points no
information is available.
The flexibility of electrolytic processes for effecting organic
reactions is shown by the fact that two other products (dis-
tinct from amido-phenol) may be prepared by the reduction
of nitrobenzene :
346
ELECTROLYSIS OF ORGANIC COMPOUNDS
(1) In dilute acid solution aniline is formed
C 6 H 5 (N0 2 ) + H 6 = C 6 H 5 (NH 2 ) + 2 H 2 0.
(2) In alkaline solution azobenzene may be obtained
2C 6 H 5 (N0 2 ) + H 8 = C 6 H 5 NNC 6 H 5 + 4 H 2 0.
A large number of similar processes, many of which have been
patented, deal with the reduction of nitro-compounds to
the corresponding hydroxylamine and amido derivatives
reactions which are accomplished without difficulty by purely
chemical methods. Any advantage which may lie with the
electrolytic process will rest rather on the greater control
of the course of the reaction which an electrolytic process
may give, or on the avoidance of waste products and conse-
quent increase of yield, than on any novelty in the reaction
itself. In certain cases, however, electrolytic reduction of
organic substances takes a characteristic course. Thus
certain nitraldehydes of the aromatic series yield, not
amido compounds, as might be expected, but hydroxylamine
derivatives. The following examples may be given :
10 kilos of meta-nitrobenzaldehyde (C 6 H 4 (N0 2 )COH)
are dissolved in 150 kilos of sulphuric acid, and electrolysed
with a current having a pressure of 4 to 6 volts and a density
at the cathode of 6 to 7 amperes per square decimetre. When
the reduction is complete the electrolyte is diluted with
water ; a colourless substance is precipitated. This is an
anhydro-derivative of meta-aldehydephenylhydroxylamine,
C 6 H 4 NH(OH)COH. In like manner para-nitrobenzaldehyde
yields a product which is an hydroxylamine derivative,
though not in this case an anhydro-derivative. Such
products are utilised in the manufacture of colouring matters
and synthetic drugs ; their production is simply a step in
a long chain of reactions which is conveniently and economi-
cally accomplished electrolytically instead of chemically.
An example may be given of the direct production of a dye-
stuff by electrolytic means. Naphthazarine (alizarin black) is
dihydroxyanthraquinone, C 10 H 4 (OH) 2 2 ; it may be prepared
by reducing alpha-dinitronaphthalene by means of zinc
347
PRACTICAL ELECTRO-CHEMISTRY
in the presence of strong sulphuric acid. Equally it may
be made by electrolysing a solution of dinitronaphthalene
in strong sulphuric acid. This solution is placed in the
cathode compartment of the cell and electrolysed with a
current having a density of 15 amperes per square
decimetre.
Elbs has studied the conditions of reduction of nitroben-
zene with intent to obtain a high yield of aniline. When
sulphuric acid is used as a solvent for the nitrobenzene to
be electrolysed it may be regarded as serving a three-fold
use : (1) as a solvent, (2) as aiding conduction, (3) as bringing
about the transformation of phenylhydroxylamine, which
may be considered as the first product of reduction, into
para-amidophenol, C 6 H 4 (OH)NH 2 . Seeing that aniline,
C 6 H 5 (NH 2 ), is obtained as well as para-amidophenol, it
seemed possible by a modification of the conditions of elec-
trolysis to obtain this substance as the main resultant and
not merely as a by-product. When acetic acid is substituted
for sulphuric acid as a solvent the yield of aniline is considera-
bly increased ; an increase occurs also when a lead cathode
is substituted for one of platinum ; under these conditions
the quantity of para-amidophenol is correspondingly re-
duced. It appears from direct experiment that para-amido-
phenol is not reduced to aniline, whence it follows that the
use of a lead cathode must aid in determining the course
of the reduction of nitro-benzene to aniline instead of to
para-amido-phenol ; it is suggested that lead itself may
effect this reduction, much as iron does in the ordinary chem-
ical process of making aniline from nitrobenzene. The lead
oxidised and transformed into a salt (sulphate or acetate)
by the action of the nitrobenzene is promptly reduced by the
current and deposited as lead sponge, which again acts as
a reducing agent. Thus the formation of aniline may be truly
said to be effected by the action of the lead, in spite of the
fact that no appreciable quantity of lead is permanently
oxidised and dissolved. A zinc cathode will act in a similar
way, doubtless for the same reason. Such results may be
compared with the various products of reduction of nitric
348
ELECTROLYSIS OF ORGANIC COMPOUNDS
acid when treated with different metals nitrous oxide
with zinc, nitric oxide with copper, nitrous anhydride with
silver, and nitrogen peroxide with tin. The current may be
looked upon merely as a convenient method of bringing
into play reactions proper to the several metals which are
alternately oxidised and reduced.
A direct instance of this is afforded by the patented pro-
cess of Castner for the reduction of nitro-compounds in the
cathode compartment of the Castner-Kellner cell (see page
313). Here the substance to be reduced is exposed to the
action of sodium amalgam formed and continually renewed
electrolytically. A large number of investigations have been
made to determine the course of reaction in similar
processes of reduction or oxidation of organic compounds
by electrolytic means, but the discussion of these pertains
to the domain of a particular branch of organic chemistry
and is not cognate with the subject of this book.
Here it is proper to observe that the course of a reaction
maybe determined by adding to the electrolytic cell some
substance which is capable of alternate oxidation and
reduction, and will, in consequence, ensure that the electro-
lytic effect of the current is applied at a fixed pressure.
A substance of this kind will act as a sort of reducing
valve ; no considerable surplus of pressure can occur ; the
voltage is automatically maintained within small limits.
Salts of chromium, manganese, and cerium have been
used. Doubtless others which are labile, such as those
of thallium, mercury and cobalt might be employed.
Anthracene is the hydrocarbon from which anthraquinone
is produced and thence alizarin. The ordinary chemical
process is to oxidise anthracene with chromic acid in the
presence of sulphuric acid ; in fact, the customary method
of assaying crude anthracene is to submit it to such an
oxidation. Attempts have been made to oxidise anthracene
by electrolysing its solution in sulphuric acid. These have
not been particularly successful, and a sort of combination
of the two methods has been effected by using chromic
acid as an oxidant and, when it is reduced, regenerating
349
PRACTICAL ELECTRO-CHEMISTRY
it electrolytically. This with so labile a substance as
chromic acid must be done in a divided cell.
The same spent liquor containing chromic sulphate and
sulphuric acid is used in each compartment ; the chromic
sulphate in the anode solution is oxidised to chromic acid,
and in addition some sulphuric acid emigrates from the
cathode to the anode compartment. When oxidation is
complete the anode liquor is used to treat a fresh portion
of anthracene, and the cathode liquor is transferred to the
-anode compartment, a fresh portion of spent liquor being
placed in the cathode compartment. The same changes
occur ; the chromium is oxidised to chromic acid and the
solution in the anode compartment enriched with sulphuric
acid, migrating as before from the cathode compartment
and compensates for the depletion which the liquor, now in
the anode compartment, suffered when it was in the cathode
compartment. Most, of the Farbwerke Meister Lucius
und Briining, has patented a process for oxidising anthracene
in which a cerium salt is used as a carrier. The electrolyte,
which it seems is used without a diaphragm, .consists of a
20 per cent, solution of sulphuric acid, containing 2 per cent,
of cerium sulphate. The containing vessel is lead, and
serves as an anode. Any unattackable metal, e.g. lead, may be
used as the cathode. The anthracene to be oxidised is
added to this bath and well mixed by an agitator. The
operation is conducted at a temperature of 70-90 C. = 158-
194 F., rising towards the end to 100 C. = 212 F. ; the
current density is 50 amperes per square foot. The cerium
changes from the cerous Ce ui to the eerie C iv state, and the
completion of the process is known by the electrolyte remain-
ing yellow from the presence of the eerie salt. The use of a
cerium salt as an oxygen carrier has also been applied to
the preparation of naphthaquinone and phthalic acid from
naphthalene as the starting point.
Sometimes, however, it occurs that a reaction can be
brought about electrolytically which cannot be directly
accomplished chemically. Thus in the normal course of
oxidation of para-nitrotoluene, C 6 H,CH 3 (N0 3 ), para-nitro-
350
ELECTROLYSIS OF ORGANIC COMPOUNDS
benzole acid, C 6 H 4 (N0 2 )COOH, is produced ; by electro-
lysis, however, it is possible to obtain para-nitrobenzyl
alcohol, C 6 H 4 (N0 2 )CH 2 OH. An illustration of the use of
electrolysis in the preparation of synthetic dye-stuffs is
afforded by the oxidation of certain hydroxy acids of the
benzene series. The Badische Anilin und Soda Fabrik has
patented alternative processes for the manufacture of a
yellow dye-stuff from meta-dihydroxy-benzoic acid, C 6 H 3
(OH) 2 COOH. According to the chemical method 10
kilos of this substance are dissolved in 200 kilos of strong
sulphuric acid and treated with 15 kilos of ammonium
persulphate, the temperature being kept below 50 C. =
122 F. The reaction is allowed to proceed for 12 hours
and the mixture is thrown into 1,000 litres of cold water.
The colouring matter separates in yellow flocks and can be
filtered off and washed.
The corresponding electrolytic operation is conducted
as follows : 10 parts of meta-dihydroxybenzoic acid are
suspended in 40 parts of sulphuric acid of 50 B. (specific
gravity T53) ; the mixture is placed in the anode compart-
ment and subjected to the action of a current of 20 amperes
at a pressure of 8 volts. The current density is 20 amperes
per square decimetre. The product is identical with that
produced chemically. It is a fast yellow colour applicable
to both cotton and wool. Other hydroxy acids e.g. gallic
acid C fi H 2 (OH) 3 COOH, cresotic acid C 6 H 3 (OH)CH 3 COOH,
and hydroxy-benzoic acid (C 6 H 4 (OH)COOH) may be
similarly treated so as to yield analogous dye-stuffs.
One of the methods for preparing saccharin involves the
use of potassium permanganate in dilute neutral solution as
an oxidant.
An electrolytic method has been devised in which oxidation
is effected somewhat in the same way as that used in pre-
paring anthraquinone from anthracene, namely, by taking
advantage of a suitable carrier and continuously rein-
stating this in its more highly oxidised condition. In
the case of saccharin the substance to be oxidised is ortho-
toluene sulphonamide and the oxidising body is the same
351
PRACTICAL ELECTRO-CHEMISTRY
as that employed in the chemical process namely potassium
permanganate. The oxidation is effected in alkaline solu-
tion and a diaphragm is used. The regeneration of the per-
manganate is not complete, but nevertheless a large saving is
effected, about J the quantity necessary to effect the oxida-
tion unaided, being found sufficient for the preparation of
the saccharin.
Another instance of the electrolytic preparation of organic
compounds is afforded by the oxidation of isoeugenol to
vanillin. Eugenol is converted into iso-eugenol by treatment
with alkalies ; its alkaline solution is then exposed to oxida-
tion at the anode, a current density of 13 amperes per square
decimetre and a temperature of 60 C. = 140 F. being used.
The reaction may be expressed thus :
/O.CH 3
C 6 H 3 (OH)( +30
\HC : CH.CH 3
Iso-eugenol. / . CH 3
= C 6 H 3 (OH) / + CH 3 COOH.
\CHO
Vanillin. Acetic acid.
Vanillin is the odoriferous principle of vanilla, and has a
high price. It can be prepared from coniferine (Ci6H 22 8 )
by purely chemical methods. The success of its electrolytic
manufacture from eugenol is obviously a question of cost
and yield by the two processes. It must be remembered
also that a synthetic product is not always as marketable
as the natural material. There may be a real difference
due to the presence of an impurity in one or the other, or the
difference may be imaginary for imagination plays a great
part in trade but however this may be, the artificial body
has usually to win an uphill fight before it is accepted as on
A par with the native substance.
