-
A mineralogical study of nodulated copper cathodes
J.E..Dutrizac and T.T. Chen CANMET 555 Booth Street Ottawa,
Canada KIA OGI
ABSTRACT
Mineralogical studies were carried out on nodulated copper
cathodes from three primary refineries to characterize the nodular
growths and to elucidate the causes of the nodulation. Nodulation
is often initiated at the surface of the starter sheet or the
stainless steel plating blank, although a layer of smooth copper
sometimes is deposited before nodulation commences. In some
instances, the "roots" of the nodules exhibit a pronounced
dendritic texture that is associated with an abundance of cavities.
Slimes particles are not usually associated with these growth
features which lead to a globular surface deposit. The globules
sometimes develop into larger nodules, and this type of nodulation
is likely caused by improper addition agent concentrations. The
nodules on most cathodes, however, exhibit "roots" at the contact
with the substrate that are associated with microcavities and large
clusters (>40 pm) of slimes particles. The slimes constituents
are commonly Ag powder, PbSO, and Cu,(Se,Te) but not AgCu(Se,Te) or
Ag,(Se,Te). The size of the slimes clusters, rather than their
composition, appears to be the important factor causing the copper
to grow into nodules. Tiny individual slimes particles themselves
do not appear to cause cathode nodulation.
Proceedings of Copper 99-Cobre 99 International Conference
Volume III-Electrorefining and Electrowinning of Copper Edited
by J.E. Dutrizac J. Ji and V. Ramachandran The Minerals, Metals
& Materials Society, 1999
-
INTRODUCTION
The Inco Copper Cliff Copper Refinery has a capacity of 170,000
t!y, and plates onto conventional copper starter sheets (1). The
Kidd Metallurgical Division of Falconbridge Limited produces
145,000 t/y of copper by deposition onto stainless steel plating
blanks (2). Although the CCR Refinery of Noranda Inc. traditionally
plated onto copper starter sheets, the company has recently adopted
stainless steel plating blank technology for its entire production
of 360,000 t!y ofcathode (3). All three refineries normally produce
high purity copper having a good physical appearance. Occasionally,
however, varying degrees of cathode nodulation occur in all three
operations, and the nodulation can affect all parts of the cathode
deposit. The nodules sometimes grow to several centimeters in size
and the presence of such large surface features makes the handling
and stacking of the cathodes more difficult. The formation of
nodules fiequently traps quantities of electrolyte and slimes
particles, thereby reducing the purity of the copper product.
Furthermore, the nodulation also causes an uneven distribution of
the current density, a reduction in current efficiency, and hence,
an increase in the operating cost of the refinery.
Several mechanisms have been postulated to explain cathode
nodulation; these include insufficient mass transfer, improper
concentrations of the addition agents, and suspended conducting
particles. In an electrolyte fiee of additives, nodulation may be
caused by insufficient mass transfer (4); in this case, gas
sparging sometimes reduces the degree of nodulation. An improper
ratio of thiourea, glue and chloride in the electrolyte can also
produce nodulation, and it is known that the optimum concentrations
of the addition agents change as the current density increases
(5,6,7). Nodulation is also reported to be caused by suspended
conductive particles, such as anode slimes (8,9). Consequently,
those parameters which enhance the suspension of particulate
matter, such as increased electrolyte density or viscosity, the
evolution of gases at the anode and a high slimes fall, could
indirectly promote cathode nodulation. It is believed that once a
nodulated surface develops, the localized current density, and
hence the copper deposition rate, increases abruptly resulting in
the further rapid growth of the nodules.
Cathode nodulation is clearly undesirable, and all three.copper
refineries would. like to 'eliminate the nodulation problem. A
first step towards the. elimination of cathode nodulation is a
clear identification of its causes. To this end, CANMET recently
carried out detailed mineralogical investigations of nodulated
copper cathodes from the three refineries to elucidate the various
causes of the nodulation and to suggest possible means to resolve
the nodulation problem. The results .of those studies are
sumrnariied in this report.
-
ELECTROREFINING AND ELECTROWINNING OF COPPER
EXPERIMENTAL Samples
Two nodulated copper cathodes were supplied by Inco's Copper
Cliff Copper Refinery. The first sample (Inco-4) was a severely
nodulated cathode obtained after 5 days of plating at a current
density of 180 AM, and1 the second sample (Inco-2) was a similar
cathode collected after 6 days of plating. Nodulated areas from the
tops and mid-sections of both cathodes were chosen for study. To
provide complementary information, a sample of the suspended slimes
was collected from about 2 cm below the electrolyte surface prior
to the removal of the Inco- 1 sample from the cell.
Three cathode deposits from the Kidd Metallurgical Division of
Falconbridge Limited were studied. The first sample (Kidd- 1) was
obtained after 16 h of plating onto a stainless steel blank at a
current density of about 250 A/m2. Many 100-500 pm nodules were
dispersed over an otherwise smooth and fine-grained copper deposit.
The Kidd-2 sample was collected after 24 h of plating. This deposit
was generally smooth, but contained several 1-2 mm nodules
dispersed randomly over the surface. The Kidd-3 sample was obtained
by plating copper for 36 h onto a piece of a milled and polished
Kidd anode. The -0.9 mm thick cathode copper deposit was smooth and
fine grained; no nodulation was evident although several pin holes
-1 mm in diameter were detected in the deposit.
