Minerals Engineering 18 (2005) 89102

This article is also available online at:

www.elsevier.com/locate/mineng

A study of gold anodic behavior in the presence of various ions

and sulde minerals in cyanide solution
a,* b
M.M. Aghamirian , W.T. Yen
a
CAMM, Queens University, 1027, Kingston, Ont., Canada K7L 3N6
b
Department of Mining, Queens University, Kingston, Ont., Canada K7L 3N6
Received 18 November 2003; accepted 11 May 2004

Abstract
Gold mostly occurs in nature in the metallic form and is usually associated with sulde minerals. In most cases, these minerals
have a retarding eect on gold leaching rates in cyanide solution. This negative eect could be attributed to dierent factors such as
galvanic eect or chemical and electrochemical interference.
In this investigation, the anodic dissolution of gold in pure and contaminated cyanide solution and in the presence of sulde
minerals was studied. The potentiodynamic method was used to study the anodic dissolution reaction. It was found that the anodic
reaction of pure gold was inhibited due to the passivation, which resulted from various ions and sulde minerals. A minor amount of
silver in gold disturbs the passivation mechanism and the reaction rate is controlled by mass-transfer mechanism. The results of this
investigation indicate that sulde ions have serious negative eects on gold anodic dissolution, accompanied by antimony ions.
Metal ions such as copper, iron and lead, did not have signicant negative or positive eect on gold anodic reaction. Among those
sulde minerals studied, stibnite had a very strong deleterious eect on gold anodic behavior. Gold anodic reaction in the low
overpotential range was almost uneected in the presence of galena. Under the same conditions, pyrite, pyrrhotite and chalcopyrite
decrease the activity of gold toward the anodic reaction.

2004 Elsevier Ltd. All rights reserved.

Keywords: Gold ores; Precious metal ores; Cyanidation; Leaching; Redox reactions

1. Introduction On the other hand, gold cathodic reaction involves

Over the last few decades, cyanide has remained the
main lixiviant for gold leaching. The advantages of
cyanide other lixiviants have been discussed elsewhere
(Fleming, 1992). As early as 1900, Christy suggested that
gold dissolution in cyanide solution is electrochemical in
nature (Mills, 1951). Thompson (1947) proved that the
gold cyanidation process proceeds via an electrochemi-
cal mechanism. As a result, an electrochemical approach
can be used to investigate the inuence of solution
compositions on gold leaching behavior. The gold an-
odic reaction proceeds according to the following reac-
tion:


Au 2CN
! AuCN2 e
*
Corresponding author. Tel.: +1-613-3849774; fax: +1-613-
5478125.
E-mail address: massoud@post.queensu.ca (M.M. Aghamirian).
oxygen reduction on the cathodic area. This reaction
may proceed through 2 or 4 electron transfer as follows:
O2 H2 O 2e ! HO

2 OH


2


O2 2H2 O 4e ! 4OH 3
Gold anodic behavior has been the subject of several
publications (Mills, 1951; Kudryk and Kellogg, 1954;
Cathro and Koch, 1964). Examining the EhpH dia-
gram for gold in cyanide solution, one will realize that a
stable gold complex, Au(CN)
2 , forms in a wide pH
range of alkaline solution. This means that gold is sol-
uble in cyanide solution; however, the kinetic of this
reaction can not be predicted by using such a diagram.
1 On the contrary, except in one case (Kudryk and Kel-
logg, 1954), it was reported that the anodic prole of
pure gold (99.99%) in cyanide solution indicated a pas-
sive region. Woodcock was one of the rst researchers
who realized that a pure gold electrode does not exhibit
a normal currentvoltage curve in cyanide solution

0892-6875/$ - see front matter
2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.mineng.2004.05.008
90 M.M. Aghamirian, W.T. Yen / Minerals Engineering 18 (2005) 89102
(Mills, 1951). He found a normal currentvoltage curve
with a limiting current proportional to cyanide con-
centration with a silver rotating electrode. However,
with a gold electrode the curve became complicated. The
curve showed normal behavior up to )500 mV, then the
current suddenly dropped. Using the potentiodynamic
method, Kirk et al. (1978) identied three dierent
maxima in the gold anodic curve at )650, 40 and 380
mV vs. SCE. This behavior was not due to the diusion
limitation of reactants or products (since in either case
the limiting current plateaus) and no current peaks
would be expected. By employing rapid sweep cyclic
voltammetry, he suggested that the adsorption of
hydroxide ions on the gold surface through the Reaction
4 was responsible for gold passivation before the last
peak (Kirk et al., 1980a,b):
Au OH
! AuOHads e 4
In his hypothesis, dierent crystal planes are susceptible
to passivation by the formation of the hydroxide layer at
dierent potentials that are responsible for three passive
zones in the anodic curve. Mills (1951) also suggested
that hydroxide ions might adsorb on the gold surface
and cause passivation. The adsorption of hydroxide ions
at potentials more cathodic than those of oxide layer
formation was also reported by Cordova et al. (1979).
However, Kirk et al. (1980a,b) reported the formation
of such a layer at much lower potential. The formation
of an oxide layer, Au2 O3 , is believed to passivate the
gold surface at the region of peak three, the most anodic
one. This mechanism of passivation at potentials higher
than the last peak is accepted almost by all researchers.
Applying cyclic voltammetry and galvanostatic
methods, Mac Arthur (1972) suggested that the reaction
at low overpotentials, up to )300 mV, would be con-
trolled chemically by the dissolution of the intermediate
product, Au(CN)ads . He believed that there was no true
passive region in his work when the gold potential was
less than 450 mV since an abrupt decrease in current did
not happen.
It is generally believed that gold dissolution in cya-
nide solution proceeds through several stages instead of
a single step, Eq. (1). McIntyre and Peck (1976) and
Kirk et al. (1978) proposed the following mechanism:


