:: r:: ri:i.

fi iri i:9 :
..:. :i'::-.
..._ - I :iti

. . .,,': .-
..
.j''. l:i1:::{
.

.' ; :ü
..1_,,,1
, . t...1

' j,¡
::)
-i
l
Jj
't
-: .j

.:
i
:: .l

il

:;
r,i
i iir: T ' r.]
,l

...i1:

'::',f ...litl: ",,.]:- i.:i,:'-.;.tl'r.,4..::1.:r:.i:;:.;,1',l,::.:.:.,..r:'.,,-..:,,_

, ....:.
. ,.i:r.i
]...::|
i.r t
:it,,l rviñni{;¡:

- vi¡JysilI\J '.j | ¿1¡ {;u,¡s :

Tucscn, P,ri¿nr.¿
 41
q.¡
4 .l Pl airi Equi pt'ten; . , ' : 'Lea:tning Acti,u'[-ty
ji . Prin:ir.v
Prin:ar.v Ccnsideral j cns
i.3. ilectrical !{uiFnent
i'lechani cal [cui Pnient

4 .? tl ectrc? ef i n i ng 'Lea::'t':'i,
1; Actit: i.t.E

7 , El e;iroch*ni i"'t y
3.':,**:":st*ere*j*i:"s f i the Lresi gn':,cf :r[quipm*r'r¡'"'. . lr',..,l,'l'j
.. .:..
..
¡,'1"':'r: . .;:: ';

4.3.Practica'][]ec¡r*reffn.]*g.F::*e*s1es:.',¡.¡.'r..
i ; ,.,,1ri1*c3 :f{ur] , , l; ties
0r jectÍve5 '
ry31'
7.. l"'lajor
l"'lajor^ Ob'iecttves _
:

{. 5p:'-i al C':T i }es i grr

5.; 5ie* an* C*st,*f C*m¡*reial. *perg*tens
4"4 Ilectrcr+innir:g
'1 ñ.
: rrt-.:r.: I ..: r.t:-,
. ::::. .",'..ri i i r- Lecz.r,ing Activ'irg
i . f'.j. iJr.Jlu-
?. Cilh';jic Fr'+l*ss
3. ,ii,;:t c FrolEsg
t

t ;' '- e h*r*j s iry an* [] *c*¡:*ch*ryi st,rg-.q{r,.tht

3',',r.tatnc** fqatei: al s , ^
,

4 . Ar.:ce l'14:,:i-i;'ls

3 :'' f,*x*¡:t*.ii *;¡.'f;*a*i'¡":¡"¡ i'1**i:ai:i sn

il ectromeiai 1 urgy Equi pment

ModuLe 4

INDUSTRIAL PRACTICTS I
IXTRACT I ON

Module 0bjectjve

The basjc electrochemistry of electrolytic proceSses has been given in

previous modules. The objective of this module js to present to you a
descriptive approach to industrjal practices 'in electrorefin'ing and electro-
winning processes. After completing your study of thjs modular rnaterial you
should be able to give a detailed descript'ion of a modern electro process,
its basic design, its matenjals of construction, and its operating problems.
 ACKNOI^ILEDGEMENT

"The author is grateful to the management and staff of the fol'lowing

organizations for ass'isting in the accumulation of thjs material."
Com'inco Ltd., Trail B.C. Oanada
INC0 Ltd., Toronto, 0ntario
il Kennecott Copper Corporat'ion, Salt Lake City, Utah
Insp'iration Copper Company, Miami, Arizona
¡t
Ranchers - Blue Bird Mine, M'iami, Arizona
fl
Bagdad Copper Company, Bagdad, Arizona
il
Magma Copper Company, San Manuel, Arizona
it
¡t- Canadjan Copper Refiners, Montrea'1, Quebec
il
ilat The Universjty of Britjsh Columbia, Vancouver, B.C.
The University of SheffÍeld, Sheffield, EnEland
The University of Mjssouri, Ro]la, Missouri
ti
ti

¡t

ll
¡l
¡t

il-
ti*t
II

[i-
II
lt _
il
it
il
¡i-
Ét

liÉl
¡t
H-
H
ti

ti
il-
tl
fi
H-
EI

ft_
n
g¡

ti*i

!l-
l'
t'
TI
Ti

ti --
H

I
il eclr*neiai 'i ';rgy [qui nment

4..l .I
 i+) '.

:''1.';: -r
¡-lqure +. 1.? Rotaiing CYIinder Contjnuous Produiii bn Cell.

] :.i

Rotati ng

[cnii nuüirs
Sheet of
.l4eial ,

:-.

Cont=ct^mi iig
An:de

ll e':trclyte

¡'lt
Figure 4..l.3 Rotating Disc Continuous Production Cell.

Rotati ng D'isc
Cathode

Powdered metal
continuously harvested
from the rotatinq
ca thode

Statjonary Anode Plate

4.1.3
 El ectrometal 1 urgy [qui pment

Both of these continuous cells were technically successful but failed to

be adopted commercially because the 'l
costs of constructjon and operation were
h i gher than for the mul ti p1 e parai e1 pl ate des'i gn .

b. Practical Current DensitY

The current density'is a term used to describe the irmourit of current
transferred to or from a given area of electrode p1ate. it has the units of
amperes per square foqt (ASF) in the Eng'lish Eng"ineerjng System ¿r,C amperes
per square meter (n/flz) or mjliiamps per square centimeter (ma/cmz) in the
metric system.
of the current density that can be used depends on several
The magnitude
faciors includjng the type of product desjred, the method of transporting ions
to the electrode surface, the actual electrochemical reactions and the crystal
growth processes as discussed in Module 3 on electrode kínetjcs. As a general
rule of thumb, the magnitude of the current density is usually ljmited to
]/3-1/2 the magnítude of the ion concentratjonjonjn solutjon. For example, if
there were 30 grams per'liter {gp1) of copper in solution, a current
densjty of l0-.l5 ASF coulci be used to deposjt smooth sheets of copper on the
cathode plates. if the current density were ra'ised to say 50 ASF, the transfer
of ions from the solutjon to the electrode could not keep up to the reaction
rate and powdery deposjts rvould result. The powders are often less pure than
the smooth sheets and so are less desirable.
c. Sol ut j on Mi x'ing 'in the Cel I

Solution flow between the electrodes was d'iscussed in some detail in

Module 3 so the reader is directed to review that material as needed. The
ma'in purpose of solution motion js to pro..'Íde a convective force to transport
the'ions from the soluijon to the electrodes, the more intense the niovemeni
of solutjon, the more rapid the ion transport. Since the practical current
dens.ity is limjted by this transport it follows t.hat the current ciensity can
be increased if ihe solution motion is intensified.
is createC by ihe electrode reactions themselves (see
Some movement
Mociule 3), but ii is not very intense. Research efforts have yielded ways
to increase the rnotion; two methocjs that have ach'ieved some cor¡merc'ial
success are shown in figures 4.1.4 and 4.'l.5. in the djrected solution flow
system in fjgure 4..I.4, the process solutjon ís pumped up the face of one
electrode and drawn down the face of the acljacent electrode, io creaie a
circulating flow. In the air sparg'ing systern depicted in figure 4..l.5, tiny
air bubbles are djrected up the cathode face. The bubbles carry tile solL¡ijon
+T' upwarcis pastthe cathode and this motion causes a downward flow at the anoCe,
tl
t1
tt
again creating a circular mot'ion between the electrodes.
ll
The praciÍ ca] curreni densi tJ' has been i ncreased by a fact,or of I .5-2
ijmes by the directed flow s-ysiem and 3-5'Li::res by ihe air sparging system.
i. The overall benefit with a'ir sparging for example, with a 3 fold increase
jn current dens'ity, js that ihe 100 tpd copper piant can be clecreased jn
size from 7BB cells to 263 cells.

r
t
i
Al tt

t
I
¡
Figure 4.1.4 Mixing Inciuced by the Directed Frow of solution.

t
!
l¡
,*'
4¿----
zr-
Soluti on out
in

Sol ut'ion i s di rected up one face and drawn down the

adjacent face.

Figure 4".l.5 lvlixing by Air Sparging.

Ca thode

4.1.5
fir¡
il
El ectrometal 1 urgy Equi pment
it
lt
il 4 .1 .2 El ectri cal Equi Pment
ilil
il a. El ectrodes
¡t
¡t
il The plates used in electrolytic cells have to satisfy certain structural,
il electrjcal and electro*metallurg'ica1 requirements. First they
'in
have to be
r"ig'id enough to hold their shape whi'le they are suspended the cell. The
il noimal method of suspension is to attach a cross rod or header bat onto the
tt boiting, riveting oi
it
!t top of the plate as shown in figure 4..l.6, by welding.
it The bar rests on the walls on each side of the cell tank.
il Idea11y, the plates would be very large so that the number that had to
¡t
ti be handled would be small. However, s'ize limitations are 'imposed by several
il factors jncluding the metal fabricatjon technjques, and the electrical parameters
Éá
*¡ Sheet metal rolling mills often limjt the width of the plates to 40-45'inches,
il while djstortion during rolling can limjt the length. Because of the narrow
spac'ing between plates 'in the cell, nominally 2" center-to-center, very little
bowing of the plates can be tolerated jn order to prevent short c'ircuiting.
Such i-¡owing'is d'ifficult io avoid if the piates are more than 40-45 inches
in iength, and thus a maximum size of 45" x 45" js seldom exceeded.
size of plates in cqmmon use are in the range 24" wide by 36" deep to
The
4?,\ide by 42" delp, or 6 ft2-12 fú per side. The largef sizes reflect the
metal fabiication limitations whjle the smaller sizes reflect the age of the
pl ocess p'lant. Those plants orig]na"l'ly .built.around the turn of the century
Itilized manual handling of the electrodes and the 24" by 30" size was as
iarge a plate as the average man could ljft repeatedly.
electrjcal design ljmitation is based on the cu¡rent carrying
The
capacity or ampacity of the thin metal piates. Depending on the metal
whethei it js tjtanium, stainless steel, aluminum or copper, the ampac'ity
is jn ihe range 400-1000 amperes per square inch'if overheating and solution
boiljng or ptáte di:stort'ion is to be avoided. The size is also restricted
by *uhe amount of current that can be passed through the contact zone between
t-he bus bar on the cell and the header bar on ihe electrode. Contact design
is djscussed in the fol lowjng section.
The electro-nie'ua11urg'ica1 factcrs that affect p'Iate selection are con-
cernecj wjth ihe material Inat:s used and any treatment given to'uhe material.
primarjiy, the material must be ine¡i to the process so'lution; to satisfy this
crjierion, aluminum piat.es can be used jn a z'inc solution because it is not
attacked úy the solutjon whereas a copper p'late would djssolveo a'ibeit very
slowly, ani contaminate the solution wiih a few parts per m'illion of copper
i ons . In contrast, al umi num p} ates woul d corrocie 'in a copper sol ut j on and
for use.
so wor-¡ld be unsuitable
b. Coniaci SYstems
Current from the rectifier is brought to the cells through bus bars,
normaliy made of copper and it must be transmjtted to the electrode plates
through a contact sysiem at the ends cf t.he header bars. The requirernents
of ihe contact sYstem are:

4.r.6
Figure 4..l .6 il ectrcde Pl ate w'ith Cross ,Rod or Header Bar.

Header ba r

Cathode p1 ate

4.1 .7
 El ectrometa'l 1 urgY Equi pment

l. Thatit is capable of passing the current.

2. That it ís easy to make repeatedly as electrodes are
withdrawn from and replaced into the ce1'1, and
3. That jt js ejther easy to keep clean and free of salt
formations, or that it is self-cleaning.
Three concepts have been employed to achieve these requirem"ents. They are
the "area", "line" and the "po'int" contact systems.
In the "area" contact the bottom face of the header bar is flat and
ljes d1rectly on another flat piece of copper restjng 9n !9p of thetocell as
shown 'in figure 4.1 .7. Aithough it would seem t,heonetically sound
js
have a

sign.ifjcant amount of area in éontact to pass the current,'it actually a

poór system since on a m'icroscopic scale on'ly a few po'ints are jn contact.
Furtheimore, salts can easily crystallize between the faces to insulate the
contact and restrict current flow, and moreover, as the p'late increases w'ith
the grow'ing depos'it, the header bar becomes deflected so that the full area
is no ionger in contact.
The ljne contact was developed as a result of the recognition of these
probiems wjth the area system. Three vensions of the line contact are shown
schematjcaliy in figure 4..l.8. The convenljonal knife edge has had wide
acceptance in tfre cópper industry; in thjs version a rectanguiar header bar
is rbsted on the apex of a trianguiar cross-sectioned bus bar, so that a "ljne"
appears to be in conract. An upside down versjon of the knife edge has been
ulir¿ Uy INCS Ltd. In this model t,he triangular component was at.tached to the
underside of the electrode header bar, and it rested on a flat plate on the
top of the cell wall. A th'ird vers'ion developed by Cominco Ltd. had a triangular
piece of copper attached to ihe underside of the header bar, and this triangular
tip woulC wibge jtself between two copper bus piates. In this version it would
appear that two "ljnes" were in contact.
I

I ii has generally become appreciated that in any of these contact systems

only a tew pó'ints were actually in contact on a mjcroscopic scale. In recog-
I

Ccmjnco Ltd. developed a "spool" contact design where

nition of ¿his fact,jnteniionally
- point contacts were formed. In this design, a round copper bar
i
was machined jnto a series of slots wiih the appearance of a thread spoo'I, and
the electrode headen bar was prov'ided wiih an inverted "V" on i,he undersicie.
The drawjng in fixure 4.].9 depicts the concept of the spool contact. When the
"V', notch in the heacier bar comes together with ihe spool shaped notch in the
I
I
i

bus bar, four djstinct point contacts are formed. This system'is self cleaning
due'uo wedg'ing act'ion of the heacier bar fitting into ihe spool notch and jn
addjtion ié sélt aligning, in that the electrode is fjxed in pasition in the
I
¡
I
cell.
t
The ult.imate test. of the contact systeni is ihe amcunt of current that
jt
can carry. Plan'u experience has shown that the "area" contact is limjted at
I
I

400-5CC ámpe.es, the "line" can be used up to 1000-2000 amperes while ihe spooi-
i-
I
po'int. system has been used at up to 20ü0 amperes.
i

I
I
I

i
i
i
Figure 4.1 ,7 Area TYPe Contact

Bar

Cel I Top
Bus Pl ate
Area Contact

4.1.9
¡i
;l
Figure 4..l.8 Line Contact Systems

ri
iri
tii
t1
il
il
¡tl

ii
il
i,i

i,i

Convent'ional Kni fe Edge INCO Knife Edge COMINCO Wedge

Contact Con tac t Contac t

4 .l .10
Fi gure 4 .1 .9 Comi nco "Spoo1" Po"int Contact SYstem

Electrode Header Bars

Spool Bar

4..l .l I
 Equi pment
El ectrometai 1 urgY

c. Electrical LaYout
Ifalltheelectrodeswereplacedjnasilgl"cellwith.parailei
3 miliion amps would be required.,
connections, an enormous current of about.¡00
at a potential oi áuout 2 volts to make tons of copper' If' on the other
to a cell' an enormous
hand, each electroáe was connected in se¡ies' one output of^about 500 amps
number of cells'wóul¿ be required with a rectjfier where 20-50 catl¡odes are
.l900
and volts. In practice a compromise-is^reached
.l0,000-30,000 amps and 100-400 volts'
used in each celi with rectifier oltputs
The actual configurations vary trom ptant to p1ant, often because of the
personal preference of the designers'
The detailed layout of the plant equipment is a comprom'ise
based on the
size, shape an¿-iosi"Ái tf,. cells, e'lecirodes, bus bar, solutjon handling
equipment, and tñá-mecrranical nanáling equ'ipment for product recoveries'
General"ly the cells are arranged adjaóent t9 another so that they can
91" These cells
share the celi iop-bus bars ui in-iiüut.4..l..l0. are then arranged
in the bus bar
in rows an¿ connácted to the rectifiár
'layouts
such a way
in
as--to
figure
keep
4'l 'll ' In the
length short. i;;-ivpióul qlult -are
shown
the rectifier
upper layout tne ceii, ur" divided lnto four groups and connected
in the middl. *ñiie in the lower layout the cells are arranged in two groups
with the rectifier at one end'
4..I .3 l'lechanical Equ.iPment

a . Sol uti on Handl j ng Equ'iPment

in almost every e'lectro'lytic plant the process solutjon is pumped into

the cells and ¡i tnán w'lthdrawn for some formtoof treatment, for example to
heat or coo1, to purify, or to add materials
jt. Thus all of the cells
must be tanks to
have to be fitt;d'wjth inlet and outlet piping and there
hold Lhe solutión, and pumps to move'it, as wáll as heat exchangers or pur"ifi-
cation equiPment.
Two types of piants will,bediscussed in the following sections, electro-
refjning and ellc¿lowinning piants, and each has a disiinctive solut';on hand'l'ing
system. in the former, impure metát js cast jnto anodes and these are refined
into cathcdes ín án *ri*niially cap'uive volume of solution' The circulation
system is shown in fiéute t+.1 .12. The reasons for soluiion makeup, reagent
additions and for a bíeed wiil be discussed in sections 4.2 and 4.3. The cir-
an electrowinning plant and
culatjon system shown in fjgure 4..l..l3 represents a leach p1ant,
it js evicjent iirái sotuiion'is regularly
íons ai*
idded from passed
"wor¡" from the solution and then
through the cell wnere the metal
bled back to the leaching plant. This type of process will be
discussed in
niore detail in Sections 4.4 and 4'5'
b. t{ateri al s of Constructi on

boxes have traditjonally been made from concrete, ejther as a

The cejl has to be protected
casting or u, un aisembly of pre-cast parts ' The concrete
have been used including
from fhe corros.ive solutions and several materialscoat'ings' In recent years'
lead, rubber, li;;r;iuir, on biiunen based irowel from sheets of mild steel
sonre p1ants have'"*áer;oánted wjth cells
fabr"icated

4.1 .12
Figure 4..l..l0 Adjacent Arrangement of Cells.