Four samples were provided by 'the CCR Refinery of Noranda Inc.
Two of the samples (CCR-1 and CCR-2) were starting sheets produced
using -22 h plating cycles and conventional copper plating blanks.
Both samples were extensively nodulated and multiple nodule growth
was common. The nodules on the CCR- 1 sample were 2-3 mm in
diameter and were often loosely attached to the copper matrix;
those on the CCR-2 sample were 1-2 mm in diameter and were firmly
adherent. The CCR-3 sample was another copper starter sheet made on
a copper plating blank; several sinall nodules were present near
the solution level, although most of the deposit was smooth. The
fourth sample (CCR-4) was a copper starter sheet made by plating
for -22 h at about 260 A/m2 onto a stainless steel blank. The
deposit was relatively uniform but contained numerous tiny
semi-globules and many pin holes dispersed over the entire
surface.
Mineralogical Techniques
Special efforts were necessary to prepare the nodulated cathodes
for study. After direct examination of the nodulated samples using
an optical stereomicroscope andlor the scanning electron microscope
(SEM) to select the areas for study, the samples were sawed to a
convenient size. The sawn pieces were mounted at right angles to
the deposit surface using liquid epoxy, and were ground and
polished to expose the contact zone between the nodule and the
underlying copper. The sections were examined using optical
microscopy and scanning electron microscopy with backscattered
electron (BSE) or secondary electron images to determine whether
any slimes particles or impurity phases were present at the "roots"
of the nodules or inside the nodules. The sections were then
lightly ground to
-
VOLUME 111
remove -100 pm of the surface and were repolished; these
sections were re-examined with the SEM. Repeated grinding and
polishing of the samples were carried out across the "roots" of the
nodules to increase the possibility of detecting any impurity
particles around the contact zones. When impurity particles were
detected, the sections were analyzed in detail using the scanning
electron microscope with energy dispersive X-ray analysis (EDX) and
the electron microprobe.
The suspended anode slimes sample, which was collected using a
suction device, was filtered, water washed and dried at room
temperature. The slimes were studied initially by X-ray diffraction
analysis to identify the major crystalline phases. Polished'
sections were prepared and ,the samples were examined with a
scanning electron, microscope equipped with an energy dispersive
X-ray analyzer to determine the chemical species and their
morphologies. In this regard, extensive use was made of
backscattered electron (BSE) images to differentiate the various
chemical species. Details of the mineralogical procedures have been
reported elsewhere (10).
RESULTS AND DISCUSSION
Suspended Slimesin the Inco Refinery . .
Figure 1 illustrates the morphologies of the suspended slimes in
the Inco refinery that are generally similar 'to those detected in
the Kidd and CCR operations. In general, the slimes occur as large
clusters or agglomerates which are composed of a diversity of
species. The tiny bright grains are Ag powder, the ring-like
particles are selenides, the triangular- shaped particle is an
octahedral crystal of NiO and the matrix is the oxidate phase,
(Cu,Ni)SO,.nH,O or Cu-Sn arsenate. Some of the bright particles are
PbSO,, but Cu20 is rare or entirely absent. Figure 2 shows another
view of the suspended slimes particles. The tiny bright grains are
mostly Ag powder, the spheroidal and ring-like particles'are
selenides, the platy hexagonal-shaped crystal is Cu-Sn-Ni oxide,
the matrix is mainly (Cu,Ni)SO,.nH,O, and the two largedark grains
are K-AI silicate. It is apparent from both photomicrographs that
Ag powder is an important constituent of the suspended slimes, and
the silver is believed to form mostly by the reaction of low
concentrations of silver ion with the abundant cuprous ions present
in the electrolyte (1 1,12).
Ag' + Cu' - Ago + Cu2' (1) Overall, the suspended solids contain
major amounts of Cul(Se,Te),(Cu,Ag),,(Se,Te), PbSO,, Ag powder and
NiO, as well as minor or trace amounts of AgCu(Se,Te), Cu-Sn-Ni
oxide, K-Ca-AI silicate, Al silicate, K-A1 silicate, Cu-Sn
arsenate, Ni-Fe-Sn oxide, BaSO,, Cu-Pb-As-Bi oxide and an oxidate
phase of Cu-Ag-Se-Pb-Ni-As0,-SO, composition. Significantly, the
Ag-rich selenides, which are a major constituent of; the bulk anode
slimes fiom the Inco refinery (139, are but a minor species in the
suspended slimes. The implication is that the suspended slimes were
recently generated and released from the anode surface. If the
suspended slimes had been in contact with the electrolyte for a
prolonged period of
-
ELECTROREFINING AND ELECTROWINNING OF COPPER
Fig. 1 - Morphology of the suspended Fig. 2 - Morphology of the
suspended slimes in the Inco electrolyte. 1,- slimes in the Inco
electrolyte. 1 - NiO, 2- Ag, 3- selenide, 4- Ag, 2- selenide, 3-
Cu-Sn-Ni oxidate phase (matrix), 5- oxide, 4- NiO, 5- PbS04.6-
(Cu,Ni)SOd.nHzO (Cu,Ni)S04.nH20, 6- K-A1
silicate
Fig. 3 - Secondary electron .micrograph Fig. 4 - BSE micrograph
showing the show,ing the general morpho- general' morphology of a
logy of the copper nodules in nodule in the Kidd-1 sample. the
Kidd-31 sample. 1- copper matrix, 2- plating
blank (removed$, 3- slimes particles
-
VOLUME I11
.time, as would 'bethe case if thezsuspended slimes originated
from the bottom of the refining cell, the selenides would ,have
been enriched in silver,. as, suggested by .the following
equations:
Cu,(Se,Te) + Ag' - AgCu(Se,Te) + Cu'
Ag' + AgCu(Se,Te) - Ag,(Se,Te) + Cut (3)
On the other:hand; the suspended slimes musthave been in contact
with the electrolyte long. ,enough for the.Cu,O .phase, which is a
minor species in the slimes attached to the ,face of the Inco
anodes (1 3), to dissolve according to the following reactions:
As will be discussed in detail later, the constituents of the
suspended slimes are similar to those found at the "roots" of many
of the cathode nodules. The Ag powder, Cu- rich selenides and PbSO,
particles are abundant in the suspended slimes, and'these are the
common constituents of the "roots" of the cathode nodules.