Au CN
! AuCNads 5 that the adsorption of thallium on the gold surface
AuCN

ads ! AuCNads e 6 would prevent the gold from becoming passivated. Nicol
AuCNads CN !

AuCN
7 (1980) also postulated that the addition of minor
2
Kirk proposed that the reaction mechanism is the same
for the rst and second anodic peak. Nicol (1980) pos-
tulated that since only one electron transfer is involved
in this reaction, it cannot explain the formation of more
than one peak. Kirk and Foulkes (1980) tried to remove
this doubt and suggested that the presence of various
peaks is due to dierent planes in the gold crystal and
each plane is active at a certain potential. Such an
argument without further clarication would be dicult
to believe. The fact that gold cyanidation proceeds
through an intermediate specious, Au(CN)ads , has also
been conrmed by other researchers (Sawaguchi et al.,
1995) using in situ scanning tunneling microscopy
(STM). This compound is known to have polymeric
structure with an innite linear chain. In conclusion, as
pointed out by Cathro and Koch (1964), the dissolution
of gold in aerated cyanide solutions would be controlled
by the diusion rate of dissolved oxygen if the surface
was unaected by the reaction products (active surface),
but would be controlled by the anodic dissolution rate
of gold if it were passive. The proceeding reactions can
be interfered with the silver ions in the cyanide solution
as discussed by Wadsworth and Zhu (2003).
Several researchers (Thompson, 1947; Kameda, 1949;
Cathro, 1964, 1965; Nicol, 1980; Deschenes et al., 1998;
Deschenes et al., 2000; Tshilombo and Sandenbergh,
2001; Wadsworth and Zhu, 2003) have examined the
eect of dierent ions, such as silver, lead, thallium,
mercury and sulde ions, on gold electrochemical
behavior in cyanide solution. Thompson (1947) reported
that an immediate white precipitate formed where lead
ions successively added to cyanide solution until nally
the gold electrode attained a potential positive to calo-
mel electrode. He suggested that this eect was due to
cyanide removal from the solution due to lead cyanide
formation. However, the formation of such a compound
and the mechanism for cyanide depletion is under
question. Kameda (1949) postulated that lead is more
electronegative to the gold in cyanide solution; there-
fore, lead ion deposition on gold surface would be
possible. Later, McIntyre and Peck (1976) and Adzic
et al. (1980) indicated that lead is able to form an ad-
sorbed monolayer on the gold surface even at a potential
which is more positive to its reduction potential. This is
referred to as underpotential deposition. As has been
pointed out by Omar et al. (1993), This phenomenon is
commonly considered as the manifestation of a stronger
bonding of the metal adatom to the foreign substrate
than to its own bulk phase. The formation of lead or
lead alloy on the gold surface has also been conrmed
by Deschenes et al. (2000) using XPS method.
Both Mills (1951) and Cathro (1964, 1965) reported
amounts of mercury or lead increased the height of the
anodic peak in the potential region )400 mV signi-
cantly. It appears from his work that the gold anodic
character in the presence of those ions, at more positive
potential, is essentially unaected. Since a passive area
still exists in such an anodic curve, depending on gold
cathodic behavior, it may not explain the pronounced
eect of lead ions on the gold dissolution rate.
 M.M. Aghamirian, W.T. Yen / Minerals Engineering 18 (2005) 89102
Anodic behavior of goldcopper and goldsilver al-
loys in cyanide have been studied by Guan and Han
(1994) and Sun et al. (1995) respectively. Guan et al.
reported that anodic polarization of dierent gold
copper alloys did not exhibit any passive region. Sun
et al. also did the same tests but with dierent gold
silver alloys in which silver content was at least 50%. No
passive area was observed in this case either.
Since gold dissolution is electrochemical in nature,
this study was undertaken so as to lead to a better
understanding of anodic behavior of pure gold or gold
with minor amounts of silver either in pure cyanide
solution or in the presence of dierent ions and sulde
minerals. Such an investigation is important since it
contributes to our understanding of the reaction mech-
anism of dierent ions or sulde minerals on gold dis-
solution in cyanide solution. In fact, foreign ions or
sulde minerals have been observed to either retard or
enhance gold dissolution rate. Consequently, they
should have a direct eect on gold anodic dissolution
reaction or oxygen cathodic reduction or both. The
objective of this paper is to investigate the eect of these
impurities on gold anodic reaction.
2. Experimental details
Gold anodic behavior was studied by using the
rotating disc technique method. In this method the disc
surface area is constant and mass transfer from this