Concrete Cel I
Box Cel I bus p1 ate shared
between two cells

4."l .l 3
Figure 4..l..l1 Examples of Plant Cell Layouts.

Rectifier
El ectrodes
In Cel I

El ectrodes i n Cel I

Rect'i f i e

4. I ..l4
Figure 4.'1.,l2 Refinery Circulation SYstem

Solut'ion
Ma keu p Reagents

Holding Holding Bl eed

Tank Tank
Stream

Heati ng
or
Cooling

Figure 4..l ".l3 hlinning Pl ant Circulation System'

Reag e nts
Leach Liquor

J
Hól di ng M iX
ank

Heati ng
or
Cool i ng
To Leach
Plant

4.1 ..l5
El ectrometal 1 urgY [quipment

lined with rubber, while others have tested self support'ing boxes of fibre
reinforced plast'ic. Both of these show promise but have not had wicie accept-
ance in the industrY
The solution handling equ'ipment must also wjthstand attack from the
corrosive solutjons. To do so the pumps are generally type 316 or higher
a11oy stainless steel. The tanks are normaily made from fibreglass or
rubbér Iined steel while the pip'ing has been made fron'¡ siainless slee'l ,
p'lasti c or I ead I i ned troughs or I aunders.
c. Mechani cal Handl 'ing Equi pment
Some machinery is required to wjthciraw and repiace the electrodes into
the cells and to recover or harvest the products. The major piece of equipment
js usually a crane runnjng over the cell and into the electrode preparaiion
and product recovery areas. Figure 4..l..l4 shows a crane in operation at a
modern p1ant. These cranes have traditionally been controlled by an operator
riding in a cab on the crane. However, in a few of the most modern plants,
conrpuier control of the crane has been 'incorporated.
In some plants, cathode blanks made of alumjnum, iitanium or stajniess
steel are placed in the cells, and periodically withcirawn for product recovery.
The metal product is separated from the cathode blank by a type of crowbar
action where a sharpedged bar is forced between the product and the blank.
This was o¡iginally a manual operation, a me'uhod which pers'ists to the present
day'in many plants. The job is very mundane and hjghly labor intensive; it
is generally environmentally unsafe as well. These factors have spurned the
devélopment of mechanical stripping dev'ices which presumably will gain universal
acceptance except jn the smallest plants. A typical nechanical stripp"ing
operation js illustrated in figure 4..l.'l5.
In coppei refineries, cathodes are often fabrjcated from thin electro-
deposi'ued sheets orig'ina1'ly grown ona rig'id b1ank. After 24 hour s in ihe
ce1i, the blanks are wjthdrawn and the product manual}y separat.ed. These thin
sheeis are fab¡icated into cathode starler sheels by attaching loops of the
same maierjal anci suspending the sheet from a header bar. This entire fabrj-
cation is performed manually as in fjgure 4..l..l6, in smali p'lants and on sheet
preparaiion machjnes like that shown in figure 4.1.17 in iarger plants.
4.4 [l eitrol Yti c Sol uti ons
.1

a . Mslal I on Concenrra'ui on

The first and most obvious component in any eleciroiytic solutjon is

the metal ion io be recovered. The concentraiion for this ion is chosen so
as to allow a reasonably high practical current density bui beiow the solu-
biliiy ljmjt. In a refinery where i.he solutjon ion conceniraiion'is maintajned
constant by the dissolving anode, a concentration of 7Ü-80 percent of the
solubil jty I jmi*u is used. For example in a copper refinery the copper ion
is 5C-55 gp1. in a wjnning plant ihe feed soluiion frcm the leach p'iant is
usualiy close io the solub'ility'limit while lhe cell +verflow solu-iion r,rould
have 5b perceni of the metal value removed; a copper'*¡innjng plant would
operate at 2Ü-30 gPi.

4.1 .16
 .l
Fi gure 4. ..l 4 Crane Serving ElecÉrode Area. The crane operaton is trans-
poried with the crane as 'it travels the length of ihe plant
carrying a full batch of electrodes to charge one cel I .

4 -1 .17
Figure 4..l..l5 Automat'ic Stripping of Cathodes.

A vjew of ihe automatic strippjnE machine at the Cominco Ltd' zinc

plant in Trail Canada. Cathocies move from left to right in the
pt'rotograph along' the "walking" conveyor toward the stripping mechanism
at lhe extreme ri ght hand edge of the photograph.

4.1.tB
F'igure 4..l. I6 Manual Preparation of Starier Sheets.

;;d

u:g{.
,-'
::

Photograph of the manual cathode sheet preparation such as that

occurring at the Ray Mines electrowinning tankroom.

4..l.19
Figure 4..l..l7 Mechaníca'l Preparation of the Starter Sheets.

Mechan'ical preparation of eiectrolytic copper startjng sheets at Magma

Copper Company's San Manuel, Arjzona copper refinery. Thin electroiytic
copper sheets weighing about i0 pouncls are fed to machine which trjms and
embosses them, attaches hang'ing'loops, inserts copper hang'ing rcd and
stacks them'into bolsters in sets for ciirect charg'ing by overhead crane
jnto cells. Starier sheets are produced in a Z4-hour electrodeposit of
copper onto a roiled copper mother blank from which they are strjpped by
hand.

4. I .20
tl ectrometai I urgy Iqti'i pment

b . Suppoi"ti ng Ei ectroiytes

Metal ions generaliy have low diffusjon coeffjcients and hence move
around s1ow1y, even under a strong electric field. For this reason, the
solutjon resjstance and cell voltage would be high if only the major metal
jon were present. in order to decrease the resistance, an ion wlth a high
mobility is added; usua'lly it js ihe hydrogen ion, added as an ac'id. The
majoriiy of the electrolytic processes emp'loy 100-200 gpl . of sulphuric
acid, although some are based on chlorjde or hydroxide solutions.
c. Sol ut'ion Add'iti ves
Additional components are present 'in the sol ution, somet'imes useful ly
and sometimes harmfuliy, and each plant will be djfferent. Examples of useful
fore"ign components include permanganate'in z'inc solutions, cobalt in copper
solut'ions and borates'in n'ickel solutions. Harmful components jnclude chiorides,
fluorides, caicjum and magnesium in many sulphuric acid based solutions.
Almost every plant makes use of parts per million levels of organic
materials such as an'imal 91ue, gum arabic, thiourea, lignin sulphonate, a1oin,
etc. in order to control the crystal growth behavior of the cathodes. The
type and concentrat'ion of each addit'ive is often peculiar to a particular
¡

pl ant.

tf

f
i.:
t

i.:
I
I
{.

4.1 .21
il eci'rcn*:a'l 'lurgy ll ecirore.f i ni ng

L[.qRilriilü ACT 2
"l!f\
t n
- , L
!:r ^^+L ^cr^ef
L i g! r ini ng

Leai^:.:i ng Acti vi ty 0b ject j ve

After compiet.'ing youn siudy cf tnis leai'¡ling aci'ívity rnater íai you should
be able i,o desc"ibe ihe gurpose fo'r" elecirorefjiring;'iis elecirochemisi.ry and
some of the des'ign cansjienaiicns ihat^ are jrnpo:"iani in pTan'r- praciice.

4.2.1 cf '¿he Process

The Purpose

As t,he naile iilpiies, elec'ircrefiaing is a process where impure meials are

refjired to a hjgher p'-rri'"y b-v an eieclroíytic method. A bonus derived from ihe
process is thai the peials that were present as impurity elenents jn the +iginal
neial are separaied and can be recovered subsequentiy in ihe"!r high prrrity
forms. Thus not on'ly js elect;"orefin.ing'capable of purifying the ;rajor metal
in a technical'ly superior way io any other process step, it js also the only
way some cf the minor metals can be separated.

4.2.2 tlecirochemistry of the Process

. a. The Sol ubl e Anocie
Bullian frorn a smelting operat'ion would con"la"!n 90-99iá of the major meta"l ,
such as copper, iead or nicke'l, and an assortment of impurity metals. The
impurities would jnclucie some of the noble metals, such as go]d, silver,
piat'inum, selenium, etc., anci pcssibly some elenents less noble than the rnajor
metals such as iron, bismu'uh, arsenic, anrj aniimony. l¡Jhen the bullion is
piaced in an electroTyi-e and ccnnected as an anode to a power source, some of
thE elements are selectively d.issoiveci into the solution. Elemenis that are
less'noble than the r*ajor neial would have the strongest tendency to dissolve
while the metals that are mole noble would rema"in in their metellic state. The
selectjvity sequeni:e which ind'ícates the degree of nobiiity of each element, is
indicated in figure 4.2.'1. Elements to the right in the table would nornrally
d'issolve in preference to those listed to the left.

F'igu'*e .1. Sel ectj vi ty Secuence of a Few of the il emenis

4 .2 .

Pt AU Se Ag Cu As Sb Pb 3i 112 Fe Ni Zn Mg AI

Increasing Decreas i ng
Nua'ii i ty Nobi'lity

,The noble impurities forma slime, which is a term used to describe the
somewhat gelatinous mass of metal residue that is left after the anodic cycle
is completed. Two distinct types of slimes have been observed in refineries.
In copper plants the d'iscrete metal particles fall to the bottom of the cell
from which they are eventuai'l,v collected, and hance are called falling siimes"
In contrast, in a lead refjneny the. slime forming eiernents const'itute a larger
fraction of the oriqinal bullion and instead of being present as discrete

4.2.1
 El ectrometal T urgY El ectrcre fi n i ng

partjcles,theyarepneSentasanetworkof¡^.ibbons:Theleadisdissolved
awa-v from the neti,rork to leave a skeleion-l'ike st,ruciure ihat has enough
-tach
stringth :o support'itself jn the.solutjo¡:. This 'is a hanging s1ime.
of tnáse sl jmes requ'ines its own process'ing iechn'!ques and equipment for
recüvery of the metal va1 ues. , , , r l

-::b"CathocjeSeiectivity..
Ár iiie caihode, the niejcr r'retaj ions are s ieccroiyticai 1y redi;i*d to
netal , and s"ince reduciion is the oppos'ite prccess. to o¡1daiion, 'uhe selecti-
v'iiy wouid wcnk in ihe cppcsite way. Ions more noble ihan the major metal '
wouid be redi¡ceC, whjle iens'iess noble would not be affected by the electroiysis.
Thus refin"inE takes piace b'eca¡.:se only ihose impur i:ies
ihe pun'ificat.'1cn or
less noble than the rnaion meial can be djss*lved f,ram the anode, whi'le oniy
those more noble than the major metal can be ¡eciuced with it at the cathcde;
4.2;3 Cansidera-uians in the Des'ign of [qu i pnrent
a. Anodes
in practice, ine impui-e ne'"al bullion is c3:t. inio rectanguilr plates
one to iwo inches ti^;'ick and is p'laced jn the cell between two cathodes. i,Jhen
cur¡ent js paEseC thr^ough th* celi the ancde dissoll'eso gradila"lly'necoming
th'¡nner. l^ihen 85-911?i of ihe nietal h¿s d.!ss*l'o'ed thl anode is wit.hdrawn from
the cell, renelied, and ces'u into a new anode. Scre recyci'!ng and reprocessing
of ancde bulljon is a normal part of the process. If more than 85-907i were
dissolved the anocie would lose'jts siructural str^ength and possibly fail
agart or form a shor*u circLlii to the ca"hocie
I
1',-+ t i,'n
uU>L ¡l\E *i.e carnoies discu:secí
-1 ¡
'in learning activ'ity'l , anodes ai e
eqr;ippeci vrjih a means of suspension ii: ihe the cell and a means of,"he
coniaciing the
rt
f

bus Lars. The conventjcnal way of producing ihe anoCe is'uo cast lugs onto
:he iop cf tne claie so ¡ha'u ii has the shape illr:st.rated in figure 4.?.?.
¡
!

i . :. ll '

b. Ca'"hodes

lhe me-"al -;har fcrtrS ái -rr';e:athoie is t:ie pu:ifiei produci fron the
process and as such mus'i nct 3e conianinated b;vt he materiai used as the rirclh9r
shest. üiher requirenenls af ihe'caihode mother sheet are that it has sufficient
ic carry the li:t-'i'ent, and
j+ *,,-+ "^tc-^*-ty'uctecj
sirength be s*'lf suprorting, encuEh co¡:auctivily
in such a way ihal *;he punifÍed product is easii-v
reccvr;red wh*n 'ui,e elec;rcae is w'i:hCrawn. Twc t-r,pes af nother sheet-q are
in coir-¡lct use 'irr industry, one is fo:^meci frcn a ¡retal such as iiianium, siain-
less st*El o:" aluminum, fit.ted w'iih a header bar and contacijng Cev'ide as shewn
ir, €{¡,,¡r, .¡1 e 2. Th.is iype is called a blank ani is reusable"
Ill l:Uu:: -":,J. ¡ii¡J -y"HA lJ !ql i!L s

so;r: ;;an:s, b;ai,:s ere p1aced jn the cells anc leii ícr ihe en*iire
in :!-
a: :- Ul' '
i ¡i3 -1-*¡siticn process, 3-15 days. in oihe: plants the blanks are j*ft
Ll¡C uCU!
:n ihe cells ri-rit't-^'¡
cniy onejÁ!day, ihe ihin meial cepss'ii
a^'-'
is re*ct'ed fr*n ihe blank
L-'- -' -:-- '* i-
¡< ¡ ir:c:- ¡-tr
*,.- T liLeC n'l;ii ^--:''-
::l'JlS ailir d tica;tl ucr c: :iiJWI¡ ¡!¡ fiCUfg 4.?.4.
Tr:ca:13 r::'=,¡i 5:al":¿r Sl,¿¿i C¿i:¡CC:S ¿ii ¿f= ile.:=d'i n the CellS f*;^ a 7-.:4

!//
Figure 4.2.2 Typical Cast Anode l,jith Suspension Lugs.

Fiqure 4.2.3 Cathode Blank Fi tted Wi th Header Bar.

Ri vets ho'l d i ng structure

together

Header Bars
sandwi chi ng the
Cathode Bl ank

Cathode Bl ank

4 -2.3
Figure 4.2.4 Starter Sheet Cathode as Fabricated.

Figure 4.2.5 Rotating Drum Apparatus for the Cont'inuous Productjon of

Lead Cathode Siarter Sheeis.

Conti nuous Sheet

io fabrication
machi ne

i,Jater cool ed
\t Drum
*-/
mol ten I ead

4 .2.4
 il ectromeial 1 urEy El eciro refi n i ng

day depos¡'tioncyc'ie. in lead refineries, starter sheets are made by casting

thjn sheeis of refined lead metal" Generaliy a cold rotating drum:s suipenáeO
in a vessel of molten meial and a coniÍnuous sheet is withdráwn from the brum
(figure 4.?.5). The sheet is then cut to iength and connected to a header bar.