Slimes-Related Cath~de~Nodulation
The Kidd-1 sample, which was obtained after 16 h of plating,
represents the earliest stage of nodulation considered in this
investigation, and was studied to try to characterize the "root"
responsible for nodule growth. Figure 3 shows the typical
morphology of the nodules which developed on the fine-grained and
flat copper matrix. The nodules are 100- 200 pm in size, and seem
to occur randomly on the copper matrix. Multiple growths of the
nodules are common, and many of the nodules in this sample exhibit
a somewhat irregular form. Figure 4 shows a cross-section of one of
the nodules, and illustrates a common morphology. The nodule and
the copper matrix appear to exist as a single mass; no physical
boundary is discemable between them. A cluster of slimes particles,
however, is present at the base or "root" of the nodule. The slimes
particles, typically Ag powder, Cy(Se,Te) and PbSO,, are randomly
embedded in a compact copper matrix. Significantly, no cavities are
detected in the copper mass, although abundant cavities are often
associated with the nodules. The thickness of the copper deposit is
about 450 pm; the slimes particles are detected approximately 150
pm from the start of the copper deposit and they extend into the
nodule itself. It appears that the attachment of a cluster of
slimes particles to a pre-existing smooth copper deposit caused the
development of the nodule in this sample.
Figure 5 shows another common morphology of the nodules in the
Kidd- 1 sample.. A cluster of slimes particles, approximately 40pm
in size, occurs at the beginning of the copper deposit and near the
contact between the copper deposit. and the stainless steel;
plating blank. No other slimes particles aredetected in the nodule
or in the copper deposit adjacent
-
ELECTROREFINING AND E ,LECTROWINNING OF COPPER
Fig. 5 - BSE micrograph of another nodule in the Kidd- 1 sample.
1- copper matrix, 2- plating blank (removed), 3- slimes
particles
Fig. 6 - ;BSE micrograph illustrating the morphologies of the
slimes particles shown in Figure 5. 1- copper, 2- Cuz(Se,Te), 3-
PbS04, 4- Ag
Fig. 7 - Secondary electron micrograph showing the morphology of
the copper deposit of the ,Kidd-3 sample and the presence of holes
at the surface. 1- copper matrix, 2- 'hole, 3- groove, 4-
striation
Fig. 8 - BSE micrograph of the interface between the cathode
deposit, the anode copper.plating blank and a hole in the 'Kidd-3
sample. 1- hole, 2- cathode copper, 3- anode copper (plating blank,
4- Cu20+Cu2(Se,Te)+(Cu,Pb,As) oxide inclusions, 5-.interface.
-
VOLUME HI
to the nodule. The nodule and the copper deposit occur as a
single mass; no boundary can be discerned between them and no
cavities are present. It appears that the slimes particles occur at
the "root" of this nodule. Figure 6 shows the detailed morphology
of the slimes cluster illustrated in Figure 5. The slimes cluster,
approximately 40 pm in size, appears to be a single entity, with
PbSO, and Ag particles closely associated with the larger
Cu,(Se,Te) structure. The slimes particles seem to be embedded in a
compact copper matrix, which superficially resembles a fragment of
Kidd anode copper (14).
