You are on page 1of 68

MTRL 358

Electrowinning and Electrorefining

2014

Metal recovery is the final step in most hydrometallurgical processes. This is


commonly practiced for aluminum, copper, zinc, nickel, cobalt and gold. (Aluminum
cannot be obtained by electrolysis from water; it is too strongly reducing. Molten
salt electrolysis is used instead.) In all hydrometallurgical processes metals are
present in solution as complexes of the metals in positive oxidation states, e.g.
[Zn(H2O)6]+2. All metal recovery processes then necessarily involve reduction.
Hence all these processes require a reducing agent. The process may be
thermodynamically favourable (like hydrogen gas reduction of [Ni(NH 3)n]+2
complexes (n = 2, for instance) or unfavourable (like electrowinning of copper in
which water is forced to be the reducing agent). When the process is
thermodynamically unfavourable (E < 0, by definition) it is categorized as an
electrolysis. Electrolysis for metal production is called electrowinning.
Briefly, a copper EW plant may have many cells (hundreds). Each cell is a
little over 1 m wide, ~1.5-2 m deep and several meters long. They contain several
dozen cathodes and the same number + 1 anodes. Metal is plated onto both sides
of the cathode sheets, while water is oxidized to form O 2 and H+ at the anodes. A
schematic illustration of a cell is shown in the diagram below. Enriched electrolyte
supplied from solvent extraction stripping is fed into the cells. It passes through a
cell once and then is returned to SX stripping as the lean electrolyte. Once the
copper has been plated to a thickness of about 0.5 cm, the cathodes are removed
from the cell and the copper is prepared for sale.

Figure 1. Schematic illustration of a copper


electrowinning cell.

Pyrometallurgically produced metals are often not sufficiently pure to be sold


as high purity products. They are usually further refined and often using
electrolysis. Molten, as-produced metal is cast into electrodes (~1 m x ~1 m) and
these are interleaved with metal sheets in cells. The cast, impure metal electrodes
are anodically polarized to electrochemically corrode them (a form of leaching),
while the interleaved sheets are cathodically polarized to plate out the dissolved
metal ions. A very pure metal product is formed. This is called electrorefining.
Background Electrochemistry
The relevant electrochemistry was developed in the Eh-pH diagram course
notes and should be consulted.
Reminder on calculating E or E: The potential difference or voltage
generated by an electrochemical cell at a certain temperature is strictly a function
of the composition of the cell, i.e. activities of the reactants and products. It does
not depend on the charge that passes through that potential (i.e. nF). The energy
associated with passage of the charge through the potential difference does
depend on the amount that is passed: Energy = voltage x charge. But, the potential
itself generated by the cell has nothing to do with the charge that is passed.
Therefore DO NOT multiply Es by n numbers to calculate E or E for a cell!

Electrowinning Equations (Faraday's Law Relationships)


(1) Faradys Law
Faradays law states that the number of moles of metal produced in an
electrolysis is directly proportional to the charge passed. The constant of
proportionality is nF, where n = moles of electrons per mole of metal produced (an
integer) and F is the Faraday which is 96485 C/mole e -; a mole of electrons has a
charge of 96485 C. (i.e. 6.02205 x 10 23 e-/mol e- x 1.60218 x 10-19 C/e-.) Taking
into account the fact that charge, q = current x time for fixed current (or the
integral of I vs. t for a varying current) and that the moles of metal produced =
mass/atomic weight leads to the formula provided.
Moles of metal plated q (charge passed in the electrolysis)

{1}

moles metal = nM = q/nF

{2}

The units of q/nF are C/(mole e-/mol metal x C/mole e-) = mol metal. Charge
passed at constant current is q = It. Moles of metal = M/AW where M is the mass of
metal plated and AW is the atomic weight in g/mol. Then moles metal plated is,
nM = It/nF = M/AW

{3}

Rearranging gives,
M = It AW
nF

{4}

From Faradays law it is obvious that the lower n is, the less electricity that
will be required per unit mass of metal plated. Some metal ions have more than
one oxidation state. For typical copper electrowinning the cathodic half reaction is,
Cu+2 + 2e- = Cu

(1)

Alternative leaching processes have been developed that form cuprous complexes:
Copper sulfides

air, Cl-

CuCl2-aq

(2)

In this case the cathodic half reaction is,


CuCl2- + e- = Cu + 2Cl-

(3)

which would use half the electricity. (No such process is currently commercially
applied.)
(2) Current Efficiency
The simple formula above determines mass of metal plated for a given
current and time, or vice versa. If the only cathodic (reduction) process operative is
metal ion reduction to metal, then the formula gives an accurate indication of the
mass of metal for a given current and time. However, other reduction half reactions
may also occur simultaneously. These unwanted side reactions also consume
electricity (current) and result in a lowered efficiency of use of current for metal
plating. This leads to the idea of current efficiency. Current efficiency (CE) for
metal plating then (or any electrolytic process), is the ratio of actual mass of metal
plated to the theoretical mass based on Faradays law. It is usually given in %.
CE =

actual mass metal plated x 100


theoretical mass expected

{5}

The theoretical mass is given by Faradays law.


CE =

100M
It AW/nF

= 100nFM
It AW

{6}

Now M is actual mass of metal plated. For instance, calculate the current efficiency
for the following conditions: 100 g of copper was plated in a copper electrolysis
experiment using a constant current of 4 A for 23 hours. What was the current
efficiency?

CE = 100 g Cu x 100 x 2 mol e-/mol Cu x 96485 C/mole e- = 91.7%


(23 hr x 3600 sec/hr x 4 C/sec x 63.546 g Cu/mol Cu)

{7}

(3) Energy Efficiency


There is a theoretical minimum energy required to electroplate a metal. This
is the thermodynamic minimum voltage times the charge passed at 100% current
efficiency. Electrical work (energy) is voltage times charge (more generally, Vdq if
the voltage varies with charge passed, which equals Vq at constant voltage). By
definition 1 VC = 1 J (1 voltcoulomb = 1 joule). In practice the actual voltage will
be greater than the thermodynamic minimum for a number of reasons, and due to
less than 100% current efficiency the charge passed may be greater than the
theoretical minimum. On both counts the energy required will be greater than the
theoretical limit.
Energy efficiency is the ratio of the theoretical energy required to the actual
energy required, in percent. The theoretical voltage is E, i.e. the thermodynamic
cell voltage. The theoretical charge required is given by Faradays law,
q = nMnF

{8}

Hence the minimum energy requirement is,


Werev = E nMnF units in J

{9}

(Work and cell thermodynamic voltage are related by,


-G = nFE = we'rev

{10}

units in J/mol; this is the difference between we'rev and We'rev.

where Werev and we'rev are the electrical work under reversible conditions.* In
electrowinning E is negative (G > 0; E < 0); the reaction as written is not
favourable so we'rev < 0. The units of w e'rev are J/mol. As per the engineering
convention, work done on a system is negative. The symbol W e'rev represents the
work in joules. The actual applied voltage is designated Eappl. The applied voltage
opposes the thermodynamic voltage and hence it is taken to be positive. For
practical rates we require Eappl > E . The current is forced to go in the opposite
direction to the natural tendency of the cell.)
CE = 100 nFM = 100 nFnM
It AW
q

{11}

* Reversible and irreversible processes are reviewed in the next section. For now suffice
it to say that if an opposing voltage equal to E , where E < 0, (an electrolysis) is applied,
then the thermodynamic tendency is just matched or just overcome and the reaction is
exceedingly slow, i.e. reversible.

Since M/AW = nM and It = q,


q = 100nMnF
CE

{12}

Then the actual energy input is,


-We = Eappl q = 100EapplnMnF
CE

{13}

(-We' is a positive number.) The energy efficiency is:


100 E nMnF
100Eappl nMnF/CE

EE = 100Werev =
We

= E CE
Eappl

{14}

Take an example again of 100 g of copper plated as above at a voltage of 2.0


V with 91.7% current efficiency. (If the cell is large and relatively little copper is
plated, the applied voltage will be about constant, as will be the current.) E =
-0.89 V (= E assuming standard conditions; 0.89 V = the necessary applied
voltage to just overcome the thermodynamic negative cell voltage. (In reality the
Nernst equation E would be required for a real cell with non-standard activities.)
EE = 0.89 x 91.7/ 2.0 = 40.8%

{15}

This is not very high. Reasons for this will be explained later.
(4) Specific Energy Consumption
This is the actual energy requirement in units of energy per unit amount of
metal plated (e.g. J/mol). The derivation above employed the energy consumption,
i.e. Eappl q.
-We = 100Eappl nM nF
CE

{16}

Divide both sides by the moles of metal (n M here) to get the specific energy
consumption in J/mol (designated -we):
-we = 100 nFEappl
CE

{17}

A more conventional unit is kilowatt-hours per tonne of metal. A kWh is 1000 watts
for 1 hour = 1000 J/sec x 3600 sec = 3.6 x 10 6 J. To convert the energy
consumption number to kWh/t metal involves only unit conversions:
-we J

1 kWh

x 1 mol metal x 10-6 g

{18}

mol

3.6 x 106 J

AW g

For copper this works out to -we J/mol x 0.0043713 kWh/t.


J/mol
For copper electrowon as above at 2 V and 91.7% current efficiency,
-we = 2 mol e- x 96485 C x 2.0 V
mol Cu
mol e91.7 x 0.01

= 4.209 x 105 J/mol Cu

= 4.209 x 105 J/mol x 0.0043713 kWh mol / J t = 1840 kWh/t Cu

{19}

{20}

In practice a typical energy requirement for copper EW is about 1900-2000 kWh/t.


(5) Metal Production Rate
Starting with Faradays law and current efficiency again,
q = 100nMnF
CE

{21}

nM =

{22}

q CE
100nF

where nM is the moles of metal produced. Only CE% (e.g. 91.7%) of the total
charge passed goes to plate metal. The charge at constant current is q = It.
nM = It CE
100nF

{23}

Metal is plated onto both sides of a cathode starter sheet (e.g. a steel sheet in
copper electrowinning). The total plating surface area for a number N cathodes is
AcN, where Ac is the surface area per cathode sheet. Taking j as the current density
in A/m2, the current being passed is j times the total plating area, i.e. I = jA cN.
nM = jAcNt CE
100nF

{24}

i.e. j (C/sec m2) x surface area (m2) x time (sec) = charge (C). Next, rearrange to
obtain:
nM = dnM = jAcN CE
t
dt
100nF

mol/sec

{25}

AcN is the total plating surface area. This may be obtained in two ways, either using
N to be the number of cathode starter sheets with A c being the area of both sides
combined, or with N being the number of plating surfaces and A c being the surface

area of just one side. Either way is equivalent. Regardless, the fact that metal is
plated on two sides is factored in. The area of the narrow sides is negligible and
little copper plates there since the electric field is rather diffuse at the sides anyway.
In practice, edge strips may be used to prevent plating there. This makes removal
of the plated metal sheets much easier.
To get plating rate in mass per unit time, multiply dn M/dt by appropriate
conversion factors. This depends on the metal being plated since it involves the
atomic weight. For tonnes per day:
dM = dnM mol x AW g x 10-6 t x 3600 sec x 24 h
dt
dt sec
mol
g
h
d

{26}

For copper,
dMCu = dnCu mol x 63.546 g x 10-6 t x 3600
dt
dt s
mol
g

s x 24 h = 5.49037 dnCu t Cu/day


h
d
dt
{27}

A typical EW cell would contain 60 cathodes 1 m wide x 1-1.2 m deep and 61


anodes. Copper is plated on both sides of the cathodes. Typical current densities
range from 200-350 A/m2. For copper the cathode production rate for a cell with 60
cathodes per cell, each with length x width = 1 m x 1m, at 200 A/m 2 current
density and 91.7% CE is:
200 C
x 2 m2 x 60 sheets x 0.917
2
dnM =
sec m
sheet
dt
2 mol e- x 96485 C
mol
mol e-

{28}

= 0.11405 mol/sec
dMCu/dt = 0.11405 x 5.49037 = 0.6262 t Cu/day
Note: the cathode sheet has dimensions of 1 m x 1 m. The plating area on one
side is 1 m2. The plating area of the whole sheet is 2 m 2; we plate on both sides.
The calculation above allows estimation of copper production based on current and
current efficiency. This aids in design of an actual EW tankhouse. The number of
cells needed to achieve a desired production per year can be readily determined.
How many cells of 60 cathodes each would be needed to achieve 50,000 t/yr of
copper production under the conditions we have been using in the calculations
above?
0.6262 t Cu/day/cell x 365 days/y x S cells = 50,000 t/y

{29}

S = 218.8

We would need 219 cells. The total number of cathode sheets that must be
employed then is 219 x 60 = 13,140. If a fully plated copper cathode is 0.5 cm
thick on average and 1 m long on each side, and given that the density of copper is
8.92 g/cm3, the average weight of a cathode would be:
100 cm x 100 cm x 0.5 cm x 8.92 g/cm3 = 44.6 kg (0.0446 t)

{30}

This means that 1.1211 x 106 sheets of copper have to be handled per year, or
3071 per day. If the current density was increased the number of cells required
could be lower. Often cathode quality issues limit the current density.
How long would it take to plate a copper cathode to a thickness of 0.5 cm?
The metal production rate equation can be rearranged to obtain time. Since the
current is constant, dM/dt is constant as well, so the mass M plated over a specified
time t equals dM/dt:
MCu = jAcN CE x 63.546 g/sec
t
100 nF

{31}

t=

{32}

100 nF MCu
sec
jAcN CE x 63.546

where MCu in g. Now Ac is the area for a single sheet (on one side), i.e. 1 m 2 in this
case. N now is 1. (We are considering the time to grow a single cathode copper
sheet.) For MCu = 44.6 kg = 44,600 g as above, plated at 200 A/m 2 with 91.7% CE:
t=

100 x 2mol e- x 96485 C


x 44,600 g
mol
mol e
= 7.385 x 105 sec
2
200 C
x 1 m x 91.7 x 63.546 g
sec m2
mol