. During electrolysis, eciges tend to aci as high currenl density s'iies, and
decrease cur:^ent to the rest of the electrode. Añ anode will dissólvu roi"
qu'ick1y near the edges and a cathode will col'lect metal more quickly at the
edges. To avoid uneven growths the cathode is usually made sfightiy wi¿ér tnan
the anode. Figure 4.2.6 shows schema'u'ica11y the resuits from a-sysiem where
the anode is wider than ihe cathode anci from a sysi,em where ttre cáthode-is
wider than the anode.
c. Cel I s

The cell box must allow for easy placement and withdrawal of electrodes
and for free moi,ion of solui1on. This 'is generaliy accompl ished by making all
cells 4 to B inches wider and 6 to 12 inchás deepei than ihe electiodes. And
as we11, free space of 6 to B jnches is often required at the ends of the cells
to accommodate feed pipes and heatjng or cooling coi1s. The space below the
electrodes js used to collect the tail;ng slimei. The cells are often fitted
wjth a sioping bottom and a bottom drain to allow the slimes to be flushed out.

- A copper refinery electrolyte is almost exclusively a sulphuric acid

solution and since lead forms a very low solubility sulpnate *i'tn;n sulfuric
aicd,'it is an ideal liner material. Rubber has récentiy been used, but it is
necessary to protect the rubber with rigid plastic platel to pnevent the
electrocjes from accidentally cutting it.
In refinerjes where sulphuric acid js not used but rather solut'ions
capable of dissolving 1ead, such as fluoroboric acid, other materials are
required. A particularly difficu'lt problem exists 'in lead refineries where
a fluorosilicic agld, capable of so1ubil'izing a vast variety of materiais,
is used. A trowelled on mastic coating has 6een w'ide1y accépted in these
p1ants .

d. Contact Systems

Refineries generally operate at low current densities, about 20 ASF so

that the current.drawn by each electrode'ís appr^oximately jSO-a.OO Amps. The
triangular (1ine) type of contact has had wide'acceptancé and jn somá plants
a triangular contact bar is used on both ends of thb electrodes to improve
the current d'istributjon. One variation of the triangular bar system, employs
a round header bar restjng on the triang'le, to achievé a single loint-.ontu.i.
Cl':anliness has been found to be of extreme importance in order to obtain
a sound contact,.so regular and thorough washing is practiced. In some plants
it has been found necessary to hanrner ihe headei bar'onto the triangie tb -
ensure a clean metal to metal contact.

e. Recovery of Soluble Impurities

As the metals less noble than the major metal 'increase in concentration

4.2.5
$

Figure 4.2.6 Cross-Sectional Views of Cathode Deposits in Anode l^lider

and Cathode l.Jider SYstems.

Anode Wider Cathode 1¡li der

Anode

Cathode

Anod e

Excessive
edge growth

Superior conirol of cathode depos i t

growth resu I ts from Cathode l^lj der System.

4 .?.6
il ectrome'.a11urgy [.iectrorefining

they'r-end ic iec:'ease the conduci"iviiy of ihe elecirolyie and increase the

cell voitage. It is desirable to recover these djssolved netals in order to
avoid wasiing power and to real'ize thejr econom'íc value. in a copper ref inery,
for exampl e , ni cke'l graduai 1y i ncre ases to 10-20 gpi and the sol ut'ion j s
treated con"i jnual iy for jts removal and r"ecovery. In a s'il ver refin€ry,
copper bu'ilds up to a level of several grams per liter and is periodicaiiy
stri pped f :^om ihe sol u'l j on .

In a large integnated refinery, a s¡'nal1 portion of the electrolyt,e'is

regularly bled awali, in order io remove foreign ions. in a cocper refinery
the lirsi siep is lc strjp the copper from ihe solutjon, normally'in I'iberator
cel ¡s equi pped wiih i nso.i uabl e anoces ; the concentrat,i on i s gradual 1y reduced
from about 50 gp1 to about 0.5 gp1. The copper free solr¡ijon can then be
passed through a crystalI'izer where nickel sulphate is precipitated and
separated from the solut"ion.
f . of insol ubl e impuri lies
Recoveny

The falling slimes formed in a copper refjnery are allowed to collect in

the botioms of 'uhe cells unt'il the anode life js over, aboul l4-28 days. They
are then collected by removing the cathodes and anodes, siphoning oíf niost of
the cell solution and then flushing through the bottom of the cells jnto a
filterjnE or centrifuging cjrcuit.
When hanging sljmes are encountered, the spent anodes are withdrawn and
brushed io scrape off attached impuritjes. These slimes are then washed'in
plant electrolyte to d'issolve entrained lead and cenirifuged to renove ihe
l iquid. In s'ilver and inciium ref ineries, the fresh anodes are placed 'in cloth
bags which reiain the slimes.
Collection of sl'imes and recovery of the'inrjivjdual metals is always an
important part of the refineries operations, jn fact the va1ue in the sljmed
metals often exceeds the entjre cosi of operating the refinery. in a sense,
the major metal is purified at no cosi.
g. Tankhouse Problems

Considerable efforts are required to mainta'in proper concentrations of

the key metall'ic and organic species jn the large volumes of solution. l^Jater
washing contributes unfavorably to the volume balance, uneven rates of dis-
solution from the anodes and deposition on the cathodes affect the metal'ion
balance, and jn addition, the organic species added to promote the growth of
smooth deposits can only rarely be analyzed by normal analytical techn"iques,
so an e¡.'erienced eye is often required.
Frc,¡il d capitai cost point of view, a refinery has a large inventory of
metal tied upin the process, not only of the major metal, but of impurities
as well. Metal entering the process does not emerge as products for l5-20 days.

h. Trends in the Desjgn and Layout of Plants

in the oid plants, starter sheet cathodes are prepared manually; this iob
js now down almost exclusively by complex machinery. Anodes once cast

4.2.7
 El ectrometal 1 urgY il ectronef i n'ing

individually are now cast as a continuous strip and stamped to shape to y'ie1d
irñ"riór ¿i*unr.ionai tolerances. A large_number of concrete boxes, each with
iii own cell i'ining are common in older plants but now 1arge, rubber ljned
steel tanks with pédestals to support tlre bus bars repiace the smaller boxes.
Anode residence has been reduced from 30 days to 15 or 7 days and overhead
cranes that once pu'l1ed the cathodes in a series of s'ing'le steps, now completely
á*piv and rechargb ti.'e cells 'in a sing'le. lrlq. short circuiting wh'ich was once
jnfra-red
detected manualli is now observed by soph'ist'icated or voltage scanning
techni ques.

4.2.8
!

'I
 il e,::,.*n* ;e i T u'r'gy iri
r ¡,- i ..: -'l ¡...Ju¡
riiLlw+¡ ;-l ^.
lrrr, -.t¡L:c)

ir- : a - j - * r 1. . - r r ', i- i ,- ! ;
:5. ^ ¡:,irL ^-
.:L*¡ '.1 ,;,J j=!,;'.'+
^- , . . -

5c;¡e *i' ;he c¡:¿i.: :f p,-ac:i.;cl cl,:,;',-r+r*i-'!-ilg trpcÉs:üs al,"a ,ijscusscú

;n ch:s *r*s:icn i,*iih t,h* p';r pcse cf inircd;-icing rnc ieacer io i¡':e ,,si-i¿ll i*ii:ts,,
Whifr
rs,¡!g¡¡
.jÉ.^r,, a--r:¡
dui t
" jrL!;
i- r+)ia ': ih=
Jil_- u'(¡c ¡r¡
uucfdl I ¡i"ccss:, T. -- j- 3n ¡p:aeci:iicn
.,^., gejn
If licu fcr:
rha ,j=.*- -l-=,..
L,¡c L":-.u. -,i.i;,' tf i,,-i ,lrci g iiCbr-e :i'; -.:1 3 lg5S lüb : jnlU,--;teS: iiC :.3.:jS:S ,f
*on-uili i*rn ca'intde ;r+wrh ans ;i,;rl*L¡ni f,;rm ¿n,3iÉ di ssciijxicít ; ;ne :;ec-ia;
¡eature :;l I ;¿s;;ns '.-nal a,.e s¡;ie :,:¡Es lse{i. inC ;h"3 ;lli;¿:: t:ies3 fac:or^s c:n
have on ihe ces;;l ai':d +üeratjün üf á t:,,: iinei;,, y0u ,"{il'j hava Jí}iers¿*od ihe
CCnCeaIS ffesenii: 3rj':r.

4.:. I ?r'i nci pal I,ipur: :i as

t¡¿
Tn *h¡ n*^
urr= ;;i'r:'Ji*ü5 I earil j g acii vi ii*s the ¡"e f ini ng pr ccesses fo:^ copper and
lead have Leen d'!scussed. .11 acd.it:üntc ihess, njc,{ei is alsü refjneC in
Tar:gg tcnnagess end virtualiy all of ihe gold, éilver, indÍum ¿nd p:aiinum group
*etal s thal ate l'ecovereC fram base ::ietal -cre cünceni?"aies, are re¡over"ed b!
rJsifiü slectrorefirring as a pr-incj*al prccess siep.
The me-r.als that ¡re iijore noble or less noble ihan ihe n,ajor meral in the
electrochemical s.er"ies 'rsijal iy have simi I ar characterisiics Íñ the smel tinq -
process that prod;Jces the crude bullion and so soine separation occurs jn the
furnace- The less noble elenlents gxi¿ized inir-tná rl*g,*
*n,l.v the;cre noble -are predominantly
;mpurit'ies to accompany the majür meial jntc the
]*gyiilg
br"¡llion. Electrcr"efining rherefare, is utJ¿ pri*ui'i':y tá sepa.ai,e ihe major
rnetals frcm ihe n',c:"e nabie irnpuriiies. Howevbr, smaii quanriiies of the less
ncble meiais do repori to the bullion and hence to the iefir"rery e;ectro1yte.
Figure.4.3.I summarizes the deportmeni cf elements found in varlour nuiii,ing
operaticns.

. By far the la:"Eest ionnage meial retined by eiectroiysis is copper; a

view of a modern-ref-inery 'is iho',+n in f"igure 4.3.2. Refeiences j-, l'*ná :
provÍde infcrnative back-up reading on tñe process. Lead is eleitrorefined
cn'ly when bismuth'is present in thé crude Uil¡lion, múst other ele*rents can
be fire nef,ined away from lead, but bismr¡th is so much like lead jn its
behav'ior during srne'ltinq that oni-v the superior separet'ing abiiiiy cf ¡he
control 1ec poiént'ial e1 áctro'tyi'ic"_ref i can separate the two métal s . Fi gure
-
4"3.,1 shows a typica'l lead ¡'e?inery. 1;1b RefJrences 4 and 5 provide useful uacÉlup
reading.
Ni*:kel bullion-is partfcuiariy useiul in its ability to collect the
p"latinuT* Sroup metals anci hence ihe refin'ing prccess serves as methori of
prociucing a conce;iiraie of ihese metals. A-view of the njckel iáiinánv is
shown in figure 4.3..4.

4 xr
 :i. ': '- .::
ti f.ri--::ii- ::
,rir: i,ii:..:r_';:
i,:,."!_r :_i::-: 1:ri,
.: j;::.'lr::: jj
lr

1, irij.l:+:r',:l
-;-l:,i.":
::;.*
:, 1;'";+:' '
r''"_ü:-i,l
;.
.. ,' ":..:' i"
-.:,..-i,
_-.,-":
l. i:'1-..
,

Summat:y
Processe
ji", :,.-,...: ..r:i

r' : l-
'::
:-,!
.:-:r,
'- .
-
,

_'jl.f .:; ii:,;.:i;::.it ,

': :: .:1.: '..'
)_ .a: . :::.
_i -.
.'j: -
r'': I '.j j; ,. :' '

it:- ': .{]:i-:l:'

¡ ; _. .. ¡" . .':(" 'r'r;

:.: _, l.r' i1
j r, :,r. I i.tr:rr'

.:'

*'
i.
L
&r,r
É..
I
&.'
$
lr
*.
E

[.]
É:

p:,
L:
;:1- .:r' 'i-:-'s
?,
Lr
i:
É.
F:,
F.,
f::
t.
F.
F-r
T.
t:

.9.,:,:.¡',!1 i :- 1:.¡ -:1.:. r', -, I

F.igure 4.3.2 A V'iew af a Modern Coppe: Refinery (Courtesy of Magma Copper
Company).

4.3.3
 Figure 4.3.3 A View of an 0perating Lead Refinery.

Photograph shows an operalor correcting a shori circujt 'in a cell. The

size and number of cells are evident. {Courtesy of Comjnco Ltd. )

¿_aL
 -o
+J
J

o
c
(+-
o
ln
OJ
.p
!-
=
o
(J
P
O
-
-('
;l- o
c) o-
= a,
-g
+
OJ
o
É. -c
+J
(ú
(J
-:¿
(J OJ
-c
z. P
O)
cO) c
+)
p
r{, U)
(u

o- nt
O .E
c tn
rú L
q- o
o +J
ro
l-
=
(U
a
OJ

o
Q

o
=
.<l .C
v\
cr)
-c
<f o-
(t
qJ L
(',
::t o
0, +J
o
tL -c
o-

4.3.5
[] ecirometa'l 1 urgy Principa1 Impurities

4.3.2 Majon Objecti ves

a. Coherent Cathodes

hjhile the fundamental characterjstics of the refining process discussed

in Section 4.2.2 virtually assure it of success, the electrochemistry can
often be hindered by easily overlooked factors. In orcler to produce a high
purity cathode, it'is usually necessary to produce a fully dense, sn:ooth
surfaced deposit such as that shown in figure 4.3.5. A solid, smooiir deposit
will not allow solution or impurit'ies which may be suspended or colioidal to
be phys'ica11y entrapped, as it grows.

A small variation in the concentration of the organic components can

qu'ickly change the cieposit from a smooth sol jd piece of metal to a wiry,
dendrj'ujc mass I jke that shown jn fjgure 4.3.6a and b. Th'is deposit would
entrap'impurit'ies easi'ly, and would have a strong tenciency to grow across
the electrode gap and short out the cell. Shorting could also be induced by
an uneven rate of impurity deposit'ion from one side of the cathode to the other
which mighi stress the deposit and cause it io warp.
Ifthe electrical contacts are not unjform throughout the ce11, then some
electrodes will carry more current than others. There will be a variat.'ion in
current density from one electrocie to anolher and sjnce the organic reagents
thai are addeci are often most beneficial jn a narrow current density range,
some cathodes nray be out of the des'irable range. These cieposjts wculd tend
to be rougher thán the others, and short circujting may also deveiop.
b. Uni formly Dissol ving Anodes

impurity segrega-uion during anode casting may produce a prcduc*u thai will
be consumed unevenly during electrorefining. This could lead to structural
problems jf, for example, one side dissolved in preference to the other sjde.
This problem was encountered durjng the early development of the Pb-Sn refining
process at Comjnco Ltd. 'in Traii, Brjtjsh Columbia. The anoCes were cast 'in
the cusiomary horizont.al plane but they had to be cast verticai'ly to achjeve
a uniform freezing rate from each face.
Casting can aiso procuce highiy oxidizeo surfaces on ihe anocie which at
times can cause very slow dissoluiion. The anodes become h'igh1y pc'larized and
operate in the "passive" region of the po'larization curve. Passiv'ity is
discussed in the module on Corrosion of Metals. This condi-tion will resul't- in
a raie of dissoll¡tion of the anode which is lower than the r ate sf depositjon
at ihe cathocie and in turn will result in a depieiion cf ihe meial ion in solutjon"
Figure 1.3.7a sholcs a group of anoejes that ciissolved uneveni-v; 'r.he amouni of
meial on each anode that was recyc'led was evicieniiy q'uii.e ciiffe¡ eni. Contrast
thjs with Figure 4.3.7b.
$ccasionally in lead refinerjes lhe impurity balance geis out of control
ciu¡ing anocie casting. The jmporiant factcr js the combjned pei cen'uage of
arsenjc and ant.imony in ihe bullion. Ideally it ts ?% if a proirer hanging s'lime
is to develop, bui if it is ¡nuch less ihan Z% the slime structure does noi form
prcper'iy anc a falling sl'ime resulis. The b'iggest probiem is ihat the slimes
canno'r- be col lected by the normal machineriy and so productjon decreases occur.