The Kidd-3 sample was generated by plating copper onto a
smoothly milled Kidd copper anode. This test was done to ascertain
whether individual slimes particles, on an otherwise flat copper
surface, caused cathode nodulation. In this experiment, the
deposited copper was consistently smooth and nodul'e-free, although
numerous pin holes were evident in the deposit. Figure 7
illustrates the morphology of two of these holes which are nearly
circular in cross-section; they seem to start as a tiny point and
gradually expand with prolonged plating thickness. Therhorizontal
striations.on the interior walls of the holes likely imply an
irregular release of gas during copper deposition. A groove
originating from the hole and pointing vertically upward (the
sample is shown 'upside down in Figure 7) is commonly present on
the surface of the copper deposit. The orientation of the groove
and the shell-like morphology of the hole with the
larger>dimension pointing upward imply that these are gas vent
holes. No impurity or slimes particles are detected in any of the
holes. Figure 8 shows the detailed morphology of a hole near the
surface of the plating blank. The lower-right portion of the
photomicrograph is the anode copper plating blank and the left
portion is the deposited copper. Clusters of
Cu,O-Cu,(Se,Te)-(Cu,Pb,As) oxide particles occur along the grain
boundaries of the anode copper. The pin hole developed from the
surface of the polished anode copper plating blank; that is, the
gas is released from the beginning of electrolysis and at the
contact between the anode copper starter sheet and the cathode
deposit. An impurity cluster in the anode copper is exposed at the
polished surface of the plating 'blank. Based on the morphology of
the copper deposited on the surfaces of the impurity clusters and
that on the surface of the inclusion-free metal, it is concluded
that the tiny individual slimes particles embedded in the surface
of the plating blank do not affect the morphology of the cathode
deposit and do not cause cathode nodulation.
Figure 9 provides a general view of the cross-section of a
copper nodule formed at the solution level of an Inco cathode
(Inco-1). Numerous slimes particles and tiny cavities are detected
at the "root" of the nodule, and these features are shown in
greater detail in Figures 10 and 11. Numerous tiny Ag grains are
embedded in the "root" of the nodule, and many (Cu,Ag),(Se,Te)
particles are present on the surfaces of the cavities which are
otherwise totally enveloped by thelmass of the nodule. Figure 12
illustrates the cross-section of another cathode nodule occuning at
the solution level of the Inco cathode (lnco-1). The "root" of this
nodule contains several tiny cavities, which are totally enveloped
by the mass of electrodeposited copper. Numerous slimes particles,
such as Ag powder, selenides, PbSO,, Cu-Sn-Ni oxide, NiO and
Cu,(Se,Te) particles with Cu-Pb-As-Bi oxide cores are present in
the cavities (Figure 13). The morphologies of these slimes
particles are very similar to those of the suspended slimes
collected from near the surface of the Inco
-
ELECTROWFINING AND EZ ,ECTROWINNING OF COPPER
Fig. 9 - Cross-section of a cathode nodule occurring at ,the
solution level of an Inco cathode (lnco-1). 1- nodule, 2- cathode,
3- impurity particles, 4- cavity
Fig. 10 - Detailed morphology of the "root" of the cathode
nodule shown in Figure 9. 1- cavity, 2- Ag, 3- Cu2(Se,Te) + FbS04,
4- copper
Fig. 1.1 - Detailed morphology of the cavity at the "root" of
the cathode nodule shown in Figure 9. 1- Cu2(Se,Te), 2- Ag, 3-
PbSO., or Cu-Pb-As-Bi oxide
Fig. 1'2 - Cross-section of a cathode nodule occurring at the
solution, level of the Inco cathode (Inco- 1). 1- nodule, 2-
cavity, 3- cathode
-
392 VOLUME III
Fig. 13 - Detailed motpho'logy of a cavity Fig. 14 -
Cross-sect.ion of a nodule in the cathode nodule shown in occurring
,at the middle of Figure 12. 1- Ag, 2- the Inco cathode (Inco-2).
1- Cu2(Se,Te), 3- Cu-Sn-Ni oxide nodule, 2- cathode, 3- slimes +
NiO, 4- ,PbS04. 5- copper particles + tiny cavities
Fig. 15 - Detailed morphology of the Fig. 16 - BSE micrograph of
a nodule at "root" of the nodule shown in the solution, level of
the CCR-3 Figure 14. 1 - NiO + Cu-Sn-Ni sample. 1- cavity, 2-
plating oxide, 2- Ag, 3- cavity, 4- blank, 3- slimes particles, 4-
copper, 5- porous copper copper dendrites
-
ELECTROREFINING AND ELECTROWINN1N.G OF COPPER
electrolyte (Figures 1 and 2). In fact, the association of
Cu-Pb-As-Bi oxide with Cu,(Se,Te), and not with AgCu(Se,Te) or
Ag,cSe,Te), implies that the "root" consists of a cluster of slimes
particles spalled relatively recently from the anode surface. In
the bulk anode slimes on the bottom of the refining cells, the
selenide species are significantly Ag-rich; e.g., AgCu(Se,Te),
(Cu,Ag),(Se,Te), or Ag,(Se,Te).
Figure 14 illustrates a cross-section of another cathode nodule
occurring at the middle portion of an Inco cathode (Inco-2). The
"root" of this nodule, which is shown in Figure 14, contains
several'cavities and many tiny impurity particles. The cavities are
totally enclosed by the nodule. Detailed examination of the "root"
revealed the presence of Ag powder, crystals of NiO associated with
Cu-Sn-Ni oxide and Cu,(Se,Te) particles with Cu- Pb-As-Bi oxide
cores embedded in the copper matrix, whereas the cavities appear to
be free of impurity particles (Figure 15). The presence of
Cu,(Se,Te) particles with Cu-Pb-As-Bi oxide cores in the copper
matrix and the association of many Ag particles suggests that the
slimes cluster, which' appears to serve as the nucleus for the
growth of the nodule, was freshly liberated fiom the anode
surface.