The electrolytic manufacture of iodoform has occupied
inventors. The ordinary chemical method for preparing
this body is by heating alcohol or acetone with caustic potash
and iodine, thus :
352
ELECTROLYSIS OF ORGANIC COMPOUNDS
C 2 H 5 OH +101+9 KOH = CHI 3 + K 2 C0 3 + 7 KI + 7 H 2 ;
Alcohol.
!(CH 3 ) 2 CO + 12 I +8 KOH = 2 CHI 3 + K 2 C0 3 +6 H 2 +6 KI.
Acetone.
The same changes can evidently be brought about by
electrolysing a warm solution of potassium iodide in the
presence of alcohol or acetone and water. Potassium iodide
electrolysed in the presence of water may be regarded as
potentially iodine and caustic potash, thus :
2KI + H 2 = KOH + H + I.
Seeing that alkali is necessary for the reaction which results
in the production of iodoform, and that it is formed at the
cathode, together with an equivalent of hydrogen which
would tend to reduce the iodoform or to combine with the
iodine to form HI, it is desirable to work with a diaphragm
and to provide a supply of alkali from without. Of course
the alkali formed in the cathode compartment can be with-
drawn and transferred to the anode compartment, the pro-
cess being thus made self-supporting. In an experiment
made by Elbs a platinum anode was immersed in a solution
consisting of 15 grammes of Na 2 C0 3 and 10 grammes of KI
in 100 c.c. of water and 20 c.c. of alcohol ; this was con-
tained in a porous cell and was thus separated from the cath-
ode compartment, which contained caustic soda solution
and a nickel cathode. The temperature was 70 C. = 158 F.
and the current density 1 ampere per square decimetre at
the anode. After a three hours' run a yield of 70 per cent,
of the calculated quantity of iodoform was obtained. The
chief by-product was sodium iodate It appears to be less
easy to prepare iodoform from acetone electrolytically.
The conditions have been studied by Abbot, who finds that
fair results are obtained if the acetone is added little by little
1 According to some authorities the reaction takes place accord-
ing to the equation :
(CH 3 ) 2 CO + 61 + 4KOH - CHI 3 + CH 3 COOK + 3 KI +3 H 2 O.
Probably the changes actually occurring are more complex than
is indicated by either statement.
353 AA
PRACTICAL ELECTRO-CHEMISTRY
to the anode compartment. In a laboratory experiment
the anode solution contained 6 grms. of sodium carbonate,
10 grms. of potassium iodide and 100 c.c. of water. To
this was added 5-5 c.c. of acetone at the rate of 0-5 c.c. per
10 minutes. The current density was 1 35 amperes per square
decimetre, and the temperature of the electrolyte 75 C. =
167 F. ; an output of 0-57 grm. per half-hour was obtained,
and the^ yield was 47 per cent, on the weight of acetone used.
Bromoform and chloroform can be prepared in a similar
manner.
For the manufacture of chloroform an apparatus has
been devised consisting of a leaden still which can be heated
by steam and contains an agitator armed with carbon plates
to serve as anode in a 20 per cent, solution of common salt.
The still itself acts as the cathode. Acetone is admitted at
the bottom of the still, and is converted by the joint action of
chlorine and caustic soda into chloroform. The reaction
may be regarded as occurring in two stages :
(1)(CH 3 ) 2 CO + 3C1 2 = CH 3 COCC1 3 + 3 HC1 ;
Chloracetone.
(2) CH 3 COCC1 3 + NaOH = CH 3 COONa + CHC1 3 .
Sodium acetate. Chloroform.
The chloroform is distilled off and collected in the usual
manner. It is stated that from 100 parts by weight of ace-
tone 180 parts of chloroform are obtained, as against a
theoretical yield of 206 parts. Assuming the substantial
correctness of this claim, it will be noted that only one of the
two methyl groups in the acetone is utilised for the produc-
tion of chloroform. 1
The preparation of an indigo vat for dyeing can be accom-
plished by reducing indigo to indigo- white by means of zinc
in alkaline solution. Experiments on the electrolytic re-
duction of indigo have shown that it takes place much more
readily when a solution of zinc oxide in caustic soda is used
1 Cf. the equations representing the reactions concerned in the
preparation of iodoform, p. 353.
354
ELECTROLYSIS OF ORGANIC COMPOUNDS
as the liquid at the cathode than when caustic soda alone is
used. In fact, it appears to be necessary to use the zinc as
an oxygen carrier, and thus ensure the reduction proceeding
to the desired point ; otherwise either the indigo is not fully
reduced or the reduction is carried a step farther than indigo-
white and the vat is spoiled.
The case is parallel to that already cited on page 348,
viz. the reduction of nitrobenzene to aniline by the aid of a
lead cathode.
There has been much systematic study of the course of elec-
trolysis and of the products obtained in the case of definite
classes of organic substances, such as the alcohols, the salts
of acids of the fatty series, salts of acids of the aromatic
series, nitro-compounds and the like, which will doubtless
form a starting point for many industrial processes in due
time. At present such work is of purely academic interest,
and special manuals such as Dr. W. Lob's Elektrolyse und
Elektrosynthese organischer Verbindungen must be consulted
for a knowledge of its details.
Sharply distinguished from this systematic enquiry are
certain processes which are almost wholly empirical, but
have nevertheless attained a sufficient measure of success
to justify a description.
In the purification of crude sugar juice, lime is commonly
used to neutralise organic acids and to precipitate albumin-
ous substances and colouring matter. It has been proposed
to accomplish this defecation by electrolysing the juice
between electrodes of zinc or aluminium. The anode is at-
tacked, giving a zinc or aluminium salt, and alkali is produced
from the alkaline salts naturally present in the juice ; the
products of the two electrodes intermingle, giving rise to a
precipitate of hydrated zinc oxide or alumina, which acts
as a defecating agent. It is stated that a few minutes'
treatment is effective.
Assuming that the defecation is better accomplished thus
than with lime, there appears to be no reason why the use
of hydroxides produced electrolytically should present any
advantage over the same substances prepared chemically.
355
PRACTICAL ELECTRO-CHEMISTRY
A similar process has been proposed in which lead anodes
are used, and the same remark applies.
Somewhat elaborate experiments by Baudry (Jahrbuch
fur Elektrochemie, 1897, 323) have shown that, when raw
juice from beets is defecated with a small quantity of lime
and then electrolysed with zinc anodes, a greater purifica-
tion is effected than with lime alone. A large consumption of
zinc and a considerable expenditure of electrical energy,
however, make the process unduly expensive. It does not
appear that a comparison has been made between the use
of zinc hydroxide made chemically as a defecating agent and
that of the same substance prepared electrolytically. Fail-
ing such data it is impossible to decide how much of the
advantage claimed arises from employing a defecating agent
other than lime and how much is due to the use of electrolysis.
Endeavours have been made to aid the purification of crude
sugar juice by treatment with ozone alone or aided by
electrolysis. The degree of purification attained is not
high, and the process offers little prospect of practical em-
ployment.
The process of tanning, which consists essentially in treat-
ing hides with an aqueous solution of tannin derived from
various barks, berries and other vegetable products, is one
of the slowest operations industrially carried out, being
comparable in this respect with the seasoning of timber or
the manufacture of white lead by the old Dutch or English
corrosion process.
This slowness is largely due to the difficulty with which
the tannin penetrates into the hide. As the penetration
progresses, the outer part of the hide becomes converted
into leather and is thereby made impervious, consequently
the rate of penetration decreases. Months of soaking in
the tan pit are, therefore, necessary for thick hides. Many
attempts have been made to hasten this absorption of tannin
by hide. The methods used include circulating the tan
liquor so that fresh portions are continually presented to the
hide, forcing the liquor through the hide by pressure, and
using strong aqueous extracts of tanning materials. It has
356
ELECTROLYSIS OF ORGANIC COMPOUNDS
been sought to attain the same object by passing a current of
electricity through the vat in which the hides are suspended.
One such process (Groth's) has been found to shorten the time
of tanning to a quarter of that necessary when no current
is used, and the leather is said to be unexceptionable. The
apparatus devised by Groth is designed to hasten tanning
by circulation of the tan liquor as well as by the use of elec-
tricity. The tan liquor is contained in a tank in which
is a frame carrying hides and capable of being moved to and
fro or rotated so as to bring the hides continuously into con-
tact with fresh liquor. Copper electrodes are placed at
the side of the tank. For a vat holding 1,500 gallons a cur-
rent of not more than 4 amperes is used.
The current density is not more than 0-1 ampere per
square foot of transverse section of the vat. With this mild
stimulus it was found that the rate of tanning was sixteen
times as fast as when the hides were simply immersed in the
tan liquor and allowed to be stationary, and four times
as fast as when the hides were moved and no current passed.
Considering the well-authenticated tests which have been
made, it is noteworthy that tanners at large will have nothing
to say to electric tanning. In the Worms and Bal process
(which was the forerunner of Groth's) the apparatus used is a
barrel of about 12,000 litres capacity taking a charge of
700 kilos of hide and 5,000 litres of oak-bark extract. The
electrodes attached to the inside of the drum are of copper. A
current of 11-5 amperes at a pressure of 74 volts is used.
Tanning is said to be complete in 48-144 hours, but the pro-
cess is somewhat violent, the leather suffering from the
mechanical pounding which it receives.
Another process, consisting essentially in passing a cur-
rent of 12 amperes at 60 volts between electrodes of nickel-
plated copper through a bath in which tanning liquor
was continually circulated by a pump, proved to be capable
of tanning heavy leather in about six days, the product being
not inferior to that prepared by the old process in twelve-
months.
Burtin dehairs the hides by suspending them in the or-
357
PRACTICAL ELECTRO-CHEMISTRY
dinary dehairing liquid consisting of size and arsenic and
passing a current for 15-20 minutes, reversing its direction
and continuing the treatment for an equal period. It is
stated that dehairing, which takes 10 days to 3 weeks by the
usual process, can in this manner be accomplished in an
hour to an hour and a half. The dehaired hide is then
electrically tanned. The inventor of the process also
prepares his tanning solution electrolytically.
358
SECTION IX
Power
Power
IN certain electro-chemical industries, such as the elec-
trolytic recovery of gold from cyanide solutions used
to extract its ores, in plating, and in refining as distinct
from winning -metals, the quantity of energy required is not
large.
The fact that in a large copper refinery some hundreds
of H.P. (or even a few thousands) may be utilized is a con-
tradiction to this statement, not real, but only apparent. The
huge size of modern copper refineries obscures the fact that
the energy needed per ton of copper handled is by no means
large. Thus on page 36 it is shown that, with a liberal
allowance for waste, a plant of 1,000 H.P., working day and
night for a year of 365 days, will give an output of 15,000
tons of copper an enormous amount of what is a relatively
costly metal.
In other electro-chemical industries, however, such as
the manufacture of caustic soda and chlorine, of sodium,
of aluminium, and of calcium carbide, the expenditure of
energy is extremely great. So large is it that a source of
cheap power is indispensable for these industries.
At present the cheapest form of power is that afforded
by moving water. A large steam plant deriving its energy
from cheap coal comes next. Water power is usually
obtainable only in mountainous regions difficult of access,
remote from supplies of raw materials and labour. Thus it
comes about that frequently when both raw materials and
labour are required in quantity it may be more remunera-
tive to use somewhat dear steam power at a spot where
PRACTICAL ELECTRO-CHEMISTRY
both are abundant than to seek cheaper power from wacer
in an industrial desert. This holds good to-day, when the
quantity of energy obtainable from coal by means of a
boiler and steam engine is not greater than 10 per cent. ;
it will apply with greater force when it is possible to extract
from coal something approaching a fair fraction of its energy
say 50 per cent. The problem of obtaining from carbon-
aceous fuel a large fraction of its total energy is the greatest
of those set before the modern technical investigator. By
present methods the loss is almost wholly in the steam
engine. The boiler gives a fair return of the heat put into it
say 70 per cent. The dynamo gives a good return of the
energy put into it say 95 per cent. The combined efficiency
of the two is 66-5 per cent. The rest of the loss, which brings
the efficiency of the combination down to something less
than 10 per cent., is due to the steam engine. Now when
suitable material, such as zinc, is oxidised in a battery, the
fraction of its energy which returns as electrical energy is
high, e.g. 90 per cent. But zinc is too costly a fuel to be
used for any but highly special purposes, where cost is a
secondary consideration. Therefore it has long been a
matter of endeavour to convert the energy of carbon
or carbonaceous fuel directly into electrical energy. There
have been many attempts to reach this goal some ill-con-
sidered and doomed to failure, others rational but unsuccess-
ful. The task is still unaccomplished and the problem
unsolved.