{33}

= 8.55 days
Obviously the higher the current density, the shorter the plating time.
The five relationships above are summarized in the table below.
Table 1. Summary of the Faraday's Law relationships.
M = I t AW in g
Faradays law
nF
CE = 100nFM
Current efficiency
It AW
EE = E CE
Energy efficiency
Eappl
-we = 100 nFEappl in J/mol
Specific energy consumption
CE
Metal production rate
dnM = jAcN CE in mol/s

dt

100nF

Thermodynamics of Electrochemical Cells


Reversible and Irreversible Processes
In thermodynamics a reversible process is one for which the direction of a
process (such as a reaction) can be reversed by an infinitesimal change. For
example, if a process is operating reversibly, an infinitesimal change in pressure or
temperature or concentration can reverse the direction of the process. Reversible
processes are in thermal equilibrium with their surroundings. They are also at
equilibrium in other respects, e.g. chemically or mechanically. How then can there
be any actual change of state? Suppose there is a chemical reaction occurring,
A = B. And suppose the system is at equilibrium. Increase the concentration of A by
d[A], an infinitesimal change. The reaction proceeds to the right to an infinitesimal
degree. Continue to increase the concentration of A in infinitesimal steps. The
reaction proceeds to produce additional concentration of B by d[B] increments. In
the limit of infinite time a finite extent of reaction will have occurred. Note that at
any stage during the process the reaction can be reversed by adding an
infinitesimal concentration of B. This is a reversible process. Truly reversible
processes are of no practical use; they occur infinitely slowly. But, they are a
condition or case that thermodynamics can use to tell us something about
theoretical limiting possibilities. It helps us to answer questions like, "What is the
minimum possible heat we can put into a process to make it go?" Or, "What is the
maximum possible work we can get out of a process?"
Naturally then, real processes are always irreversible. (Irreversible does not
mean that it cannot be reversed, but, rather the opposite of thermodynamically
reversible.) They have a finite (not infinitesimal) driving force to proceed in one
direction. Stopping or reversing the process requires a finite change in a variable.
Real processes sometimes can approach, but never truly attain reversibility. A
reversible process always has associated with it the minimum possible heat flow.
(This can be qualitatively understood from an example. If you very slowly and
gently set down a large rock on a surface there will be little or no perceptible
change in temperature of the rock and the surface. If you drop the rock it will hit
the floor with substantial force and generate a substantial rise in temperature. Both
cases involved the same change in gravitational potential energy. The latter was the
most irreversible case.)
Real processes involve conditions that are far from equilibrium. They move
spontaneously towards equilibrium. If a process is exothermic, for instance, the
heat loss is larger than would be the case under reversible conditions. Heat loss
from the system is negative and q irrev < qrev, i.e. qirrev is a bigger negative number
than qrev (qirrev > qrev).
With respect to galvanic electrochemical cells (favourable reaction) under
hypothetical reversible conditions, the heat flow is the minimum, while the work
that can be done is the maximum. Recall that the change in internal energy for a

change of state (e.g. 1 mol A aq 1 mol B aq, as above) is the sum of heat flow
minus work flow, (U = q - w; work done by the system is positive by definition and
heat exiting the system is negative.) Internal energy is a state function, i.e. it
depends only on the final and initial states, not how you get from one to the other.
Then, the less heat evolved for a given change, the more work that was extracted
from that change. For a real cell, operated under real conditions, the process is
necessarily irreversible and the heat flow is greater than in the reversible case, so
the work obtainable is less. The farther from reversibility (or the more irreversible
the process), the more the heat and the less the obtainable work. Extending the
idea that reversible processes run infinitely slowly, the faster the process is run, i.e.
the more rapidly the battery is discharged, the more irreversible the process, and
the more of the energy that is lost as heat.) The reversible case defines the limiting
possibility.
Some examples of irreversible processes include:
1. Flow of heat from a hot body to a cold one
2. Water flowing downhill
3. Hydrometallurgical leaching reactions
4. The conversion of chemical energy into electrical energy in galvanic cells.
5. An electrolysis.
Once the final equilibrium state is reached the capacity of the system to do
further work is exhausted. In example 1, the two bodies reach the same
temperature, and in example 4 the battery goes dead. Real processes involve finite
changes of state with finite energy changes. There is a driving force, or potential for
the process to occur. If the change is thermodynamically favourable then the
process is spontaneous and irreversible. (Recall that for a spontaneous process, G
< 0. The Gibbs free energy function expresses the requirement that spontaneous
processes must increase the net entropy of the system plus its surroundings.) If the
process is not favourable it is not spontaneous and does not naturally tend to occur.
The reverse (opposite) process is actually favoured and naturally does tend to
occur. The non-spontaneous process can be forced to occur by input of sufficient
energy, and when this is done the real process is also irreversible.
A Review of Some Relevant Thermodynamics
Next, a refresher on some aspects of the thermodynamics related to
electrochemical cells is needed. Enthalpy is the sum of internal energy + PV, where
P = pressure and V = volume.
H = U + PV

{34}

10

U = q w, staying with the engineering convention that work done by the


system is positive and heat flow out of the system is negative.
dH = dq dw + d(PV)
{35}
(H is very similar to U, but more convenient at constant pressure.)
For a reversible process the heat flow is denoted dqrev. Then,
dqrev = TdS

{36}

where T is the absolute temperature and dS is the entropy* change of the system
(this from the definition of entropy). Under reversible conditions, the system can do
its maximum possible work, and the heat flow is the minimum possible, i.e.
-dw = -dwrev = maximum work possible

{37}

G = H TS by definition

{38}

where G is the Gibbs free energy. At constant temperature, reversible conditions:


dG = dH - TdS

{39}

dG = dU + d(PV) TdS

{40}

dG = dqrev dwrev + d(PV) - TdS


dG = TdS dwrev + d(PV) TdS

{41}
{42}

Generally, the work is comprised of pressure-volume work and non-PV work, such
as electrical, gravitational etc. (the former is of interest here). The work term is,
-dwrev = -PdV dwrev

{43}

where dwrev is the non-PV work (electrical work here) under reversible conditions.
At constant pressure,
and,

dG = -PdV dwrev + PdV + VdP = -dwrev

{44}

G = -w'rev

{45}

since dP = 0 (constant pressure). (The cancelling of the PdV terms is what makes
the enthalpy function convenient.) Hence the maximum non-PV work (w rev) is equal
to -G (at fixed P, T), and this is obtainable only under reversible conditions. This is
a limiting case. Electrochemical cells commonly do operate under conditions of
constant temperature and pressure. However, real cells cannot operate under truly
reversible conditions. Sometimes real cells may come moderately close.

11

* Entropy can be thought of as the inverse of the "concentration" or "quality" of energy.


Energy naturally tends to disperse: heat flows to cooler bodies, unequal concentrations tend
to equalize, light moves away from its source, and so on. All this occurs naturally without
having to be forced. It just happens. Thus the "concentration" of energy always tends to
drop; energy wants to become more diffuse. This is the entropy effect. Thermodynamically
speaking, entropy naturally tends to increase; the dispersal of energy increases. THIS IS
WHAT DRIVES ALL SPONTANEOUS PROCESSES. If a process is spontaneous (or favourable)
it means that overall there is a net increase in entropy; a degradation of the "concentration"
of energy. It may occur within the system of interest (a reaction in a cell, for instance) or it
may occur in the environment surrounding the system (the "surroundings") or both.
Regardless, it is the inviolable requirement for any process to be spontaneous.
What is so marvelous about the Gibbs free energy function is that it accounts for
both the change in entropy in the system and in its surroundings. To provide a brief and less
than rigorous rationale for this, consider that G = H - (TS) = H - TS at constant
temperature. Then G/T = H/T - S, where S is the entropy change for the system. And,
H/T is q/T when the pressure is constant (U = q - w and at constant pressure w = P V, i.e.
pressure-volume work, such as the expansion of a gas. Then H = q - PV + PV = q =
heat flow at constant pressure, often denoted q P.) Under reversible conditions H/T = qrev/T
= the heat flow into the surroundings over T. Then H/T is the entropy change in the
surroundings. The minus sign in G = H - TS accounts for the fact that heat flow out of
the system into the surroundings is the opposite of heat flow in the surroundings to the
system; it takes care of the sign convention issues. Thus if G < 0 there is a net increase in
entropy within the system + surroundings, and the process is spontaneous. If G > 0 the
reaction is unfavourable (not spontaneous); if it were to occur there would be a net
decrease in entropy. This cannot naturally occur, though it can be forced with energy input.

Energy relations for electrochemical cells


Recall the first law of thermodynamics, which states the energy of the
universe is constant, or energy is neither created nor destroyed. In any system
energy can be transferred to or from the surroundings as heat or work. These are
forms of energy in transit, i.e. both are flows of energy. This is expressed
mathematically as,
U = q - w

{46}

The equation follows the engineering sign convention where,


q < 0 means heat flows out of the system into the surroundings (exothermic)
q > 0 means heat flows into system from surroundings (endothermic)
w < 0 means work flows into the system from surroundings
w > 0 means work flows out of system into the surroundings
Heat flow is considered from the perspective of the system, while work flow is
considered from the perspective of the surroundings. (In the SI convention both
work and heat flows are considered from the perspective of the system. Either
means of energy flow into the system is positive; either means of energy flow out

12

of the system is negative. Most chemistry texts follow the latter convention.) By
definition,
G = U + PV TS = H TS

{47}

If a system undergoes a change from state 1 to state 2,


G2 G1 = U2 U1 + (P2V2 P1V1) (T2S2 T1S1)

{48}

Since U2 U1 = q - w

{49}

G2 G1 = q - w + (P2V2 P1V1) (T2S2 T1S1)

{50}

and at fixed P and T,


G2 G1 = q - w + P(V2 V1) T(S2 S1)

{51}

w is the total work, including work other than pressure-volume work (PV work; e.g.
expansion of a gas against some external pressure P). For an electrochemical cell
where electrical work is also possible (by means of electrons flowing through an
external circuit),
w = w + P(V2 V1)

{52}

where w represents the electrochemical work. Substituting this into the preceding
equation yields,
or

G2 G1 = q w T(S2 S1)

{53}

G = q w - TS

{54}

Again, this applies to a process at fixed P and T. Since,


G = H - TS

{55}

It is then apparent that,


H = q w

{56}

which is equal to the heat flow at constant pressure. For a change of state achieved
reversibly, at constant temperature and pressure, we would have the equation,
G = qrev wrev - TS

{57}

An electrochemical cell for which G < 0 (favourable or spontaneous) operated


reversibly will generate the maximum possible amount of electrical work,
-wrev = -wmax = G

{58}

13

as per equation {45}. Substituting this into the equation above,


G = qrev + G - TS

{59}

or, as we would expect,


qrev = TS

(at constant temperature)

{60}

This is the flow of energy as heat in a cell operated reversibly and represents the
minimum possible heat loss (again for a cell where G < 0 which does electrical
work). Real cells, operated irreversibly will generate less work and more heat.
For an electrolytic cell (G > 0; not favourable), the minimum possible work
that we can input to force a reaction to go in the unfavourable direction is again
-w'rev (< 0). To make the reaction go at a practical rate, making the process
irreversible, the actual work input will be > -w'rev.
Definitions
The cell potential is also called the EMF (electromotive force). It is the
voltage of a cell under reversible conditions (no current is flowing, or, the current is
infinitely small). It represents the driving force for electron transfer. When E > 0
the cell reaction is spontaneous. When E < 0 the reaction as written is not
spontaneous, i.e. is not favoured. (Recall that G = -nFE.) In this discussion, in
order to distinguish charge from heat flow, the charge will be symbolized as q c.
Work (in joules) = voltage (V) x charge (C). The cell potential E and the reversible
electrical work wrev have the same sign (for the engineering convention, not the SI
convention) and are related as follows,
-G = nFE = wrev = qcE

{61}

Types of Electrochemical Cells


Now different types of electrochemical cells can be compared along with their
energy relations. The four common types of cells are illustrated in the Figure 2
below. A piece of zinc is suspended in a solution of ZnSO 4 and H2SO4. The other
electrode is a piece of platinum. Hydrogen gas is bubbled over the platinum
surface. The H+/H2 half reaction is rapid on platinum. (Rates of electron transfer
depend strongly on the surface at which they occur.) The half reactions are:
2H+ + 2e- = H2

E = 0 V

(4)

Zn+2 + 2e- = Zn

E = -0.76 V

(5)

There are two possibilities. The favourable reaction may occur, for which E > 0.
Alternatively the reaction can be forced to go in the opposite direction by applying a

14

suitably high opposing voltage. This is electrolysis. The favourable reaction involves
oxidation of Zn to Zn+2 and reduction of H+ to H2. In fact there is no thermodynamic
reason why the reaction should not spontaneously occur directly on the zinc
surface, as illustrated in Figure 3 below. However, the reduction of H + on very pure
Zn is very slow, whereas it is quite rapid on Pt. Because the reduction of H + on pure
Zn is so slow, the cell can be set up with Zn metal in direct contact with H +.
Otherwise two half cells with provision for ionic conduction would be used.

Figure 2. Schematic depiction of four different types of electrochemical cells [1].

(a)

(b)

15

Figure 3. Schematic illustration of hydrogen evolution on (a) a zinc surface (slow)


and (b) catalyzed by Pt metal in contact with the Zn (fast).
1. First consider the short-circuited cell. There is no load (no electrical work is
extracted) in the system. The reaction proceeds spontaneously since E >0 (or
E > 0 for non-standard conditions) and the process is favourable.
Zn = Zn+2 + 2e- E = -0.76 V

(6)

2H+ + 2e- = H2 E = 0 V

(7)

The overall reaction is:


Zn + 2H+ = Zn+2 + H2

E = 0 - (-0.76) = 0.76 V

(8)

This is really equivalent to the situation in Figure 3 (b). Since the reduction of H + on
pure Zn is slow, all that is required for the reaction to proceed at a substantial rate
is a catalyst, which is the role of Pt.
G = -nFE = -2 mol e-/mol x 96485 C/mol e- x 0.76 V
= -1.47 x 105 VC/mol = -1.47 x 105 J/mol = -147 kJ/mol

{62}

H = q - w and w = 0

{63}

Since,

(no electrical work is being done; there is no load),


H = q

{64}

All the energy is dissipated as heat. Heat flows from the system (the cell) to the
surroundings, and so by convention is negative. The process is exothermic. (If you
have ever short-circuited a battery by connecting a wire across both ends, you
know this from experience; the wire can get red hot and the rate of discharge of
the cell can get dangerously fast.) Since there is no work being extracted, the
potential difference is zero (w = 0 = Eqc. Therefore E = 0). Note that the cell in
principle is capable of manifesting a voltage E, but when short-circuited this is not
realized. Cementation reactions and corrosion processes are examples of shortcircuited cells. Any redox reaction occurring directly between reagents without
running the electron transfer through an external circuit is a short-circuited cell. So

16

is a cell where a wire is connected across the poles. Other examples include
oxidative leaching processes of sulfides and combustion reactions.
2. Open-circuit cell. The half reactions are the same as in (1). However, if the
electrical connection between the two electrodes is broken, no current flows. A
voltmeter can be used to measure the potential difference. This may employ a very
large resistance inside the meter. Then the current flow is so small that the rate of
reaction also is extremely slow. This very nearly approaches the reversible case (a
process carried out infinitely slowly). (Alternatively, an opposing voltage can be
applied until the current is zero. The opposing voltage slows the reaction until a
point where the reaction stops. At this point the opposing voltage is equal to the
cell voltage. These devices, called potentiometers, are not much in use anymore.)
Under standard conditions (298 K, PH2 = 1 atm, unit activities of the ions) the
measured potential difference is 0.76 V. This indicates the thermodynamic potential
difference, or driving force for the reaction. If the conditions were non-standard the
potential difference would differ from E, as per the Nernst equation, and this
would indicate the driving force under those conditions. In principle, open circuit
cells can be used to measure thermodynamic potentials. In practice, it is often not
so easy for many reasons.
An open circuit cell is essentially a cell working reversibly; the rate of the
reactions is infinitely slow by virtue of the open circuit. The cell voltage equals the
thermodynamic potential. The cell can do its maximum possible work. For all
practical purposes, however, we can't extract work from such a cell in a finite time.
It's of no practical use as far as obtaining electrical work. It is of use for measuring
cell potentials. For practical work we use the cell galvanically or electrolytically.
3. Galvanic cell (battery). In this case the electrodes are connected to a moderately
high resistance device that uses electrical energy as work, such as a radio. The half
reactions are the same as in (1). The reaction is favourable. The current passes at a
fairly low, but, finite rate. This might be close to reversible conditions of operation,
though it is irreversible. Hence the heat loss must be somewhat greater than the
minimum reversible process heat loss, and the work obtained must be somewhat
less than it would be under truly reversible conditions. Since the same charge is
being passed as it would be under reversible conditions (2e - per Zn+2) and the
available work is less, the cell voltage (denoted V) must be lower:
(w = V qc) < (wrev = E qc)

{65}

Therefore V < E (E is the cell voltage under reversible conditions; the maximum
possible voltage). The greater the current, the faster the process and the greater
the extent of departure from reversibility. Then the heat loss is greater and the
work that can be extracted is lower. One way to rationalize this is that as the
current gets high, the process is getting closer to operating like a short-circuited
cell, where all the energy is dissipated as heat.
The thermodynamic potential for the reaction is given by the Nernst
equation:

17

E = E - RTln aZn+2 PH2


nF
aH+2

{66}

Say PH2 is kept constant at 1 atm. As the reaction proceeds [Zn +2] increases, while
aH+ decreases. The ln term thus increases as the reaction proceeds and E drops.
Eventually equilibrium is reached and no further reaction occurs. Then E goes to
zero. E then is a measure of how far away from equilibrium the system is; how
great the driving force is for chemical reaction to occur.
The process will continue until chemical equilibrium is reached, at which point
G and E for the cell both go to zero. At equilibrium there is no more driving force
for the reaction to proceed; no further change occurs. At this point the battery is
dead. All real galvanic cells operate at less than the thermodynamic limit of
efficiency. However, if the cell is discharged slowly, the efficiency approaches that of
a reversibly operated cell. This is why fuel cells are naturally quite efficient. A fuel
cell is simply a galvanic cell in which the reactants are continuously replenished,
and the products are continuously removed.
The thermodynamics can be conveniently represented on a diagram as
shown below. Recall that H is a state function, meaning that for going from a
specified initial state (e.g. the left side of reaction 8 at a given temperature,

Figure 4. Summary of thermodynamic effects for galvanic and short-circuited cells.