41n
Figure 4.3.5 View of a Smooth Surfaced Cathode.

Photograph shows an operator about to discharge a cel'l load

of cathodes onto a chajn conveyor. The pat-uern ernbossed onto
the start'ing sheets for rigidity is still evident on the
cathodes after the seven day deposition cycie. This is an
indicat.ion of a well controlled cell operation.

4.3.7
Figure 4.3.6a Sketch of the Types of Cathode Deposits That Can Devel op
l.lith Vary'ing Amounts of Additíon Agents.

De5..iiPlr,r, Se[l¡,]¡i - Laihú€

'. ; ,., A.:t ts

if:S9. ieari .,€r'i:! iies

5rollh SurÍaL€. 5r ali [[9es

!:j:l r,,. srlphljlt-e ¿!J

:f.:ial Alte5
H¿rd. iiaiPe¡, \'¡¡rs:],
i il fafi¿ i¡arrplg

Source: Reference 5.

AlA
 Figure 4.3.6b An Example of Dendr jtic Cathodes,

The deposits with their rough edges are an example of a',poor1y

controlled" cell where the conceñtration of ad¿it;on ug.nii-*ái
not at the optímum level. (Courtesy of Cominco Ltd. )

4.3.9
 FÍgure 4.3.7a An Example of Uneven Anode Dissolution.

The surface of this Pb-Sn a'l1oy anode d j ssol ved preferenti a'l 1y
!-- due to a non-homogenous as-cast structure. (Courtesy of Cominco
I Lrd. )

4.3. l0
Fi gure 4.3.7b An Exampi e of Uni fo:^m Anode D'issol ution.

A smooth surface is evident on this lead refinery anode after

its term of residence in the cell.

4.3. il
El ectrometal i urgy Addition Agents

in a faliing siime process the solution flow rate is carefully adjusted

so that solutjon turbulence does not suspend the metal particles. Such
suspensjon would favor phys'ica1 entrapment sf the impurity particles into the
growing cathode, w'ith subsequent loss of slime value and loss of cathode
purity. Too I ittle sol utjon movement, however, would lead to stratified
layers of electrolyte which would also adversely affect the cathode purity.
The optimum rate of solutjon flow through the cells'is three to five gallons
per m'inute .
4.3.3 The Act'ion of Addition Agents

a. The Rol e

Addjtjon agents are generaily organic compounds, usually not very pure
and often derjved from natural products, for example, lignin sulphonate
derived from the black liquor waste product formed in pu'lp mi"l1s, cedar extract
from the cedar tree, and alojn extracted from cascara trees. Some manufactured
compounds are used which are reasonably pure such as thiourea and polyacrylimide.

These organic ma'uerial s were probab'ly fi rst adcied accidentai 1y to the

electrorefining solutions and were found to have a very strong jnfluence on
the way the cathode deposits grew. Concentrations on the order of a few
milljgrams per'liter can change a hopeless'ly wjred iead deposjt into a very
solid, dense piece cf metal. Thjs effect js jllustrated jn fjgure 4.3.5 and
4.3.6 presenteci earlier. The precise actjon of the organ'ic compounds is not
completely understood, but it is believed that the molecules become absorbed
onto the high poÍnts on the growing cathode and block these spots to prevent
me-ual jons from reacting. The met,al ions can only deposit at the lower sites
and the surface tends to become smoother. Thjs is illustrated schematically in
figure 4.3.8.
Some of the unusual features of addjtion agents are that it'is often
necessary to have ? ar 3 additive agents present at the same time in order
to achjeve useful effects. The materials are very d'ifficuli to analyze and
somet'imes the'important component is on'ly a part of the riaterial that is added.
For exampTe, in lead refjnjng an aloes material is used but jt is believeci that
only 40-6Ciá is'inthe useful alojn forn. Reference 5 gives a descripiive
account of lead refinery addiijve behavior and points out the requirement of
us'ing both aloes and i ign"in sul phonate. Reference 3 ind-icaies thai in a copper
refinery giue, orzan A and avitone are used s.imulianeousiy.
b. üonirol
the acjdi t j ve nater j al s are vi i iaul'ly impossi bl e ro analyze
Al though
chemjcally when they are in ihe elecirolyte, an experienced plant operator
can cften rietect when one or another componen-u is out of balance. The usual
sign is -r-he appearance of the cathode deposit, anci soine opei^ators can actuat'ly
preiici a shcrt.age or an excess of any of the materials just by visualiy
inspecting the Ceposit.
A mare scjentjfic methoC is tc piot a poiarizatjon curve for ihe cathode
in a sanple of "uhe elec'urclyte. The siope and shape of i.he curve can shcw jf

4 .3.i?
Figure 4.3.8 Model of the Cathode Growth Surface Showjng the Way Long Chain
0rganic l,lolecules Become Attached to the Asperities and Lead
to a More Uniform Growth.

fuM
-L¿J- +¿
I
M. M. M. M' M' M'

i1
l,lithout organic tlith organic Result of organic
addi ti ves metal addi ti ves , the I ong act'ion 'i s to y'iel d
ions diffuse to chai n molecul es a smooth surface.
hi gh spots. attach to hi gh
spots and b'lock
them off.

4.3. I3
El ectrometal I urgy Addition Agents

t.he proper concentrations are present. Figure 4.3.9 shows typical polarization
curves that have been measured in lead refinery electrolyte. The optimum
concentrat'ions produce a fairly straight line wjth a voltage range of 50-'l00 mV
at 20 ASF.
Reference 5 gives an account of how the analysis of polarization curves
has led to a fajrly s'imple and direct method of assessing the lead refinery
eiectro'lyte. A continuous measurement is made of the cathode polarization, and
adjustments are made if the voltage js outside of the 60-100 mV ranc!.
4.3.4 Special Cell Design
Mostrefineries use a conventional cell design with 40-50 vertjcaliy
suspendeci cathodes in each cell box, however, prec'ious meial refining Ís pet"-
formed'in equ"ipment of a different design. An examp'le is in the refining of
the s'i'lver by-product of a lead refjnery. The anode slimes are treated in
furnaces to separate the bismuth, arsenic and antimony from the Dore bullion
contain'ing the silver, gold and copper. This bulljon is cast into anode slabs
rneasuring 1? inches by 6 inches by 1 inch, and the slabs are placed jn cloth
lined trays suspended in the horizontal piane above the cathode plate whici¡
serves also as the cell bottom. This single anode-s'ing1e cathode ce11, called
a Thum cell is shown in figure 4.3..l0.
During the refinÍng operatjon, copper in ihe bullion is djssolved jnis ihe
silver nitrate-nitric acid solution, whjle the sjlver dissolves from the a*ode
and deposiis on the cathode in the form of a mass of dendrjtic crystals. Gold
and the other more noble impurities form a slime which is contained in the
anode cloth diaphragm. A view of an operating sjlver refinery is shown in
fi gure 4.3. I I .
4.3.5 Size and Costs of Commercjal 0perations
Smelters of various capacities, between 80-600 tons per day, arr,: in operation
while mos-u refineries have a capacity of at least 200 i.ons per day. [lectrolytic
piants are built both as part of a compTetely integrated-smelter complex, and aiso
as independent facilities to hancile materjal from several smelters. The Insp'iraiion
Consolidat.ed Copper, Mjami, Arizona, is an example of a small iniegratecf p'lant
wjth a capacity of about 200 tpd, whereas the Canadian Copper Refiners plant in
Montreal , with a capac'ity of about 2000 'upd, is an exampTe of an independe*i
refi nery.
in a lypicai plant ihe cathccies are about 3 ieet square, so that each
electrode has about i8 square feet of exposed area anci each electrode operates
at abou'L 20 ASF or about 360 anrps total.
.l7,000There are usually 48 caihodes in each
cell wiih each cell recujring about Amps. Sjnce one thousand amperes will
y'ieid 53"5 lbs of coppel per day, an individual cell can produce abaut l00S lbs
.l000
per day, and a plant prcduc'ing 500 ipcl would have cells in operai'ion. The
cel I s woul ci occupy an area of abou'L 72 square feei each, and ihe pi ant area
rvould be aboi:t 707á occupied by cells. As you can ca1culate, the total plant
area is very 1arge.
La¡ge pi qnl s i ze and -uhe I aigc ft";r-irber of el ectnodes i nvol vei, (e.g . I GÜt
cel I s eaih ,*i ¡h 48 ca Lhocjes anci 47 anodes , preparati on equ'ipmeni 5 cranes , etc. )

4.3.14
 F'igure 4.3.9 Polarization Curves for the Deposit of Lead.

.¡60

l. Addition - Free
Solut'ion

124
2. Lignon 5ui phonate
Addition
3. Aloes Addition
4. Al oes pl us Goul ac
rl (mv ) 5. Aloes plus Goulac
6. Aloes plus Goulac

40

0 200 400 600 800

Current Densi ry (tln2)

Note: Al oes (gp'l )/Li gn'i n (gp¡) for each tes'u:
1. 0:0
^.
¿ 0:4
3. l:0
4, 0.5:4
5. l:4
6. 2:4

4.3.15
Fi gure 4. 3 . l0 Vi ew of a Thum Si I ver Ref i ni ng Ce'l 1

The shape of the cel I 'is evident. The horizontal ly d'isposed

anode slabs are shown 1yíng in the slimes collecting cloth
diaphragm. The cathode deposit which forms as crystals on
the bottom of the cell is shown collected at the right hand
end of the cells. (Courtesy of Cominco Ltd,)

4. 3. l6
Figure 4.3..l.l View of an Operating SiIver Refinery,

The cells are arranged in rows as with vert'ical electrode cells,

but'it js evident that only one anode and one cathode is used.
This refinery, of which about two-thjrds is shown, is capable of
pro.lrlcing ten mill jon ounces of silver per year. (Courtesy of
Comfnco Ltd. )

4.3.17
El ectrometai 1 urgy Additlon Agents

requ'ires a high capital investment in facilities and handl ing equipment. In

1975, a 200 tpd piant had an estimated cost of 40-60 million dollars. For
compari son the accoupl i ng sme'lter compl ex wou1d cost 80-l 00 mi I I íon dol I ars.

4 .3. 18
 .: :-.:

Liectromelal iurgy Purpose

LIARNiNG ACTIViTY 4
':' I

. ', n 4, [Jectrcwi:ining'
,
',, ,,
.,' ":
Leain'i'ng ActÍvity Sbjectiyg
",

of ihi s I earn Í ng act.i vi ty

FoI'lowi'.rg your siudy you should be able to
describe the electre¡vinning process and io describá iome af the índustrial
prob'lems encountered in the process :

'Purpose
4.4.1 of the Process
- .,: .'

. to pyrometal'lurg'icaily srnelt the material to an impure bullion andr

.:','-,.one way.'is
- then to eiec¿rorefine to h'igh purity by the methods discussed in Sections 4.2
and 4.3, and the other way is to leach or d'issolve the metal values in an
aqueous soluiion by the so-cailed hydromeiallurgy ancj then recover ihe- metal
- .: ,,,,ffom.,so]gt]on by elgglr9wín1íng. As !!ie term imp'lies, the meta], !s "wb'n" by
" eiecirolysi.s., in ,thjí leach-electrow'in process n' the aqueous solution cycles ''
..'-"''.,'.fiom thá -I.each p'lant where the metal values'are depleted, and then. back. to r"' ,:j
' ,the lea'ch plant. In order for this cyciing to work effeótively, each step
must prepare the solution for the subsequent step. l

l
,.,f,..';,.:-,,..¡n the winninE step, three essential functions are carried out to prepare
'the solution for leaching. Meial values are rernoved by the cathod'ic process
¡-n: a
can be gene¡atízed"6y-,'the fol lowihg eqüation whiáh' is,,tfie
jti-::lvhich jn ¡-ame is'woulQ
occui the refining cell.

This equation represents the esse_ntíal d'ifference between refining and winning.
Whereas in refining the anode is'soluble and yields the metai ions in solution,
in winning the anoáe is jnsoluble and decompoles water. The energy required
to carry out the anode reactíon is of the order of .l.25-1.75 volts which would
compare io a voltage of about 0.25 volts for a copper anocie dissolving'in a
refining celi and thus the energy requireci for electrowjnning is much higher
r
The th'ird funct'ion is the formation of acid by combin'ing the sulphate
-a.;

'ion from the cathode reaction wjth the hydrogen ions frorn the anode reaction
according.totheequation,'''.... '-!l¡

' - . '' :', 2H* + S0o= + Hzs04

, This acid'serves as the'leachant to dissolve the metal values ín the'leachiÍig"

\'
\
\
\

"i .._
 . :. : . . i: r' i.: .,::r'

tl ectron'etal l urgy Cat,h,ldi; F¡^ocess

- lt
; '

and-,thus,.thet'overe'l:l'cyc.1e ian'be summafjzed.as,,fajlows..i':.ln leach!ngu:,acidi,,.,,'..,1'

i s :'c'onsumed i"metalt. is dissol ved and water:'j si pro_dr;ced; whi'¡ e . i n' él ecirowi Í1fli ng,i,-,
metal:is producedn the'waleris decomposed, the oxigen is evolved ánd aejd is' -,.

noble ihan the major metai, was discusseci'!n the session on eiectrcrefining.
1t wasluf,ó*n'thai"ihe''refiñing'process uiilizes the selectivity in oná *ry",-..=,t,,,,
at the anode ánd ir¡ the opposite way at ihe caihode,' jn order- to prodúcé',,a'hig¡ '.
purity caihode depas'it:',, Tnis selectivity can not be uied to'such arivantage'in
winning, because the anode reaction does not involve the metal values. Thus
tne sel"áciivity"!s limjteci ro t;,e cainocÍe ploress where cniy those ions *orá
rnóble thán ihe"majar metal can be deposited with the desi"*á *ujár *éiál .--t¡.'-'i' .

less noble ions are i,herefore e:.clucied, and in a sense a partial refining is.' ',
accompl ished.

'.,, " This partia'l refinjng aciion is not enougn to allow tiie priJ,;;ti**
usable higir purity metal ,"however, sÍnqe the ns¡re rroble metali tl¡at wu¡qipf.s'ent',,'',
"- "i.-t.ti,"i'
:in the solution cou'ld significant'ly affect the cathode pu¡ ityr;, Foi,.,.éiampi¿;,i;f .

cCIpper.were bgin,g eiectrói3'z?d, inetals. such al s] jy*",. goid,- sejer¡iu* und

i*llurium could be incl..¡ded in ihe cathoce and while the former twc metels would
simp'iy be iost tc t.he ccpp=r, and.couli not be sepanately ccilecied fcr saie, the
latier iwo wculd seriousiy I'incier ihe subsequsn-u mechani"cal r,¡orking cf th¡ -
copper prociuct. Such hincirance occu!"s w1th concentraiions of selenium o¡,..,, l
te]luriumofoniyafewpartspermilljoninthecopper.:,
-_''
':..,.rj In the case v¡here one of, the less noble"netals I ike nickel , 2inc'or'rnangan-
ese was being electrowcn, the rnore noble impurity eIements har'e an even more
serious effect. The irnpurii,y e'iements ihat deposit ai the cathode along with
tiie'major metal form a galvanic couple wh¡ere hycirogen can act as the cathode
reacti on anci the ma jor meial cii ssol ui,i on as ihe anóde reacti on . For i-. exanpl e , :'
in a zinc cel I, jf antimany were present', bcth':inc and antimony would precip'itete _ i_-:

t,o fo¡m a bi-metaliic deCIoiit cf the tws neiels. Tiren hycircEen ions wculci b!
':; a ,..: : :: -
. :1:¡.r_'::,r.t:i:i-. r._-
.i , r!;rr: -.: .:7:, ,: ;.

la¿ql,
This cathade' r'eacEicn r.{eulC require a source of electrcns anci rhese cculc be
provÍded by the zjnc gcing back intc soluijsn ES
¡. .L4
ft't--.u
t1 ."! c'l I
' : '..: : .:' .-1.