The fundamental difficulty in the way of constructing
a primary cell which shall yield electrical energy by the
oxidation of carbon instead of zinc depends on the fact that
carbon will not dissolve in any electrolyte by simple dis-
placement of the positive ion of that electrolyte. The sort
of reaction which must be sought if carbon is to be utilised
as the positive element in a primary cell may be stated as
follows. Suppose a carbon electrode immersed in fused
silica, and opposed to a platinum electrode immersed in
fused lead oxide. One may conceive the carbon being
dissolved at one end of this chain and lead being liberated
362
POWER
at the other, the balance of energy represented by the
difference between the heats of combination of oxygen with
carbon and lead appearing as electrical energy. If the car-
bon were immersed in fused lead oxide and opposed to a
platinum electrode it may be assumed that the combination
would be less effective, because of the oxidation of the carbon
being chemical and local instead of electrolytic. Such a
condition is comparable with a cell consisting of zinc and
platinum immersed in strong nitric acid. No doubt a
portion of the energy of the dissolving zinc would appear
as electrical energy, but the greater part would appear as
heat. Separate the zinc from the oxidant, as in a Grove's
cell, and the combination becomes efficient. If a successful
carbon cell is to be constructed on the lines of ordinary
primary cells using zinc, it must have the carbon dissolving
in a non-oxidising electrolyte and it or its equivalent being
oxidised at the other electrode by an oxidising electrolyte.
The difficulty of devising such a cell is enhanced by the fact
that the only practicable oxidant, air, is a gas, and the pro-
ducts of the oxidation of carbon, carbon monoxide and
carbon dioxide, are gases. These and like considerations
make the task of devising a rational carbon cell so difficult
that one may well believe that the solution of the problem
of converting the energy of carbonaceous fuel direct into
electrical energy will be on lines totally different from those
furnished by the analogy of the zinc primary cell.
The difficulty of using carbon as the attackable electrode
in a primary cell is not unique. It occurs with most non-
metals. Thus it is not easy to scheme a cell in which sulphur
shall furnish energy smoothly and completely by virtue of
its heat of combination with oxygen ; the same holds for
phosphorus. It is true that both these elements, and in-
deed carbon itself, will dissolve when made the attackable
electrode in an electrolyte consisting of hot concentrated
sulphuric acid, but the reaction in all cases is more or less
local and confined and does not yield a favourable return of
electrical energy.
Becquerel in 1855 seems to have been the first to observe
363
PRACTICAL ELECTRO-CHEMISTRY
that when a rod of carbon was immersed in fused nitre at
such a temperature as to cause its oxidation a current was
produced if an unattackable electrode was present, e.g.
the platinum vessel containing the nitre. This observation
was repeated by Jablochkorf in 1877, who constructed a cell
consisting of a cast-iron pot serving as the unattacked elec-
trode and containing fused nitre, in which hung a basket
of iron wire containing coke. The coke was oxidised at the
expense of the nitre and a current was produced ; the com-
bination is said to have given a pressure of 2-3 volts a
somewhat doubtful statement. This apparatus had the
considerable defect that the inevitable and wasteful local
chemical oxidation of the carbon was enhanced by local
electrolytic attack, due to the iron basket used to contain
the coke. It was, however, better than some inventions
of later date, in that it attempted to use coke instead of
plates of artificial carbon of impracticable cost.
Before proceeding to a further discussion of the carbon
cell a calculation of its possible output may be usefully
made. Carbon in being oxidised to C0 2 gives 96-96 Cal.,
i.e. 24-24 Cal. per gramme equivalent. This corresponds
with 96,540 coulombs at the pressure of 1-04 volts. There-
fore a cell in which carbon is oxidised by air cannot have a
higher E.M.F. than 1-04 volts. Zinc similarly oxidised will
give current at a pressure of 1-86 volts. In this respect the
carbon cell is inferior to one burning zinc, because it is
generally convenient to obtain current at a high pressure
to avoid the necessity of multiplying units, i.e. cells. But
when the total electrical energy, as distinct from the pressure
at which it is delivered, is considered, the superiority of a
carbon cell becomes manifest. One kilo of carbon gives
8,080 Cal. as against 1,329 Cal. for 1 kilo of zinc, i.e. a given
weight of carbon will give more than six times as much
energy as an equal weight of zinc. Zinc is at least twenty
times as dear as carbon in the form of coal, wherefore a given
quantity of energy could be produced from carbon in a prim-
ary cell for T ^ () of the cost of the same quantity of energy
from zinc, assuming identical efficiency. The disadvantage
364
POWER
of a slightly lower voltage is insignificant compared with
this great economy.
In all the early experiments and there are many on
the production of electrical energy by the oxidation of carbon
and other non-metals there is a sad lack of quantitative
records. The voltage of a given cell is generally stated, but
the output of current for a given consumption of electrode
almost never. It is, therefore, impossible to say how far
the experiments approached towards a practicable cell ;
it is certain that they never came within reasonable distance,
as otherwise that cell would be in use now.
Other oxidants than nitre have been used in the carbon
cell. Barium peroxide will serve, and has the advantage
of being capable of regeneration by air from the barium
monoxide to which it is reduced. Copper or lead oxide
would not act in this manner apparently, because the metals
which are produced by their reduction establish direct
metallic conduction between the electrodes and prevent the
progress of the electrolysis. When, however, the carbon
is not directly in contact with the oxide, but is covered with
a layer of fused salt, e.g. potassium carbonate, the con-
ditions necessary for electrolytic dissolution are re-estab-
lished.
One of the latest attempts to devise a practicable carbon
cell has been made by W. W. Jacques. The chief features
of the cell proposed are shown in Fig. 69. A is a carbon
electrode, immersed in fused caustic soda contained in an
iron pot B, set in a furnace (not shown) so that the alkali
may be kept liquid. The pot serves as the unattacked
electrode. Oxygen is supplied in the form of air blown in
through the pipe c, ending in the perforated ring D. Sur-
plus air and the gaseous products of oxidation escape by the
vent E in the cover F, which is of insulating material,
e.g. fire-clay.
The carbon is said to be oxidised to carbon dioxide by the
finely divided air issuing from the ring and to yield its
energy as current. It is claimed that from a battery of 100
cells a current of 16 amperes at a pressure of 90 volts was
365
PRACTICAL ELECTRO-CHEMISTRY
obtained for 18} hours with a consumption of 8 pounds of
carbon. This corresponds with an efficiency of 79 per cent.,
reckoned on the amount of carbon consumed. Even
accepting these figures the true efficiency of the cell cannot
be stated thus, because a large quantity of heat is required
to keep the electrolyte fused and a good deal of energy
is needed to drive air through the molten mass. But the
fact of the matter is that the cell is a chimera. Various
elaborate calculations and experiments have been published
tending to attack it in detail ; they are* unnecessary, because
FIG. 69.
the device is wrong in principle. There is no evidence that
the current is due to oxidation of carbon ; such evidence
as there is goes to show that it is due to a thermo-electric
action and occurs as well with a non-consumable electrode.
Next, if it be supposed that the energy is produced by the
oxidation of carbon it may be rightly concluded that the
product of oxidation, C0 2 , will be absorbed by the electrolyte,
caustic soda, which will be speedily spoiled. Thirdly, the
carbon proposed to be used is battery carbon, i.e. carbon
in the form of expensive manufactured electrodes. These,
366
POWER
even if consumed economically, would be a costly form of
energy. The best proof of the correctness of these strictures
is found in the fact that the Jacques cell, although much
extolled at the time of its invention by the lay and the less
intelligent part of the technical press, is extinct.
The only other cell which attempted with any plausibility
to convert the chemical energy of carbon directly into elec-
trical energy is that devised by Borchers. This was the
outcome of a luminous and exact dissertation by Ostwald,
and was in its inception an honest attempt to follow the
principles laid down by that great chemist. Ostwald's
pronouncement is sufficiently fundamental to demand re-
production here. He indicated with clarity and precision
that direct chemical action is not adapted for the production
of electrical energy ; that if the reaction on which the pro-
duction of energy ultimately depends is caused to occur on
the spot where is the source of energy, e.g. the dissolving
electrode, the energy evolved will be as heat and not as
electricity. An experiment illustrates this point fully.
Two vessels are filled with a solution of potassium sulphate
and are put into electrolytic connection by means of a syphon.
In one vessel is placed a rod of zinc and in the other a rod
of platinum. On connecting these electrodes through a
galvanometer a current passes momentarily and then ceases
because the zinc cannot continuously dissolve in such a
medium and give up its energy. In order to make the
current continue it is necessary to provide an acid which
will dissolve the zinc. Now comes the question : into which of
the two vessels shall the acid (e.g. sulphuric acid) be poured ?
Obviously (and erroneously) into that containing the zinc ;
correctly (and evidently when the evidence is weighed)
into that containing the platinum. The zinc dissolving
from the zinc electrode traverses the electrolyte and ap-
pears in the form of its equivalent of hydrogen at the plat-
inum electrode. The zinc may be regarded as becoming
ionised, each of its ions bearing a positive charge, and trans-
ferring this charge through the electrolyte from ion to ion,
ultimately neutralising the charge of a hydrogen ion negative
367
PRACTICAL ELECTRO-CHEMISTRY
to its own, deionising the hydrogen, and causing it to appear
in the ordinary molecular state as a gas at the platinum
electrode. The fact that the connection between the elec-
trodes consists of an electrolyte containing ions neither of
zinc nor of hydrogen is immaterial ; the fate of the zinc
at one end, and the ultimate product (hydrogen) at the
other, alone need to be regarded for the purpose of the present
case. It will be observed that when, as in this experiment,
the acid is in the compartment remote from the zinc, dis-
solution of the zinc is dependent on the passage or production
of a current, and is not local and wasted in the liberation
of heat.
The broad fact that the action on the attacked electrode
should be, as it were, at a distance, leads to the conclusion
that cells of the Jablochkoff type, consisting of carbon,
opposed to an unattackable electrode in a strongly oxidising
electrode, such as nitre, are wrong in principle. The carbon
should dissolve in a non-oxidising electrolyte, and it, or
its product, should be oxidised by an oxidising electrolyte
at the other electrode. To return to our old illustration :
it is no doubt possible to obtain a current from a couple of
zinc and platinum in strong nitric acid, but the combination
is absurd. The nitric acid has to perform two functions :
(1) that of a simple solvent at the surface of the zinc, and
(2) that of an oxidant of the zinc or its equivalent (a depolar-
iser in the old phraseology) at the surface of the platinum.
Incidentally there is tumultuous and wasteful local chemical
action of the nitric acid as an oxidant on the zinc. For the
proper understanding of such questions nothing is needed
but a sound chemical instinct ; this is, unfortunately, rare,
and its absence accounts for many errors. Ostwald has
gone beyond his negative criticism of the carbon cell as it
is, and has indicated the lines on which its construction
should be attempted. " The carbon cell of the future,"
he says, "should have the oxidising agent in the place
where the carbon is not " ; this oxidising agent must be
either the oxygen of the air or some carrier thereof. Such
A cell will work precisely like an ordinary furnace. On one
368
POWER
side coal will be thrown in, and on the other air will be
introduced, energy and C0 2 being the products. Between
the coal and the oxygen must be an electrolyte which will
suffer no permanent change, and can be used continuously
to bring about electrolytically the oxidation of the carbon.