The sign of q is negative to account for heat flow from a system being taken to be
< 0, while work done by a system is > 0. Note that in this case heat flows from the
system. In principle qrev could be positive as well. By definition, a galvanic cell is
one for which w'rev is positive; work is done by the system.
pressure and concentrations) to a specified final state (e.g. the right side of
reaction 8 with specified temperature, pressure and concentrations) the change in
enthalpy is the same no matter how the change is effected, be it reversibly,
galvanically or as a short-circuited cell. In this case H < 0; energy both as heat
and work leave the system. What does depend on how we run the cell is q and w',
but the sum, q - w', is always the same. The limiting case is the reversible cell;

18

qrev is the minimum possible heat flow and w rev is the maximum possible work.
Galvanic cells operate at finite rates, are irreversible in the thermodynamic sense
and exhibit larger heat losses and lesser capabilities for work; q > qrev and w' <
w'rev. In addition, the faster the cell is operated (the greater the current) the more
irreversible it is and the greater the heat flow and the less the work. The quantity
q is the difference bewteen w'rev and w':
q = qrev + q

{67}

G = -w'rev = -nFE

{68}

Based on equation {54},


q = G + w' + TS

{69}

q = G + w' + qrev

{70}

q = -w'rev + w' + qrev

{71}

q = -w'rev + w'

{72}

-q = w'rev - w'

{73}

Then,

The heat loss is the sum of q rev + q (which is < 0). Thus the additional heat
loss arises from inefficiency in operating the cell, relative to the limiting, reversible
case. If the cell is short-circuited no work can be extracted and all the energy
output is lost as heat. This is the other limiting case.
Another possibility for a galvanic cell is when qrev > 0. This is depicted in the
alternative diagram below.

Figure 5. Alternative summary of thermodynamic effects for galvanic and shortcircuited cells. The sign of qrev is positive to account for heat into a system being
taken to be > 0, while work done by a system is > 0. Note that the net heat flow
depends on how much work is extracted from the cell. The net heat flow is
q = qrev + q. For a galvanic cell q < 0 always; some potential work is lost as heat.

19

4. Electrolytic cell. This is what is employed in electrowinning. The electrodes are


connected to an external power supply such that the voltage exceeds and opposes
the thermodynamic cell voltage. In electrolysis the reaction is being forced in the
opposite direction of its natural or spontaneous direction. The half reactions now
are,
Zn+2 + 2e- = Zn

(9)

H2 = 2H+ + 2e-

(10)

The overall reaction is how the reverse of reaction (8):


Zn+2 + H2 = Zn + 2H+

E = -0.76 V

(11)

This is NOT favourable and will not occur naturally. To overcome this, a voltage of
>0.76 V (under standard conditions) is applied externally to force the reaction to go
as written above, i.e. Eappl > 0.76 V. The flow of electrons is reversed and so are
the electrode reactions relative to the galvanic or short-circuited cases. The H 2/H+
reaction now becomes the anode and the Zn +2/Zn process becomes the cathode.
The reaction will reach equilibrium when the thermodynamic cell voltage reaches
Eappl. Then it will stop. The thermodynamic potential for the reaction is given by the
Nernst equation:
E = E - RTln aH+2
nF aZn+2PH2

{74}

Say PH2 is fixed at 1 atm. As the reaction proceeds aH+ increases and aZn+2
decreases. Hence the log term increases and consequently E decreases (becomes
a larger negative number) as the reaction proceeds. This will continue until E and
Eappl are equal. Then the thermodynamic cell potential is just balanced by the
applied potential and there is no net potential difference between the electrodes.
The reaction stops. To make the reaction proceed further still, one would have to
increase Eappl. Under conditions of fixed external potential, the reaction rate would
decrease as the thermodynamic cell voltage decreases (because the driving force,
which is the difference between Eappl and the thermodynamic E , decreases). In
practice, electrowinning is carried out under conditions of controlled current, rather
than controlled potential, as was explained in the section on Faraday's Law
relationships. As reactants are depleted, the applied voltage must increase to
maintain the constant current. In practice, electrowinning usually takes less than
50% of the desired metal ion from the solution. The barren electrolyte after EW is
recycled to increase the metal ion tenor. Considering that E changes by,
2.303RTlog(PH2 aZn+2) = 0.02958log(PH2 aZn+2)
2F
aH+2
aH+2

{75}

even a 50% change in concentrations has only a small effect on the cell voltage.

20

Now work is being done on the cell by the surroundings; an external voltage
is applied. Then w < 0, and the work done on the cell is given by,
w = -nFEappl

{76}

where Eappl is a positive number. The work done on the cell is directly proportional
to the applied voltage. The energy changes are summarized in the diagrams below.
H = q - w

{77}

Note that w' < 0, w < w rev and -w' > -w'rev. The minimum work required to drive

(a)

(b)

21

Figure 6. Summary of thermodynamic effects for an electrolytic cell. (a) q rev > 0.
Once -w' becomes large enough excess heat is dissipated to the surroundings. Note
that q is always < 0; excess applied energy is lost as heat. (b) q rev < 0. Additional
heat over and above qrev must be dissipated to the surroundings.
the reaction against its favourable direction (backwards) to effect electrolysis is
-w'rev. In practice more electrical work is required to obtain reasonable rates;
-w' > -w'rev. The net heat is the sum:
q = qrev + q = qrev -w'rev + w'

{78}

If qrev > 0, the sign of q depends on the magnitude of w', which in turn depends on
the magnitude of Eappl (Figure 6a). Once Eappl gets large enough there is a net
heat flow from the cell into the surroundings. Practical electrowinning usually uses
Eappl >> -E (the thermodynamic cell voltage) so that heat will be evolved. If
qrev < 0 the sign of q is always negative, as indicated in Figure 6b. Referring again
to the equation,
q = G + w + TS

{79}

Substituting in,
w = -nFEappl and w'rev = -nFE

{80}

then,
q = -nFE - nFEappl + qrev

{81}

q = -nF(E + Eappl) + qrev

{82}

<0

>0

When E = Eappl q = qrev, as we would expect; the applied voltage then just
matches E and the cell operates reversibly. As Eappl exceeds E then excess heat
begins to be evolved.
This discussion assumed an isothermal system. For an actual industrial cell
the electrolyte temperature will rise and reach some steady state (although it may
fluctuate with environmental conditions). There will be heat loss to the
surroundings, but also heating of the electrolyte. Depending on the metal being
electrowon, excess heat may need to be deliberately withdrawn by heat
exchangers. The temperature of the solution will depend on the heat generated,
loss to surroundings and the heat capacity of the solution.
Electrical work is supplied to the cell in order to overcome the cell's natural
thermodynamic tendency. Some of that supplied energy does work on the cell,
some of it ends up as heat. Some of the supplied work energy results in an increase

22

in chemical potential energy. This is the net effect of breaking bonds (e.g. H-H
bonds and Zn-O bonds in [Zn(H2O)6]+2, and forming new ones (e.g. Zn-Zn metalmetal bonds and H-O bonds in H 3O+) and, finally, changes in electrostatic
interactions in the solution due to changes in composition ([H +] increases; [Zn+2]
decreases). Electrostatic interactions involve charged ions (Zn +2, H3O+, SO42-) and
dipoles (such as partial charge separation in H-O-H, the oxygen being more
electronegative and developing a negative charge; the hydrogens having a positive
charge).
Rates of Electron Transfer and the Effects of an Applied Voltage
The rate of metal plating is directly proportional to the current through the
cell, since,
I = C/sec moles e-/sec moles Cu+2/sec reacted, etc.

{83}

In electrowinning we set the current and allow the voltage to adjust accordingly (as
governed by V = IR). Thus the higher the current, the higher the applied voltage
must be. When the applied voltage precisely matches the thermodynamic cell
voltage, Eappl = -E, the cell operates reversibly and the reaction is infinitely slow.
When Eappl > -E the reaction proceeds at a finite rate, and the greater the
difference the greater the rate.
Electrode polarity
For a spontaneous reaction electrons flow from (-) to (+); repelled from the
negative electrode and attracted to the positive one. Hence the cathode is positively
polarized and the anode is negative. This accords with the fact that the
thermodynamic cell voltage is positive,
E = Ecathode - Eanode > 0

{84}

In an electrolysis the applied voltage opposes the thermodynamic voltage (E < 0)


and is greater than -E. Thus power supply (+) goes to the cell (+) and likewise the
(-) of the power supply goes to (-) of the cell. This forces the cell to run in the
opposite direction, so that now the cathode is negatively polarized and the anode is
positive. In other words, the electrodes retain the same polarity in either the
galvanic or the electrolytic cases, but the flow of electrons is opposite, as is the
direction of the chemical reaction.
Electrowinning Fundamentals
After suitable solution purification a quite pure, concentrated electrolyte may
be available for electrowinning. (In the case of copper, solvent extraction is the
common method for solution purification.) Most commonly electrowinning is
performed using sulfate solutions with oxygen evolution as the anodic half reaction:

23

M+2 + 2e- = M

(cathode)

(12)

H2O = 0.5O2 + 2H+ + 2e- (anode)


MSO4 aq + H2O l = M s + 0.5O2 g + H2SO4 aq

(13)
(overall reaction)

(14)

In such cases acid is generated during electrowinning. Electrowinning may also be


carried out from chloride solutions in some instances. Then Cl - is oxidized to Cl2.
The electrowinning reaction for copper is,
CuSO4 aq + H2O Cu s + H2SO4 aq + 0.5O2 g

(15)

The reduction reaction is:


Cu+2 + 2e- = Cu

E = 0.34 V

(16)

The oxidation half reaction is the reverse of the oxygen reduction half reaction, i.e.,
0.5O2 + 2e- + 2H+ = H2O
Then E = 0.34 - 1.23 V = -0.89 V

E = 1.23 V

(17)
{85}

The reaction is not thermodynamically favourable, so energy must be


supplied to make it go. The minimum energy required corresponds to the
thermodynamic potential difference. In practice higher voltages are used to attain
practical rates. The rate of metal plating is directly related to the current; the
higher the current, the more e-/sec transferred, and the greater the metal plating
rate. The nature of the relationships between the applied voltage and logI is
illustrated in the figure below. This depicts the current-voltage relationship for a
single half reaction. Of course another half reaction has to be at play as well; we
cannot run a half reaction in isolation. However, the graph focuses on the
log(current)-voltage graph for a single half reaction of interest *. As the current
approaches zero (logI -) the cell runs increasingly slowly, approaching
reversible behaviour. Then the measured voltage, in principle, corresponds to E, the
half reaction potential, relative to some other half reaction (e.g. the standard H +/H2

24

Figure 7. Schematic illustration of a polarization curve for a half reaction plotted as


voltage vs. logI.
* A "three-electrode" cell can be used for this. The half reaction of interest occurs at a
"working" electrode. The potential is measured relative to a reference electrode. The current
is measured between the working electrode and a "counter" electrode.

half cell.) At this point a minute change in potential can reverse the direction of the
half reaction (again, consistent with reversible behaviour). For the upper branch the
oxidation occurs, e.g.
M s = Mn+ + ne-

(18)

For the lower branch the reduction half reaction occurs, e.g.
Mn+ + ne- = M s

(19)

It all depends on how the electrode is polarized (how the external potential is
applied). Consider the cathodic branch. As the potential is decreased M n+ is
reduced. At first a substantial increase in reduction current results from relatively
small decreases in applied voltage. Eventually a nearly linear region occurs, where
logI is virtually linear with applied potential. This is called the Tafel region. Its slope
is directly proportional to n, the number of electrons/mol of metal plated. (The form
of the curves is well understood from electrochemical theory, but that is beyond the
scope of this introduction.) The curve indicates that to obtain higher currents,
higher voltages, beyond the reversible value E, are needed.
The overpotential
This leads to the idea of the overpotential. Overpotential (or overvoltage) is
the additional potential needed beyond the thermodynamic potential E required to
make the half reaction go at the desired rate. It is given the symbol . As indicated
earlier, the value of n will have a direct effect on how big the overvoltage is. For a
given n value (2 for EW involving Cu+2, Ni+2, Co+2 or Zn+2), the shapes of the curves
are often quite similar (there are subtle differences). However, what really matters
is where the curves lie horizontally. This is illustrated in the figure below. The
reduction of Cu+2 on Cu is considerably faster than the reduction of Ni +2 onto Ni.
This appears as a shift of the Ni +2/Ni curve further to the left (to lower currents)
relative to Cu+2/Cu. As a result, at a given current Ni+2/Ni is greater than Cu+2/Cu. This
has implications for the energy consumption in electrowinning of the two metals. At

25

a given current a greater overvoltage necessarily implies greater energy


consumption. For electroplating simple metal aquoions three broad classes of
processes must occur:

Deformation of the aquocation complex as it approaches the metal surface


and loss of coordinated water molecules.

Electron transfer to the metal cation.

Migration of the metal atom on the surface to a suitable crystallographic site.

Loss of water molecules may occur in concert with electron transfer. Each of these
processes contributes to the overpotential; they all have an activation barrier, or
energy hurdle that must be overcome. Sometimes one of these steps can be a
major contributor to the overpotential.

Figure 8. Schematic illustration of reduction branches of polarization curves for


Cu+2/Cu and Ni+2/Ni couples. For a given current the Ni +2/Ni couple requires greater
overpotential than does the Cu+2/Cu couple.
The anode half reaction also has an associated overpotential. Oxygen
evolution is the most common anode half reaction in electrowinning. This is
ubiquitous in Cu, and Zn EW, and common for Ni and Co. Chloride oxidation to Cl 2 is
also employed in Ni EW. In gold EW from cyanide solution, the anode half reaction
is oxidation of CN- to CNO-.
Oxygen evolution has some advantages:

No additional costly reagents are needed; H2O is the reactant.

26

Relatively less corrosive sulfate medium is suitable, and sulfate medium is


the least expensive.

Lead anodes may be used, which are inexpensive.

However, the main disadvantage is that it contributes to high energy consumption.


First the E for the O2/H2O half reaction is 1.23 V. This contributes to highly
negative thermodynamic cell voltages:
Cu+2 + H2O = Cu + 2H+ + 1/2O2 E = -0.89 V
Ni+2 + H2O = Ni + 2H+ + 1/2O2

(20)

E = -1.46 V

Zn+2 + H2O = Zn + 2H+ + 1/2O2 E = -1.99 V

(21)
(22)

In addition, water oxidation on most electrode surfaces is very slow. (This is an


area of considerable economic and technical importance for fuel cell development
too.) This results in a high anodic overpotential. Much work has gone into trying to
find ways to lower the oxygen evolution overpotential as it is a significant factor in
the operating costs of an EW plant. Lead has one of the highest overpotentials for
oxygen evolution! However, lead is commonly used due to its low cost.
Lead anodes are commonly alloyed with minor amounts of other elements to
try to lower the overvoltage. The intent is to shift the polarization curve further to
the right. This is illustrated in the figure below. At 300 A/m 2 current density the
approximate O2 evolution overpotentials under different conditions are provided in
the Table below. The most common anode material in use today in Cu EW is a PbSn-Ca alloy (~1.5% Sn, 0.1% Ca). There are numerous effects of solutes and alloy
elements on anode performance and longevity. Calcium is added to increase
strength. The anodes are cold rolled for the same reason. Tin lowers the oxygen
evolution overpotential and improves corrosion resistance. There may be other
effects as well. DSA anodes are of interest because the platinum group metal

27

Figure 9. Schematic illustration of oxygen evolution polarization curves. Various


amendments may be applied to shift the polarization curve to lower .
Table 2. Oxygen evolution overpotentials from sulfuric acid solutions under several
conditions at 300 A/m2 current density [2].
Anode material
Other conditions Overpotential (V)
Pb/Sb (e.g. 6% Sb)
~0.7
+2
Pb/Sb (6% Sb)
100-150 mg/L Co
~0.6
Pb/Ca(0.1%)/Sn(1.5%) 1
~0.6
2
DSA
~0.35
1
Preferred due to better mechanical and corrosion properties compared with Pb/Sb.
2
DSA = dimensionally stable anodes; rigid titanium sheet or mesh coated with
platinum-group metal oxides.
oxides used to coat the anodes (usually a titanium substrate) are very good at
promoting O2 evolution.
Cobalt addition is commonly practiced in Cu EW. It is believed that Co +2
oxidizes at the anode and catalyzes H2O oxidation [3]:
Co+2 = Co+3 + e-

E = 1.9 V - very high!