Figure 16 shows the tiny copper nodules present.at the solution
level of the CCR-3 sample. The nodules exhibit distinctivebanding
and radiating growth textures. Numerous cavities arepresent at the
"roots" of the nodules, and the "roots" occur at the contact
between the deposit .and the copper plating blank. The radiating
copper texture seems to have developed fiom the very inception of
electrodeposition. Slimesparticles such as Ag powder, Cu,(Se,Te),
AgCu(Se,Te), PbSO,, a Pb-Sb-Bi-Cu-S-0 phase, Cu,O and ,(Cu,Ni)SO,
are present in the cavities or are.embedded in the copper deposit.
Figure 17 illustrates a cluster . .. . , , . ,.
.* ..
ofslimes particles detected at the:"root" of one.ofthe nodules.
Silver powder and various .. ..., :
. .
selenides are abundant, but Cu,O, PbSO, Sn0,and (Cu,N,i)SO,.are
.also detected:: The . . . ,... ,.. . implication is thata
largecluster of floating slimes particles became attached to
the,top of .. . ..,": .?:
. . .-I: the copper starter sheet very early in the re'fining
cycle and caused the observed nodulation ._ ,. ..+> , .. . .
...
. ,..
The adhesion of "large" clusters of slimes particles on the
cathode surface results in enhanced localized copper deposition
that can lead to cathode nodulation. Large clusters of slimes
particles, suspended in the electrolyte or floating on the surface
of the electrolyte, can be transported to the cathode surface by
the electrolyte flow or by gas bubbles. Flloating slimes clusters
may lead to nodulation at the solution level, whereas suspended
slimes clusters may cause nodulation anywhere on the face of the
cathode . The cavities at the "roots" of the nodules may indicate
that the slimes clusters were transported by gas bubbles, or they
may indicate copper deposition around the porous slimes clusters.
There is little evidence that individual slimes species, such as
discrete particles of Cu-Ag selenide, isolated grains of Ag powder
or crystals of PbSO,, cause cathode nodulation. In fact, many such
slimes particles are present as individual grains, and even as tiny
agglomerated grains, on the surfaces of the cavities, but copper
nucleation and growth did not occur on such particles. Likewise,
the cross-sections of the nodules show no evidence of the
development of individual nodules or radiating copper features fiom
the individual slimes particles. Rather, a single nodule seems to
develop from the entire slimes cluster.
-
394 VOLUME IIJ
Fig. 17 - BSE micrograph of a slimes Fig. 18 - General
morphology of a cluster at the "root" of a cathode nodule on the
Kidd- nodule in the CCR-3 sample. 2 sample. 1- Ag in (Cu,Ni)S04, 2-
Ag in Cu20, . 3 - SnOz, 4- Cu2(Se,Te), 5- PbS04, 6- copper, 7-
cavity, 8- Ag
Fig. '$9 - BSE micrograph showing the Fig. 20 - Detailed
:morphology of the contact zone between a slimes particles shown
,in nodule and the initial c.opper Figure 19. 1- Cuz(Se,Te), 2-
deposit for the Kidd-2 Ag powder, 3- PbS04, 4- sample. 1- slimes
particles, 2- cavities cavity zone, 3- nodule, 4- initial copper
deposit, 5- stainless steel plating blank (removed)
-
ELECTROREFTNING AND ELECTROWINNING OF COPPER
The nodules obtained from different parts of the cathode or from
different cathodes exhibit the same general morphology and contain
the same impurity species. Often, the arrangement of the slimes
particles at the "roots" of the ilodules (e.g., Figures 10 and 55)
is superficially similar to that of the grain-boundary inclusions
in an uncorroded copper anode (cf. Figure 8). In fact, it was
initially postulated that sinall fragments of the copper anode were
transported to the cathode and contributed to the nodulation. This
belief was further strengthened by the abundance of silver powder
and Cul(Se,Te), rather than AgCu(Se,Te), particles at the "roots"
of the nodules; these species are prevalent in the anode and at its
surface, but are relatively rare in the bulk slimes. Subsequent
studies, however, did not support this hypothesis. Firstly, the
slimes particles in the "roots" are smaller than those in the
grain-boundary inclusions in the anodes, and the copper grains
which they appear to delineate are also smaller than the copper
crystals in the anode. Secondly, the slimes particles detected in
the "roots" of the nodules are generally devoid of Cu,O whereas
copper oxide is an abundant constituent of the grain-boundary
inclusions. Lastly, the electron microprobe- determined composition
of the copper metal, even at points between the individual slimes
particles, is not that of anode copper. Specifically, low Ni
contents (
-
VOLUME 111
close association of the slimes particles with the cavities, and
hence an earlier surface of the nodule, 'may imply that these
slimes particles were simply trapped on the surface of the nodule.
On the other hand, they may represent part of a larger cluster of
slimes particles which initiated the growth of the nodule.
Figure 21 illustrates the general' morphology of the CCR-I
sample, which is a starter sheet formedon a copper blank. The
surface of this deposit is extensively covered with 2-3 mm nodules
and many of the nodules consists of multiple growths. Most of the
nodules sit in "craters" formed in the starter sheet, and inany of
the nodules appear to be only loosely attached to the starter
sheet.