As the reac'iions cc¡"¡tinued, antimony anci ¿jnc'v*oulci contir:ue to depesit, vthich

wauld üncúül"age hyercgen ic ci*posji anc ¿jnc i* diss*lve lravir¡g a firi* coilcen-
tr aied ani'imcny depcs'!r uhjch w*uld accelerate ihe h-vCrogen carñaCic and zjnc
anoC'ic reactjon. [v:nt'¡all.v the zinc would riissolve as fast c¡'f¿si,e¡^'¿han ii
cculci be d:pcsiieC, and viriualiy no zjn: wctlci remain cn ihe cathode. Thjs
raiher drastic :et *f c¡r cu¡n:tances '*¡jil *ccii; wher¡ ihe ani'i¡:ionv ccncen+-r"ai'icn
has ths value of cnl]'a feu¡ rnjlligl^árr.s per i'ite: in:he ple:,ng soirricn. The
m¡re nsblc ele¡i*n-"s iheref,o)-e are ver¡l inpo¡ ial¡i jn tha¿ iheir !i esence almcst
Figure 4.4.1 Simplified Block Diagram Indjcating the Cyclicai Naiure
of the Leach-Electrowin Process.

l'letal 0x'ide

Leach i ng Leach Res í due

tl,ohate (MS04)
IMetal
L t,tater (Hzo)
Recyc i e 0xyg en
Acid Gas (02)
(Hrsoo,

t'l ectrowi nn i ng

4.4.3
il ectrometal 1 urgy Anodic Process

always causes serjous, unwanted effects.

These jnteraciions of the more noble elements with the major metal are
used to advantage to purify the solution prior to electroiys'is, to prevent the
problems described above. In the zinc process, jt was found that if fine'ly
djvided zinc dust were mixed'in the acid free solutjon produced in the leaching
p1ant, the more noble elements would deposit on the dust particles, while the
particles themselves would dissolve. The process, termed "cementdt:s¡¡" can be
expressed as

4zno + zsb+3 -> (Zno - sbo) + 3Zn# 14.4.7)

and this step has been developed to the po'int where h'igh purit-v solutjons are
obtained which in turn allow the product'ion of very high purity cathocie deposits.
4.4.3 The Anodic Process

Since the anode serves only as the pcsjtive electrode and generaies oxygen
raiher ihan generating metal ions, it must be made from a materjal wh¡'ch w'ill
conduct eJectricity but which will not djssolve. ilost leach-eiectrowin iype
processes are conCucted in sulphuric acid solutions and s'ince lead metal and
lead al'loys have a very low solubiljty in this acjd, they have been used exten-
sively as anodes. Actualiy the behavior of the lead anode'is rather complex.
The metal is very slight'ly soiuble jn sulphuric acjd so that divalent lead ions
(fU*+1 form at the metal surface, and in turn the ions precipitaie as lead
sulphate on the anode faces. The anoeiic curreni axidizes the lead ions to the
tetravalent state and converi the lead sulphate to lead dioxide (Pb0e), and so
after a period of a few hours of electro'lysis, the anode becomes coa¡ed wjth a
lead d'iox'iCe film.
Thjs film js porous to water molecules and hydroxy'l rad'icals (GH-) which
allows ihe laiter to migrate through the film to the bare metal surface. The
hydroxy'l radicals can dissolve the lead to divalent jons which will be converled
by ihe anodjc current io the tetravalent staie and hence to lead dioxide. This
series of events leads to a thickening of 'r.he anode film and this th'ickening
s'lons the transport of hydroxyi through the film, thus a stabie, long.lasting
film with a hiEh deqree of corrosion resis'uance is cieveloped.
l,Jhile this pr ocess is going on, some of the d'ivaleni lead ions escape
from the vicinjty of the anode and remajn'in solutjon as partiaily charged lead
suiphaie particles. The solubjljty'is Ijmited to about 10 mgp'i of lead but this
is sufficjent to cause a decrease jn ihe purity of the cathcde eieposit as scme
codeposits wjth the major meia'l. Ten io one hundred pai-ts per
of the lead'ieaci
mÍllion of are commcn in zinc and copper cathodes.
In an attenpt to
preven-L this contamjnatjon, some lead free anades have
been deve'loped. The main theme has been to depcsit a p'laiinun gl'oup metal or
a combinaiion of these me'uals oni.c a ijianjum subs'Lrate, and whjle the scheme
is technjcally ssund, ihe cosi of such anade has b*en prohjbjiive tc ciate.
rh^ -*^ri^
I ¡3 di¡Utr
I i L process cf generaijng oxygen invclves tne fo¡mation of very
fine bubbles of the gas ai manjy pa'ints on the faces af the anode which rapid'Iy
rise and leave the cell sciuticn. As the bubcles íorn ihey pick up sniall

+.4.4
El ecirometal 1 urgY
5ol uiion Mixing

ciuantitjes of the acjdic cell solut'ion and then as ihey reach Thjs t'he surface ihey
ihe cell ' m'ist will
break and cause an acjd m'ist to be sprayed up from
usually exist throuSi',out t1",. piant atmosphere -yielding-quite obiectionable
workjng canditioni,-inJ co"roi'ive conditions f-o:^ ihe plani equipnrent'
The
probl*ñ'i, intensiijrd ur the operatinE curreni densjty is increased'

4 .4 .4 5ol ut'ion Mi xi ng

solut.ion movement js generated in several ways. As the metal

at
deposits at
the cathode, a u*iv thjn ráv"r oi-iorution in metárthe
ions exists the cathorie
hence jt
surface wh'ich has a lower dens1i' tf1.n the rest of so1ution and
jons' This
tends to riru uná ¡. r.placed by"fresh solution rich'in the metal 4'4'2'
creates a slow upward natural cánvect'ive motjon ai ihe cathode, figure
The anode generaged oxygen bubbles produce an upward mot'ion
at the face
of the anode which becomes *or" lnt.ntu as the current dens'ity is increased'
jn froni'
As seen in figure 4.4.3, the solutjon rjses up the anode and descends
ofthe cathode.
of the mot'ion actuallY occurs as edd'ies of solution
Most are circulating
at the toP half of the electrodes while very little motion is generated near
the bottoms.
solut'ion entering (rich in metal ions) and leaving (d9Pleted
New
in the
metal)-iauiei a tenerat móvement of solution in the cell but th'is does not
contribute greatiy io-tirá solutlon movement between electrodes because of the
baffling effect of the electrode plates.
whenmetals such as nicke], zinc and manganese, all of which are less
noble than hydrogán, are deposiied, there js á natural tendency to reduce some
hydroqen ions to -*t hyárogen aioms. Ínut. react to form gas bubbles (generally
;il;i í';; ,irá l i.r, move very qui ckiy up the cathode face . Thi s .moti on . occurs
imme¿iateiy aájalent to the cáthode and creates local intense mixing to the-
solutjon, and ir-irr. móst s"ignificant factorin solution mix'in9. Mixing allows
the cells to run at higher cúrrents than if the solution were stagnant'

4 .4 .5 Control of Aci d Mi st
Several methods have been devised to eliminate acid mist from the the
tankroom
atmospñer., th.ru can be grouped'into two generai typest .3) scrubbjng
acid m.ist at the vessel surface, and b) movilg a'ir above the cells to
prevent
Urii¿up of object'ional concentrátions of acid mist in the surroundings'
An effective method for mist depression can be achieved by form.ing
a

blanket of foam on the surface ói tf'l cell solut1on, using a proprietary solu-
t.ion of Dowfax 2A-.|. Th'ls reagent added as a few drops to each cell spreads-
out like an ojl sljck and js t.ótfreá into a blanket byplants
the rising anode bubbles.
and to a lesser
It has been used-.*i.niive'ly in copper electrowinning
áxtent jn zinc plants. Unfórtunatbiy, the organ'ic frothing solution reacts
unfavorably wÍth some solutions such ás coppei solvent extiactÍon solutions and

most zinc el..troiyies, renderiné it unusabie, however, where'itand prevent'ing

can be used'
the foam branket is very etrectiie ut-Jupr.rsíng the acidmist,
it from leaving the cell, figure 4.4'4'

4.4.5
 Figure 4.4.2 Natural Convection at the Cathode.

Cathode Bulk Solution

Boundary Concentrated
Su rface Layer i n Metal ions
Depl eted
i n metal
ions.

The solution within the boundary layer is

I ess dense than the sol ution i n the bul k
solution and this density difference causes
the sol uti on adj acent to the cathode to ri se,
thus inducing solution mot'ion in the gap
between the anode and cathode.

4.4.6
 Figure 4.4.3 Convection Induced by the lvlotion of Anode Gas Bubbles-

Solution
Su rface

Anod e
Cathode

Anod'icai'ly
Generated
Oxygen Bubbl es

The bubbles act as a pump to move

solution up the face of the anode.
As the bubbles emerge from the surface
they burst and spray a smal I amount of
acid solution as a mist while the majority
of the sol ution ci rcul ates between the
el ectrodes.

4 .4.7
 Figure 4.4.4 Use of a Foam Blanket to Suppress the Acjd Mist.
( Pnoto Courtesy of Kennecott Copper Corporation.)

Phctograph of an operat'ing electrowinning ce1'l at the Ray

Mines Division piant. The cell as viewed from 5-.l0 feet
away to show the cell walls and the electrocje header bars
as well as the foam blanket used to suppress the acjd mjst.

il- -

ti

ii
,t 4-+-ó
r!
¡
I

ii
il ectrome'ra11urgy Contro'l of Aci d lli si

l¡ihen the foam iype scrubber can nci be useci, a blanket of hollow or soljd
polyprcpylene balls 4-5 iayers deep has been successful'ly used. Thjs blanket
depresses ihe acid mjst and retajns the acjd in the cell, but the sma'll balis,
usually 1/4 ta 1/2 jnch in djameier ¡i*aj/ cause serious problems by enrering
pumps, ianks, fjlters, etc. throughout the piant. Figure 4.4.5 shows a view
of a piant equipped with plasijc balis.
A third:hrouding method is io simpl;r cover ihe cells with a sheei of
polythene as jn figure 4.4.6. this does prevent acid mjsi from getiing into the
plant atmosphere, but'it allows the condenseii acid solutjon to fall cnio the
header bars, bus bars, and coniacts caus'ing fairly rapid corrosjon of these parts.

is to allow the acjd to enter the plant atmos-

The second conceptuai niethod
phere. Air is horizontally along the iops of the cells jn a uniform laminar
moved
flow and is expelled from the piant through a scrubbing tower, figure 4.4.7.
This me'uhod'is the most generally appljcable but can be expensive, since major
plant equipment is involved.
A modi fication of the laminar air flow method is to place cover plates 6
to 12 inches above the cells. Then a current of air is directeci beiween the
cel I s and the cel I cover.

4.4.6 Problems of Heat Generation

The resistance of the €lectroiyte causes heating in the celi, and in
add'itjon heat is also generated at the electrical contacts, and other electrode
resistances. Some of this heat'is lost from the cell by vaporization processes
wh'ile the rema"inder raises the temperature of the electroiyte. In most electro-
winning plants there is an optimum solution temperature of 30-45'C. If the
temperature js too high, the anode corrosion rate increases, leading to a higher
lead content in the cathode, and a shorter anode l'ife. The cathode-deposit óan
be deleterjously affected by temperatures both above and below the optimum level.

solutíon heating is required, jn excess of that generated in the ce11,

If
shell and tube heat exchangers are genera'l1y used while if cooling is required,
cooling coiis placed jn each cell or an external atmospheric coo'ling tower can
be used.

4.4.9
F'igure 4 .4 .5 Acid Mjst Suppression by a Blanket of Severai Layers
of Plastic Spheres. (Photo Courtesyof Bagdad Copper
Company)

4.4.10
 Figure 4.4.6 Acid M'ist Suppression by Covering Cells wjth a Flastic
Box Connected to an Exhausi Fan. (ehoto Courtesy of
Bagdad Copper Company)

4.4.1I

i1e:a.ry4:
 of the Ac'i d M'ist bY APPf ication of a Laminar
qure
I r'sr
Fi Y 4'--4 .7 El i mi nati on
the tló-ót ir,.' bul I t. (Courtesy of
Flow of Air across
Powesj and Engi neering' Toronto, Canada)

'-Ñt
\r
3?+ ,,lll

:.,

4.4.i?
 lleia,l ñeccve:^1r

+"5 il;::,;:l :l:tr,':,, :rl;lg ?.*;ces-:s

i'-L!.¡:r'-j-.-i) l¡.J :.:i
at*i
,j-
,! ,^,!ic¡*i,¡:
L -L. JC
- ¡ tJ,.

*f ihe speci;'r'c :e¡trre s ¿i ccnn;rcja.l ¿i¿cl:^civje íl.inq cp*raiiGns are

So*:e
C'!scus:ec in ¡h.is secirDn. A.í;er siud;,ri-'rg;t yau should have a Eenerai under-
siandi,lE *f ih* i l¡ierac:.íc*s i n sal uii oi.r ar:d cn the el ect,-+de , tna i ccn:l i caie
ihe pr*i::s anü nECe ssi:-ate car€f:i: ¡teral;*n. Ycu sh¡,.,¡lci rr:áo¡ 'vnich re¡al s
;!^O
q¡s rpa:r;¡::
JsLUJ-:;i -'.
Júy ¡hi:
Lí,i) iFn,-ic:
J;.JLCJJ :,1 ,1 .,
Gi.J:y :r :-l-¡
u,:.- .--.:lc
Jir-J -€
ul +;^,
Ll;t: ',-. a ,^
/c'í -: ¡uk) ¡lJicli:uJ
=¡=: r'¡r1
Cí!"

4. 5 . I l4ei:l Rec:,¡er"-y and Sj ze of ihe uaperai j o¡':s

The :ri n:: :a'j :reta I s l-ecrve:'ed ;y el el::.c¡,i nni nq i ncl uce zi nc , coppe].,
nicrel ;.-: ri,i;i.i:ra J .'¡rl; I ¿ c:c:,Jr, iiti :¡lel i 3re ¡,^cc,,ieC .1 S )lr-p,.lC;C;S c i
ihese c:i:r- ne :¡l s. Zinc p1ar.:is tnrcijEnou; the wcrl d rar:ge in size ;i.cn l5l
to 8ÜÜ tú¡s per:a3'v;;th i.he by-crcciuci camiun i¡e¡al be"ing prüd.icEC"!n ihE
range 5üS ¿o 3ÜCü pounds pe:" da;v" i,ihere ccpper is elecirüwün, the pian"Ls ranEe
jrl size bf cne o) two tons oer da;,r uD io ZCil tc 3üC tons per da-v, witil a numbi'^
of plants being in ihE crde: of lü lo l5 icns pe: day. liickel is electrcwon
in on"iy a ferv p'ianis, witn 'rates of several huneireC pounds up io about 2Cü i.ens
per ia,v a:i i n tnese pl anis cobal t ;iieial '!s al so re.overec, ccbal',- ; s al rlost
exclusiveiy íound as a b¡¡-produci cf either nickel ol" copper. Manganese is
prcduced:as a pure metai exclr¡sively by elEcirowínning, with the pianis ranging
n'n.'''*fromafewhundredpouncistoseverali.onsper-day
: 4.5 .3 Chen.isi-ry and il ectrccnenisiny of the Zi nc Cel I

,, ' Z'inc js 'lhe highest tcnnage metal processed by acueous Elect,rawinn jng and
yet it ísone of the mos¿ technjcal.iy difficult. A study of the varjous
featureE,¡f the zinc cell'¡¡il1 provide an understanding lf sorne,:f the compiex
'interacij,¡ns ihar ceil be enccuntered jn elecirow'inning p!^ocesses in gener"a'l .

:: ,

Soluiions fnom a zfnc leaching plant undengo a thorough pLlrificatjon to

remove ur¡¡¡anted inipur ity elernenis, but ihe pur.!fied electroiyte siill conta.ins
a variety ci co¡¡pon¿nts. ilenenis less nclle than zinc ihai are usuali¡r "in
the soJutjcn include rf;anganese and magnesi;m ai up ¡ü 1C gpi and sma'iler
quantiiies of calcii:m" chrcmii.¡m and alumini¡n:. Species ;uch as antimony,
arsen'ic, ladniium, coba'lt, iron, njckel , seienium, copper and lead are more
noble th:* zjnc anci aye re*roved aimosl compieie"ly by lhe zinc dust purificaii0n
treatme¡:i, but are usual]y stiii present :n ttre b"l-10 mgil !ange. Non-metallic
ímpurit"lgsljkechloridesandfluor.ideswijlalscbepresentasw'ill5oine
organic rarerjals'trike gun arabjc or animai giue.
': i' ', ' .'