2Co+3 + H2O = 2Co+2 + 2H+ + 0.5O2

(anode)

(23)
(24)

Co+3 is a powerful oxidant and rapidly oxidizes water. It also lessens anode
corrosion and improves stability of the PbO2 layer. However, cobalt is expensive. It
is added at 100-200 mg/L, beyond which it has little beneficial effect.
Resistance losses
Conductors that carry electricity to the electrodes have an intrinsically low
resistance. However, these extend over the length of the tankhouse and overall the
resistance is enough to cause a moderate voltage drop. This is lost as heat.
Likewise, contact between anodes and the current distribution conductors (busbars)
and between cathode sheets and the busbars also has a resistance loss (contact
resistance). This is necessary since anodes and cathodes must be removed (anodes
in order to be replaced, cathodes for Cu metal harvesting). The other substantial
resistance in the circuit is the solution resistance. In solution the current is carried
by migration of ions. This allows for a complete circuit, without which there would
be no current. Cations move towards the cathode (negatively polarized) and anions
move toward the anode (positively polarized). Thus electrolyte conductivity is an
important technical consideration in EW.
Ions in solution are capable of conducting electricity. (This was one of the key
observations that lead scientists to conclude that some compounds were comprised
of discrete cations and anions.) A strong electrolyte is a salt that is soluble in water

28

and which fully dissociates into ions. Ions vary in their ability to conduct electricity
(per unit concentration). The hydrogen ion, H +, is by far the best conductor,
followed by OH-, then other cations. H+ is at least 3 times more conducting (per unit
concentration) than other cations*. Thus the presence of H 2SO4 in the electrolyte is
quite
beneficial.
Copper
EW
electrolytes
contain
on
the
* The reason is believed to be due to the "proton jump" mechanism. H + is present as H3O+.
An H+ in an H3O+ ion can "jump" to a neighbouring water molecule as shown below. This
allows it to move through solution much more rapidly than the diffusion or migration of
other ions. The same mechanism may be in effect for OH -, though it is somewhat less
effective, as indicated by its lower conductance.

order of 180 g/L H 2SO4. Much beyond this and corrosion of the steel starter sheets
and other parts of the plant becomes problematic.
EW is operated at constant current, so based on Ohm's law the voltage drops
due to resistance losses can be summed up:
VIR = IRi

{86}

The resistance of a resistor is a function of its geometry and its inherent resistivity
(the resistance of a standard geometry). The longer the distance through which the
electrons (or ions) must travel, the greater the total resistance; R L where L =
length. The greater the cross sectional area of the resistor, the more current it can
support (more pathways for electrons/ions to move through). Hence R 1/A where
A is the cross sectional area. Then,
R L/A

{87}

R = L/A

{88}

and

where = resistivity. This is illustrated in the figure below. Thus in EW large surface
area electrodes are used with a minimum practical spacing between cathodes and
anodes. Limitations on these features are discussed later. Clearly the lower the
resistivity the better. Thus a more concentrated solution of ions, and where feasible,
high concentrations of H+ are desirable. In electrowinning of divalent metal ions
from aqueous solution there is a complicating factor and that is ion pairing. This
essentially the formation of a complex, e.g.
M+2aq + SO42-aq = MSO4 aq

(25)

The extent to which this happens depends on the metal ion, temperature and pH;
recall that SO42- is also a weak base and may form HSO4- as well. A neutral ion pair
does not contribute significantly to the conductivity of the solution. Some example

29

Figure 10. Illustration of the dependence of resistance on conductor length and


cross sectional area.
data for simple CuSO4-H2SO4 electrolytes are shown in Table 3. Note that for a
given acid concentration (50C column) the resistivity increases with increasing
[Cu+2], i.e. [CuSO4] (as well as decreased [H+] due to formation of HSO 4-), while
higher acid concentrations lower the resistivity. Iron is also commonly present in
copper EW electrolytes at concentrations of up to 5 g/L. It will be present in its two
common valence states: Fe+3 and Fe+2. This also will affect the resistivity of the
electrolyte. It is also common to consider the inverse of resistance or resistivity
instead. The inverse of resistance is conductance. The specific conductance is 1/ in
-1cm-1. Now higher conductance (lower resistance) is desirable.
Table 3. Illustrative example data for resistivity of simple CuSO 4-H2SO4 solutions at
different temperatures [4].
Resistivity cm
Cu+2 g/L H2SO4 g/L
30C
40C
50C
25
71.7
3.75
3.52
3.26
35
143.4
2.29
2.09
35
191.2
1.92
1.74
40
163.4
1.50
40
182.6
1.42
50
163.4
1.71
50
182.6
1.57
55
163.4
1.76
55
182.6
1.63
60
167.3
2.37
2.15
As an example, take the cathode-anode gap to be 4.8 cm, and the immersed
plating area of 1.2 m 2 (100 cm x 120 cm). A typical electrolyte might contain 40
g/L Cu+2, 180 g/L sulfuric acid and have an operating temperature of 50C
(although there may be a significant range of conditions in practice). The
corresponding resistivity then is about 1.42 cm. The resistance then is:

30

1.42 cm x 4.8 cm / 12000 cm2 = 0.000568

{89}

The voltage drop across the resistance is V = IR. For a 300 A/m 2 current density,
0.000568 x 1.2 m2 plating area x 300 A/m2 = 0.204 V

{90}

The four contributions to applied cell voltage


Putting the pieces together we have the following relationship for the applied
voltage:
Eapplied = -E + C + A + IRi

{91}

The first term is the thermodynamic potential (E is negative and must be


opposed). Next are the two overpotentials. The last term is the sum of the IR
voltage drops. Note too that increasing the current increases this term. Likewise the
overpotentials are a function of current (or current density). -E is constant for a
given electrolyte composition and changes only a little through the cell as metal is
plated. Ballpark estimates for the four terms in copper EW are given below:

-E ~ 0.9 V
Anodic overpotential (A) ~ 0.5 V (depends on current density)
Cathodic overpotential (C) ~ 0.1 V (depends on current density)
IRi ~ 0.5-0.6 V (depends on current); solution resistance ~0.3 V and
i

rectifier + cell hardware resistances ~0.2 V (The rectifier converts AC


power to DC power.)
Total applied voltage ~2V in the example above. Depending on the source
cited, there is some variability in values quoted for solution resistance and cell
hardware resistances. However, a total voltage drop dues to resistance losses of up
to 0.6 V or so is typical. In general industrial plants employ cell voltages of 1.8-2.2
V. Note that this significantly exceeds -E. The actual cell voltage (measured
between anode and cathode) and the total applied voltage (including the power
source and current distribution network) may differ somewhat, depending on
whether the hardware and rectifier resistances drops are included or not. In the
end, what matters is the total applied voltage and the total current. This is what
determines the energy consumption and associated cost.
Limiting and practical current densities
From Faraday's law and related relationships, it is clear that plating at higher
current densities is desirable; for a given plating surface area production increases
with increasing current density. However, there are two important limitations. The
first is a fundamental limit. Referring back to Figure 7, at higher currents the

31

polarization curves start to approach infinite slope (tend toward vertical lines). As
metal ion is plated, the concentration drops near the surface, relative to the bulk
concentration. This creates a concentration gradient, which induces mass transport
of Mn+ towards the cathode. The greater the plating rate, the bigger the
concentration drop, i.e. the lower the concentration at the surface. This is illustrated
in the figure below. At some critical current density the concentration of metal ion
at the surface goes to zero. This is the diffusion limited current and represents the
maximum current that can be sustained under the given conditions. At this point
the polarization curve approaches infinite slope. Further increases in current cannot
be sustained by Mn+ reduction. Then the next available half reaction will begin to
occur. Reduction of H+ to H2 may then occur, for which E = -1.23 V. An increase in
the applied voltage then would also necessarily occur (at constant current). This
wastes electricity and would never be attempted in a normal electrowinning
situation. For copper EW the diffusion limited current density is about 500 A/m 2
[5].

Figure 11. Schematic illustration of concentration profiles for a metal ion being
plated at an electrode surface. Distance is measured from the cathode surface. The
dashed line represents the boundary layer thickness.
In practice significantly lower current densities are employed for conventional
EW. This is related to a practical limitation. At very high plating rates the copper
atoms deposited on the cathode surface do not have time to migrate to a suitable
crystallographic site. The net result is that large numbers of new crystals form on
the surface, rather than growing existing crystal faces. This makes for very crumbly
deposits that adhere very weakly to the cathode surface, making cathode
harvesting difficult and costly. Fine grained copper particles would spall off the
cathode and have to be collected, filtered and washed. At lower, but still too high
plating rates too many crystals are still forming; the deposit may be more
adherent, but it will be porous as crystals grow together rapidly and trap solution
between their faces. This will increase the sulfur (from electrolyte sulfate), oxygen
(from water, etc.), iron, etc. impurities content in the cathodes and make them offspec. Thus in conventional copper EW maximum current densities are about 350
A/m2 [6].

32

The 500 A/m2 limiting current density is for a solution without intentional
agitation. The vertical line in Figure 11 demarcates the point closest to the
electrode surface where the solution metal ion concentration is equal to the bulk
solution concentration. This is called the boundary layer. If the boundary layer is
made thinner by agitation then the concentration gradient steepens and higher
limiting (and practical) current densities are possible. However this requires energy.
A typical Cu EW tankhouse may have several hundred cells, each with as many as
60 cathodes and 61 anodes, closely spaced together. Agitation then becomes
difficult. However, if electrolyte is directed up between cathodes and anodes using a
header, this can impart some additional agitation allowing current densities to be at
the higher end of the practical range. The minimum thickness of the boundary layer
is about 0.01 mm, which requires intensive agitation [7] and is not achieved in EW.
There are two features in an EW cell that result in a measure of natural
agitation (actually convection) [5]. At the anode surface oxygen gas is evolved.
This gas pushes up solution as it rises. The bubbles accumulate at the surface
before rupture. Overall this displaces solution in an upward motion in the vicinity of
the anode. At the electrode surface Cu +2 is depleted, lowering the [Cu +2] and
decreasing the solution density. This causes solution to naturally rise near the
surface of the cathode. The net result is two counter-rotating loops, as shown in the
figure below, which also helps to thin the boundary layer near the cathode surface.
The boundary layer thickness achieved by this natural convection is about 0.1-0.2
mm [7]. Directing electrolyte up between electrodes using a header further thins
the boundary layer, though not to the limit of ~0.01 mm.

Figure 12. Illustration of natural convection in an EW cell based on the rising of


lower density, copper depleted electrolyte near the cathode and rising oxygen gas
bubbles near the anode.
Current distribution and protrusions
Anodes and cathodes are large plane sheets that are kept parallel. This
promotes a uniform distribution of the electric field over the surfaces (other than at
the edges). This in turn promotes uniform plating rate over the surface, which is
important for growing a smooth, compact deposit. This lowers porosity and helps
prevent occlusion of electrolyte with the attendant increase in impurities in the

33

deposit. However, deposits are polycrystalline and surface roughness will eventually
develop. Then there are high and low spots on the surface.
Electric field lines tend to converge at points and diverge at depressions. This
leads to the situation shown in the diagram below. This makes the tips grow faster
and the depressions grow slower, further increasing surface roughness and
promoting growth of protrusions. Further, the high points are closer to the anode
and the depressions are further away, decreasing the IR voltage drop due to
solution resistance for the high points and increasing it for the depressions. Finally,
high points extend out into the boundary layer where the [M n+] is higher, making
electroplating easier. All totaled these effects can result in more rapidly growing
protrusions. As these extend further out from the surface their growth rate
increases. Then a polycrystalline agglomeration of copper called a dendrite will
extend out towards the anode, eventually making contact and causing a shortcircuit. At this point all the electrical energy in the cell is lost as heat. The current
follows the path of least resistance through the short circuit, rather than through
the solution to plate copper. This in part acts to limit how close cathodes and
anodes can be to each other.

Figure 13. Schematic illustration of the effects of protrusions and depressions on


electric field distribution. Due to differing distances IR" > IR > IR'. The figure at
right is a schematic illustration of a short-circuit caused by a dendrite.
Plant operators take pains to minimize this problem. This can be done by
using "leveling" agents. These are often complex chemical mixtures that act by
selectively adsorbing on the fastest growing sites. This increases the resistance of
the protrusion and helps slow down their growth [8]. It is common practice to use
infrared scanners to detect hot spots in cells where a short circuit has occurred. An
operator will then go to the cell and use a bar to dislodge the dendrites to resume
plating.
Another approach to controlling surface roughness is periodic current reversal
(PCR). The polarity of the cathodes and anodes is switch briefly at frequent
intervals. The copper electrode now becomes the anode for a short time. The

34

protrusions being the most active sites, as indicated above, are most rapidly
oxidized, dissolving them. This effectively prevents growth of protrusions on the
cathode surface. Copper metal will plate on the lead electrode, briefly, but this will
again be completely oxidized off. The duration of the polarity switch is brief. This
may obviate the need for addition of leveling agents and associated costs, however,
it does cost electricity.
Copper Electrowinning Practice
Specifications
LME (London
specifications:

Metals

Exchange)

grade

copper

has

the

following

Pb < 5 ppm
S < 15 ppm
Other impurities <49 ppm
Cu >99.993%
A major source of sulfur impurity is sulfate from occluded electrolyte.
Cells and hardware
A diagram depicting a simplified cell is shown below in Figure 14. Cathodes
and anodes are connected in a parallel arrangement. Cathodes are interleaved
between pairs of anodes. A cell will have n cathodes and n +1 anodes. Then metal
may be plated onto both sides of every cathode, making the most of the available
plating surface area. A typical Cu EW cell will have up to 60 cathodes and 61
anodes. Cathode copper is often plated onto stainless steel blanks or starter sheets,
usually ~3 mm thick and 1 m X 1-1.2 m surface dimensions [6]. Stainless steel is
used to minimize corrosion. Sometimes copper starter sheets are used instead.
These are plated separately and are 0.5-1 mm thick. The starter sheets are
removed from a cathode substrate (such as titanium, to which copper adheres very
weakly) and placed into the main EW cells. A schematic illustration of a cell is
shown in Figure 15 below.
Cell dimensions are on the order of 6 m long x 1.4 m deep by 1.25 m wide
[9], with some variation. Long conductive bars called busbars conduct electricity to
the electrodes. Previously the electrolytic cells were made of concrete and had to be
lined to prevent corrosion. PVC was a common liner. Modern cells tend to be made

35

Figure 14. Schematic illustration of a copper electrowinning cell electrode


arrangement.

Figure 15. Schematic illustration of a copper EW cell.


of polymer concrete (aggregate in a plastic binder). Cells are electrically insulated
from ground to minimize stray currents that lower current efficiency. Regular
cleaning is required to remove lead sludge due to slow anode corrosion (mainly as
PbO2). Excessive build-up can eventually compromise cathode quality [10]. A
typical tankhouse will have hundreds of cells.