Fired by these beautiful and exact ideas, Borchers attempted
to devise a cell for obtaining electrical energy direct from car-
bon, or at least carbon partially oxidised. Carbon monoxide,
in being oxidised to C0 2 , yields about two-thirds of the total
quantity of energy obtained by the complete oxidation of
carbon to CO 2 . CO is soluble in cuprous chloride, forming
therewith a loose compound (Cu 2 CLCO). Oxygen in the
presence of an acid, e.g. HC1, is capable of oxidising cuprous
chloride to cupric chloride. Here, then, are all the ele-
ments of success. A cell consisting of two carbon electrodes
immersed in an acid solution of cuprous chloride and sup-
plied, the one with CO and the other with O (or air), might
be expected to yield a current at the expense of the CO and
O, and with no permanent change of the electrolyte. A
cell constructed on these lines gave a feeble current, which
was slightly increased by increasing the surface of contact
between gas and liquid by surrounding the electrodes with
coke. When copper electrodes were substituted for elec-
trodes of carbon, a somewhat better result was obtained,
but the results put forward tend to show that the current
was produced by the dissolution of the copper electrodes
rather than by the oxidation of CO. Borchers brings for-
ward almost no quantitative .evidence, especially concerning
the consumption of CO and production of C0 2 . He claims
an efficiency of 27 per cent., but this claim appears to be
based on an observed maximum voltage of 0-4 volt, as
compared with 1-47 volts, the calculated maximum for the
equation = CO + = CO 2 . Seeing that no data are given
respecting the consumption of CO necessary to produce
the feeble current (0-008 ampere) which could be maintained
at this pressure, it is evident that the claim for a 27 per
cent, efficiency is groundless. Throughout the investigation
the evidence adduced is weak and inconclusive from the
369 BB
PRACTICAL ELECTRO-CHEMISTRY
chemical and quantitative side. There is, for example,
no attempt to measure the consumption of CO, to prove that
it is actually oxidised to C0 2 , or to show that the source
of current is not merely the oxidation of Cu 2 Cl 2 by air. Later
attempts, by the use of copper electrodes and the like,
to attain a better result are still more indecisive, because
they import questions (such as the dissolution of the copper)
other than the plain issue, " Do CO and O unite electrolytic-
ally with the production of current when supplied to two
unattackable electrodes immersed in a solution of cuprous
chloride ? if so, what is the efficiency of the combination ? "
Direct experiments by R. Mond with two carbon electrodes
immersed in a solution of cuprous chloride and supplied,
the one with CO and the other with air, showed that the vol-
tage of the combination was only 0-0015 volt. This most
destructive observation has never been explained or refuted
by Borchers, and until it is his cell must be considered as
based on an illusion. This is the" situation of the only
earnest attempt to follow a course of enquiry consonant
with Ostwald's dicta, and at the present time the
Borchers cell may be dismissed as a mistake.
It is clear that, if it is attempted to obtain electrical
energy direct from carbon by methods analogous to those
used for obtaining electrical energy direct from zinc in a
a primary cell, some plan must be found whereby carbon
can be dissolved in an electrolyte in such a way as to form
ions. The balance of evidence goes to show that carbon
has not been thus dissolved to form -ions, but, nevertheless,
some ground exists for maintaining a contrary opinion. Dr.
Coehn has called attention to the work of Bartoli and Papa-
sogli, and has extended the line of enquiry there indicated.
Bartoli and Papasogli observed that when a current is
passed between carbon electrodes in dilute sulphuric acid
the anode is not quite unattacked, but takes part in the pro-
cess of electrolysis, as is witnessed by the fact that CO and
j0 2 , as well as 0, appear as anode products. By varying
the concentration and temperature of the acid and the
density of the current, Coehn succeeded in obtaining con-
370
POWER
ditions in which the carbon was consumed, with the pro-
duction at the anode, no longer of oxygen, but of a mixture
containing 70 per cent. CO 2 , about 30 per cent. CO, and
not more than 1 per cent. O. During the electrolysis the
acid became red-brown in colour, and evidently contained
carbonaceous matter in solution ; the gradual destruction
of the anode is due, not to mere disintegration, but to actual
dissolution of the carbon. When electrolysis is continued,
using in such a solution a carbon anode and a platinum
cathode, a black deposit appears on the cathode. Coehn
has succeeded in collecting a small quantity of this, and
finds that it consists of carbon, with hydrogen and oxygen
in proportion to form water. He is disposed to regard it
as an hydrated form of carbon, and to consider that he has
succeeded in effecting the electro-deposition of carbon ;
hence that carbon ions are formed under the conditions
of his experiment. These interesting observations may be
recorded, but the deductions drawn from them must be
received with some reserve. Even if the deposit is an
hydrated form of carbon, it by no means follows inevitably
that carbon ions are present in the electrolyte and are de-
prived of their charges and deposited in the usual way as
elementary carbon. It is quite as likely that the dissolution
of the carbon anode forms complex organic substances, which
by reduction at the cathode yield highly condensed car-
bohydrates of the general form C m H 2n O n , such as the body
Ci 2 H 6 3 , said to be left in the carbonaceous residue from
the dissolution of highly carburetted iron (e.g. white cast
iron) in cupric chloride solution. It will be observed that
there are here two distinct questions. The first is whether
carbon will dissolve in sulphuric acid to form ions ; it is
indifferent for the purpose of this enquiry whether the ions
are formed by the spontaneous dissolution of the carbon
with the production of current, or by the enforced dissolu-
tion of the carbon by the impression on it of a current from
without. This question must be considered undecided ;
the balance of evidence is on the negative side. The second
question is whether carbon under these conditions dissolv-
371
PRACTICAL ELECTRO-CHEMISTRY
ing in sulphuric acid can (whether it forms simple ions or
not) act as a positive plate and produce electrical energy.
Direct experiment by Coehn goes to show that this is pos-
sible. When a plate of carbon is opposed to one of lead
peroxide in sulphuric acid it gives a constant current until
the lead peroxide is reduced or the carbon consumed. No
data are available as to the output of this combination per
equivalent of carbon consumed. The efficiency is probably
not high, and in any case the combination is not a practi-
cable means of consuming carbon for the production of
electrical energy on a large scale. There have been many
other attempts to devise cells which shall dissolve carbon
and render its energy electrically. With none of them has
any real success been attained. In the greater number
there has not been even an attempt to show success ; all
inventors have shrunk from recording the two factors needed
to judge of the efficiency of the cell, viz. the consumption
of the carbon per unit of current and the pressure at which
the current is delivered. Many investigators seem to think
that, if they show their cell to have a voltage of 0-7 on open
circuit, or through a high resistance when the calculated
voltage is approximately 1, the cell has an efficiency of 70
per cent., the current per unit of material consumed being
ignored. The fallacy is the converse of that frequent in
the description of electrolytic processes, in which it is com-
mon to find the efficiency stated in terms referring solely
to the output per unit of current, irrespective of the pressure
at which that current is delivered. In either case the error
is sufficiently obvious and gross.
Gas cells of the type of Grove's gas cell have also been
tried. In the Grove gas cell, hydrogen and oxygen are fed
to platinum electrodes, which are platinised and partly
immersed in acidulated water. By reason of the power
of platinum, especially when finely divided, to condense
gases in its pores, the two gases are brought into such inti-
mate contact at once with the electrode and the electrolyte
that they unite electrolytically and produce a current,
The possibilities of the cell are great, and an attempt has
372
POWER
been made to realise them by Mond and Langer, who have
striven to improve the cell mechanically so as to economise
platinum and to use purified water gas as a source of hydrogen.
It was found possible to construct a cell, having 700 square
centimetres of active surface and containing only 0-35
gramme of sheet platinum and 1 gramme of platinum black,
which yielded a current of 2 to 2- 5 amperes at a pressure of
0-73 volt, and gave an energy efficiency of 50 per cent.
Although ingenuity and perseverance have been lavished
on it, the Mond-Langer cell has failed to achieve any practi-
cal success.
The roundabout conversion into electrical energy of the
chemical energy of carbon is represented by all ordinary
primary cells using zinc, which metal has been reduced from
its oxide by coal. The energy efficiency is very low, say
2J per cent., and the money efficiency greatly lower, e.g. less
than 1 per cent. Now it may be possible to utilise in some
circuitous way the energy of carbon more efficiently than
can be done with zinc as an intermediary, and Reed has
sketched such a method, which may be summarised thus.
A current is obtained from cells supplied by a solution of
sulphur dioxide (S0 2 ) opposed to one of sulphuretted hydro-
gen (H 2 S) ; the electrodes are of inert material, e.g. platinum
(or carbon). The combination of S0 2 and H 2 S gives as its
chief products sulphur and water, thus :
SO 2 + 2 H 2 S = 2 H 2 + S 3 ,
the energy evolved being obtainable as electrical energy.
A constant supply of S0 2 and H 2 S can theoretically be ob-
tained by a cycle of reactions, needing for its realisation
nothing but a limited stock of sulphur and water, on which
is impressed at intervals the energy represented by the
oxidation of carbon. The requirements of the cycle are
that sulphur shall be burned in air, the S0 2 sent to the elec-
trolytic cell, and the heat used to induce the formation of CS 2
from C and S, and H 2 S from CS 2 and H 2 O, the carbon being
thereby oxidised to C0 2 . The H 2 S is then sent to the elec-
trolytic cell, where, reacting with the S0 2 , it regenerates
373
PRACTICAL ELECTRO-CHEMISTRY
sulphur; this is collected and again burned at the first
stage of the cycle. For the details of the idea, the reader is
referred to Reed's paper, " The Transformation of the Energy
of Carbon into other Available Forms," appearing in The
Electrical World, xxxviii., 1896, page 44. The various
steps mentioned above lead ultimately to the formation
of C0 2 as the end product of the circuitous oxidation of
carbon, with the calculated production of 61 per cent, of
the total energy thus liberated as electrical energy. The
sulphur and water are perpetually oxidised and decom-
posed, and are merely intermediaries. The CS 2 , H 2 S,
and S0 2 are still more ephemeral intermediaries. The
whole scheme is sound and philosophical, but hardly to
be realised in practice.
It will be seen from this brief sketch that the present
position of the problem of converting the energy of carbon
into electrical energy by means other than the boiler engine
and dynamo is one of attempt, not of achievement. Much
has been done to prepare the way for final success ; of
practical success at present there is absolutely none. The
enormous importance of the solution of this problem must
be my excuse for the space which I have given to its consider-
ation.
Returning from the possibilities of the future to the
accomplished facts of the present, let us examine the ques-
tion of the cost of electrical energy under different local
conditions.
WATER POWER
A large waterfall is the cheapest source of .power. An
artificial fall of water, such as may be obtained by impounding
the head waters of a river and conveying the collected water
to a lower point in a closed channel, such as a steel pipe,
comes next in order of merit. The power station at Niagara
Falls is a type of the first. Here a canal is cut from the
river, above the falls, to the power house. In this canal are
the intakes of large steel pipes which descend to the bot-
374
POWER
torn of the turbine pit, which has a depth somewhat less
than the height of the falls. The water passes from these
pipes through the turbines to the tail race, which is carried
out at a point below the falls. Thus the whole head of
water represented by the height of the falls is utilised without
the employment of any great length of steel main. A typical
example of the other mode of construction is afforded by
the power station at Brieg, on the Swiss side of the Simplon
tunnel. Some miles above Brieg is the glacier from which
the Rhone issues. The river flows torrentially down the
valley, but there is no definite waterfall. A portion of the
river is impounded at the glacier end, and is conveyed in
steel pipes along the course of the river and delivered to
turbines at the power house. The head is of course repre-
sented by the difference in level of the upper and lower end
of the pipe. The turbines are used to drive dynamos which
supply electrical energy representing a large fraction of the
total calculated energy of the falling water. Thus, if the
efficiency of the turbine is taken at 70 per cent., and that of
the dynamo at .90 per cent., the joint efficiency of the plant
will be 63 per cent, at the terminals of the dynamo. It is
often found necessary to transmit current to some distance,
and for this purpose that supplied by the dynamo maybe sent
into a step-up transformer, transmitted at a high pressure,
and reconverted into current at a low or moderate pressure
suitable for the work in hand by means of a step-down
transformer. The expenditure for capital sunk in the trans-
formers, together with that represented by their joint losses,
is smaller than that needed to cover the interest on the
capital sunk in a copper conductor of large section at
least when the distance of transmission is considerable.