[ach of the above species can and usually cioes have an 'importani effect
on the wiirning of zinc, for example ihe less noble elemenis will general'ly
increase the resistance of the eiecirolyte and h*nce inc;ease ihe heat that
is generaied by the passage of current. Sjnce the solution will have been

/1 t1
--,¡^ t
 [] ectranetai iurgy l4cta j s Rec*very
purif ieio ihe more noble metals be pr esqnt a! conceni¡-atior:s below those
1;il'!
at which 'r,hey can be harn¡fuJ n but they witt-;;to;ir-;;-;i,;';;it;;_-;_i"c*."*_r_
rhe purit-v of the zinc met,al:

b. React"ions in the Soluiion

The mosr promineni feati:re 0f,, an operai'ir:g_zinc cejl
has a deep red coloration due io the p"*uun.. Jr-*unsrn*r*-uu"lir* is ihat the soluticn
'ion ani ii usuallv índiceies i¡i-ai
mosi;;-;ñ; iar*iiuí elem*nis a:-*;"r*"*;;#;_;;
r;nder c*ntr.ol
The logic of ihi: reatL;re can be ünüe rstcoc b¡ the fci lcu¡inc
ov= more of ihe mcre noble "ímpurities ís preser:t
argurri,in*u. If one
ln a n: ji.rer=tiiñ no"*ul ;;-;;;_
trat¡on, for examp'le"if an¿imor:y was present at_o.z-mllí";;r;;;;;;;ü¿;;;ü;
value *l !.ü4 mg/l= ir wculd have an advelse effeci by deposiiÍng
along v¡'ith ihe z-iric. A qelvanic celI i¡culc be c: ea;ed vuh*r*b-v at the ca;hcce
ih*
rris¡qlve and hydr'cgen ioñs wci¡ici be rec¡";c*c tá *yirogln-'-*i-l;;"" z.inc i,vould
rha h,¡¡.{-^-<
would reciuce the rérj permanEanate ion to the c*loriess mangunou!"iorJ,'finlJílt?f;--
hydrogen would return to iii ionic ;;ri;, -uJ.o*¿ing
to the eqi:ation.: ,,, , ,, ,,,,
l+. s. I .:'
This rea¿ticn waul d cor:t-inue u*ii I gll of ihe :,ermanganate was
i
i.he red color was dissipated, and t.hen untfl t*e zini caihace e*asui"¡ed and.
t dÍssol veC and the pr"rduc:,ion los;. Thus if the red colo.'itu*i] conJi*t.iu-.*_,
p'iani cper¿tcr shoul d rake act jan ic prevent ii.lis sei.jes to face the
*f ev*ni.s
c. Reaciion a-u the Ansde

i:call fram ihe previous section iha'u the anode js a iead al'loy, specífÍ-
cally'it'!s a Pb - c-.7a%,Ag ai1oy, ihat fn ie'^,-¡c* fcrms-atiuuc djoxice ccatincr}'
idhile the nai¡': r eacti*n is-the cjióonpositicn-o: $-É:er Íntc ¡xu¡.raÍ-= r::< ¡n¡.,{
ii;'cir:-:sn icns, a secc::c3l'iJ reac:'ion invoiviig-tt",;-,-ic.ti.n'li'ü*-Y"in"ill*"
alsc oc,:i;rs ¡:i:n a pcrtion of ihe Mn02 clinqínq;r-;h;-;;rju"iuri...']r'''l'30,.¡
rilm jnte*.ucven w'rh the pbOz firni. ii,u *uñsu;.;; j;;-il""i:ii u..r
prctscti've baryier tc h'¡'nder*th*.orroiion-ráiu * u.
rhjckens jn oí the;eac'ino¿"-
t"ine.to rhe pcint. r+he:e tli anc.* i¡n.:juctii,iiy h¡rr il ,...'
.?>n-{¡c +ec .r'^'igt e:^p :,;.-i.t:ti{.i-;-l u, < i,.-?-::) ¡Lfi:l¡rir¡llv
:, .-:.-^\.:
d;;i-;:Joj"
!h_:.
,n
'- -'"- in : --' +i.;^i,
aa,.li.l:rr 'J.

C" F:*aci'icils a¡ th: iatncre

T"e.J t:*:lljii

ln zirc ir'i;;:':i¡:c ihe caihl;ie nair¡-iar ¡E ex.rrus:,.:ri, al;ll;l;u;i ¿:-:i as

:.n",.1-_-^,1
i ,u / Lr'-=u .:c\,ü.
-L^.^
;:i€ ;:e :ül i:;"rs i pl-ar::i; \r: ür.c: ii;::"
gtierali¡ :i:ií; ;Ii .--G;,::,irai- p;i.cus, -i: u::^y inrr,ri:n_i i* tleTr,is filn'*¡hich .!s
d:::rs,,ii*r ijr*c*!s
!l ectncil¡e ra'l 1 L; rgy Cathode lvlater iats

for the iol lovr.ing reasons. if the film wera abseni, rr,e z.inc atcms wouli
becone tighily bonceC to the aiuminum atoirs anci ii lvculd ¡e'impossible io
separaie the zinc product from ihe cathode blank. 0n the other ha*d jf the
film'¡¡ere r^e'jatjvely thick anC impervious, the electrjc curreni could noi
be ccnductei through jt and a zinc deposjt would noi be obtajned ai all.
Boih sf these problerns can occur, for exa;r:pie, if fluoriCe 'ions are allovleC
to aliack the su:face the cxide would be conrpietely stripped off ani if a
catharie i{ere coní-recied as an anoCe by misrake, oxidaiion of the surface would
proceed until a thick imperv.ious film haC for ed"

In otner processes, other maier.ials have been used, for insiance jn

copp€r winning, blanks of iitairium, stainless steel or ccpper itself have been
successíu1iy empiolred. S'imjlariy in nickel vri:'rning, iiianium, sia'ínless sieel
and n'ickel meial sheets are pract j ca.l . T'!tani um and stai nl ess steel have al so
found appiicat'ion in manganese and cobalt solutions while alumjnum has been
suitable for cadmium. Pretreaiment invclving mechanical abrasjon cr chemical
elchÍng have become commcnplace in order io periodjcaliy restore ihe mosi
des'irabl e oxi de f i I m to the bl anks .

4 "5 .4 Anode l4ateri al s

As mentioned above, the Pb - 0.75% Ag alloy has been found to give the
best performance in zjnc cells. In contrast, leaci with either 0.06?á Ca or
6.0% Sb are most ccmmonly used jn copper cells. The aniimony a11oys are mos'u
popu'lar, predominantly because of the structural strength of the material.
The zinc plant anodes are often pretreated in a fluoride-sulphate electroiyte
to form a stable Pb02 film prior to usein the piating ce'I1s, while the copper
plani anodes are treated during the electrolys'is by cobalt ions added to
improve the anode fjlm characterist'ics.

4.5.5 Cell Design Cons'iderations

In order to avojd contamination problems resulting from electrode short
circr¡iting, electrode spacers are often employed which can be.in the form of
slotted boards jn the bottom of the cei1, piastic insulators affixed near the
bottor¡l of ihe anodes, or plast'ic'insulators mounted aiong the tcp of the cells.
These are shown jn the drawings of fjgures 4.5.1,4.5.2, and 4.5.3.

A sludge of lead sulphate, lead dioxide and manganese dioxide generaliy

forms jn the cells and cell cleaning is necessary every 30-60 days" The sludge
is vacuumed out in plants where the cells have flat bottoms, but can more
conven.iently be flushed out through bottom drains where sloped bottom cells
are employed. These sludges are aiways treated for recovery of lead and
mangancse and where appropriate for silver entrained 'in the sludge.

4.5.3
Figure 4.5..l Use of a Spacer Board Fixed to the Bottoms of the El ectrodes
to Provide Spacing.

Cathode Anode Cathode Anode

Solution

Spacer Board.

a-qú,
Figur-e ú,.5.2 iise cf ?las;ic Ins,-il:ioi-s lixe; ¡c ihe
t.^^)^-
,11¡uLr:: ic Space the il ec-i_'rodes.
Bci'iom cf Indi vi dual

Spacer
Button
Spacer
Clip

¿qc
F'igure 4.5.3 Use of Plastic Insulator on the Cell Top Bus Bar to provide
Electrode Spacing.

El ectrode
Header Bar

Cel I
Top Bus Bar
Plastic
Insulator

4tr,F.
 1.6 Reienences

'.,.1',,,,,. J'..'. Mag¡l?á' Cl cses the 11i ne io FJarkei Gap, fngineerf *g anci lnli ni ng Jor.,rna1 ,
April i972, pp. 73-33.
':
s ai Canadian Ccpper RefínerÍes l-ínii¡d, G- tsridgstcck,
trr anü 5. 5. Fcrb-e,-s, AIME Transactions, Vol. LXIii, 1960,
38, ,, r, ,.:',.'

khouse at Onah¿Ea Snelier" and Refinery, l{. iúal=

¡¡\suu :n-.1 Vt.
q¡¡v

,r [xiractive MetaI 1 urgy of Copper Vol. J, Chapier 30, pp. 588-

t'! i ur^gi cai Soci ety of AiME, N. Y.

tic Leaci 'Refining

wÍth
Refining
w'iih l-arge
l-árge Cel1 and Its Mechanization aat
Cell,,,ánC'l¿s: ,.-:l '..
gefjnery'j, [.'Nomürd, M. Arámski', i. Nishimura. ,i.#; ñr=i*nt.d=' : .'
nuáI'lleei.ing of the AIi'iE in Dailas, ig73 .:'. 1','',.:,': ,.:,,, ,"' ., ,-' ,1
eposit contr"oJ in Lead Elec'urorefining, c. J. Krauss, Journal
, November 1975 pp. 4-g

.'.....'.''....'-:..'.-.l'.u]...j.,...''..,..'.'.
ju i
¡*ni" ; r ;'''
llelin,.,iuly JglT,
,, ii;l'i*itilÍi*?ffi;¡T;ri]
pp."179-:186. .' i..-,-t
¡-'iir;it-'
.'ii,, ---- '
:.i. .,:., .,:.,
,l ,... ,, , :;

:,;.".-,:r.i:...rt.,- ;,::._-:li,:',.;:,.r,.;rl¡ -.'..:

.,''É*.i', '.'].:'..'.¡|.i'^^.

i*¡tf#r#;:,:ri,¡*,:;j;¡*;:i!i*¡
.,¡.'.¡i¡¡fi*$$j**; ?;'+i;{i ji:il,¡i;:**;á*r;i,#
 LEARI'iiNG ACTIVITY 6

metal..-to .be deposited, e.g.'i,Cu, Co, Ni, üd from zinc sulphate electro'iyte.
The, mbst coiffnon industri'al appl ications are summarized in Table 4.6.1 .. ',-,.,;:j .:.: r
'

..a).MetalRecovery:,.,.:.'']'.
.'i;.';;"-cápperbyIion(acrd.sulphatesolution)
.:
:.. "":" "J

4.6.1
 [] ectrorneta'l1u: gy
.;:,'
'
.':l':'
lln and LoDper by Indium from Indium tlectrolyte
Copper by Lead fron Lead Ref inery Electrolyte" -. ''1¡ ','
n Nickel, Cobalt, Antimony A.r^senic, etc. by Tinc from
o
t' L ¡ectro lyte t...
_
b
!,
F
!.: -
t: 'C-éméniation.¡áaitions ,cqn bE piedicted t,herrnodynamically on the !ái!s.,:.,
t!.:.
f,' of the eiectrode poteniials for the various half-reactions involved. The metal
*:tñ in" rnrr. poriiiuu iox'!cjai,ion) pot'entÍai, as given in the electri:motive: -
t
r

series (Table 4.6.2), will pass into solution and displace a meial ,w'!th a-less
positive potentiai.
'
i", ' ',S'b ráuiü
l;uiü 4.6:.? Potentialss at t5'"¡.,,',
standard üxidation Pátential
4.6".?' Standard es"r.
[";.voits r

A'l .' -rioor'. :*

'.,,, . :.-':,,.,:r,. ,.:
\.::).

L{t 0.76r '. ,

Fe 0.44 ,

Ld 0.40
. t' :'-.':,

Co 0.28'
NJ G. 25
(n n'!á
u. t$
n Tn
Pb U.I J
... tl
n7 ',q.00
Sb -0"t
. .0 .;a

.-+ As + ,¡.s,*_3 + 3e- -U. JU

thcugh cenen¿aii cn reacii ons can be pred'i ,._:.d . on ,*"he bas i s cf el ectrode
Ai
noieniials of ihe ccn:pon*ni haif-reactions" aciuat yjelds anci pl ccess efficiencies
u"u-J=t-rniie¡ by :'óte facicrs. lience a xncwlecge ái cemeniaiion reaction
tlinetics anC nechanisns is essantial fcr ihe ncst efficieni and economical
utilizaiion cf'-uhe prücess. A-r- ihe same -uine jt is jmporiant ts realize i,hat ' ','
cenenta¿ion react.'ion kineijcs can be conp'iicated by a number cf sicie reaciions,
includjng hyci: *iyser ancS other cxi,jario¡': ¡"eaciisns, and also by'ih* cemEni '

* ^^ -' 4
-.i^
ü=IJUS r L.

Á real:za;.icn cf :ne iÍ-,r:rúrtance ¿i reac:-icn k-inetics ii c:li:n:a:jon

¡^eseárch in ihis a]'*a in receni
r!'-r >i-iLicies h.','e bi^cugi: a'l:'i;i s:gr,ific¿n-" develc¡n:n:s'in :he
j-.,,..-'--i i,;c:l
r=di\,
.;,'ri:-- iiirl :i:,1 ---^irar,i:l t:*c:li¡ns
'^'L:^*^ ui^z LYiil:¡jLC.L¡uii i_=;-v:J:: such as, :hE i.e nrr accit,
f,l!.',c*;Jfi>
-^*-*:- Á-:^* ^**: fc;: :^cca;i,*l^inc c+ii3r^ frc;¡ dilul= leach liqr,,¡,;1 "

Lr. /
 4.5.2 Genera'l A.spec¡s.

Cementation i"eactions are redox react.ions 'invol ving netal "!ons, immei sed
metals and dissolved gases. These reactjons may be grcüped as follows:
I' Cementaijon reactjons in acjd solu'uicn (su"lohur-ic, perchlcrjc) wjth
^ no significan+- ccmpiexing of reacrani and/o.r prcduci ions.
2- Cementaiicn reactions in acid soluiion wiih sirong compiexing of
reaciant and/on product ions, €.g., by cNS or hali'ce ions.
3. Cementaiion react.ions jn basic solutions .in which reaciant and product
ions are strongly coilplexed to preveni hydroiysis, €.g., cNS-, bl¡- o"
dill i ile5.

The injtjal basjc cementatjon sludies for the mosi part involved reactions
from Group 1 because Lhe k jnetics anC mechan jsm are '!ess corlpl icated owing to
the absence of compi exi ng ef fec¿s . Reaci j ons 'in th'ís group rlray Ue subrii vi ded
acccrd"i!9 io thejr pos.ition in the elecircchemjcal seríes relaiive to tn* nilUr
couple (see Table 4.5.?):
l-A. Reactions occurring without interference by H+ reduciion, e.g.,
2Ag++Cuo*2Ag"+Cu++ [4.6. I ]
l-8. ,
Reaciions occunring wÍth sjmultaneous reduct ron of_ _-+
H', €.9.,
Cu+Z+Feo*Cuo+Fe+Z 14.6.2)
2YÍ +Fe"*H2 +Fe+2 [4.6. 3]
l-C. Reaciions occurring with diffjculty because H+ i s reduced by both
reactani and product metal, €.g.,
Co+z*lno*Co"+Zn+Z [4.6 .4]
2H+ +7n"-HZ +Zn+Z [+. o. si
2H+ +Coo*H2 *Zn+Z [+.o.oi
. andIn subgroup l*A, jn which the standard redox potentials.of the cementing
ion reagent metal are more negat'ive than that of the H2/H+ couple, the
amount of reagent d'issolved will be stoichiometric wiih thE amount of metal ion
cemented. Irr subgroup 1-8, jn whjch the redox potential of the cement'ing ion
is below that of the H2/H+ couple, but in whjch that of the reagent metaí is
above, th'amount of réagent metal dissolved wjll be'in excess ó¡ tfre sto-ichio-
metric re'lujremenis for cementation. The excess metal dissolved will vary with
solut'iofi;rH and the presence of other rgducíble species in solution, e.g.-,
dissolved oxygen or aaníome such as N03= or As04='. In subgroup l-C, tf.ré iedox
potentials of both the cementinQ-ion añd reagent metal are"above that of the
HZ/U+ couple. Hence the amount of metal disiolving wiil be considerably in
excess of the. stoichiometric requirements for cemeñtatjon.