36

The cathode-anode gap needs to be as small as possible to minimize the IR


loss due to solution resistance. Cathodes are grown to a thickness of about 0.5 cm
and due to misshapen electrodes and cathode protrusions there is a limit on how
small the gap can be. The cathode-cathode centre-line spacing is commonly
9.5-11.4 cm [11]. With 0.3 cm thick steel cathode sheets and 0.6 cm thick anodes,
and cathodes that grow to 0.5 cm thickness, the anode-cathode gap is about 4.35.3 cm at the start of cathode growth and 3.8-4.8 cm at the end. Steel starter
sheets, being more rigid than thin copper starter sheets, have allowed for the
narrowing of the cathode-anode gap, which lowers resistance losses.
Anodes are typically made of a lead alloy as mentioned earlier, although the
DSA's are starting to appear. Typically anodes are ~0.6 cm thick [6] and about 3
cm shorter on each side to promote uniform current distribution on the cathodes
[12]. A PbO2 layer forms on the anode surface, which is also conductive. With tin
alloy electrodes a SnO2 layer also forms. Due to the formation of the PbO 2 layer the
corrosion of lead is very slow; PbSO4 is also quite insoluble. Gradually small
particles of PbO2 spall off the anodes. This can lead to some lead contamination of
the cathodes, if say, a PbO 2 particle lodges on the cathode surface. Lead anodes are
chosen for their longevity and insolubility. Lead contamination of cathode copper
can be kept to <1 ppm [5]. A cold-rolled Pb alloy anode can last 5-10 years [13].
Anode consumption is 0.013-0.17 anode/tones copper plated (~1-12 kg/tonne Cu)
[11]. Their disadvantage is a very high overvoltage for O2 evolution.
Electrolyte
Keeping the Cu+2 concentration in the solution high helps with sustaining a
high current density. In order to keep the cathode quality similar throughout the
cell the total drop in electrolyte copper concentration within each cell is typically
quite small (2-5 g/L). Then the cathodes at the front of the cell experience nearly
the same composition as those near the exit. The lower the rich electrolyte copper
concentration, the less tolerance there is for change in copper concentration. Hence
lower rich electrolyte copper tenors may limit a plant to lower current densities
[12]. Rich electrolytes can have as little as 25 g/L Cu +2 [12]. Most plants have rich
electrolytes with 30-38 g/L Cu +2 [6] and concentrations of up to 50 g/L are
possible. Low rich electrolyte Cu+2 concentrations and low concentration changes
between electrolyte entering and exiting the cells may necessitate smaller numbers
of cathodes per cell (<60).
Other compounds in the electrolyte include 140-150 g/L H 2SO4, up to 3 g/L
Fe+2/+3, up to 150 mg/L Co +2 and 30-40 g/L Cl-. A common leveling agent in Cu EW
is a derivative of guar gum [8]. Guar gum is a polysaccharide; a polymer of
mannose and glucose. Again, this coats on the fast growing protrusions and slows
their growth. Addition rates corresponding to 150-300 g/tonne of copper produced
are typical.
As mentioned previously Co+2 is added to lower the O2 evolution
overpotential. However, another beneficial effect is that it seems to promote the
adherence of PbO2 to the anode surface [8,13].

37

Iron enters the electrolyte from solvent extraction. In part this arises from
chemical absorption of Fe+3 by the hydroxyoxime extractants, and in part it comes
from entrainment of aqueous phase in the loaded organic. In a good plant the ratio
of Cu:Fe entering the electrolyte is 1000:1 on a mass basis. Iron up to a point has
some beneficial effects. It helps to promote a smooth copper deposit [14].
However, iron also lowers current efficiency. Ferric is reduced to ferrous at the
cathode and ferrous is oxidized back to ferric at the anode. Thus Fe +3/+2 cycles back
and forth and consumes current, lowering copper plating current efficiency. This is a
major source of current losses in copper EW. Up to about 3 g/L iron appears to
provide a good balance between promoting smooth Cu deposits and unduly
lowering current efficiency [8].
Chloride at 30-40 mg/L promotes smooth copper deposits [15]. Excess
chloride is harmful as it promotes corrosion of the steel cathode blanks.
Manganese may enter the electrolyte from solvent extraction by entrainment
of aqueous solution in the loaded organic if the PLS is high in manganese. This can
cause various problems in EW. Mn+2 can be oxidized at the anode to form MnO4-:
Mn+2 + 4H2O = MnO4- + 8H+ + 5e-

(26)

Permanganate is a powerful oxidant that can oxidize organic extractant and solvent.
It can also oxidize ferrous to ferric. In either process solid MnO 2 may form:
MnO4- + 3Fe+2 + 4H+ = MnO2 + 3Fe+3 + 2H2O (27)
Manganese dioxide is itself a strong oxidant. In addition the fine particles may foul
cathodes and cause crud formation in the stripping operation of SX [16].
Plant operations
Copper is usually plated to a thickness of about 5 mm (40-60 kg each) over
the course of 5-10 days. Much beyond this thickness and the mass becomes too
great and the cathode may dislodge from the steel sheet under its own weight. (It
is important that the copper not adhere too strongly to the cathode substrate,
otherwise it would be too difficult to remove.) Copper plated onto copper starter
sheets is plated to about 100 kg of mass. An average annual production rate is
50,000 tonnes per year. The largest plants produce about 100,000 t/y. Current
efficiency is typically 85% - 95% . Typical energy consumptions are 1900-2000
kWh/t Cu [6,9]. Figure 16 shows photographs of a modern EW tankhouse.
Either 1/3 or 1/2 of the cathodes are harvested from a cell at any given time.
(Hence the number of cathodes per cell must be evenly divisible by three or two.)
The rest remain in the cell so that it can keep operating [9] (called live stripping).
Cathodes are mechanically stripped from the steel sheets.

38

Cells are operated at 45-50C solution temperature [8]. This arises in part
from the electrical resistance heating directly from the EW process. The electrolyte
heated by the waste heat from electrolysis (solution resistance) is used to preheat
incoming electrolyte [10]. Good control of temperature is important for obtaining
smooth, dense deposits (fine-grained) [6,10]. It may be necessary to cool the lean
electrolyte returning to stripping to minimize organic degradation reactions [12].
Oxygen evolution at the anodes forms an acid mist when the gas breaks the
surface. This is corrosive and toxic and much effort goes into controlling it. It also is
a small loss of copper values. Control measures include polyethylene balls on top of
the surface to promote coalescence, surfactants that produce a foam layer and
prevent mist escaping and very good cross ventilation.
In many older plants electrolyte was introduced at one end of a cell and
exited from the other without any deliberate agitation. Sometimes poor quality
(rough) cathode deposits were formed. It was found that introducing the electrolyte
via a distribution manifold led to consistently better cathode quality with the added

39

(a)

(b)

Figure 16 [17]. (a) A copper electrowinning tankhouse. This example uses steel
starter sheets. Each parallel bank of electrodes is a cell. Openings along the right
wall are for fans for ventilation. Notice that the tankhouse accommodates two
parallel rows of cells. (b) An illustration of live stripping where a fraction of the
cathodes are removed and electrowinning continues in the cell.

40

advantage of allowing higher current densities (around 320 A/m 2 at the plant where
this was initiated) [5]. The manifold directs electrolyte flow upwards between the
electrodes from all across the bottom of the cell. This is illustrated in Figure 17
below. A typical good flow rate is 0.12 m 3/h/m2 of plating area (number of cathode
faces x area per face) [5,8]. The maximum is about 0.14 m 3/h/m2 [6] (too high
and PbO2 particles can stay suspended). If the electrolyte flow into the cells from
SX is too low to sustain this then additional recirculation pumps may be needed.
Roughly half of the worlds EW plants in 1999 had electrolyte inlet manifolds [13].
Strips or moldings or wax are placed on the edges of the cathodes to prevent
copper plating on the edges [5]. Plating on the edges would bind the cathodes on
each face together, making cathode stripping difficult. Uneven current distribution
at the edges produces dendritic growths and this copper is not readily recovered.
Spacers are also used to keep anodes and cathodes from coming into physical
contact. Another means of preventing contact is anode buttons [12]. These may be
made of PVC and attached to anodes at the corners and in the middle. Spacers on
various sorts keep a sufficient gap between anodes and cathodes so that exfoliating
PbO2 does not as readily come in contact with the cathodes [9].

Figure 17. Illustration of electrolyte circulation manifold developed at San Manuel


copper plant [5,8]. P.E. means plant electrolyte. Electrodes are not shown.
Bleeding for impurity control
Inevitably impurities will build up in the EW electrolyte [5], whether by
entrainment of aqueous solution in the loaded organic or chemical extraction. This
occurs because the flow of electrolyte between EW and SX stripping is a closed
loop. This leads to accumulation of iron especially, and depending on the ore being
leached, also Cl-, NO3-, Mn+2 and others. Several of the deleterious effects of
impurity elements have been mentioned previously. Removing impurities is
commonly done by bleeding. A small fraction of the lean electrolyte is removed
from the circuit (commonly 1-3% [8]).

41

As mentioned, in a good plant the Cu:Fe ratio entering the electrolyte from
SX is 1000:1, i.e. the concentration of iron in the electrolyte is 1/1000 that of the
copper, which is very good selectivity. If we take as an example 150 t/day copper
production (54,750 t/year) the iron entering the electrolyte is 0.15 t/day. If the iron
concentration that can be tolerated in the electrolyte is 2 kg/m 3 the bleed rate can
be calculated:
(0.15 t Fe/day x 1000 kg/t) / 2 kg Fe / m3 = 75 m3/day = 3.125 m3/hr

{92}

For instance, consider a 50,000 t/yr copper plant with Cu from SX = 10 g/L
(difference between rich electrolyte and lean electrolyte [Cu +2]). The electrolyte
flow rate then through EW is 571 m 3/hr. The bleed rate then represents only about
0.55% of the electrolyte. Obviously this should be kept to a minimum. The higher
the iron content that enters EW, the higher the bleed rate must be. Complete
removal from the electrolyte by bleeding is not feasible; then the whole electrolyte
would have to be treated. Then too a little iron has beneficial effects on cathode
quality.
The bleeding process necessarily removes cobalt from the electrolyte. Since
cobalt is quite expensive this adds a significant cost. Obviously, again, it is
important to keep the bleed stream flow to a minimum. This is best achieved by
having a very well operated SX plant to keep the Cu:Fe ratio high.
The bleed solution may be treated in a number of ways. It can be returned to
leaching or to solvent extraction loading. The acid concentration is quite high, so it
may have a mildly negative effect on extraction. Where iron, nitrate or chloride are
at high levels in the PLS, the loaded organic may have to be treated to lower the
impurities concentrations [8]. Then another stage is added in SX prior to stripping.
This is called scrubbing [18]. High copper and somewhat higher acid than in the
PLS can displace iron and with it Cl - or NO3-. Some of the bleed (diluted) can be
used for this purpose.
An ion exchange process for iron removal (ferric only) from EW electrolytes is
also available [19]. The resin selectively removes Fe+3. Once the resin capacity is
saturated the resin is stripped using a Cu +2/Cu+/H2SO4 solution; then the resin can
be reused. The great advantage is that no cobalt is lost from the electrolyte.
Plant configuration and electrolyte storage
Rich electrolyte from SX goes into an electrolyte storage tank. The flows to
and from this tank are shown in the figure below. It is common practice to split the
lean electrolyte flow between storage and return to SX [6]. This is necessary
because the change in copper concentration between rich and lean electrolytes may
be greater than the change in copper concentration between the inlet and outlet of
the cells. Note that electrolyte makes one pass through the cells and then is
returned to stripping as lean electrolyte.

42

The fraction returning to SX ranges from 3-100% (100% means no splitting


of the flow), but in most cases it is about 20-40% [6]. The storage tank provides
for a consistent composition electrolyte in each cell in the EW plant. The copper
concentration drops by 2-5 g/L as it passes through each cell. It depends on the
total copper plating rate in the cell and the electrolyte flow rate through the cell.
Also as noted previously, the specific flow rate of electrolyte to each cell should be
around 0.12 m3/h/m2 plating area, at least when higher end current densities are
used. The arrangement results in small change in copper concentration across the
cell (about 2-3 g/L), ensuring that cathode quality is uniform across the cell. The
copper delta for stripping (difference in copper concentrations for the rich and lean
electrolytes) is often quite a bit higher (up to 15 g/L) [6].
Note that the total mass flow through EW must equal the total mass flow
through stripping due to mass balance. (We are omitting the lean electrolyte bleed
at this point. When that is considered then mass flow through stripping = mass
flow through EW + bleed.) However, the concentration changes through stripping
and EW may differ significantly. This requires that the flow rate through EW
(designated E) must differ from that through stripping. The mass flow through
stripping is As([Cu+2]rich elec. - [Cu+2]lean elec.), and this must equal the mass flow
through EW, i.e. E([Cu+2]into EW - [Cu+2]lean elec.). This in turn means that [Cu+2]into EW
[Cu+2]rich elec.! Thus the rich electrolyte cannot be run directly into the EW cells! A
flow configuration that accommodates these requirements is shown in Figure 18.
The lean electrolyte exiting EW is used to dilute the rich electrolyte from stripping
so that after passage through EW, its concentration equals that of the lean
electrolyte.
The strip solution flow from/to SX is designated A s m3/h. The rich electrolyte
copper concentration is [Cu+2]R, and for the lean electrolyte it is [Cu +2]L. The
difference in concentrations [Cu+2]R [Cu+2]L = copper . The electrolyte flow to EW
is E and its concentration is [Cu +2]t. The change in copper concentration across the
cell can be represented as [Cu+2]tL. Designate the specific flow rate as fs.
The copper production rate is set by the net amount of copper taken up by
the organic in extraction, all of which is transferred to the electrolyte during
stripping. For a plant producing 50,000 t/y copper the hourly production rate is
5707.76 kg/h. Take as an example a rich electrolyte with 45 g/L Cu +2 and a strip
copper of 15 g/L. Then the lean electrolyte [Cu +2] is 30 g/L. The strip solution
flow rate is then given by the copper production rate and the copper :
As = dMCu/dt =
Cu+2] strip

5707.76 kg/h = 380.518 m3/h


15 kg/m3

{93}

(mass per unit time / mass per unit volume). The copper production rate from
equation {25} and equation {26} is:
dMCu = (j CE)(Ac N S) mol x 63.546 g x 10-3 kg x 3600 sec
dt
100nF
sec
mol
g
h

{94}

43

= 1.185499 x 10-5 (j CE)(Ac N S) kg Cu/h

{95}

Figure 18. Schematic of flows between SX and EW and how electrolyte storage is
used. The [Cu+2]t is such that after EW the electrolyte [Cu +2] = lean electrolyte
[Cu+2]. Some of this returns to stripping, and some to the tank. This allows for
[Cu+2]stripping to be > [Cu+2]EW, and [Cu+2]R to be > [Cu+2]t into EW.
where Ac is the total plating area per sheet (both sides), S is the number of cells,
and values for j, CE, Ac, N and S are to be chosen. Similarly, the flow rate to the
cells, E, is given by the copper production rate divided by the change in copper
concentration across each cell:
E = 1.185499 x 10-5 (j CE)(Ac N S)
[Cu+2]tL

{96}

Ideally, the flow rate E should be such that the specific flow rate is
0.12-0.14 m3/h/m2 plating area. The total plating area is Ac N S. Then,
fs =

E
= 0.12 (or a similar value)
Ac N S

{97}

With these two equations we can solve for [Cu+2]tL:


E = 0.12Ac N S = 1.185499 x 10-5 (j CE)(Ac N S)
[Cu+2]tL

{98}

[Cu+2]tL = 1.185499 x 10-5 (j CE)

{99}

0.12

44

Take as an example j = 300 A/m2 and CE = 95%. Then for each cell,
[Cu+2]tL = 1.185499 x 10-5 x 300 x 95 = 2.816 kg/m3

{100}

0.12
If fs is smaller, then the concentration change will be larger (lower flow, longer
residence time for Cu+2 to be depleted). The mass balance for copper in Figure 18
is:
dMCu = As([Cu+2]R [Cu+2]L) = E([Cu+2]t [Cu+2]L)
{101}
dt
Hence, E = As [Cu+2]strip

{102}

[Cu ]tL
+2

E = 350.518 m3/h x 15 kg/m3


2.816 kg/m3

= 2027.22 m3/h

{103}

Assuming a conventional cell with 60 steel cathode starter sheets, with 1 m x 1 m


plating area per side, the number of cells then is given by using either equation
{96} or equation {97}:
S=