Thus it comes about that the process of converting low-
pressure current into its equivalent of high- pressure current,
transmitting the current at high-pressure, and re transform-
ing it to low-pressure current, complicated as it sounds, may
be rational and economical.
The cost of water power naturally varies according to
local circumstances. Where the engineering difficulties in
375
PRACTICAL ELECTRO-CHEMISTRY
impounding the water and utilising it are small, the cost
per H.P. year, allowing fof interest on and depreciation of
plant, may be as low as 2 to 3. It must not be concluded
that power to be acquired at the rate of 2 per H.P. year is
necessarily twice as cheap as power at 4 per H.P. year.
The value of the power clearly depends on its prospect of
being commercially utilised, and since the ordinary object
of these large water-power plants is to manufacture some
chemical product, it is evident that the value of a given
plant depends not only on its inherent cheapness, but on its
accessibility. Raw materials must be brought to the spot,
and finished goods must be taken away ; local labour must
be obtained. Generally speaking, the cost of all means of
doing the same thing becomes ultimately identical. Power
from a waterfall is at present cheaper in money than power
derived from coal first, because its inherent value is less
understood ; secondly, because its utilisation involves a
heavy expenditure of capital, a return on which is depen-
dent on the establishment of novel industries, and thirdly,
because it has to offer some attraction to the user of power
to induce him to leave a known manufacturing centre for
a wilderness, access to which for his goods is difficult and
expensive. An estimate based on actual expenditure is
afforded by the calculated cost of power from the Lachine
Rapids on the St. Lawrence River, near Montreal. The
power house is designed for the production of about 20,000
H.P. The total capital cost is taken at 222,653, i.e.
11 3s per H.P. Interest and depreciation on this at 10 per
cent, will equal 1 2s., and to this must be added a sum for
operating expenses of 9$., making for the H.P. year 1 11s.
This estimate rests on the assumption that the whole of
the 20,000 H.P. will be needed day and night for 365 days
per year, a condition of things obtaining in electro-chemi-
cal manufacture. For intermittent supply, such as that
required for lighting and traction, the cost would be greater,
because interest and other permanent charges run on while
no return takes place.
With steam the cost per H.P. year is higher. A modern
376
POWER
plant of not less than 1,000 H.P., using coal of fair quality
costing 85. per ton, may succeed in producing power at about
5 per H. P. year (reckoned at the engine shaft), correspond-
ing with about 7 per H.P. year of electrical energy at the
terminals of the dynamo. A plant to work at this low cost
must be exceptionally well placed ; under less favourable
conditions the cost of an electrical H.P. year will approach
10. In all these cases the cost is inclusive, due allowance
having been made for interest, depreciation, and the like.
Broadly it may be taken that with water power a normal
figure is 4 per H.P. year ; a good figure may be taken
as 2 10<s. per H.P. year, and an unusually good figure as
1 105. per H.P. year. In all cases it is assumed that the
plant will be driven day and night for seven days a week,
and for as nearly 365 days a year as need for cleaning and
repairs will admit. Under modern conditions the com-
fortable, old-fashioned plan of periodical pauses is as ob-
solete as the ancient military method of going into winter
quarters. It is probable that for large installations a power
plant consisting of gas engines driven by producer gas will
be more economical than a good steam plant. In this
case a portion of the nitrogen of the coal used in the pro-
ducers may be recovered as ammonium sulphate, and this
turns the balance of advantage on the side of the gas engine.
Failing such by-product, the advantage is less certain. The
case is different when the gas is ready made as occurs
with blast furnaces and coke ovens. There the gas engine
is certainly the better.
Since writing this chapter some five years ago, I have little
to add. The carbon cell is still in nubibus. The costs given
for water and steam power are representative. The tendency
now is to use for large powers dependent on fuel either
steam turbines or gas engines fed with producer gas ; the
day of the gas turbine is not yet.
377
Index
ACETYLENE from Calcium Carbide,
209, 225
Acheson, Mr. E. C., discoverer
of Carborundum,
his process for
manufacturing,
226 et seq.
cited on Artificial Graphite, 232,
and on Siloxicon,
230
Aciertype, definition of, 281
Acker Cell, process using, for Elec-
trolytic manu-
facture of Caustic
Soda, etc., 298,
324
Alizarin, Electro-chemical produc-
tion of, 349
Black, see Naphthazarine
Alkali, Chlorine, and their Products,
289-339
Electrolytic Manufacture of
Cost of Plant for, and Standard
for judging Pro-
cesses of, 291
Present position of, summar-
ized, 323-4
Processes using Dissolved Salt
as an Electrolyte,
301 et seq.
Bell gravity system, 305
Castner-Kellner, 312
Electro-Chemical Co., 302
Hargreaves-Bird, 305
Le Sueur, 320
Rhodin, 319
Solvay, 318
Processes using a Fused Elec-
trolyte, 292 et
seq.
Acker, 298
Borchers, 299
Hulin, 295
Vautin, 293
Products
Caustic Soda, Processes for,
290-1 et seq. Cost
in relation to, and
Energy required
for, 290, 291
Alkali
Electrolytic Manufacture of
Products other than Caustic-
Soda, etc., and
Processes for, 324
et seq.
Bleaching Liquor
Kellner's, 334
Schlickerts', 330
Siemens-Halske, 329
Caustic Potash, 324-5
Chlorates, 326, 332
National Electrical Co.'s
cell for, 337
Hypochlorites, 327
Ozone, 338
Potassium Chloride, output
and cost of, 325
Perchlorate, 338
General Chemical considerations
concerning, 289
Alkaline earth Metals, Silicides of,
326
Alloys, (see Magnalium, and Vana-
dium), Electro-
deposition of, 205
Production of constituents of
by the Electric
Furnace, 193-206
Gold, Silver, and Copper
Electrolytic Refining of, Pro-
cesses for
Borchers, 111
Dietzel, 6.
Silver and Cadmium, Electroly-
tically deposited,
268
Sodium and Lead, Borchers'
apparatus for
making, 299
Alumina, fused in the Electric
Furnace, pro-
able uses for, 238
Impurities in, difficulties caused
by, 166-7
Aluminium
Chemical methods of Winning
Castner's, 158
Deville's, 157-9, 172
others sought for, 172 et seq
379
INDEX
Aluminium
Commercial,
Cost of Production, 174
Impurities in, 166-7, 179
Specific gravity of, 177
Uses for, as
Constituent of Alloys, 178
Industrial Metal for Small
Works, etc., 177-
8
Material for Electrical Con-
ductors, 178
Reducing agent, 177
Difficulties in Zinc-Plating of, 280
Electrolytic Production of, 157
Cost of, 174-7
Drawbacks to, 171
Electrolyte for, ib.
Energy requisite, 174-5
Plant for, 174-5
Processes for,
Hall, 167, 170
Heroult, 159, 169, 170
Minet-Bernard, 171
Tucker and Moody's, (ex-
perimental), 174
preferred to Magnesium for
various uses, 183
in the Production of Chromium
203-4
in relation to Copper, as to price
in weight, 176,
and in bulk, 177
Rules for manufacture of, formu-
lated by Hunt,
169
Specific Gravity of, compared
with that of Cry-
olite, 170n
in Zinc Amalgam, 151, 152
Aluminium Bronze, production of
by Cowles' Fur-
nace, 195
Sulphide, difficulties in manu-
facturing, 172-3
Aluminium and Copper Wire, Spe-
cific gravity of,
compared, 178
Aluminium - Industrie - Aktien - Ge-
sellschaft, Her-
oult process used
by, 159, large
output of, 164-5,
sulphide method
(alleged), of, 173
Amalgamation process of Gold-
extraction, 195
Andreoli's process for Electrolyticre-
covery of Gold, 99
Aniline, Electrolytic production
of, 348
Anions, defined, 4
Anode(s) (see also Electrodes),
defined, 4
for Electrolytic Deposition of,
Nickel, 113,
(Foerster's), 117
for Electrotyping, 259
for Gold Plating, 269
for Silver Plating, 267
for Refining of Copper, 32
for Winning of
Aluminium, drawbacks to, 171
Antimony, 128, 130
Electrolytic Iron (Burgess and
Hambuecher's process), 283
Zinc, 142
Anode Sludge of
Copper, 36
Composition of, 41
Wor king-up of, 56
Gold, 101, 102
Impurities in, 106-7
Lead, composition of, 90
Nickel, 115, 118-9
Anthracene, Electro - Chemical
manufacture of, for
production of Ali-
zarin, 349
Electrolytic Oxide of, 350
Most's process for, ib.
Antimony,
Analyses of
Dry refined, 129 '
Electrolytic refined and un-
refined, ib.
Aspect of in (a), Ingot form, (6),
Stripped from Ca-
thodes, 130
Chief ore of, reduction of, to
metal, 128
in Electrolyte, productive of
Spongy Zinc, 136
Electrolytic Winning of,
Processes for
Borchers', 130
Izart's, 131
Siemens-Halske, 128
results of, 129
defects in existing processes,
130, 13f
Aqueous Solutions, Winning and
Refining Metals in,
31-153
380
INDEX
Arsenic in Electrolyte, productive
of Spongy Zinc,
131
Artificial Graphite, production of,
by the Electric
Furnace, 231-3,
the Acheson pro-
cess for, 232
Uses of, 231-2
Ashcroft process for Electrolytic
Winning of Zinc
from mixed ores,
142, as worked at
Cockle Creek, 144-5
and Swinburne process for the
same, from Sulphide
ores, (partially Elec-
trolytic), 145-6, ap-
paratus for, 150
BADISCHE Anilin and Soda Fabrik,
electrolytic process
for manufacturing
yellow dye stuff,
351
Balbach Smelting and Refining
Co., Nickel refining
by, thick sheets ob-
tained, 113 ; com-
position of Crude
and Refined Nickel
employed, 120
Barium Carbide, 225 n.
Bartoli and Papasogli's observa-
tions on behaviour
of Carbon Elec-
trode under Elec-
trolysis, 370
Bath for Electroplating and Elec-
trotyping, 261-3
Gold plating, 269
Silver plating, 261, 266
Cyanide in, 268
Baudry's experiments in Electro-
lytic purification of
Sugar, 356
Bauxite, as a source of Aluminium,
165
Composition of, ib.