Complete removal of metal ion from solution by cementation occurs only

with reactions ofsubgroups i-A and l-8. To achjeve this compiete removal

4.6.3
of ions, the reagent metal jmmersed'in the solution must be in excess of the
stoichiometric requ'irements. The amount of excess metal required ciepends on
the specific surface area available for reaction, the porosity and adherance
of the cement iayer, and the rate of H+ reduction. In subgroup l-C, complete
cement.ation is not l'ike1y to be achieved unless spec'ia1 additives are present
jn the solution because both the cemented and reagent metals can be oxjdized
by H+. An exception is the cementation of Cd++ Oy Zn powder which qoes 'largely-
to completjon most probably because both Zn and Cd metals exh'ibit rather high
hydrogen overvoltages which makes H+ reduction more difficult.

4 .6 .3 Cementaii on Reacti on Mechan'ism. r

In general the cementatjon reaction may be represented by the equation:

*,*', -1*, + F1i *|vr*az l+.a.t1

The reaction'is a heterogeneous one in which ions from solutjon are reduced to
zero valence at a soljd metallic surface. Microscop'ic observations of cementa-
tion reaciions show that the cemented metal is deposited onto the grow'ing surface
while the dissolving metai is ionized and enters the solutioñ from separáte
anodic areas. So, provided that the deposit is porous, the cementation reaction
can be considered as occurrllg via five steps, whjch are shown schematica'l1y in
fígure 4.6.1. Using the Cu¿*/Fe" system as an example, these steps are:
I. transport of CuZl to the i ron surface
2. adsorption of CuZ+
3. el ectron transfer
4. desorpiÍon of Fe2+
5. transport of Fez+ from the surface

Figure 4.6.1 Schematic Representation of the Cementation of Copper by Iron.

Dotted line is original iron surface.

i1 an
 Sieps I, 2 and 3 ccmpi ise the cathodjc :eac¿icn
^2+ + le * LU",
LU* [+. e. a]
anci sieps 3,4 and 5 the anocjc;-eacijon
Feo-FeZ++2e. i¿ a ql
if step I or 5 is reacijon rate controll ing t.hen rhe react'ion is ciiffus'ion
conti"oiled; if s'uep 2,3 oi 4 is the siow step ihen the reaciion .is said to
be chemical 1-v control I eC.

Ceneniaiion is essentiai'1y an elecirochemical pi ocess jn whjch the overall

reaction is the sum of numerous shcri-c'ircuited electrolytic m'!crocells. For
this reason the appiication of electrochemical parameter! is l'im'iied since the
anodic and cathcd'ic processes cannot be separaied unar¡biguously by acljusiment
of electrcde poieniials. in spite of thjs cjl'fficul iy ti is pois:'-nle, by means
of po]arizat.ic,n diagrams, to develop a criterion tc decicie wheiher or not a
given cementation reaction will be diffusion controlled. A detaileC discussicn
of thjs criterion is not wiihin the scope of this article but a useful approxi-
mation of thjs criterion'is ihat if the standard e'lectrode potential djfference
between the two half-reactjons, AE" > 0.36V, then the cemen!at'ion reaction would
be expected io be djffusion controlled. i-u turns out that most of ihe cementa-
tion reactions which have been siuciied are in fact diffusion-controlled and
have lEi > 0.36V. An exception is the Cu/N.i system whjch has been found to be
chemicaily-controlled; this "anomalous" behavjor observed for this system has
been attribuied to the formation of a passivat'ing film of nickel oxiie. The
PdlCu sysien and the PblFe reaction in strong bríne have aE! < 0.36V and have
been found to occur under chem.ical control.

Experimental criteria for distinguishing between possible rate-determining

steps in cementation react'ions will be discussed in the next sect'ion under the
kinetíc aspects of these react'ions.
4.6.4 Kinetic Aspects of Cementat.ion Reactions.
in the absence of comp'licating deposit effects, the cementation rate can
be descrjbed by a pseudo-first-order rate equation, viz.,
-¿[uz+]/ot = kA[Mz*J/u [4.6.10]
where [Nz*] js the concentration of the metal ion, A the area of the metal
surface, ang V the volume of solution. When A and V are expressed'in cmZ and
cmJ resep. iively, and the time t in sec., the rate constant k has the dimens'ions
cm/sec. rhe above rate equation is generaily valid for heterogeneous reactions
in solutir)n,, r€gdrdless of whether the rate control 'is by jon'ió transport across
a djffusion layer adjacent to the metal surface or by chám'ical reaction at the
surface or by a mixture of both.
Integration of equation 4.6..l0 yields:
v, 1og ([-Mt*]o / [¡,tt*]r) = kAt/2.303 [4.6.1 I ]

4.6.5
¡t
¡I

rl
:lIt

where Vr'is the volume of solution gt sampling time t*, [Mz*]t the cementing
I ion conóentration at t, IMz+]o the injtjal concentration of cdmenting ion, "
qld,8-the geometrjc area of tñe precipitant metal surface. A plot oi V¡ íog
([M*zlel[M+zJ¡) vs t'ime should be linear with a slope equai to kA/Z.303"if
I equalion 4.6.'l applies during a cementation experiment. Although this is
ofien the case, particularly for dilute solutions, cementation reactjons have
been studierj for which the V¿ 1og ([Mz+]o/[Mz+]t) vs time plots e.,hjbit both
positive and negalive deviatjons from l'inearity. Such deviations have been
f
attributed to the influence of the cement deposit on the reaction kinetics.
These deposjt effects will be commented on in more detail in a separate section
f; of this artjcle. Figure 4.6.2 shows, in schematic form, the various types of
i experimenial'1y observed V¿ 1og ([Mz+]o/[Nz+]t) vs time p1ots. The rató
constant k js obtajned fnom the relationship:
T
t k = b x 2.303/A 14 .6.121
where b is the slope of the V¿ 1og ([Mz+]ol[¡,tz+]t) vs tjme piot.
T
I
As indicateci jn figure 4.6.?, the simplest case 'is that illusirated by
curve type A, since the rate constant k does not vary over the duration of
c

* ihe experiment. The type B plot is typ'ica1 of a cementation reaction rate

which increases with tjme whereas the type C plot jndicates that the rate
t"
F
I
of cement,ation decreases with time.
¡
t
:
t It'is oíten possible to djstjnguish beiween d.iffusion-controlled and
chemicai'ly-controlled cementaiion reactions by studying (i) the temperature
."

I ciependence of the rate constant, k and (ij) the dependence of k on the

{ ¡
{ I
:|
roiatjonal speed of the precip"itant metal.
i
! i ) tffeci of Temperaiure on the Raie Constant
t
,!
¡Í
{ The apparent activation energy, ta, for a given cementatjon reaciion can
t
I be obtained from the iemperature dependence of the rate constant, k, by using
ihe fami I i ar Arrhenj us equati
É

i!
r¡ on
.*
j:
k = A exp{-talRT) [4.6. I 3]

* The ¿erm Vt accounts for the ciecrease in toial soluiion volune which
it resul is fron ihe removal of sanpl es for ana'iyses. Hov;ever, i f the
.-f
I ictal number of samples iaken js small ancj the samsle volume js also
snall :her,rthjs correc'uion may be iEnored, in wh'ich case the raie
equaiion will have ihe form;
lcg ([tlz-]o /itz"]t] = kA."'/2.3c3 v
rvhere \i'is ¡he ioial solL;t'ion vcluii:e. lhis same ra'ue equatjon has alsc
o:en wri tien as
I oE ([Mz+ jo / [r'{z+]r ) = kAt/2.3c3 vave

,,iIre:.e Vave = avei'age sOlr.¡iion volun:e.

tGc
 Fisure 4.6.2 or ixperimenially 0bserved '/¿ los ([Mz*]o/[Nt*].)
X:"1.;;"1::*=

C-J
CJ
o- {J
c-
F F
lco
IO
<lc4
*
vt -
l"'.'
i¡

-o
{J
co o-
o
CJ
o- th
F ^

úr
-t'
.oc(J
OJ

CJ
E
F

v. 1os ([Mt*]o/[Mz+]t) (.*3)

4.6.7
,É
l'

where k 'is the rate constant, A is a pre-exponenti al factor, Ea js the

apparent activatjon energy, R the gas constant, and T the absolute temperature
( " K) . I^Jri tj ng the equati on i n 1 ogari thmic form gives

iogk=-Ea/2.303RT+ iog A [4.6. r 4]

Thus if 1og k 'is p'lotted vs l/T, a stra'ight iine, should result, the s1 ope
of which is proportionai to Iu, thus
s'lope = Eu/?.303R [4.6.rs]
The actjvat'ion energy of the process is an important parameter for deter-
mining the rate-control'ling st,ep. If boundary layer diffusjon of the aqueous
spec.iás js rate-control'iing, [u is genera'lly.6 kcal/mole whereas adsorption
of the species on the reactjon-surface and subsequent chemical reaction and
desorpiion of products involves much larger Eu vaiues, usually t l0 kcal/mole.
Figure 4.6.3 shows, in schematic form, the various types of experimentally
observed Arrhenius plots for cementation reactions; these may be classified as
fol I ows :
Type A: Ea is generally between 2-6
'in
kcal/mole and the reaction rate is
cóntrol led by diffusion a I iquid phase.

Type B: An abrup'u change in E¿ is the temperaiure is increased. This

change usually occurs at 35-40"C. A"L T<35 C, Ea t l0 kcal/mcle
and at T>35oC, Ea .6 kcallmole.

Type Simjlar to Type B but 'uhere is a more gradual change in [u as

the temperature is jncreased ; in the transjtion reg'ion, there is
mi xecl cii ffusi on and chemi cal control of the reaci'ion rate.

Type [ntjre Arrhenjus plot js curved; mixed control of reaction rate.

D:

Type t: Arrhenius plot is'lYlinear but E¿ > l0 kcal/mole; the react'ion

rate i s ciremi ca1 conlrol I ed .

Mosi of rhe cemen-¿ati on reacti ons sludi ed to dat.e ex.hi bi t Type A Arrheni us
plots; i .e. , ihe rates are conirol led by a cli ffus jon p!^ocess. Several syst-ems
exhibji a rather large i¿ value at 1ow iempev'aiure but shjft to lower [¿ vaiues
as i,he isnperature ii inireased* (e.g., Cu cemenraiion on Ni discs).. 0n1y.
relativeiy few cemenia'lion reactions exhjbji chemical ccnirol over the eniire
iemperaiure i ange stucied.
ji) tffect ot Siirring on ihe Raie Consianl
The efíect, of tne rctaijonal specd of t.he spinning base metal on the raie
ccns'uani,, k, can also oe used ro derernine wneiher a cenentation reacrjon rare
* i.ypc B Árri''enius Pi ots found undei certai n ccrrdi i'io¡rs , für exanpl e i n copper
cenenta'¿i on on j ron and al urni num, have been at-uri bured ic the ef fect of
i ¿r¡¡: '-: + trc an I hP roughness of ¡ne copper deposii.
r

/co
i

l:
L.

i
Figure 4.5"3 Forms of Experimentall-v Observeci Arrhenius Plots.
i
l,
iiti
tll.

i¡--
ti

d.oJ
fa
OO
CDE
(\
.o
(g()
trl l¿
lr @
O)
CJ

o
¡t
U1

I
o

F
ce
O

1og k

4.6 .9
rp
á
F

'is diffusion or chemica'l1y controlled' As a genera'l ru1e,jsif the rate constant

need then lhe reaction diffusion controlled;
ii.r.utát w'ith i ncreased sti rriorngstirri' ng speed, then the reactjon rate is'likely
;;;'tl;;;;.;, k is indePendent
chemicai 1Y control I ed '

Two.of the most commonly used geometric.arrangements

for obtain"ing kinetjc
data on cementatión reaction! are tño*n in figure 4 '6'4' These arrangements and
provide two types-óf surfaces fot if.'á ienientjñg ions' (a) peripheral surface
have been used in the stud¡r
(b) c.ircular rrriuó..--Although'p..ipi érul suriuc"s
surface geonetry i'capable of
of many cementatión reaction., oñiy Lé.ircular
jnto deta-ir this analys'is leads to the
mathematicar ana.lys.is. without-going
equati on
c-
.-nF-=-<L aTlt ,r;! - [4.6. r 6]
'" co l '4 vul/b
diffusion of a species,to a p"rane rotating disc surface
wh.ich descr.ibes raminar D is 'uhe diffusion coefficient, v
(b jn figure 4.6:;'j'.'*rn-uquitión [4.6.rb], of the disc
the kjnematic vjscosity of tire lolution, i'th" angglar
velocjty and
it is seen t'ha'u 'if djffusion
A the jnjtiat ,rriá.. á."u. rrá* equatÍon.[4.6..i6]
is the only ratá controlting meiúanjsm, and v and D are constants' the rate
the square root of the rotat'ional speed'
constan-u, k, is directly proportiqnal to yield a siraight iine passing through
ü.). Therefore, a plot of k vs.,1/z-tno;ld
the ori gi n.
4.6.5 DePosi -r' Iffects
Ithasonlyrecen.u.lybeenrecognize-dthatthenatureofthecementmetal under certajn
cleposit contror! tire icjnéti.,..pánie of the cementation reaction
-pirticularly behavior
circumsrances. in-tñe case of copper cement.ation anomalous
consiant on the concen-
such as fhe depená"n.* of t.he firsi-orcien reaction-rate
jn aciivaijon energy with i,ncreasing
t.ration of cupric ions, the sudden-dÁcrease in reaction rates with ciurat'ion
i.emperaiure and it,u .núun..rr.ni-or retardatjon monpholog)'of the cement depos'it'
of the experi**ni have all beeñ related to the
Thus, t,he appar.*t increase in the firsi-orcjer rate conilant for most cementation
jnci"ease jn ihe rough-
is ncw Eenera-r1y accepied as b.;;; due-io a s'inurtaneous
nes s of 'uhe dePos i i .