E
Ac N fs

2027.22 m3/h
2 m2 x 60 x 0.12 m3/m2/h

= 140.8

{104}

There will be 141 cells in use. In a real tankhouse, there will be more cells in order
to allow for downtime, maintenance and cleaning. (Typically the tankhouse is split
in two, with a central cathode harvesting area. Then a tankhouse will have an even
number of cells.) The nominal copper production rate will be:
1.185499 x 10-5 x 300 x 95 x 2 x 60 x 141 kg/h
= 11433.2 kg/h = 50,078 t/y

{105}

The flow rates through EW will be:


As = 380.5 m3/h
E = 2027 m3/h
E As = 1647 m3/h
The fraction of spent electrolyte going to SX stripping is:
Fraction to SX = As = 18.8%
E

{106}

45

The strip [Cu+2] is 15 g/L (as specified) and the [Cu+2] from the tank to the cells is
32.8 g/L.
Higher current efficiency, higher current density and higher cathode surface
area increase the copper plating rate in the cells. This necessitates a higher
circulation rate through the cells (E). The lower the copper concentration drop
across the cell, the higher the specific flow rate must be, and this too requires a
higher rate of flow of electrolyte to the cells. However, there is a limit to how high
the specific flow rate can be; too high and spalling PbO 2 particles from the anodes
may remain suspended in solution longer and contaminate the cathodes. Higher
flow rates also incur greater cost for pumping. Sometimes there are two electrolyte
storage tanks. Electrolyte clean-up processes may be conducted on solution flowing
from the first tank (e.g. column flotation to remove entrained organic) [16]. The
lean electrolyte should probably not have <30 g/L copper. A high concentration is
needed to allow for good quality copper plating at high current density.
Recent developments
Dimensionally stable anodes (DSA) have been in development for many
years. As mentioned previously, these are composed of a titanium substrate coated
with platinum-group metal oxides. They dramatically lower the oxygen evolution
overpotential, which in turn lowers the applied voltage, and results in cost savings
in energy utilization. Of course, the coatings are expensive. Nevertheless, the cost
savings in energy consumption appear to more than compensate for the cost. Thus,
there has been a move towards adopting DSA's into EW plants [20,21]. These
anodes do not required CoSO 4 addition either and labour and down time to clean
lead corrosion sludge out of cells is no longer necessary, both of which are cost
savings. Removing lead metal and compounds from the plant also improves worker
health and safety.
Current densities of up to 450 A/m2 are being employed in some modern
plants [21]. However, this will tend to produce rough, nodular deposits with
consequent increase in short circuits and loss of current and energy efficiency. What
is required is a decrease in the boundary layer thickness to steepen the Cu +2
concentration gradient and increase the limiting current density. (Recall that
practical current densities are roughly 2/3 of the limiting current density.) Thus air
sparging of the electrolyte in the cells is also practiced to improve mass transport of
Cu+2 to the cathode surfaces. Another problem is that traditional additives used to
promote smooth plating (guar derivatives) begin to break down more rapidly as
current density begins to exceed 300 A/m 2. This has necessitated development of
new, synthetic additives. Maintaining evenly spaced and parallel cathodes and
anodes is also increasingly important as current density increases.
There has been a move to add hoods or covers to copper EW cells [21].
These have been in use for many years in nickel EW plants where chlorine gas
evolution occurs at the anodes. The gas is much too toxic to allow it to escape into
the tankhouse and the atmosphere; it is also captured for use in leaching
operations. In copper EW hoods are starting to be used to prevent acid mist

46

accumulation in the tankhouse. This becomes all the more important at higher
current densities and with air sparging. It obviates the need for foaming agents to
reduce acid mist. Mists are directed to scrubbers. However, it also prevents manual
inspection of the cells to look for problems. New types of sensors (such as online
monitoring of cell temperatures and voltages) are needed to detect short circuits.
Electrowinning other Metals
Figure 19 shows potential/molal ion activity relationships. These E values are
essentially plots of E for the half reactions relative to the standard hydrogen
electrode (as per the Nernst equation) versus ion activity (on the molal scale). The
values at the right hand axis are E values for the various metal ion reductions
(where a = 1 m):
Mn+ + ne- M

(28)

The slopes are derived from the Nernst equation at 25C:


E = E - (0.05916/n)log(1/aMn+)

{107}

Note again that orders of magnitude change are needed in ionic activities to effect
large changes in E. In electrowinning it is desirable to have high electrolyte
concentrations to maintain high energy efficiency (lower applied voltages). From
equation {107} as the metal ion concentration drops, E Mn+/M also decreases, and E

47

Figure 19. Potential-ionic activity relationships for several metal ions and H +/H2 on
Pt and O2/H2O on Pt at 25C.
becomes a bigger negative number (Eappl has to increase):
E = EMn+/M - EO2/H2O; EO2/H2O is near 1.23 V

{108}

In addition, lower concentrations will necessarily result in lower limiting current


densities, making the operation inefficient. Most metal sulfates are soluble to a
maximum of around 1 M, so electrolysis on concentrated solutions tends to be
feasible.
Note the positions of the H+/H2 and O2/H2O lines. Generally, metal ions with
E > EO2/H2O cannot electrowon from aqueous solution. They are too strongly
oxidizing and the simple cations do not persist in water; water would be oxidized by
them. So, for instance, Au + (EAu+/Au = 1.69 V) cannot be formed in water at any
practical concentration. (Some complexes of gold have lower E values and can
persist in aqueous solutions.) Most metal ions have E Mn+/M < 0 V, so that hydrogen
evolution is the favoured process, thermodynamically speaking. Aluminum and
magnesium cannot be electrowon from aqueous solution. The potentials are so
negative that only H2 evolution could occur.* Relatively few metals have
E Mn+/M > 0 V. The higher this E value, the more noble the metal is said to be. For
these metals hydrogen evolution is not a concern.
For metals with EMn+/M < 0 V the rate of hydrogen evolution relative to metal
plating rate is a crucially important question. Plots of rate of hydrogen evolution on
the surfaces of various pure metals are shown in Figure 20. The values plotted are
the exchange current densities (symbol Jo). The exchange current is the current
when both the anodic (H2 = 2H+ + 2e- here) and cathodic (2H+ + 2e- = H2 here)
half reactions for a given couple have the same rate. Referring back to Figure 7 this
is the point where the linear regions of the polarization curves intersect, usually at
a potential close to E, the thermodynamic potential for the couple (0 V when
aH+ = 1, PH2 = 1 atm, or as per the Nernst equation under other conditions). This is
illustrated in Figure 21. If one were to set up a cell with the H +/H2 half reaction
occurring on Pt electrodes, and with both half cells under identical conditions, there
would be no net effect. No voltage would exist between the two electrodes and
there would be no change in composition in either half cell. However, H 2 would be
undergoing oxidation and H+ would be undergoing reduction to H 2, but, both at
identical rates. The rate at which electrons are being exchanged between H + and H2
is the exchange current density. It is this current that is estimated by the

* But, why then can aluminum be put in contact with water without noticeable effect?
Because of a very strong, adherent oxide coating that forms on the Al surface and protects
it from corrosion. This is another case of kinetic stabilization, making the thermodynamically
favourable process very slow. In a sufficiently acidic or basic solution, that oxide coating
gets broken down, and aluminum corrodes. The same considerations apply to titanium

48

(ETi+2/Ti = -1.6 V), which also forms a stable TiO 2 layer on its surface. This layer is stable in
an acidic and oxidizing environment, but degrades under reducing conditions.

Figure 20. Plots of log (exchange current density) for H + reduction vs. atomic
number. See text for explanation of exchange current. Data from [22].
intersection of the Tafel regions in Figure 21. The system just described is at
equilibrium, but equilibrium is a dynamic situation; reaction is still occurring, but
the forward and backward rates are identical. In essence, a higher J o means that
the overvoltage for the hydrogen evolution half reaction is smaller.*
Note the enormous range of current densities for the H +/H2 exchange current
density. They range from about 10 A/m 2 to 5 x 10-9 A/m2. This spans almost 12
orders of magnitude! Platinum has the highest exchange current density. It is a
very good catalyst for promoting hydrogen evolution. (This is why fuel cells have
used platinum electrodes; unfortunately it makes the technology very expensive.
Much research has gone into searching for alternative electrode materials that are
cheaper. Rhenium has a very similar Jo to Pt, but it is even more expensive.)

* There are correlations that provide something of an explanation for the variation in
exchange current densities. For instance, the exchange current densities can be at least
partially correlated with the strength of the metal-hydrogen bond. To make sense of this,
consider the nature of the H+ reduction process on a metal surface. H+ may adsorb onto the
surface, then accept an electron. This forms an adsorbed H atom. Two of these H atoms

49

must then combine to form H2 gas. If the metal-hydrogen bond is very weak, this inhibits
the electron transfer; a high-energy intermediate state (adsorbed H) involves a highly
endothermic reaction that is not very favourable. If the metal-hydrogen bond is fairly
strong, then the electron transfer is favourable. But if the metal-hydrogen bond is very
strong, then the formation of H2 is inhibited, and again this would slow the overall reaction.

Figure 21. Extrapolation of linear Tafel region lines to obtain the exchange current
density, Log Jo. The lines usually meet near E, the thermodynamic half reaction
potential. At this point the anodic and cathodic currents are identical. Both half
reactions proceed at this same rate so that there is no net change in composition.
Consider the example of the Zn +2/Zn couple. EZn+2/Zn is -0.76 V, which is well
below H+/H2. This is illustrated from the Eh-pH diagram in Figure 22. At all pH
where Zn+2 can predominate H2 evolution is the thermodynamically favoured
process. Beyond pH ~6.2 Zn+2 is no longer dominant, but rather solid ZnO is. But,
the rate of hydrogen evolution on pure Zn is very slow (J o = 3.2 x 10-7 A/m2) which
indicates a high overvoltage for H2 evolution. Thus it is practically feasible to
electrowin zinc metal from ZnSO 4 solution. The rate of Zn plating is much greater
and outruns the much more thermodynamically favourable, but much slower H +
reduction. Indeed, Zn EW is practiced from quite acidic solutions containing on the
order of 100 g/L H2SO4. (This helps with conductivity of the electrolyte and also is
important for the leaching process, which uses acid generated in EW.) This same
phenomenon allows Mn EW as well, albeit with at best about 60% current
efficiency, the rest going to make H2. And in this case high acidity cannot be
tolerated. The solution must have a relatively high pH, though not so high as to
precipitate an Mn(II) hydrolysis product, Mn(OH)2.
Purity of the electrolyte is a critical issue in Zn EW. Small amounts of Co +2,
Ni , etc. will readily deposit the parent metals on the zinc surface; E Ni+2/Ni and
+2

50

ECo+2/Co > EZn+2/Zn. Then regions of say Co metal form on the zinc surface. The rate
of H2 evolution on Co and Ni is quite high, so now hydrogen evolution has a site on
the Zn cathode surface where it can readily occur, thus opening up a kinetically
facile pathway for the thermodynamically favoured process. Then current efficiency
for zinc EW plummets and hydrogen gas is produced. Hence one of the biggest
concerns in zinc hydrometallurgy is electrolyte purification.

Figure 22. Partial Eh-pH diagrams for three M-H 2O systems. The H2 pressure is low
(0.01 atm), and in practice, it would be kept low to avoid the danger of explosion.
Metals in the Group VIII block of the periodic table exhibit especially high
rates of hydrogen evolution. The Ni, Pd and Pt column have the highest rates of any
group. For instance, ENi+2/Ni = -0.24 V and the rate of hydrogen evolution on Ni is
relatively high as well. So how can we electrowin Ni metal and those like it?
Consider again the Eh-pH diagram in Figure 22. The H +/H2 and Ni+2/Ni lines
intersect at pH 3. Below this pH H+ is easier to reduce to H2 than reduction of Ni+2
to Ni. Above this pH Ni+2 is the stronger oxidant; EhNi+2/Ni > EhH+/H2. This means that
Ni+2 can be electrowon above this pH. However, above pH 6.4 Ni(OH) 2 s precipitates
and this would foul the cathodes, introduce hydrogen and oxygen impurities into
the solid and passivate the electrode surface. Hence the pH must be kept at 3 < pH
< 6.4 (at these ionic activities, at least). In practice, the cathodes and anodes are
physically separated using porous cloth bags that surround either the cathodes or
the anodes. This is illustrated in Figure 23. Here bags surround the cathodes. Nickel
sulfate solution is pumped into the cathode bags (called the catholyte). This
maintains a small hydrostatic head in the bags, relative to the level in the anode

51

compartment (the anolyte solution). Hence solution flows from the cathode
chamber to the anode chamber. This prevents anolyte solution from coming into
contact with the cathodes. In the anode chamber acid is produced via oxygen
evolution (in sulfate medium):
SO42- + H2O = H2SO4 + 1/2O2 + 2e-

(29)

The increasing acid concentration is prevented from coming into contact with the
cathodes using the cloth bags. The anolyte is pumped out at the same rate that the
catholyte is pumped in. The anolyte is continuously neutralized with NaOH. Nickelenriched, neutralized catholyte is pumped back into the cathode bags. All this adds
considerable expense. The conductivity of the electrolyte is also lowered by having
to avoid H+ (which is the best ionic conductor) and use Na 2SO4 instead. This adds to
the energy costs. However, nickel is valuable enough that this can be tolerated.
Another variation on Ni EW is to use chloride medium. This has advantages in
leaching, though it is more corrosive. Then Cl 2 gas is formed at the anode. Now the
anode chambers must be enclosed and the Cl2 gas must be collected; it is highly
toxic and it is very useful in leaching. Again, it is imperative that powerfully
oxidizing Cl2 not come into contact with the cathodes, or EW will cease. Hence again
separate cathode and anode chambers are required.

Figure 23. Schematic illustration of a nickel electrowinning cell. Numerous other


configurations are also possible. Nickel may also be electrowon from a chloride
electrolyte. Then the anode reaction is 2Cl - = Cl2 g + 2e-. Toxic Cl2 gas must be
collected and is used in leaching operations.

52

In practice some degree of hydrogen evolution occurs in nickel


electrowinning. Although this represents a loss of current efficiency, one side
benefit is that it produces a great deal of mixing in the vicinity of the cathode.
Small bubbles growing and dislodging generate significant mass transport. The
reduction of H+ to H2 at the electrode surface necessarily raises the local pH there.
This can lead to precipitation of Ni(OH) 2 with attendant passivation. To overcome
this B(OH)3 (boric acid) is commonly added. This weak acid reacts with OH - to help
buffer the local pH in an acceptable range.
The Eh-pH diagram in Figure 21 also illustrates why copper can be so easily
electrowon. The Cu+2/Cu reduction potential lies well above the H+/H2 line at all pH
where Cu+2 is predominant. Cu+2 is simply a much stronger oxidant and so copper
can be electrowon from acid solution.
Electrorefining [23]
Most of the worlds copper is produced pyrometallurgically. This is generally
not pure enough for most applications, despite being about 99.8% pure. The most
viable method for purification of copper is electrorefining (ER). In this process, the
impure copper is cast into anodes and placed into electrochemical cells. The impure
copper is anodically dissolved by means of an electric current. The dissolved Cu +2 is
plated onto high-purity copper cathode starter sheets. In the process the impurities
present in the anode copper either stay with the anode or are precipitated as
insoluble salts or are removed from the electrolyte by taking a bleed stream and
purifying it. The process produces high purity copper (>99.99% Cu; <0.004% S +
traces of other metals). Hence the great majority of the worlds primary copper
production (i.e. neglecting scrap) involves either electrorefining or electrowinning.
Electrochemistry figures very prominently in metal production. Nickel and cobalt
also may be purified by electrorefining.
Staying with the example of copper electrorefining, copper metal is oxidized
at the anode,
Cu = Cu+2 + 2e-

E = 0.34 V

(30)

E = 0.34 V

(31)

and reduced at the cathode,


Cu+2 + 2e- = Cu+2

The cathodes are negatively polarized and the anodes are positively polarized.
Clearly E for this process is 0 V. The activity of copper in the anodes is only
slightly less than one, and the copper concentration in the electrolyte is typically on
the order of 1 m. Hence the thermodynamic cell voltage in practice is actually quite
close to 0 V. However, there are other voltage drops that result in a cell voltage that
is somewhat less than 0 V. (The cell voltage is E cathode - Eanode < 0.) The following also
contribute to the required cell voltage:

53

Electrical resistance (0.01-0.02 V. Leads used to connect the electrodes to


the power supply - called busbars.)