Bayer's Patent for Producing Nitro-
compounds, 346
Becquerel's observations on Carbon
and fused Nitre,
363-4
Bell gravity method for making
Alkali and bleach,
305
Mercury cell, principles of, 320
Bixhof and Thiemann, experiments
of, in Electrolytic
preparation of pure
Nickel and pure
Cobalt, 114, 115
Bleach, see Alkali, Chlorine and
their Products
Bleaching-liquor,
Electrolytic processes for
manufacture of,
Schiickert or Elektrizitats
Aktiengesellschaft,
330, 336
Siemens and Halske, 329
Powder,Electrolytic manufacture
of, cost in relation
to, 291
Bolton, Messrs. Thomas and Sons,
Nickel refining by,
thick sheets ob-
tained, 113
Borchers' Apparatus for Electro-
lysis of Fused Zinc
Chloride, 149-50
Carbon Cell, 367, 369
Mond's expei-iments on, 370
Process for Electrolytic
Manufacture of
Alkali, Chlorine, etc., 299
Refining of
Alloys of Gold, Silver, and
Copper, 111
Lead, 92-4
Winning of
Antimony, 130
cited on his process for Electro-
lytic manufacture
of Alkali, Chlorine,
etc., 300-1
Suggestion by, as to Electric
Furnaces, 217
British Aluminium Co., Foyers,
Aluminium manu-
factured by, on the
Heroult process,
159, raw material
for, methods of
treating, 165
Bronze, see Aluminium Bronze
Bucherer's patents for producing
Aluminium Sulph-
ide, 173
Burgess and Hambuecher, manu-
facture of Electro-
lytic Iron by, 2&3
Burtin's process for -Electrolytic-
dehairing and Tan-
ning, 357-8
381
INDEX
CADMIUM, in Alloy with Silver, 268,
for Plating, 270,
286
Calcium Carbide
Electro-chemically produced, 24
Produced by the Electric Furnace,
208-225
Cost of, 222-5
Energy requisite for, 220- 1
Raw material for, 220
Specific gravity of, 209
Calcium Silicide, and its probable
uses, 237
Calculating Output in Electrolytic
Processes, method
of, 17
Canadian Copper Co., process used
by, for Electrolytic
Refining of Nickel
from bessemerized
matte, 121-2
Government Commission on Iron
and Steel processes
in Europe, report
of, 242
Carbides, produced by the Electric
Furnace,
Artificial Graphite, 231-3
Calcium Carbide, 208-25
Silicon Carbide, 208, 225-30
Siloxicon, 230-1
Carbon,
Borchers cell for obtaining Elec-
trical energy from,
367, 369-70
Cells, see under Cells
Dissolution of, in an Electrolyte,
Coehn's
experi-
ments on, 370-2
Electro-chemical transformation
of form in, Moissan
cited on, 199
Energy of, Reed's method for
utilizing, 373-4
Carbon and fused Nitre, Becquerel's
observations on
363-4
Carbon Boride, hardness of, and
possible uses, 234
Uisulphide, production of, by
the Electric Fur-
nace, 237
Electrodes, high quality essential
in, 119
Monoxide evolved in Carbide
production, 217
38:
Carborundum, (Silicon Carbide,
q.v.), Electro-chem-
ically produced, 24
Carmichael process for Electrolytic
Winning of Copper,
Carnaltite, source of Magnesium
Chloride, 180
Castner chemical process for ob-
taining Sodium,
158, 185
Castner- Kellner Dissolved Salt pro-
cess for Electro-
lytic manufacture
of Alkali and
Bleach, 312
Cathocle(s), (see also Electrodes),
denned, 4
for Xickel Plating, 274
Cathode cages for, Delval
and Pascali's, 275
Refining Copper, 32
Winning Zinc, in the
Hcepfner process, 146, 152
Mond process, 147
Cations, defined, 4
Caustic Potash, Electrolytic manu-
facture of, 324-5
Soda, (see also Chlorine and),
processes for, using
a Fused Electro-
lyte, 292 et sea.
Cell(s)
Carbon 362 et seq.
Becquerel's observations on
363-4
difficulty in devising success-
ful, 362-3
of the future, Ostwald's views
on, 368
output (possible) of, 364
oxidants used in, 364, 365 et seq.
various kinds of
Borchers, 367, 369
Mond's experiments on,
370
Jablokhkoff, 364
Jacques, 365
Mond-Langer, 373
for Chlorate manufacture, used
by the National
Electrolytic Co., 337
Deacon, 304 & n.
Gas,
Groves, 372
Mercury
Bell, principles of, 320
INDEX
Cell( s) used in Dissolved Salt Electro-
lyte processes, see
also under names of
Processes
Acker, 324
Bell Mercury, 320
Greenwood, 323
Le Sueur, 320-1
Moore, Allen, Ridlon a,nd
Quincy, 312 n., 324
Outhenin-Chalandre, 321
Chemical and Electro-Chemical
processes, relative
value of, 23
Chlorate(s)
Electrolytic manufacture of, 326,
332
Cells for, 337
Potassium manufacture of, 332
difficulties in, 332
Kellner's proposals, 334-6
Chlorine, (see also Alkali, Chlorine,
etc.), Chemical me-
thod of prepar-
ing, 24
Electro - chemically produced,
(Hargreaves - Bird
process), 312
Chlorine and Caustic Soda, chem-
ically and also
Electro - chemically
prepared, 24
Chlorination process of Gold-Ex-
traction, 95
Chloroform, Electrolytic manufac-
ture of, 354
Chromium, production of, by the
Electric Furnace,
201, another me-
thod, 203
Properties of, and uses, 203
Specific Gravity of, 202
Cobalt, not as yet prepared on a com-
mercial scale, 123
presence of, in Nickel or Nickel
ores, 112 cfc note,
117, 120, how elim-
inable, 118
present uses of, 123
Pure, Electrolytic preparation of.
experiments of Bis-
chof and Thie-
mann, 115
Cobalt Plating, 275
Coehn, experiments of, on Carbon
dissolution in an
Electrolyte, 370-2
Cohen's process for Electrolytic
Winning of Copper,
82 et seq.
Commercial Electrolytic Nickel, see
Nickel
Conductors for producing Heat,
21-2
Conversion of Electrical Energy
into Heat, for
Electro - Chemical
processes, 21
Copper
Alloys of, processes for Electro-
lytic Refining of,
110-11
Chemical method of preparing, 23
Electrolytic Refining, Practice of,
32, et seq.
Anode Sludge of, 36
Composition of, 41
Working-up of, 56
Cost of, 61
of Energy for, 38
Electrolyte for, composition
of, 32, 42, vats for,
how arranged, 33,
42, electrodes in,
how connected, 32,
50
Mode of working the process,
52
Principles of, 31
Product of,
Quality of, 54
Special methods of Deposit-
ing, 58
Cowper-Coles, 60, 61 n.
Dumoulin, 60
Elmore, 59-61
Graham, 60
Thofern, ib.
Raw Material, for 39
Source of Power in, 37
Electrolytic Winning of
Energy requisite for, 63-5
Cost of, 65
Processes for, 63 et seq.
Carmichael, 82
Coroda, ib.
Cohen, 82 et seq.
Douglas, 84
Hcepfner, 72
Illinois, 82
Keith, 81
Siemens-Halske, 66
Electroplating, Iron and Steel
with, 260
383
INDEX
Copper
Use of, in Electrotyping, 253-03
Copper and Aluminium, compared
as to price,in weight,
176, and bulk, 177
do. Wires, Specific gravity of,
contrasted, 178
Copper, Silicide of, and its uses
Coroda's process for Electrolytic
winning of Copper,
82
Cost (see under various processes),
as de t e r m i ni ng
choice of Chemical
or Electro-Chemical
methods, 26-7
Coulomb, defined, 20
Cowles' Electric Furnace for the
Winning of Zinc,
133, 193-8
Cowper-Coles' process for Electro-
lytic Refining of
Copper, 60, 61 n.
do. for manufacture of Search-
light Reflectors,.
283-4
views on the Cost of Electro-
Zincing, 279
Crampagna, experiments at. in
Electric Reduc-
tion of Zinc, 133-4
Cryolite, Specific gravity .of, as com-
pared with Alum-
inium, 170 n.
Cupric and Cuprous ' Sulphide and
Sulphate, see Sie-
mens-Halske pro-
cess under Copper
DARLING process for Electrolysis of
Nitrate of Sodium,
188
Deacon cells, 304 & n.
Process for the manufacture of
Alkali, Chlorine, and
their Products, ib.
Delval and Pascalis' Nickel-plating
Cathode cage, 275
Deville's Chemical method for
Aluminium produc-
tion, 172
Diaphragm, ordinary porous, draw
backs to, 315
in Electrolytic manufacture of
Saccharin, 352
in Hargreaves-Bird apparatus,
308, 311
Diaphragm,
in Schiickert apparatus, 336
Cell,
Greenwood's, 323
Le Sueur's, ib.
Dieffenbach, see Duisberg
Dietzel's process for Electrolytic
refining of Alloys
of Gold, Silver, and
Copper, 111
Dissolved
Electrolyte, condition of, 14
Non-Electrolyte, normal condi-
tion of, 11
Salt as an Electrolyte for manu-
facture of Alkali,
Chlorine, and their
Products,
Processes using, 301
Bell, 305, 320
Castner-Kellner, 312
Electro-Chemical Co., 302
Hargreaves-Bird, 305
Le Sueur, 320
Rhodin, 319
Solvay, 318
Dorsemagen's Electric Furnace
(proposed) for Zinc
Winning, 133
Douglas's process for Electrolytic
winning of Copper,
84
Dry winning of Antimony, 129
Duisberg process (Dieffenbach's)
for Electrolytic
winning of Zinc,
148-9
Dumoulin's process for Electrolytic
refining of Copper,
60
Dye-stuffjElectrolytic production of,
Naphthazarine, 347
Yellow, Electrolytically produced
by Badische Anilin
und Soda Fabrik,
351
EFFICIENCY in Electrolysis, how
stated, 10
Elb's investigations into Electro-
lytic production of
Aniline from Nitro-
benzine, 340
Electric Furnaces
Application of, to the production,
etc., of
Alumina, 238
384
INDEX
Electric Furnaces,
Borides, 234
Carbides, 207-8
Artificial Graphite, 231-3
Calcium Carbide, 208-25
Types of furnace used,
illustrations of,
210-19
Silicon Carbide, 208,
225-30
Siloxicon, 230-1
Class of current used for,
210
Phosphorus, 237-8
Silicon and Silicides, 235-8
Steel, 246-7, 248-9
Zinc, 133
Principles of, 21, 193
Various makes of,
Cowles, 133, 193-8
Crampagna Co., 133-4
Dorsernagen, 133
Gin, 249
Heroult, 246
Horry, 216
Keller-Leleux, 242
Kjellin, 248
Memmo, 218
Moissan, 198-200
Spray, 213
Stassano, 242
Taylor, 237
the two common forms of, 22-3
Winning and Refining Metals
and their Alloys in,
193-238
Electrical Energy, conversion of,
into Heat, for Elec-
tro - Chemical pro-
cesses, 21
Elektrizitats-Aktiengesellschaft, see
Schiickert
Electro-brassing, 205
Electro-Chemical Co.'s process with
Dissolved Salt Elec-
trolyte for manu-
facture of Alkali,
Chlorine, etc., 302
do. do. for the manufacture of
Chlorates, 332
Electro-chemical processes,
Conditions, relative to preference
for, 23, 24 et seq.
Conversion of Electrical Energy
into Heat for, 21
Production of Alizarin by, 349
Sodium prepared by, 24
Electro-chemical processes,
Transformations of form in Car-
bon, due to, Moissan
cited on, 199
Electro-Chemistry, Principles of, 3
Electro-Deposition
of Metals, etc.,
Alloys, 285-6
Iron, 281
Palladium, 283
in Various processes
Cobalt Plating, 275
Electrobrassing, 205
Electrogravure, 284
Electroplating, 253, 254,
261-2, 264
Electrotyping, 255, 259, 261-2
Electro-zincing, 276
Gold Plating, 269
Nickel Plating, 113, 114 et seq.,
271
Silver Plating, 265
Electrodes, defined, 4
Carbon, 119
Graphite, ib.
for manufacture of Alkali, etc.,
with Fused Electro-
lytic processes, 292,
293, 295, 299, 300
Electro-Gilding, see Gold Plating
Electrogravure, 284
Electrolysis,
Definition of, 3
Efficiency of a process of, how
stated, 10
Faraday's law on, 7, 9, 10
Mechanism of, 1 1
Nature of, 4-5
Quantitative relations of process
of, regulated by
Energy required,
17-18
Electrolysis of
Fused Zinc Chloride, by Borchers'
apparatus, 149-50
Salt for manufacture of Alkali
and Chlorine, con-
ditions essential to
success, 289
Sea-water, Hermitprocess for, 328
Electrolyte(s)
Carbon dissolution in, Coehn's
experiment on,
370-2
Constitution of, 1 1
Normal condition of Dis-
solved, 14
385 cc
INDEX
Electrolyte(s)
Definition of, 3
Dissolved Salt used as, 301 et
seq.