1oy formati on ciuri ng cementai on reac-ui cn shoul ci al so

j
Ti^¡ nnc "; b j I i ty cf a'l
óan be deposiied on others at
uu ,".1[n#;¿:' It is known t'nut iome metals pr edicleo for deposjrjon of
pcientials ccnsiáerably 'iess .uinácic t'hanlcihose
the free energy of alloy format'ion'
rhe pure meiais, ine ci'ríferenc!-¡iing due
4.6.6inf-tuenceofimpurii.iesonCemeniationKjneiics
in an-y, cenentarion reaciion, side reaciions can occui jn jn
which reactant
ár praduceC. For exanpie, the cenentation
or produci species aie.on.ur,*á-uhe i red on j
á, Irpi., wi ¡n i r^on , re
ai tnough
cies ac'¡-i s

^..+?
LU - f r r¡o .,,. ,
i-C +- vu ¡u*Z

A num;er of oiher reaciions a-isÜ occur inciuding

 Figure 4-6.4 Common Gecme¡rjc Arrangemenis Used io 0btajn Raie Data for
Cemenlaiion Reeciions.

tiil
;rr¡
lllr ll
rl
T--f
|---l
---v*_-

LUCITE
CLAMPING
CYLINDIR

Z INC c.) PIRIPHIRAL SURFACI

CYLI NDER

STIRRING
SHAFT

b) CIRCULAR SURFACE

4.6. il-
 zFez+ + zH+ *br* zFe3+ + Hzo [4.6.18]
ZFe3++Cuo.-?FeZ++Cu?* [4.6. I e]
?Fe3+ +Feo.3Fe2+ 14.6.201
These reactjons jncrease the'iron consumed per unit of copper recovered. Sjde
reactions involving dissolved oxygen a'iways result in increased usdge of pre-
cipitant metal. [xtremely 1ow oxygen potent'ials can be maintainec by using
solutions purged iith high purity hydrogen; however, no measurable djfference
has been observed between the data obtained with hydrogen purging and ihat
obtained with nitrogen purging. 0n the other hand, the presence of oxygen
changes the reactjon stoichiometry and cjrasi'ica11y modjfies the deposii
L morphology. For example, copper deposits formed in the presence of even small
amounts of oxygen are smooiher and the rate enhancement caused by roughness
I
effects is substantially reduced. At high oxygen potentia'ls, the deposit
consists of basic salts and hydroxides.
Side reactions jnvolving either hydrogen or hydroxide ions are possibie
with all cementation reactjons so that scme control of pl-i appears desjrable.
I
I
S'ince jonjc transport depends on the electrjc field produced by a1l ions
t
I present, and sjnce the diffusivjtjes of the hydrogen and hydroxide ions are
nuch greaier than those of other ions, the simul'uaneous transpcrt of either
can strongly influence t.he mass flux of the cementa¿ion reactant caiion.
However, some surface process appears to be rate controlljng for surfaces such
as those present in cementation systems so that ihere is relaiively little
movement of hydrogen ions through the mass transfer boundar-v layer and hence
I iitle hydrogen ion effect on the reactant cat'ion díffusiv'ity. Notwjthstanding
this, with moderate to high hycirogen or hydroxicie ion concentraiion, theoreti-
caiiy ihere coulcj be a ieductjon jn the reactjon raie. in pract'ice, hcwever,
1i sone degree of acidit-v is ofien necessary to djssolve any protective oxide
¡i film that may be preseni on ihe precipitant metal surface.
r
¡
I Cementation reactions have been studiec using buffered solutions jn
itii or¡jer to controi ihe pH. However, since'r.he componenis jn a buffe:^ing mixture
can have a r¡ar ked effect on tne depos'ii morphol og.v and hence on ihe reactjon
k j ne-r-i cs , care must be iaken tc ensure that al I i ts effects are thoroughly
ev--,r¡':'nerl

linl jke iabora'uory solurjons: mosi leach f iquors being p)^ocessed in

cernen¡aiion piants have a vely conp'lex chemistry because these soluiions
contain h'ign ccncen'u¡^ar"ions of nonce:entable species, e.g., aluminum and
r-nagnesium. it is'iniel^est'ing to no¡e thar one of ine effecis of these
dissolved salts js io t'uffer ihe solu'uicn against pH changes;'in fact, the
buiierinE action increases in'ith an jncrease in sali conceniraiion. Data
obiajned us-ing acr-ua'l piani solui-i¡ns indicate ¿har fcr coiaper cementaiion
cn i rcn , ihe i ron consrnpi-ion Dj/ ac-i c "increases as i.he corrcenirati on of
d'Íssolvec salts ol^ buffering iricieases. ThE caia also sugEests tha'u ihe
üresgnce of alui-uj!iun and nagnes'ium reduces ine rare of cementaiion as well.
t,.6.7 0u'-line cf a T_vrical .a5ola:cry C:ren-aiion Stuciy
Ir: th js seciicn, a ¡r-ief acco,Ji,i is girren af iire experÍnenrai procedure

/. ; \.,
ani ti-re process'inE of the iata obrajne'd fron a iypical labrr-atcry cenen¡alion
study. The exanpie chosen is ba:ed on an actual s'uudy on ihe kjneiics of
coDper cemeniaticn on al urn'ini_¡i.:i.
.\
I ) [xperimenial

The experjmenial apparalus ccniprjsed a Luciie (or Teflan) cianoíng

cylinder assembly arranged such that only the peripherai surface or a pure
aluminum disc (see tigure 6.4.4a) was exposed io ihe solution and a 2-l'iter
covered reaction vessel. The cylincier was rot¿ied .in the experjmental solution
wjih a varjable-speed rnoior. The reactjon vessel was sei jn a constant-
temperature waier' baih jn which the temperaiure was con¿:^ol leci to wiih'in
t 0.05oC with a Haake "Un'itherm." The temperature y/as measured lv'iih an ASTI'1
precis'ion thermometer. Purjfied nitrogen was sparged through the solut'ion io
mainiain an oxygen*free environment.
Reagent-gracje chem.ical s anc reci'isi'il I ed deoxygenaied ,+iater irere useci 'in
ihe preparai jan of ihe experimen-ua'l anci s-"ock sbl ut jons. The coppe:^ sui phaie
siock solution was prepared from CuS01.5H20 crystals meet'ing ACS standards.
To follow the course of the ceme¡tatjon reaction, solution samples were taken
periodically and analyzed for Cu¿* with a Varian Techtron l,lodel AA-3 atomic
absorption spectrophotometer. The initial solution volume was I lÍter; the
samples 'uaken for analysis were 5 ml .

The alumjnum discs (surface area 23.9C cmz) were machined from pure
aluminum (99.914%.Ai ) obtained from Alcan (The Aluminum Company of Canada,
Kingston, Ontario). To obtain reproducible rates, the reacijon surface of
the Al discs was chemjcai'ly polished at 65"C for one minute in a solution
coniaining 700 m] H3P04,250 ml HzS0q and,50 ml HNOr. immediately afier
po'lishÍng the d'iscs, they were washed thoroughly in"a streanr of tap wa'uer
and then were spun for two minutes in redjstilled water. The discs were
transferred while still wet to the reactjon vessel and there they were spun
at a preselected speed for ten minutes. During this time, they were equ'i1i-
brated to the bath temperature, The testwas begun by introducing the required
amount of copper sulphate solution to the reactjon vessel.

Calculations of the react'ion rate, pcF unit of Al surface area, were

made with a programmable desk calculator (0livetti Programma l0'l ). The rate,
at any time, was calculated by evaiuating the fjrst derivative of the rate
equation. The jntegrated form of the rate equation is (see equat'ion [4.6.'l'l])
v,
1os ffcuz*f
o/[cu*Z]t) = kAt/2.303 lq.a.nl
The rate curves obtajned by plotting Vt log (lCr*2)o/[Cu+Z1t) vs.time were
fitted t,y regression analysis to an equation of the form:
y=a+bt
where a and b are constants, t is the reaction time, and y is a functjon
involv'ing the amount of material cemented,

4.6.13
 ji) Results

Typical re¡u1ts frgm these tests are presented in the following calculatjon
of V¿ 1og ([Cu+Z]o/[cu+z]t) an¿ rate constint k from experimental ¿áta.
Condi ti ons : Volume of solution = I liter which contains 5 x l0-3 MOu++ and
0.5 MH+; Temperature = 70"C.
Al uminum d j sc , area = 23 .90 cmZ, rotated ai periphera'l vel oci ty
= 500 cm/sec
Solution samp'le volume = 5 ml

Data*

Samol e Standard v¡ ioE ([cr**]o/lcu**lt

Samp"le Ti me Absorbance Conc. AbsoIEáncE-Conc.
(No. ) Min. (N x ro2) (M x lo4) cm3 )
I 0.636
.l.0
0 0.51 5 0 .618 0
2 1.0 0.631 0.51I ??7
3 5.0 4.627 0.406 0. 545 0.8 7.58
'I0.0 0.606
4 .480
0 30.10
5 I 5.0 0.595 0.465 0.428 0.6 4? L7
6 20.00 4.577 0.443 63.77
7 25 .00 0.565 0.430 0.31 7 4.4 75.99
ó 30.00 0.555 4.417 88.00
9 40.0 4.525 0. 385 0 .250 0.3 12A.66
t0 55 .0 0 .485 0.345 I 68.00
0. 165 4.2
l0-?'
-¡--¡-
Icu l-
lO- 0. 5.l5 x M

Slope of V¡ 1og ([cr**]o/[cu**]t) vs.t p'lot, b = 0.05 cm3/sec

ra'ue constant k=bx ?.3A3/ A

,r=-v^ 0.05 cm3 2.303

'
'' sec ^ 23.9 cn¿
k=0.48?xl0-2cm/sec

* Thjs data 'is plotied jn fjgure 4.6.5.

¿. 6 1L
Figure 4.G.5 plot of v¡ 1og 1¡cu+2JolIcu*2]r) vs. Time.

(J
vl
Í c.)
EE (J
cr)
ELo
(JO
aeo
ilil
-ttr
t
--1
I

I tn
c)
+J
I
=
c
I
E
I
CJ
I
E
F
I

OOO
o@(o
tCLta
v, 1 og ( [cu''Jo/[cu''].) (cm')
4.6.15
 lqlqulation of {rrhenius plots from Rate Data*
Cementation of Co r on Alcan Aluminum Cannin Sheet
Data l*
Conditions: [Cu+Z]o = I x 10-4 M; [H*]o = 0.5 M; peripheral velocity = 500 cm/sec
Surface area of Al strip = 23.48 cmZ
Test Temperature lO3/I kx 102
(No.) ("c) ('r:l ) icm/sác) loq k
I 80. 3 2.83
2 76.2
4.34 -l .363
2.88 3.32 -1 .479
) 60.7
J 2.99 2.45 -1 .612
4 55.7 3.04 2.?4 -1.650
5 50.0 3.t0 2.09 -1.681
6
1
45. 5 3.t4 1 .92 -1 .717
44.7 3.1B I .48 -1.830
o
o 34.7 3.?9 1.23 -1.gil
.|.25
Slope of Arrhenius piot =
E¿ = slope x 2.303 R (R = ? kcal/mole)

Ea = I .?5 x 4.6 = 4.8 kcal/mole

Data 2*
Condjtions, [Cu+Z]o = I x l0-3 M; [tt*]o = 0.5 M; per.ipheral velocity = 500 cm/sec
Surface area of Al sirip = 23.48 cm2
Tes t Temperature io3¡i k x ]02
No. 'c) ("K-l cmlsec loo k
1
I 79.5 2.83 7 "57 -l.tt6
74.7 ?.87 E72 -1 .243
J 74.2 2.91 3.26 -1 .487
+ 60. I 3.01 1 .37 -1.864
5 3 .09 0.87 -2.A59
+U. J 3.19 0.30
3t .3 ?9a
J.LU 0.08
??o -?.738
z1 .3 0.06 -3.245
S'iope of Ar^rhen'ius pl ot = 47

Eu = slope x 4.6 kcal/mole

= 4.7 x 4.6
= 21 .6 kcal /mol e
* Deta plotfed in figure 4.6"6.

4.6.16
 5i^,..-^
rtctiJre 4.b.b
4 . . \
Ari"nenius plois; taia from page 4.6.i6"

c.)
CJ
OJ

= C)

= Lr)
(\ f6
(J

ro
il (\

co
aa

C\O
cr)
F
C

cn

O)
c\J

T
(\

O O
NJ
I
I
cf)
I

1og k

4.6 .17
 Data for Dependence of Rate Constant on

t. Conditi ons : [Cu+21o i x l0-4 M; [H+]o = 0..¡ M; T = 70oC

=
Test k x .l02 peripheral Velocity (pV)
(No. ) (cm/sec) (cmlsec)-
I 0.591 125
2 1 .07 250
I .55 500
4 2.32 754
( 3. 35 I 000

+¿j
Condi ti ons : ICu Jo =5x lo-3 ¡,t; IH*]o = o.l M; T = Tooc

te5L k x lo2 Peri pheral Velocjty (pV)

No. (c sec c sec l
I 0.250 125
2 0.283 25A
)
J 0.253 500
4 a.?77 750

The data presented_in figures 4.6.5 and 4.6.7 indicate that the rate of
cÜpper cementation on alumjnum is,diffusion co¡trolled at
=-l ,-iO=4 ¡';
but_chemjcally controlled at [Cu++]o = 5 x l0-3 M. However,-ur'f,.nt.iol.á" Itl- [Cu++j^
earlier,-!his apparent rate dependeñce on initial .oppur ion .on.entration
is most f ikeTy caused by changes in the cieposit morpirblogy.
4.6.8 pl ant Methods and tqui pment.
Follcwjng is a brjef c'escription of some of ihe methocis and equ.ipment
ccnmonly used by cenientai'ion p'lants ; in particul ar, by coppÁr*ce,rentat jon pi
anis.
i ) Launders
There are narrow tanks about 500 feet'long,4-8 feet wide, and l-4 feet
ceep, -inclined at a slope of abaut 2%. The laüÁ¡ers ur*.i.rurgóa *ith scrip-
iron (e.g., ietinned, jron cans, scrap auion:cbile paris,;i;.Í-an¿ the copoer-
bearing soir.¡i.'ion. The sclution en-uers e'ii,her by gr"avity ai tÁe upper enij or
else is injetted into ihe launcers by nieans of noizles-- In a bat,ch process,
ihe copoe¡-bearing sc1u'uion is alIowá¿ to ílow on the scrap untii
ccnsured- The flol+ is ihen stopped and the launcjer fs draineJ-and ihe iron is
washed with
t';aier. The cemen'u ccpper is 'unen excavaieci by bi.ickets. in a ccnt jnuous proi"rr,
fresh'iron is cantjnuousTy added tc r^eplace tÉ¡ai consu¡¡red in t.he reacrion.
hcses are used ic agiiaie ihe charge, knock the cenent copper locse and forceFjre

* D¿ia pior-ied jn fjglre 4 .6.7 .

1.6.t8
 Fígure 4.6.7 P.lois cf k vs. Perjpheral Veioc"ii.y.
O
O
O

(-J
o
s- O
¡
oc) o
-:i
>-
xr\O
¡
O
lt
,<f x O
{ Lt)
+ rt) |-\
+ ¡t
=
(-)
o
U (J
A + OJ
T +
I
CJ = E
t{ L-l U
I *
\l +J
\\ al
o
c)
(f
o=
r.l) r e
CJ
-c
o-
s-
o_

O
Ll)
c!

O
Lr)
F\ O rr) a
C\¡
cn c\l

k x ]02 (cmlsec)

4.6. 19.
 jt along the launders to settling tanks where it is then collected. The
flow rate of solution through the launders is important and requ'ires some
control in order to prevent the formation of low-acid stagnant areas which
favor depos'ition of hydrolysis products and to mjnimize excessive jron
dissolutjon with proper operation of launders a copper recovery of about
97.5% can be achieved.

ii ) Rotatjng Drums

The Duisburger Kupferkütte in Germany carries out cementation reactions

in pear-shaped rotating drums which are charged from time-to-time with scrap
iron. Because of the rotation, the cement copper flows out with the exit
solutjon and is collected by filtratjon. This operation is shown schematically
in fjgure 4.6.8.

Figure 4.6.8 Cementation of Copper by Iron jn Rotating Drums (Duisburger

Kupferhütte, Germany) .

S croP
5l¿el she¡l
[
Eitumeñ insulolion

Cement

\
Co
c0PPer

pper - frcr
soiufron
--

__
I botfiJm
.-t

i1 I

\' I
I

Filtrolion unil
r-
I
I

I
I
iii) Inverted Cone Precjpiiatcrs
This apparatus was developed by Kennecoit Copper Gorporaiion and the
conpany has since replaced all its commercial launders with this iype of
¡ eacicr. in fact, ihere are iu¡o types, shown schenatical'ly in figures 4.6"9
and 4.6..l0, one of wh-ich js called a column prec'ip'itaior (figure 4.6.9); ihe
cone precipji.aior (figure 4.6..I0) js ihe more popular of che iwo.

Briefiy, ihe ilo iypes can be characierized as semi-cont"inuous reactors

4^b.¿u
 employinE high ílo-,v.raies, a h.igh deg:^ee of solut'!cn turbulance and shsri
solut.ion reteniicn times relaijie to-ihose found in typiiij-iur.¡er
Light guaEe scrap iron js adcec io the uniis as required piants.
copper is cjischarged ai intervals of 4-.l2 houi s. Élow and precipitate
is interrupted for a few minutes during each oiicirarge. or-p..r.nunt s,_-rlution

Fi gure 4.6.9 Kennecoit Col umn precj pí tator

Detinncd Sc rap Iren
n*'
/
/+^
E fflu e::r
Discha¡g.
f**:
r l* **
/.N

T r ough ll*--
--1
l^--*
t--.--
l+ *+
l-
lu -
ln** -
t^
ln**
l-
l-L -L
l--
l'* --
I

__4

Hcad Solution
Ente r s Frorn
Wate ¡ Jackc¡
He¿ d
Through Jets
Soluii or'¡

Durnp
Valve
Detin ne d
Scrap
I¡ on
Ef f iu ónt
Discharge
Trough

I
1
Fi gure 4.6. I 0 Kennecott Cone precipitator 9
Sc reen
Solr¡t i on
Cop pe r Injccti<.rrr
Sett Iing Sys[cnr
Zone

FIead
Solu ti on

Coppc.r
Di scha rge

4.6.21