More than 70 nodules in the CCR- 1 sample were examined, and
many of them contain slimes particles at the contact zone between
the nodule and the copper matrix or between the nodule and the
plating blank. Figure 22 illustrates the typical morphology of
these nodules as seen in cross-section. Multiple growths of the
nodules are evident. The dark region with the sharp straight
boundary at the bottom of the figure represents the original
location of the copper plating blank from which the starter sheet
was stripped. The dark spaces between or within the nodules are
voids originally present during electrolysis, and such cavities are
common in this sample. The nodule in the center developed directly
on the surface of the copper plating blank. Slimes particles are
commonly present on the surfaces of the cavities, but are rare
within the nodules themselves. Theldetailed morphology of a cavity
which is located near the contact zone between the nodule and the
copper plating blank is shown in Figure 23. The dark regions are
cavities near the surface of the nodule; the tiny bright particles
are Ag powder, the bright grains are PbSO, and the ring-like
particle is Cu,(Se,Te). The slimes species are similar to those
identified in the Kidd and Inco cathode nodules. Based on the
mozphology, it appears that the slimes particles may have been
trapped on the surfaces of the nodules, especially in the cavities
between the surfaces of the nodules and the copper starter sheet.
On the other hand, the nodule may have developed from a large
slimes cluster part of which remained on the surface of the nodule.
There is no indication that any of the individual slimes particles
caused the nucleation and growth of the copper nodules, despite the
obvious abundance of such individual slimes particles on the
surface of the copper during the early stages of copper
deposition.
Non-Slimes Related' Nodulation
The CCR-2 sample is extensively nodulated. The surface of this
sample is covered by nodules which are 1-2 mrn in diameter, and are
similar in habit to those shown in Figure 21. In contrast to the
nodules on the CCR-1 sample, however, these nodules adhere firmly
to the starter sheet and no "craters" are evident. Figure 24 shows
the general morphology of this sample, and the multiple growth of
the nodules is evident. Nodules grow on top of other nodules,
creating spaces (between thein. In contrast to the CCR-1 sample,
where most of the nodules developed directly from the sudace of the
plating blank, nodulation in the CCR-2 sample began after a small
amount of smooth copper deposition had taken place. In most cases,
a layer of copper approximately 50- 100 pm thick was plated before
nodulation
-
ELECTROREFINING AND E IEECTROWINNf.NG OF COPPER
Fig. 21 - General morphology of the cathode nodules on the CCR-
1 sample.
Fig. 22 - .BSE micrograph showing the typical morphology of the
nodules in ihe CCR- 1 sample. 1- nodule, 2- copper plating blank
(removed), 3- open space, 4- slimes particles in cavity zone
Fig. 23 - Detailed morphology of the cavity zone shown in Figure
22. 1- copper, 2- cavity, 3- Ag powder,, 4- Cu2(Se,Te), 5-
PbS04
Fig. 24 - BSE micrograph showing the typical morphology of the
nodules in the CCR-2 sample. 1- nodule, 2- open space, 3- copper
plating blank (removed), 4- beginning of the nodular growth
-
occurred. Despite repeated grinding and polishing of this sample
to try to expose the "roots" - of the nodules on the copper matrix,
no impurity particles were detected in this sample. More than 60
such nodules were examined in detail, but no slimes particles were
detected in any of them.
The CCR-4 sample was plated onto a stainless steel blank. In
this instance, the copper surface is more or less smooth; it
exhibits a uniform globular morphology with -0.5 mm globules
covering the entire surface. Numerous~circular pin holes are
present (cf. Figure 7) and these originate both from the plating
blank and from the deposited copper itself. No inclusions or slimes
particles are detected in any of the holes. It is believed thatthe
pin holes represent vents which allow the escape of the gases
evolved at the cathode during commercial electrolysis.
Figure 25 illustrates the globular morphology of the nodules on
the surface of the CCR-4 sample. In the vicinity of the globule
(left half of photo) there are abundant cavities and an extensive
radiating-dendritic copper texture, which is characteristic of this
type of growth in the vicinity of the plating blank. Small copper
globules are present in the cavities and also in the nodulated
matrix. Particles of Fe-Cu-Cr-Ni sulphate and Cu-Ni sulphate are
detected in some of the cavities; the presence of Fe and Cr in the
sulphate phase implies the superficial leaching of the stainless
steel plating blank. The non-nodulated part of the deposit (right
side of the photomicrograph) is compact and more uniform; the
surface in contact with the plating blank is smooth. Figure 26
illustrates the morphology of the globule in greater detail, and
its radiating-dendritic texture is evident.
Figure 27 shows the surface of the deposit in the CCR-4 sample
which was in direct contact with the stainless steebplating blank.
Abundant tiny radiating sphemles are evident, and the implication
is that excessive copper nucleation occurred at the surface of the
stainless steel. The result is a globular-dendritic texture with
abundant cavities between the growing copper sphemles and
dendrites. Although trace amounts of Fe-Cu-Cr-Ni sulphate or Cu-Ni
sulphate are present in the cavities, significantly no anode slimes
particles such as Ag powder, PbSO, or Cu,(Se,Te), were detected in
this sample. Occasionally, particles of Cu,O were evident, and in
this regard, Figure 28 shows a semi-globular copper growth with a
strongly dendritic texture. Patches of Cu,O are present between
some of the copper dendrites, but the morphology of the Cu,O is
very different from that detected in either the CCR anodes or CCR
anode slimes 61 5). The conclusion is that the Cu,O formed directly
at the cathode, presumably by the reaction of electrochemically
produced cuprous ions with water.