Equivalents of Energy im-
pressed on, 6-7
Fused, see also Electric Fur-
naces
Processes using, 292
Acker, 298
Borchers, 299
Hulin, 295
Lead-refining, 92-4
Vautin, 293
Zinc-winning, see that head
employed in dealing with
Aluminium, 171
Copper, 32, 42
Electro-deposition of
Alloys, 285-6
Iron, 282, 283
Nickel (Foerster's), 117,
(others), 119, 273
Electroplating, 261-3
Electrotyping, 259
Electro-zincing of Iron, 278
Plating of
Gold, 269
Nickel, 273-4
Silver, 265
Parting of Silver and Gold, 103
Recovery of Tin, 124, 126, 127
Refining of
Cobalt, 115
Copper, 32, 42
Nickel, 114, 115, 118, 121
Winning of
Antimony, 128, 130, 131
Lead, 85, 87, 88
Magnesium, 180
Zinc, see processes, under
Zinc
essentials in, 135 et seq.
fused, the best, 146
Zinc Chloride as, 137
Electrolytic
Bath, see Baths
Chemistry, advantages of, over
Chemistry proper
343-4
Extraction of Gold, 95
difficulties in devising a process
for, 96-7
Manufacture of Alkali, etc., see
Alkali, Chlorine, and
their Products
386
Electrolytic
Manufacture of
Search Light Reflectors, Cow-
per-Coles process,
283
Steel, 241
Furnaces used in
Gin, 249
Heroult, 246- 7
Kjeltin, 248
do. and Treatment of Organic
Compounds, and
Fine Chemicals,
343-58
Chloroform, 354
Dye-stuffs, 347, 351
lodoform, 352 et seq.
Saccharin, 351
Sodium Acetate, 344
Benzoate, 345
Vanillin, 352
Reduction of
Nitro-benzine, (Bayer's pro-
cess), and Nitro-
compounds, 346-7
Aniline resulting from, 348
Reaction Electrolytically de-
terminable in, 349-
51
Oxidation of
Anthracene, 350
Isoeugenol, producing
Aniline, 352
Para-nitro-toluene, 350
Parting of Gold and Silver, 103
Electrolytes for, 103
Moebius' apparatus for, 104,
how modified, 107
Processes, method of calculating
Output in, 17
Purification of Sugar juice, 355
Recovery of
Gold, processes for,
Andreoli, 99
Kendall, 160
Pelatan-Clerici, 100
Siemens-Halske, 97
Tin, from Scrap Tinned Iron,
124, 126
Scrap Tinned Lead, 126-7
Reduction of
Indigo, 354
Nitrobenzine, 346
Refining of Copper, Gold, Lead,
Nickel, Silver, etc.,
see those heads
Tanning, various processes for,
356-8
INDEX
Electrolytic,
Winning of Copper, and Lead, see
those heads
Electro-metallurgical Production
and Treatment of
Iron and Steel,241-50
Electro-plating, 253, 254, 261-2,
264, uses of, 268
Electrotyping, 255, 259
Baths for, 261-2
Definition of, 255
Discoverers of, 253
Early methods of, ib.
Electrolyte for, 259
Essentials to, 254, 256 et seg.
Moulds for, materials for, 256-8
Films used in coating, 258-9
Precautions requisite in, 259-60
Electro-zincing, 278-80
Advantages of, 276
of Iron, 278, 280
Processes for, 277
Appearance of result, after hot
and cold do., 279
Cost of, (Cowper-Coles), 279-80
Difficulties in, 278
Electrolytes for, 278
Uses of, 276, 280
Elements, and their calculated
Output, table, 20
Energy
of Carbon, Reed's method for
utilizing, 373-4
Electrical, Conversion of, into
Heat, for Elec-
tro-Chemical pro-
cesses, 21
in Electrolysis, outcome of, 6, 10,
17-18, Faraday's
law relating to,
7-9, 10
Equivalents of, impressed on
Electrolytes, 6-7
do. in Volt pressure, 20
Requisite in various processes,
see under names of
processes and sub-
stances.
Quantitative relations of, 17-18
Eugenol, see Vanillin
Extraction of Gold
Processes for
Electrolytic, aims of, 95-6
Non-electrolytic, by
Amalgamation, 95
Chlorination, ib.
Potassium cyanide 96
FARADAY'S law, 7-9, 10, 19
Farbewerke vormals Friedrich
Bayer, see Bayer
Ferric and Ferreous Sulphide and
Sulphate, see Siemens-
Halske process, under
Copper.
Ferrosilicon, manufacture and uses
of, 235-6
Films or coatings for Moulds for
Electrolysis, 258-9
Flashlighting, uses of Magnesium
and Aluminium for, 183
Foerster, Dr. F., researches of, on
Electrolytic deposition of
Nickel, 115 et seg.
and Giinther, investigations of,
intoElectrolysis of Zinc, 137
Furnaces, see Electro do.
Fused Electrolyte, processes using,
for Manufacture of Alkali,
Chlorine, and their Pro-
ducts, 292
Acker, 298, 394
Borchers, 299
Hulin, 295
Vautin, 293
Refining Lead, (Borchers), 92-4
Nitre and Carbon, Becquerel's
observations on, 363-4
Zinc Chloride, Electrolysis of, by
Borchers'apparatus, 149-50
GALVANISING, see Electro-Zincing
Gas cells, Grove's, 372.
Gases, ionisation of, 17 n.
Gin process for manufacture of
Steel in Electric Furnace,
249
Glue (prepared) for Electrotype
moulds, 257-8
Gold
Alloys of. and of Copper
Electrolytic refining of, 110
Electrolytic
Parting of, from Silver, 103
Recovery of, method for
Andreoli's, 99
Kendall's, 100
Pelatan-Clerici's, ib.
Siemens-Halske's, 97
Refining of, 101-2,110
Anode sludge, from, Gold in,
101, 102, impurities in,
106-7
Platinum Process for, Nord
Deutsche Affinerie, 101
3?7
INDEX
Gold
Extraction of,
Electrolytic, 95
Difficulties in evolving a
process for, 96-7
Non-Electrolytic processes for,
Amalgamation, 95
Chlorination, ib.
Potassium Cyanide, 96
Gold and Silver, see Gold, and
Silver under their names
Gold-plating, (Electro-gilding), 269-
270
Colouring the plating, 270
Water gilding of, 269
Graetzel's apparatus for Electro-
lytic manufacture of
Magnesium, 180-1
Graham's process for Electrolytic
refining of Copper, 60
Graphite Electrodes, 119
Greenwood diaphragm cell, 323
Groth's process for Electrolytic
Tanning, 357
Grove's gas cells, 372
Guggenheim Smelting Co.'s process
for Electrolytic re-
fining of Silver, 109
Gutta-percha for Electrotype
moulds, 256
HALL process for producing Pure
Aluminium, 167,
170
Hard materials produced by the
Electric Furnace,
(see Borides, and
Carborundum), the
scale of, 234
Hargreaves-Bird Dissolved Salt pro-
cess for the Electro-
lytic manufacture
of Alkali, Chlorine
and their Products,
305
Heat, conductors for producing, 21
Conversion into, of Electrical
Energy for Electro-
Chemical processes,
21
and cost, 172, 174
and Energy, high demands on,
in chemical winning
of Aluminium,
157-8
Hermite process for Electrolysing
Sea water, 328
Heroult Electric Furnace for
Aluminium Bronze, 197
Steel, 246-7
Process for Pure Aluminium, 159,
169, 170, 173
Apparatus for, 162-3
Hoepfner process for Electrolytic
winning of Copper,
72, and of Zinc,
146-7
Holland and Richardson process,
see Electro-Chemi-
cal Co.'s process
Horry Electric Furnace for Carbide
production, 216
Hulin process for Electrolytic
manufacture of Al-
kali, Chlorine and
their Products, 295
Hunt, rules formulated by, for the
manufacture of
Aluminium, 169
Hydrogen in Electrolytic Iron, 272,
281
Evolution of, in Zinc Electrolysis,
134, 135
Hypochlorites, Electrolytic produc-
tion of, 327
IGNEOUS SOLUTION, winning and
refining of Metals
by Electrolytic
means in, 157-189
Illinois process for Electrolytic
winning of Copper,
82
Indigo, Electrolytic reduction of,
354
lodoform, Electrolytic manufacture
of, 352 et seq.
Ionic Electrolysis, theory of, 14 et
seq.
lonisation in Electrolytes, 14 et seq.
Ions,
Anions and Cations, 4
not necessarily Atoms, 14 n.
Carbon, formation of, 371-2
Zinc, 135
Iron,
Alloys, special, manufacture of,
by Electric Fur-
nace, 250
in crude Zinc Sulphate, 145
deposited with Nickel, 117, 120,
how eliminable, 118
Electro-deposition of, 281-4
Uses, 281
388
INDEX
Iron,
Electro-plating of, with Copper,
260
Zincing of, see that head
Electrolytes, impurities in, 282
Electrolytic, hardness of, in rela-
tion to Hydrogen,
272, 281
Refining of, 283
Nickel plating of, 263
Smelting processes for, drawbacks
to, 242
Keller Leleux and Co.'s Electric
- Furnace, 242
Stassano's, ib.
in used " Tins," market value
of, 125
Iron and Steel, see Iron, and Steel,
under names
Izart, J., process of, for Electrolytic
winning of Anti-
mony, 131
JABLOCHKOFF'S Carbon cell, 364
Jacques' Carbon cell, 365
KELLER LELEUX AND Co.'s Electric
Furnace for smelt-
ing Iron from the
ore, 242
Kellner's views on Electrolytic
manufacture of
Chlorate, 334 et seq.
Kendall's process for Electrolytic re-
covery of Gold, 100
Keith's process for Electrolytic
winning of Copper,
81, and of Lead, 87
King Electric Furnace, 214-16
Kjellin process for Electrolytic
making of Steel, 248
LEAD,
behaviour of, in indiscriminate
distillation of Zinc,
134 n.
De-silverising by Zinc, and the
formation of Zinc
Amalgam, 151
Electrolytic
Refining of,
Processes for, using Fused
Electrolyte
Borchers, 92-4
Winning of
Anodes and Cathodes for, 85,
87
Anode Sludge of, composi-
tion of, 90
Electrolytes for, 85, 87, 88
Low Cost an essential to,
86
Processes attempted for,
Keith's, 87
Niagara Falls, 85
Tommasi's, 88, 91
Spongy, cost of Electrolytic Re-
fining of, 90-1
Tinned Scrap, Electrolytic re-
covery of Tin from,
126-7
Le Sueur, cell, 320-1, 323
Dry Salt process for Electrolytic
manufacture of
Alkali, Chlorine and
their products, 320
MAGNALIUM alloy, composition of,
and advantages, 184
Magnesium,
Commercial use for, 183-4
Electrolytic production of, 180
Apparatus for, 180-2
Heat of combination of, and
critical voltage for,
182
Points of interest in, 183
Electrolytic reduction of, 180-3
Raw material of, 180-2
Manganese, Aluminium preferred
to, for various pur-
poses, 183
Zinc-producing Ores, how dealt
with, 145
Pure, unproduceable by ordinary
smelting, 172
Mathieson Alkali Co., Niagara,
working of a Cast-
ner-Kellner Electro-
lytic Plant at, 317
Memmo's Electric Furnace, 218
Menne & Co., electrolytic Nickel
made by, 221
Mercury, functions of, in Castner-
Kellner process of
Electrolytic manu-
facture of Alkali and
Bleach process, 313
Cell, the Bell, principles of, 320
Metals, Winning and Refining, and
their Alloys in the
Electric Furnace :
Carbides, Borides,
Silicides, 193-238
INDEX
Metals
Winning by Electrolytic means in
Aqueems Solution, 31-151
Igneous do., 157-89
Produced or Refined by the
Electric Furnace,
Chromium, 201
Molybdenum, 204
Titanium, 206 & n.
Tungsten, 205, 206 & n.