Electrolyte resistance (0.11-0.13 V. Resistance to electrical conduction by


means of ionic migration; lower than for EW electrolytes.)

Cathodic overvoltage (0.04-0.08 V. This is due largely to organic coatings on


the cathode copper surfaces. Organic additives are added to promote a
smooth, continuous deposit. Alternatively periodic current reversal may also
be employed to promote smooth deposit growth.)

Anode and cathode connections (0.03-0.06 V. Electrical resistance of the


connections.)

Anode polarization (up to 0.01 V. Voltage required to corrode the anode


copper.)

The total cell voltage is typically 0.25-0.3 V. The greatest contributor to the
voltage is the electrical resistance of the electrolyte. Sulfuric acid concentrations are
kept high to keep the resistance of the electrolyte low. Since the solution resistance
of the electrolyte increases directly with the separation between anode and
cathode, a narrow anode-cathode spacing is desirable. (ER tankhouses also involve
very large numbers of cathodes and anodes, so minimizing the gap also is good for
lowering capital costs of the associated infrastructure.) However, cathodes need to
grow to a specified thickness (about 1.4-1.9 cm in electrorefining), starter sheets
(the substrates onto which the metal is plated) and anodes may not be perfectly
planar nor vertical and polycrystalline protrusions called dendrites can grow out
from the cathodes towards the anodes, causing short circuits and loss of metal
plating efficiency. For all these reasons the gap between cathodes and anodes in ER
is roughly at least 2.5 cm. Note that in ER as the cathode grows, the anode gets
thinner. The conductivity of typical ER electrolytes is 0.5-0.7 -1 cm-1. Typical
current densities are 190-260 A/m2 and cathodes typically are 1 m x 1 m. Given a
cathode-anode spacing of 2.7 cm, a current of 250 A and a conductivity of 0.7 -1
cm-1, the resistance of the electrolyte would be,
R = (1/0.7 -1 cm-1) x 2.7 cm x (1/104 cm2) = 0.000386

{109}

From V = IR the voltage drop would be,


V = 250 A x 0.000386 = 0.097 Volts

{110}

which is close to the lower end of the typical range of voltage drops due to solution
resistance listed previously.
Impurities

54

The table below lists the common impurities in anodes and the purer
cathodes. The main impurities are As, Bi, Ni, Pb, Sb, Se and Te. If an element has
an E that is less than ECu+2/Cu (0.34 V), then it can dissolve from the anode. (A
lower standard reduction potential means that the element is more easily oxidized.)
For an impurity metal in the anode,
Manode = Mn+aq + ne(32)
The reduction potential (for the reduction half reaction) is given by,
E = E - 2.303RT log aManode
nF
aMn+

{111}

The activity of metal impurity in the anode will be low. If the activity of the ion in
solution is also low, the reduction potential will be close to E.
Table 4. Impurities in copper before and after ER.
Impurities Anodes, % Cathodes, %
As
0-0.3
<0.0002
Bi
0-0.001
<0.0001
Fe
0.002-0.03
<0.002
Ni
0-0.5
<0.001
Pb
0-0.1
<0.0005
Sb
0-0.3
<0.0002
S
0.001-0.003 <0.001
Se
0-0.02
<0.0002
Te
0-0.001
<0.0001
Ag
trace-0.1
<0.001
Au
0-0.005
<0.00001
PGM*
trace
*PGM = platinum group metals (Rh, Ir, Ru, Os, Pd, Pt)
Iron, for example, has EFe+2/Fe = -0.41 V. It is much easier to oxidize than
copper. Now, however, since the reduction potential for Fe +2 + 2e- = Fe is
substantially lower than that for Cu+2 to Cu, copper will preferentially plate over
iron. (Fe+2 having a lower E than Cu+2 means that Cu+2 is a stronger oxidant; Fe+2
is harder to reduce.) On the other hand, elements with an E greater than that of
Cu+2 will not dissolve from the anode. Thus electrorefining is very selective. Only
species with reduction potentials similar to that of E Cu+2/Cu will dissolve from the
anode (at least by anodic oxidation) and plate at the cathode. Even then, kinetic
factors also would come into play. If a species has a substantial cathodic
overpotential for reduction, its tendency to reduce and co-plate with copper at the
cathode would be diminished.
A partial table of standard reduction potentials is reproduced below. The first
column indicates the principal species that would form in solution. Inspection of the
data indicates that mainly Fe, Ni, Pb, Sb, As and Bi should dissolve.

55

The other way that impurities can enter the cathodes is by occlusion (or
entrainment). Cathode surfaces are polycrystalline and not perfectly smooth.
Microscopic cavities form, which contain the electrolyte. These may grow over with
Table 5. Some standard reduction potentials relevant to Cu ER.
Oxidized
Half reaction
E V
form
Au+3
Au+3 + 3e- = Au
1.42
+
Ag
Ag+ + e- = Ag
0.80
+2
+2
Cu
Cu + 2e = Cu
0.34
BiO+
BiO+ + 2H+ + 3e- = Bi + H2O
0.32
HAsO2*
HAsO2 + 3H+ + 3e = As + 2H2O 0.25
SbO+
SbO+ + 2H+ + 3e- = Sb + H2O
0.21
+
+
H
2H + 2e = H2 (for reference)
0
Pb+2
Pb+2 + 2e- = Pb
-0.13
Ni+2
Ni+2 + 2e- = Ni
-0.23
Fe+2
Fe+2 + 2e- = Fe
-0.41
* Or perhaps H2AsO3.
copper metal and trap small amounts of electrolyte within the cathodes. Impurities
in the electrolyte, including copper sulfate and sulfuric acid are then occluded. This
is the principal origin of impurities in the cathodes. Notes on specific impurities
follow.
(a) Gold and platinum group metals do not dissolve from the anode. Their reduction
potentials are too high. These metals stay in the residues of the anodes. Anodes are
not completely dissolved. They are removed and replaced after about 85% reaction.
Precious metals may be a significant source of revenue.
(b) Silver does partially dissolve,
Aganode = Ag+ + e-

(33)

but a low concentration of sodium chloride (~0.05 M NaCl) in the electrolyte


precipitates most of this as insoluble AgCl:
Ag+ + Cl- = AgCl s 1/Ksp = 5.6 x 109

(34)

The net reduction reaction is,


Aganode + Cl- = AgCl + e-

(35)

In fact, E for Ag+/Ag is 0.80 V. This is considerably higher than that for Cu +2/Cu
(0.34 V). This should preclude silver dissolving in the first place. But,

56

AgCl + e- = Ag + Cl-

E = 0.22 V

(36)

Since this is less than ECu +2/Cu, and chloride is present in the electrolyte,
conversion of anode silver to solid AgCl is favourable. Most of the AgCl reports with
the anode residues. Whatever does end up in the cathode results from occlusion of
suspended particles in the electrolyte. Silver may be present as a copper-silver solid
solution and as compounds with sulfur, selenium and tellurium, depending on
relative concentrations. These compounds are too stable to corrode. It is evident
then that mechanisms of dissolution may be complex.
(c) Sulfur, selenium and tellurium form stable copper compounds with many metals
(e.g. Ag2Se, Cu2Se, Ag2Te4, Cu2S etc.), which are not electrochemically oxidized.
These elements mainly report with the anode residues.
(d) Lead and tin (tin was not mentioned in the tables above) are readily corroded,
but form quite insoluble precipitates of PbSO 4 and a basic tin(II) sulfate. As a result
these elements do not dissolve to any significant degree in the electrolyte. Some
cathode contamination by occlusion occurs.
(e) Arsenic, bismuth, cobalt, iron and nickel are all less noble than copper (i.e. they
are more easily oxidized) and can enter the electrolyte as solution species (e.g.
HAsO2, BiO+, Co+2, Fe+2 and Ni+2, respectively; see table of reduction potentials on
previous page). Much of the antimony (~60%) and some of the arsenic (~25%)
end up staying in the anode residues, possibly as a result of forming stable
compounds with copper (e.g. arsenides) during anode casting. In principle the
impurities could plate at the cathode, but many have E < E H+/H2 (e.g. ECo+2/Co =
-0.28 V). The potential at the cathode is around E Cu+2/Cu = 0.34 V, i.e. the potential
for copper reduction is too oxidizing to allow less noble ions to reduce. (Besides,
even if they could plate, the strong acid of the electrolyte would redissolve them in
some instances.) This illustrates the beauty of electrorefining. The impurity metals
that are more easily corroded than copper (E < 0.34 V) cant electrochemically
plate at the cathode, while those that have E >0.34 V cant be corroded from the
anodes. The purification method is thermodynamically tuned to be very selective
for copper.
The case of bismuth is different. The E for BiO +/Bi is 0.32 V, which is quite
similar to E for Cu+2/Cu. The differences in the two potentials are quite small and
this might allow some Bi to be electrochemically deposited, in principle.
Nevertheless, all the corroded species build up in the electrolyte over time
and eventually would contaminate the cathodes by occlusion. This is the main
mechanism by which impurities enter the cathodes. Because of this they have to be
removed from the electrolyte. A bleed stream is taken from the electrolyte and
subjected to purification to remove the impurities from the circuit. Since the
impurities concentrations in the anodes are low, the bleed rates are also low (0.10.5 m3 of electrolyte per tonne of copper metal plated.)
Electrolyte purification

57

There are a number of methods available for purifying the bled electrolyte.
One method is to first plate copper electrochemically from the electrolyte. This is
done in three stages. These are called liberator cells. The first stage produces
quite high purity copper cathodes. This is essentially an electrowinning operation.
In the second stage an impure copper product is formed that can be recycled to the
furnace for generation of copper anodes. In the third stage most of the arsenic,
antimony and bismuth are removed as well. The copper product is highly
contaminated with these elements. The cathodes may be used for production of
arsenic or recycled to the smelter. (In the smelter these elements are partially
rejected as slags and dusts. This avoids build-up of these impurities in the process.)
The third stage in particular is quite energy inefficient due to the low concentrations
of ions in solution. In order to maintain high plating rates, high voltages are
required. This results in strongly reducing conditions at the cathode and some
arsine gas (AsH3) is also evolved. Arsine is extremely toxic and must be collected.
In arsenic-containing EW electrolytes there is also a danger of arsine generation.
After the third stage the solution contains mainly sulfuric acid, nickel, cobalt
and iron sulfates. Water is evaporated and the sulfate salts of the metals
precipitate. The precipitation is driven in part by the resulting high H 2SO4
concentration (source of high sulfate concentration). What is left is a high
concentration H2SO4 solution (1000 g/L or ~10 M). This is mostly recycled to
electrorefining to make up for lost acid.
An alternative to the first EW stage is to remove copper as CuSO 45H2O. This
is done by adding copper metal (shot) and reacting it with oxygen:
Cu + 1/2O2 + H2SO4 = CuSO4 + H2O

(37)

Evaporation and cooling of the solution precipitates much of the solution copper as
CuSO45H2O. The product contains a little iron and nickel as well. It may be sold.
Plating conditions
Organic additives are introduced to the electrolyte for a number of reasons.
One typical additive is bone glue, a mixture of natural proteins. It is added at
1-10 mg/L. It is added to minimize the growth of protrusions on the cathode
surface. Glue acts by adsorbing particularly on the tips of protrusions. This results
in an insulating layer that inhibits further deposition of metal. The adsorption of
glue to varying degrees on the cathode surfaces results in a slightly higher overall
resistance and a consequent voltage drop, as was noted previously. The same effect
can be attained by a technique called periodic current reversal (PCR). For a brief
interval, repeated regularly, the cell polarity is switched so that plating occurs
briefly on the anodes and corrosion of copper from the cathodes. The protrusions
are the most rapidly dissolved, while the indentations are the slowest to dissolve.
The result is that protrusions are prevented from growing. A side benefit is that
higher current densities can also be employed, resulting in greater overall plating
rates. Other additives are also added for various reasons. For instance, flocculants

58

may be added to promote flocculation and settling of fine precipitates and solids
derived from the anodes. This helps prevent impurity occlusion in the cathodes.
The electrolyte is heated by steam coils to 60-65C entering the circuit and
leaves at 55-60C. Some heating of the electrolyte occurs as a result of the solution
resistance. The electrolyte is circulated at a slow rate (~0.02 m 3/min). This ensures
that Cu+2 in solution is transported from anode to cathode. It also carries away
impurities and aids in distribution of additives to the solution.
A cathode substrate is needed for plating copper. These tend to be thin
sheets of copper metal (0.5-1 mm). They are plated onto titanium blanks. They are
plated for about 24 hours, after which they are mechanically removed from the
titanium substrates. Due to an adherent, conductive film on titanium, the copper
sheets are easily separated from the substrate. Titanium is expensive, but the
process is easier than previously practiced alternatives. Electrorefining cathodes are
plated for about 10-14 days, whereas the starter sheets are plated for one day, so
the number of titanium sheets needed is a small fraction of the total number of
cathodes being plated.
In principle, the higher the current density, the faster the copper cathode
production rate. However, much above 250 A/m 2 (or 300 A/m2 with PCR) anode
passivation occurs. The resistance of the anode greatly increases and copper
corrosions ceases. This has been attributed to rapid buildup of CuSO 4 in the
electrolyte at the anode surface and resultant precipitation of CuSO 4xH2O. This
would occur if the solubility of copper sulfate was exceeded. The precipitate is nonconductive and forms an insulating coating on the anodes. Passivation is prevented
by keeping the copper concentration in the electrolyte well below saturation
(typically around 40-45 g Cu+2/L), operating the cells at higher temperature
(~60C, which increases the solubility of copper sulfate), maintaining good
circulation and keeping the current density low enough. In addition, very high
current densities result in rough cathodes with more occlusion of electrolyte.
Efficiency and energy consumption
Current efficiency for ER is on the order of 90-96%. The main sources of loss
are stray currents to ground (1-3%), anode-cathode short circuits (1-3%) and
reoxidation of copper cathode (1%). Stray currents arise from fortuitous conduction
pathways, such as spilled electrolyte. Copper dendrites growing from cathode to
anode cause short circuits. These may be detected by infrared scanners. The
nodules are broken off to resume plating. The cathode is only weakly negatively
polarized (the thermodynamic cell voltage is almost zero). Hence oxygen in solution
can attack the copper in acidic solution and oxidize it. This occurs to a small extent.
Cu + 1/2O2 + H2SO4 = + H2O + CuSO4

(38)

Oxygen is not generated at the anode; copper metal corrosion is much more
favourable than water oxidation. However, oxygen is slightly soluble in the aqueous

59

solution and enters from the air. Oxygen may also oxidize ferrous to ferric to a
small extent, and ferric can corrode copper:
2Fe+3aq + Cu = 2Fe+2 + Cu+2

(39)

Quantities such as energy requirement and copper production rate may be


calculated from the Faradays Law relationships. The energy requirement for copper
ER with 96% current efficiency and 0.25 V applied potential is 220 kWh/t Cu. In EW
the energy requirement is 1900-2000 kWh/t Cu. The difference is due mainly to the
low cell voltage in ER compared to the higher voltage in EW (~2 V). Hence ER for
copper is not all that energetically demanding. In addition to the ER process itself,
the purification process for the bleed stream requires some energy (roughly 25
kWh/t Cu produced). This is about 1/10 the energy for copper cathode production
(including the starter sheets).
Flowsheet
Once the anodes have been about 85% consumed they are removed from
the cells, washed to displace the residues (slimes) and recycled back to anode
casting. The residues may be collected and processed for precious metals recovery,
which can be a substantial source of revenue. A flowsheet is presented in Figure 24
that schematically depicts an ER process with the bleed stream being purified by
electrowinning as described previously. Blister copper is the term for primary
copper from pyrometallurgical processing. Approximate mass and volumetric flows
are included. Note the substantial recycle of copper anode material and that a small
amount of high-purity copper metal is produced from the bleed stream purification
process. The number of starter sheet cells is relatively small. Starter sheets are
quite thin.
Other operational aspects
Anodes and cathodes are set in cells in a parallel arrangement with each
anode between two cathodes. The electrolysis is carried out in a parallel circuit, not
in series. This keeps all anodes at the same potential and all cathodes at the same,
slightly lower potential. Each cell contains 30-50 anodes, depending on the plant.
Cells are operated in series and can be isolated for purposes of installing or
recovering cathodes, spent anodes and residues, etc. Electrical contact is made
through a system of copper metal contacts. Cells are made of concrete with
polymer liners or aggregate in plastic. The electrical contacts are profiled to keep
the cathodes and anodes at the right spacing. Care is taken to ensure that cathode
starter sheets and anodes are planar and aligned properly in the cells. This ensures
good uniform corrosion of anodes and plating of cathodes, as well as avoiding direct
short circuits between cathodes and anodes.
Some statistics for typical copper electrorefining operations are provide in
Table 6 that follows below. The process takes on the order of 10-14 days per set of
cathodes, which is about twice as long as electrowinning of copper. At the start of
the process a set of anodes and a set of cathode starter sheets are placed into the

60

cells. For n anodes there are n+1 cathodes (opposite of electrowinning). After
10-14 days the cathodes are removed, then a new set of cathode starter sheets are
installed and the process continues. Each set of cathodes is removed and washed.
Cathodes are melted and cast into now very pure form. After two sets of cathodes
are recovered from one set of anodes the electrolyte is drained away. Spent anodes

are washed to remove residues and the cells are washed out to recover the
residues. These may be treated for precious metals recovery. The spent anodes are
then recycled to anode casting, which is done by melting the copper metal.