The extensive radiating dendritic copper growth, the abundant
cavities and the heterogeneous copper growth morphologies at and
near the contact with the stainless steel
-
ELECTROREFINING AND ELEClXOWINNING OF COPPER
Fig. 25 - Cross-section of the copper Fig. 26 - deposit (CCR-4
sample) showing its semi-globular morphology. 1- cavity, 2-
radiating dendritic copper, 3- plating blank (removed), 4-
particles of Fe-Cu-Cr-Ni sulphate
Cross-section of the copper nodule (CCR-4 sample) showing
radiating dendrites. 1- cavity, 2- radiating dendrites, 3-
Fe-Cu-Cr-Ni sulphate, 4- plating blank (removed), 5- tiny copper
spherule
Fig. 27 - BSE micrograph showing the Fig. 28 - BSE micrograph of
Cu20 morphology of the copper particles in a semi-globular
originally in contact with the copper mass of the CCR-4 stainless
steel' plating blank of sample. 1- Cu20, 2- cavity, the CCR-4
sample. 1- .Fe-Cu- 3- Fe-Cu-Cr-Ni sulphate, 4- Cr-Ni sulphate, 2-
tiny copper copper dendrites, 5- plating grains, 3- radiating
copper blank (removed) spherules. 4- organic, 5- cavity
-
VOLUME TtI
plating blank indicate that the globular morphology commences
fiom the very beginning of electrolysis. Slimes particles were not
detected in this sample, and this observation conf im that the
uniform globular morphology which extensively covers most of the
surface of the copper deposit is likely caused by variations in the
electrolysis conditions, such as improper ratios of the addition
agents, locally high current densities, poor electrolyte
circulation or even the superficial corrosion of the stainless
steel plating blank.
CONCLUSIONS
Mineralogical studies, carried out on nodulated cathodes from
three primary copper refineries, suggested two causes of cathode
nodulation. The more prevalent is the attachment of large clusters
of slimes particles on the surface of the cathode; the second and
less common cause seems to be related to improper ratios of the
addition agents. Nodulation is often initiated at the surface of
the copper starter sheet or the stainless steel plating blank,
although less commonly a thin layer of smooth copper is /deposited
before nodulation commences.
The nodules on many of the cathodes show a "root" at the contact
with the substrate that is associated with micro cavities and large
clusters of slimes particles. In these instances, it is concluded
that the'nodulation is initiated by the slimes clusters. For
example, a deposit fiom the Kidd Metallurgical Division of
Falconbridge Limited, obtained after only 16 h of deposition
ontotstainless steel blanks in the commercial circuit, showed the
presence of clusters of Ag powder, Cu,(Se,Te) and PbSO, particles
at the roots of the nodules. Likewise, two nodulated cathode
samples fiom Inco's Copper Cliff Copper Refinery, after 5 or 6 days
of plating onto copper starter sheets, consistently exhibitedl
micro cavities and large slimes clusters at the roots of the
nodules. The slimes particles are Ag powder, NiO, Cu,(Se,Te),
'PbSO,, Cu,(Se,Te) with Cu-Pb-As-Bi cores and Cu-Sn-Ni oxide. One
of the CCR deposits obtained after 22 h of plating showed several'
nodules at the solution level which seem to have originated fiom
large clusters of slimes particles. The individual slimes particles
in the clusters were Ag powder, Cu,(Se,Te), PbSO,, Cu,O, (Cu,Ni)SO,
andl SnO,. Significantly, Ag-rich selenides and Cu,O, which are
abundant in the slimes h m the bottom of the refining cells, are
not detected at the "roots" of the nodules. In fact, the
compositions and morphologies of the slimes particles at the
"roots" of the nodules are similar to those of the slimes clusters
suspended in the electrolyte that were sampled at the time the
cathode deposits were obtained. The occurrence of micro cavities
and large clusters of slimes particles at the "roots" of the
nodules implies that this type of nodulation is caused by
nucleation and growth on the large clusters. That is, the size of
the slimes clusters, rather than their composition, appears to be
the more important factor leading to nodule growth. Tiny individual
slimes particles likely do not cause nodulation, and this
conclusion is supported by the observation that smooth deposits are
formed on milled anode copper plating blanks despite an abundance
of individual slimes particles at the surface of this material.
In other instances, however, the nodulation does not seem to be
caused by the presence of slimes particles or large clusters of
slimes particles. In' this regard, one of the
-
ELECTROREFINING AND ELECTROWINNING OF COPPER
starter sheets provided by the CCR Refinery of Noranda Inc.
contained abundant globule-like surface growths.and circular pin
holes dispersed across the surface. Although more than 60 globules
were examined, no slimes particles were detected, and no slimes
particles were found in the. tiny holes: In the regions of
the.globular morphology, the.copper surface in contact with the
stainless steel plating blank is rough, and. there amabundant
cavities. The texture of the copper is that of radiating dendrites
or micro spherules, and it is believed that the dendrites
initiated,the globular growths. Nodulation starts from the plating
blank, and no slimes particles are detected in the deposit,
effectively excluding the possibility of nodulation initiated by
suspended slimes. Another starter sheet provided by the CCR
Refinery was extensively covered by tiny nodules up to 3 mm ,in
size. Multiple nodule growths are common and the nodules adhere
firmly to the copper sheet. For this.sample, a small amount of
uniform copper deposition occurred prior to the development of the
nodulation. Although over 50 nodules were examined, no slimes
particles were detected, and the conclusion is that the nodulation
in this .sample is caused by variations in the electrolysis
conditions, such as improper ratios of the addition agents.