Vanadium, 206 & n.
treatable with ' the Moissan
Electric Furnace,
199-200
Minet-Bernard process for produc-
ing Aluminium, 171
Moebius apparatus for Electrolytic
parting of Silver
from Gold, 104, and
modification of, 107
Process for Electrolytic refining
of Silver, 109, see
also 104, 107
Moissan's Electric Furnace, 198-200,
and work in con-
nection with the
above, results of,
197-200
Views cited on Carbide production
by the Electric Fur-
nace, 207-8, on Cal-
cium Oxide, 220-1
Molybdenum, production of, by the
Electric Furnace,
204, its uses, 205
Specific gravity of, 205
Mond, experiments by, on Borchers'
cell, 370
Process for Electrolytic winning
of Zinc, 147-8
Mond-Langer Carbon cell,unsuccess-
ful, 373
Moor, Allen, Ridlon, and Quincy
cell, 312 n., 324
Most s process for oxidising Anthra-
cene, 350
Moulds for Electrotyping, composi-
tion of, 256-8
Mylius and Fromm, results of their
investigations into
Electrolytic win-
ning of Zinc, 135 et
seq.
NAPHTHAZARINE, (alizarin black),
Electrolytic pro-
duction of, 347
National Electrolytic Co.'s process
for manufacture of
Chlorate, 337
Nickel,
Atomic weight of,
Winckler's determination of,
114
Castings of, use of Magnesium
in, 183-4
Commercial
Electro-deposited, hardness
of, 272, advantages
of, 282
Electrolytic refining by
Balbach Smelting and Re-
fining Co., 113,
from crude ore, 120
Bolton and Sons, 1 1 3
Canadian Copper Co. from
bessemerized matte,
121-2
Foerster, 115 et seq.
Menne and Co.'s results, 121
others. 113
Refining
Anodes used in, 114, 115,
118, 119, 121
Sludge resultant, 115,
118-9
Metallurgy of, and impurities
in, 112
Pure
Electro deposited, Bischof and
Thiemann's process
for securing, 114
Nickel plating (see also Electrolytic
deposition), 113
Difficulties in, 114 et seq.
Electrolyte for, 273
Empirical nature of the art, 274
of Iron, 263
Process of, precautions essential
in, 272 et seq.
Uses for, and advantages of, 271
Nitre, fused, and Carbon, Bec-
querel's observa-
tions on, 363-4
Nitrobenzine, Electrolytic reduction
of, 346-7
Aniline Electrolytic production
from, 348
Nitro-compounds, Electrolytic re-
duction of, 346,
347, Bayer's patent
for, 346
Non-Electrolyte, Dissolved, Normal
Condition of, 11
390
INDEX
Nord-Deutsche Affinerie of Ham-
burg, process for
Electrolytic refin-
ing of Gold, 101
Normal Condition of a Dissolved
Electrolyte, 14, and
Non-Electrolyte, 11
OSMOTIC pressure, mode of ascer-
taining, 11
Ostwald's vjews on Borchers' Car-
bon cell, cited, 367
et seq.
Outhenin Chalandre cell, 321
Output in Electrolytic processes,
method of calcu-
lating, 17
possible, of Carbon cells, 364
Oxidants used in Carbon cells, 364,
365 et seq.
Ozone, Electrolytic production of,
338
PALLADIUM plating, 283
Para-nitrotoluene, Electrolytic oxi-
dation of, 350
Parkes' process, working up
Zinc Amalgam
from, 151
Pelatan-Clerici's process for Electro-
lytic recovery of
Gold, 100
Pennsylvania Lead Co.'s process
for Electrolytic re-
fining of Silver, 108
Perchlorate, Electrolytic manufac-
ture of, 338
Persulphates, Electro - Chemically
produced, 24
Phoenix process for Winning Zinc
from mixed Sul-
phide ores, (Ash-
croft and Swin-
burne), 145-6, ap-
paratus for, 150
Phosphorus, manufacture of, now
entirely Electrical,
237-8
Pittsburg Reduction Co.'s process
for producing Pure
Aluminium, 167, see
also Hunt
Plaster of Paris for Electrotype
Moulds, 257
Platinum process for Electrolytic
refining of Gold,
101
Potassium Chloride, output and
cost of, 325
Power, 361-77
Gas (producer) possibilities of, 377
Steam, 361
Cost of, per h.-p. year, 326,
376-7
Turbine, Steam or Water, 377
Water power, 361, 374
Cost of, per h.-p. year, 375-6
REED'S method for utilizing the
energy of Carbon,
373-4
Refining, see Electrolytic do. and
Winning and Refin-
ing
Relative value of Electro-Chemical,
and purely Chemi-
cal processes, 23
Rhodin Dissolved Salt process for
Electrolytic manu-
facture of Alkali,
Chlorine, etc., 319
Roseleur's Bath for Electrotyping,
262
Rossler's modification of the Parkes
process for Electro-
lytic refining of
Silver, 110
SALT, i.e., Sodium Chloride, (see also
Dissolved Salt), at-
tempts to produce
Alkali and Chlorine
by Electrolysis of,
conditions essential
to success, 289
Schiickert apparatus for manufac-
ture of Bleaching
Liquors, 330-1, 336
Sea-water, Electrolysing of, Hermit
process for, 328
Search Light Reflectors, Electro-
lytically made,
(Cowper-Coles pro-
cess), 283-4
Siemens Electric Furnace, 212, 219
Siemens-Halske process for Electro-
lytic
Manufacture of Bleaching Liquor,
329
Recovery of Gold, 97
Winning of
Antimony, 128, 129
Copper, 66
Zinc, from mixed ores, 140
391
INDEX
Silicide of Copper, uses of, 236
Silicon and the Silicides, produced
by the Electric Fur-
nace, probable uses
of, 235-8
Silicon Carbide (Carborundum)
Acheson's discovery of, and pro-
cess for, 226 t seq.
Electro-chemically produced, 24
Hardness of, 234
Uses of, 229
Siloxicon, produced by the Electric
Furnace, composi-
tion and uses of,
230-1
Silver, see also Cadmium and Silver
Electrolytic
Extraction of from
Zinc Amalgam, 151
Zinc-producing ores, 140,
141
Parting of, from Gold, 103
Electrolyte for, 103
Moebius apparatus for, 104,
how modified, 107
Refining of, processes for
Guggenheim Smelting Co.,
109
Moebius, 109
Pennsylvania Lead Co., 108
Rossler's modification of the
Parkes' process, 110
Plating, 265-8
Care, essential in, 265
Electrolytes for, ib.
Smelting Processes for Iron, draw-
backs to, 242
Electric Furnaces for
Keller-Leleux, 242
Stassano, ib.
Sodium
Bad effects of, in Commercial
Aluminium, 179
Chemically, and also Electro-
chemically pre-
pared, 24
Electrolytic production of, pro-
cesses for
Ashcroft, 189
Castner, 186
Darling, 188
Former non-Electrolytic method
of producing, 185
Uses of, 188
Sodium and Potassium Chloride
compared, for re-
sult and cost, 325
Sodium Acetate, Electrolysis of, 344
Benzoate, Electrolysis of, 345
Chloride, Electrolytic decomposi-
tion of, 289
Salts, Fused, Electrolysis of, 185
Solvay process for manufacture of
Alkali, etc., with
Dissolved Salt Elec-
trolyte, 318
Special Steels, see Steel
Specific Gravity of
Aluminium (commercial), 177
(cf. 170n.)
Calcium Carbide, 209
Chromium, 202
Copper and Aluminium Wire,
compared, 178
Molybdenum, 205
Tungsten, 205
Sponginess in Electrolytically-de-
posited Zinc, 134,
135-7
Spongy Lead, 85 et seq.
Cost of Electrolytically Refining,
90-1
Spray Electric Furnace, 213
Steam Power, see under Power
Steel
Electrolytic manufacture of, by
the Electric Fur-
nace,
Principles of, 244
Processes for
Gin, 249
Heroult, 246-7, 250
Kjellin, 248, 250
Special Steels
Processes for, 250
Electroplating of, with Copper,
260
Suitable Alloys for, produced by
the Electric Fur-
nace, 203, 205, 206
&n.
Sugar Juice, Electrolytic purifica-
tion of, 355
TANNING, Electrolytic, various pro-
cesses, 356-8
Taylor's Electric Furnace for manu-
facture of Carbon
Disulphide, 237
Temperature in relation to the
choice of Electro-
Chemical or Chemi-
cal processes, 24
et seq.
392
INDEX
Temperature
in Electrolytic Nickel plating,
115
Thofern's process for Electrolytic
refining of Copper,
60
Tin
Commercial, impurities in, of,
124
Electrolytic recovery of, from
Scrap tinned Iron,
Cast-off " Tins," 124
Cuttings from " Tins," 126
Scrap Tinned Lead, 126-7
Metallurgy of, 124
Refinement of, usual processes
for, 124
Titanium, production of, by the
Electric Furnace,
206
Tommasi's process for Electrolytic
winning of Lead,
88, 91
Tucker and Moody' s experimental
processes for Elec-
trolytic production
of Aluminium, 174
Tungsten, production of, by the
Electric Furnace,
205, its uses, 205-6
Specific Gravity of, 205
ULKE, TITUS, Cyanide process sug-
gested by, for Elec-
trolytic preparation
of Nickel, 120
VANADIUM, use of, as an alloy for
Steel, 206 & n.
Vanillin, Electrolytic production of,
from Eugenol, 352
Vautin process for Electrolytic
manufacture of Al-
kali, Chlorine, etc.,
293
WATER plant for Aluminium pro-
duction, 175
Power, 361, 374
Cost of, per h.-p. year, 375-6
in Steel manufacture, advan-
tage given by,
241-2
Weston Point, working of the
Castner-Kellner Al-
kali, Chlorine pro-
cess at, 316-17
Willson's discovery of Calcium Car-
bide, 209, his Elec-
tric Furnaces for
producing, 210 et
seq.
Winning and Refining Metals and
their Alloys in the
Electric Furnace :
Carbides, Borides,
Silicides, 193-238
do. by Electrolytic means in
Aqueous Solution, 31-153
Igneous do., 157-189
Wood's metal for Electrotype
moulds, 257
Working up " Zinc Amalgam " from
the Parkes' process,
151
Worms and Bal process for Electro-
lytic Tanning, 357
ZINC
Electrolytic
Deposition of
Difficulties in, 134, the two
chief, 135
Principles of, 134
Sponginess resulting in
Zinc so deposited
134, 135-7
Reduction of by
Cowles' Furnace, 133, 193-4
Refining of
best effected non-Electro-
lytically, 132, 134,
152
Winning of, 132, from
Aqueous Solutions,
drawbacks to, 135-40
do. by
Electrolysis of Fused Zinc
Chloride, Borchers'
apparatus for, 149-
50
Electrolytes for, essentials in,
135 et seq.
Fused do., the best, 146
Furnaces for,
Cowles, 133
the Crampagna, 133-4
Dorsemagen, (proposed),
133
Process for, usual aim of,
140, and backward-
ness of, 152
393
D D
INDEX
2inc
Winning by
Ashcroft, 142
as worked at Cockle Creek,
144-5
Ashcroft and Swinburne,
(partially chemical
Phoanix method),
145-6
Duisberg, 148-9
Hcepfner, 146-7
Mond, 147-8
Siemens-Halske, 140
Winning of by non-Electro-
lytic Processes, 132,
wastefulness and
costliness of, 132-3
Pure
Chemical methods of preparing,
23
Zinc
Pure
How easily obtainable, 134
Zinc Amalgam
Extraction and recovery of
Silver from, 151
Reduction of Zinc from, ordi-
nary, and proposed
Electrolytic pro-
cesses for, 151-2
Working up, from the Parkes
process, 151
Chloride as an Electrolyte for
Zinc, 137
Oxide, injurious in Electrolvsis of
Zinc, 136
Zinc-producing ores, extraction of
Silver from, 140
Butler & Tanner, The Selwood Printing Works, Frome, and London.
394
University of Toronto
Library
DO NOT
REMOVE
THE
CARD
FROM
THIS
POCKET
Acme Library Card Pocket
Under Pat. "Ref. Index File"
Made by LIBRARY BUREAU