Figure 24. Simplified flowsheet for a copper electrorefining process. Approximate


distribution of copper mass is indicated. The bleed rate to purification is
intermediate (about 30 m3/100 t of copper produced).

61

The circulation rate is high enough to ensure good migration of copper.


However, circulation rates need to be low enough to allow dislodging particles of
anode residues to fall to the bottom of the cell, and not be occluded in the
cathodes. Energy consumption refers to DC voltage requirements. If power is
converted from AC to DC there is a small loss here as well.
Table 6. Some typical copper electrorefining plant statistics.
Copper production (t/d)
400-1100*
Number of cells
800-2400
Cell inner dimensions L x W x D (m)
4-5 x 1.1 x 1.2
Anodes: %Cu
99.5-99.7
L x W x thickness (m)
~1 x ~1 x ~0.05
Weight (kg)
310-380
Center-line spacing (cm)
10
Lifetime (days)
20-28
Anode consumption (%)
80-87
Cathodes: Lifetime (days)
10-14
Starter sheet weight (kg)
5-6
Weight (kg)
130-165
Total metallic impurities (%)
0.001-0.003
3
Electrolyte composition (kg/m ):
Cu
43-50
H2SO4
180-200
Ni
2-10
As
<0.01-10
Fe
0.2-1
Sb
0.15-0.35
Bi
0.04-0.15
Cl
0.03-0.04
Temperature (C)
60-65
Circulation rate (m3/min per cell)
0.015-0.02
2
Power: Cathode current density (A/m ) 190-280
Current efficiency (%)
90-96
Cell voltage (V)
0.25-0.3
Energy consumption (kWh/t Cu)
220-270
* 150,000-400,000 tonnes/year!
References
[1] Adapted from Ettel, V. and Conrad, B., Fundamentals of Electrochemistry, Short
Course Notes.
[2] Ettel, V.A. in Hydrometallurgy: Theory and Practice Course Notes, Centre for
Metallurgical and Process Engineering, UBC, p. X-7.

62

[3] Nguyen, T., Guresin, N., Nicol, M. and Atrens, A., Influence of cobalt ions on the
anodic oxidation of a lead alloy under conditions typical of copper electrowinning,
Journal of Applied Electrochemistry, 2008. 38, 215-224.
[4] Price, D.C. and Davenport, W.G., Densities, electrical conductivities and
viscosities of CuSO4/H2SO4 solutions in the range of modern electrorefining and
electrowinning electrolytes, Metallurgical Transactions B, 1980, 11B, 159-163.
[5] J.G. Jenkins and M.A. Eamon, Plant practices and innovations at Magma
Copper Companys San Manuel SX EW plant, Proceedings of the International
Symposium on Electrometallurgical Plant Practice, 20th Annual Hydrometallurgical
Meeting, P.L. Classens and G.B. Harris, eds., Montreal, Que., Canada, Oct. 21-24,
1990, pp. 41-56.
[6] J. Jenkins, W.G. Davenport, B. Kennedy and T. Robinson, Electrolytic copper
-leach, solvent extraction and electrowinning world operating data, Proceedings of
the Copper 99 Cobre 99 International Conference, S.K. Young, D.B. Dreisinger,
R.P. Hackl and D.G. Dixon, eds., Phoenix AZ, Oct. 10-13, 1999, Vol. 4, pp. 493-567.
[7] Hydrometallurgy: Theory and Practice Study Guide, Module 5, Kinetics of
Heterogeneous Reactions, University of Capetown, p. 19.
[8] C.R. Merigold, Lix Reagent Solvent Extraction Plant Operating Manual, Cognis
Corp., 2nd edn, 1996. Retrieved Feb. 18, 2009 at:
http://www.cognis.com/NR/rdonlyres/97088921-AD31-461E-A1209C2B791ACD26/0/lixsolve.pdf
[9] W.G. Davenport, M. King, M. Schlesinger and A.K. Biswas, Extractive Metallurgy
of Copper, 4th edn., Pergamon, 2002.
[10] C.L. Pfalzgraff, Dos and donts of tankhouse design and operation, in Copper
Leaching, Solvent Extraction, and Electrowinning Technology, G.V. Jergensen II,
ed., Society for Mining, Metallurgy and Exploration, pp. 217-221.
[11] M.S. Prasad, V.P. Kenyen and D.N. Assar, Development of SX-EW process for
copper recovery-an overview, Mineral Processing and Extractive Metallurgy Review,
1992, vol. 8, pp. 95-118.
[12] N.T. Beukes and J. Badenhorst, Copper electrowinning: theoretical and
practical design, Hydrometallurgy Conference, South African Institute of Mining
and Metallurgy, Feb. 24-26, 2009, pp. 213-240.
[13] J.B. Hiskey, Principles and practical considerations of copper electrorefining
and electrowinning, in Copper Leaching, Solvent Extraction, and Electrowinning
Technology, G.V. Jergensen II, ed., Society for Mining, Metallurgy and Exploration,
pp. 169-186.

63

[14] G. Kordosky, S. Olafson and C. Hahn, Iron control strategies in copper solvent
extraction plants, the case for a wash stage, Randol Conference, 2000. Retrieved
Feb. 23/09 at:
http://www.cognis.com/NR/rdonlyres/B159035F-B210-47AD-81ED96771D437E82/0/Iron_Control_Strategies_Randol_Conference.pdf
[15] G. Kordosky, M. Virnig and M. Mackenzie, Solvent extraction - reagents and
selectivity control, Cognis Corp. Retrieved Feb. 18, 2009 at:
http://www.cognis.com/products/Business+Units/Mining+Chemicals/Literature/Gen
eral+Publications/
[16] G.M. Miller, D.J. Readett and P. Hutchinson, Experience in operating the
Girilambone copper SX-EW plant in changing chemical environments, Minerals
Engineering, 1997, Vol. 10, No. 5, pp. 467-481.
[17] W. Armstrong, The ISA process and its contribution to electrolytic copper,
Rautomead Conference, Scotland, August, 1999, 22 pp. Retrieved Feb. 25/09 at:
http://www.isaprocess.com/downloads/ipt_contribution_WA_Rautomead_99.pdf
[18] (a) J.R. Spence and M.D. Soderstrom, Practical aspects of copper solvent
extraction from acidic leach liquors, in Copper Leaching, Solvent Extraction, and
Electrowinning Technology, G.V. Jergensen II, ed., Society for Mining, Metallurgy
and Exploration, pp. 239-257. (b) A very slightly different version of this article can
be accessed at (Feb. 21/09):
http://www.cytec.com/specialty-chemicals/pdf/practical.pdf
[19] Shaw, R.; Vance, S.; Illescas, J.; Dreisinger, D. and Wassink, B., Ion exchange
for iron impurity control in the base metal industry, Iron Control Technologies,
Proceedings of the International Symposium on Iron Control in Hydrometallurgy,
3rd, Montreal, QC, Canada, Oct. 1-4, 2006, 757-769.
[20] Sandoval, S.; Clayton, C.; Dominguez, S.; Unger, C. and Robinson, T.,
"Development and commercialization of an alternative anode for copper
electrowinning," Proceedings of Copper 2010 International Conference, Hamburg,
Germany, Jun. 6-10, 2010, Vol. 4, pp. 1635-1647.
[21] Schlesinger, M.E.; King, M.J.; Sole, K.C and Davenport, W.C., Extractive
metallurgy of copper, Elsevier, 5th edn., 2011, Chpt. 17, pp. 349-372.
[22] S. Trasatti, Work function, electronegativity, and electrochemical behaviour of
metals. III. Electrolytic hydrogen evolution in acid solutions, Electroanalytical
Chemistry and Interfacial Electrochemistry, 1972, 39, 163-184.
[23] A.K. Biswas; W.G. Davenport, Extractive Metallurgy of Copper, 2nd edn.,
Pergamon, 1980.

Glossary of Some Common Terms

64

Acid mist Gas bubbles breaking the surface of the electrolyte in the cells forms a
fine mist in the tankhouse.
Activity Effective concentration. Aqueous solutions are highly non-ideal, meaning
that there are strong interactions between solutes and water. This makes a solute
behave as if its concentration is either higher or lower than its analytical
concentration.
Anode The electrode where oxidation occurs (loss of electrons).
AW Atomic weight in g/mol.
Bleed A solution taken from a process stream, often to prevent excessive build-up
of impurities. In EW it is taken from the lean electrolyte to minimize build-up of
harmful impurities in electrowinning. See also bleeding.
Bleeding A term for the action of removing a fraction of a process solution stream.
Boundary layer The stagnant film in the vicinity of a solid in a solution (here an
electrode). The thickness of the boundary layer is determined in large part by the
extent of turbulence (agitation). The thinner it is, the higher the limiting current
density can be.
Cathode The electrode where reduction occurs (gain of electrons).
Conductance, conductivity

See resistance.

Coulomb The unit of electric charge. By definition, 1 C of charged passed through


a potential difference of 1 V involves 1 J of energy, i.e. 1 VC 1 J.
Current The flow of electricity in C/sec. 1 C/sec = 1 A (1 A = 1 amp).
Current density The current in A per unit area. May refer to nominal area (as in 1
m2 for a 1 m x 1 m sheet) or true area (which requires knowledge of actual surface
area).
Current efficiency (CE) The fraction of the current (in %) that goes to the
specified electrochemical half reaction. If all the current goes to only one half
reaction, CE = 100%. Side reactions that consume some of the electricity lower CE
to <100%.
Dendrite Same as a protrusion; a polycrystalline mass or relatively fast-growing
metal crystals arising from a metal cathode surface and growing towards an anode.
Upon contact with the anode a short-circuit occurs.
Electrorefining An electrolytic refining process where impure metal is anodically
corroded and replated onto cathodes.

65

Electrowinning

Electrolysis for primary metal production.

ER Short for electrorefining.


EW Short for electrowinning.
Faraday The charge of a mole of electrons = 96484.6 C/mol e -.
Electrolysis An electrochemical process forced to go opposite to
thermodynamically favourable direction by application of an external voltage.

its

Electrolyte A solution that contains ions. In EW, the solution to be electrolyzed.


Favourable Same as spontaneous. Thermodynamically speaking, the naturally
favoured direction of a chemical reaction. A reaction is favourable when there is a
net increase in the entropy of the system plus its surroundings, which is expressed
mathematically as G = H - (TS) < 0. At constant temperature this becomes
G = H - TS < 0.
Galvanic Describes the operation of an electrochemical cell when current flows
spontaneously.
kWh Kilowatthour a unit of energy (not power): 1 kWh = 1000 J/sec for 3600
seconds = 3.6 x 106 J.
Lean electrolyte (or LE) The copper-depleted (somewhat depleted, not
completely), acid enriched electrolyte solution that comes from electrowinning and
returns to SX stripping.
Limiting current density The maximum possible current density for an
electrolysis. At the limiting current density the concentration of the electroactive
species at the cathode surface goes to zero. It is a function of physical properties of
the metal cation (in EW), such as diffusion coefficient and ionic mobility of the
cation, as well as mass transport factors such as the bulk concentration and the
thickness of the boundary layer.
Live stripping See Stripping.
E = E - RTlnQ where R is the ideal gas constant (8.314
nF
J/mol K), T = absolute temperature, n = moles e-/mol of reaction, F = the Faraday
constant (96485 C/mole e-) and Q is the reaction quotient (has the same form as
the equilibrium constant, but does no equal K; represents a disequilibrium
condition). The Nernst equation allows calculation of the voltage for a cell under any
conditions; E must also be known at T.
Nernst equation

n The number of moles of electrons passed for a half reaction or a cell reaction
per mole of reaction. A mole of reaction means that the molar quantities of
reactants form the molar quantities of products for a reaction as written. If the half

66

reaction is written H+ + e- = 1/2H2, then n = 1 and 1 mole H + is reduced to form


mole H2; n = 1 mole e-/mol of reaction = 1 mole e-/mol H+ = 1 mol e-/0.5 mole H2.
If the half reaction is written 24H + + 24e- = 12H2, then n = 24 and 24 mole H + is
reduced to form 12 mole H 2; n = 24 mole e-/mol of reaction = 24 mole e -/24 mol
H+ = 24 mol e-/12 mole H2.
Noble metals Metals whose cations have relatively high reduction potentials. The
higher E is, the harder it is to oxidize the metal. Au is the most noble of all metals;
EAu+/Au = 1.69 V.
Overpotential Same as overvoltage; the additional voltage required to make a
half reaction proceed at a measurable rate (As always half reactions never run in
isolation; cathodic and anodic half reactions must be coupled. However, there are
ways to study one half reaction while not considering the other.)
Overvoltage See overpotential.
PLS Pregnant leach solution (from leaching, fed to solvent extraction).
Polarization A potential difference, causing one electrode to be positive relative to
the other (rendered negative).
Potential (electric) See voltage.
Power Energy per unit time, e.g. J/s or kW (kilowatts).
Protrusion See dendrite.
Reduction potential A voltage. A measure of the tendency for a chemical species
to accept electrons and be reduced relative to the standard H+/H2 half reaction. The
standard reduction potential is the special case where all species are present at unit
activity.
Resistance The constant of proportionality in Ohms law, V = IR, relating current
to voltage. Unit = ohms, symbol . Also the extent to which current flow is
impeded. Resistance is the inverse of conductance or conductivity.
Resistivity The intrinsic resistance of a material or solution; resistance per unit
length per unit area: /cm/cm2 = cm. Specific conductivity is the inverse of
resistivity in W-1cm-1, also called Siemens.
Rich electrolyte (or RE) The copper-enriched, acid-depleted (somewhat depleted,
not completely), electrolyte that returns from stripping to EW.
Short circuit Direct electric contact between anode and cathode, e.g. by growth of
a protrusion from a cathode. Current flows directly between cathode and anode and
is lost entirely as heat.
Spent electrolyte

Same as lean electrolyte.

67

Stripping In EW the process of mechanically dislodging cathode sheets from


starter sheets. When 1/2 or 1/3 of the cathodes are harvested at a time from a cell,
the cell can continue to be operated. This improves productivity and is called live
stripping.
Tankhouse The building where EW or ER is conducted.
Voltage Voltage, energy and charge cannot be defined other than in terms of each
other. A voltage is the electrical potential energy per unit charge. The "per unit
charge" term is crucial. Voltage in itself is not a potential energy. One can think of
voltage as a potential for passing a charge. If the potential is high, the energy
involved in passing the charge is high. Whether energy is released or input depends
on whether the charge is attracted or repelled. The units of voltage are volts (V). By
definition 1 VC = 1J.

68

You might also like