REFERENCES
1. J.M. Schloen and W.G. Davenport, "Electrolytic Copper
Refining - World Tankhouse Operating Data". C o ~ ~ e r '95 - Cobre
'95: Electrorefinine and -, W.C. Cooper, D.B. Dreisinger, J.E.
Dutrizac,H. Hein and G. Ugarte, eds., Canadian Institute of Mining,
Metallurgy and Petroleum, Montreal, Quebec, 1995,3-25.
2. J. Lenz, Personal communication. Kidd Metallurgical Division
of Falconbridge Limited, Tirnmins, Ontario, 1999.
3. 0. Pogacnik, M. Guilbert, H. Persson, Y. Fiset and C.
Belanger, "Permanent Cathode Modernization at the Noranda-CCR
Refinery". Copper '99 - Cobre '99: Electrorefining and
Electrowinning of Copper, this volume.
4. D.A. Uceda, J.D. Scott and T.J. O'Keefe, "Electrochemical
Evaluation of Gas Sparging Effects on Copper Deposition in the
Presence of Additives". Comer '91 - Cobre 91 : Hydrometallur!g and
Electrometallursg of Copper, W.C. Cooper, D.J. Kernp, G.E. Lagos
and K.G. Tan, eds., Pergamon Press, New York, New York
1991,303-327.
5. D.F. Suarez and F.A. Olsen, 'Wodulation of Copper Cathode by
Electrorefining Addition Agents Thiourea, Glue and Chloride Ion".
The Electrorefining and W i n n i n g , J.E. Hofhann, R.G.
Bautista, V.A. Ettel, V. Kudryk and R.J. Wesely, eds., Minerals,
Metals and Materials Society, Warrendale, Pennsylvania, 1987,
145-170.
-
C. DeMaere and R. Winand, "Study of the Influence of Additives
in Copper Electrorefining, Simulating Industrial Conditions".
Copper '95 - Cobre '95: Electrorefinin? and Hydrometallurw of
Copper, W.C. Cooper, D.B. ~Dreisinger, J.E. Dutrizac, 'H. Hein and
G. Ugarte,.eds., Canadian Institute of Mining, Metallurgy and
Petroleum Montreal, Quebec 1995,267-286.
0. Forsen, A. Kotzschmar and A.E. Antila, "The ,Role of Organic
Additives in Copper Electrodeposition". Galvanotechnik 86, 1,955,
3580-3590.
T.N. Andersen, C.H. Pitt and L.S. Livingston, '?Nodulation of
Electrodeposited Copper'due to Suspended Particles". J. Applied
Electrochemistry 13, 1983,429- 438:
R. Sh. Navasardyan, V.F. Korovin and 'B.N. Smimov, "Technical
Re-equipment of the Existing Electrolytic Copper Foil Industry".
Tsvetnve Metally 29, 1988, 32-33.
T.T. Chen and J.E. Dutrizac, "Practical Mineralogical Techniques
for the Characterization of Hydrometallurgical Products". Process
Mineralow a, W. Petruk, R.D. Hagni, S. Pignolet-Brandom and D.M.
Hausen, eds., Minerals,. Metals and Materials Society, Warrendale,
Pennsylvania, 1990, 289-309.
T.T. Chen and J.E. Dutrizac, "A Mineralogical Study of the
Deportment and Reaction. of Silver During Copper Electrorefining?'.
Metall. Trans. 20B, 1989, 345-36 1.
H. Bombach, K. Hein and D. Schab, "Studies on the Cu+ and 0,
Content of Copper Sulphate Electrolytes DuringlCopper Refining
Electrolysis". Errmetall 48, 1995, 703-7 1 1.
T.T. Chen and J.E. Dutrizac, "Mineralogical Characterization of
Anode Slimes - 11. Raw Anode Slimes fiom Inco's Copper Cliff Copper
Refinery". Can. Metall: Duarterlv 27, 1988, 97-105.
T.T. Chen and J.E. Dutrizac,"Mineralogical Characterization of
Anode Slimes from the Kidd Creek Copper Refinery". The
Electrorefininp and Winnine of Copper, J.E. H o h a n n , R.G.
Bautista, V.A. Ettel, V. KudrylLand R.J. Wesely, eds., Minerals,
Metals and Materials Society, Warrendale; Pennsylvania, 1987,
499-525.
-
ELECTROREFINING AND 'ELECTROWINNING OF COPPER
15. T;T. Chen and J.E. Dutrizac, "The Behaviour of Silver During
the Electrorefining of Copper". Precious Metals '89, M.C. Jha and
S.D. Hill, eds., Minerals, Metals and Materials Society,
Warrendalej Pennsylvania, ,1~988,377-390.