geometry has been well established. Two dierent gold
discs were used. The rst one was cut from gold foil
having the purity of 99.99% and was supplied by the
Aldrich chemical company. The surface area of this disc
was 1.10 cm2 . The second disc was a gold silver alloy
which was made by melting a mixture of pure gold
(99.99%) and silver (99.99%) foil at 1120 C and an-
nealed at 700 C for 2 h. The silver content of this disc
measured by atomic adsorption spectrophotometry was
0.31%. The area of this disc was 1.05 cm2 . Before each
experiment, the gold discs were wet polished with 600
grit silicon carbide paper and then with alumina pow-
ders (3 and 0.05 lm). After that, the gold discs were
washed with chromic acid and then nally rinsed with
distilled water. These discs were immersed in the cyanide
solutions, 2.6 cm below the solution surface.
The gold discs were embedded in a resin, which was
attached to a teon rod. The teon rod was threaded on
to the main steel shaft. The wire, which was soldered to
the gold electrode, was strongly attached to the steel
shaft by using a clamp. A carbon brush was used to
measure the current and potential when the working
electrode was rotating. Total resistance of this circuit
was measured to be 0.25 X.
All the solutions were prepared from analytical grade
reagents and distilled water. Sulde minerals such as
91
pyrite, pyrrhotite, chalcopyrite, galena and stibnite, as
well as quartz were supplied from Wards Natural Sci-
ence. These samples were crushed and sorted to remove
all impure particles. A size fraction of )53 +38 lm was
selected for all experiments. These samples were stored
in a freezer to avoid further oxidation. Chemical anal-
ysis was carried out using an atomic absorption spec-
trophotometer to conrm the purity of the mineral
samples. Except for pyrrhotite, all mineral samples were
about 95% pure, with the remaining 5% being mainly
quartz. In the case of pyrrhotite, about 3.43% of pent-
landite was nely disseminated in the pyrrhotite matrix.
Each sulde mineral was tested individually. A known
amount of each sample was added to the cyanide solu-
tion and suspended by agitation.
A CMS 100 corrosion measurement system supplied
from Gamry Instruments, along with CMS 105 DC
corrosion testing software, were used for electrochem-
ical measurements. CMS 100 hardware was a PC3
potentiostat. CMS 105 DC software made it possible to
control the potentiostat, doing dierent DC experi-
ments and recording the data directly into the com-
puter.
A three cell electrode made from plexiglas was used to
perform electrochemical tests. Potentials of the working
electrode were measured in respect to a calomel refer-
ence electrode. A luggin capillary tube housing the ref-
erence electrode was placed within approximately 2 mm
of the working electrode. Platinum was used as an
auxiliary electrode.
All experiments were carried out at 24.3 C by
immersing the cell in a water bath. Before each test, the
cyanide solution was deoxygenated with high purity
nitrogen gas (99.9995% N2 and O2 < 1 ppm) for one
hour followed by immersing the electrode in the solu-
tion. Open circuit potential of the gold disc was mea-
sured within the designated time. It was found that in
less than 3 h, the open circuit potential became almost
constant. Consequently, the sweep voltammetry test was
started after 3 h conditioning. Nitrogen bubbling
through the solution was continued up to the end of the
experiments.
Scan rate was maintained at 0.2 mV/s to make sure
that the systems reached its equilibrium at any potential.
Unless otherwise stated, all the potentials are reported
against a SCE reference electrode. Sodium cyanide
concentration was maintained at 0.01 M (0.5g/l) and pH
10.5. The gold electrode was rotated at 500 rpm unless
otherwise stated. These are the standard conditions
which were followed through out all of the experiments.
Unless otherwise stated, gold disc no. 2 was used all
through this investigation.
The concentrations of the ions in cyanide solution in
this study were adjusted at the levels corresponding to
the concentrations of the ions released from a mild
refractory ore into the cyanide leach solution.
92
3. Results and discussions
3.1. Gold anodic behavior in pure cyanide solution
From direct dissolution tests with dierent gold discs
containing varying amounts of silver, it was observed that
the gold dissolution rate increased with the silver content
of the gold disc up to a level of 0.5% (Liu, 1993). Beyond
this level, the gold dissolution rate started to decrease. For
this reason, gold discs having a low silver content were
used to conduct this study. Linear sweep voltammograms
for the pure gold disc, sample 1, and the gold disc con-
taining 0.31% silver, sample 2, in 0.01 M sodium cyanide
solution at pH 10.5 are presented in Fig. 1. The voltam-
mograms for a pure gold electrode in stationary condition
vs. being rotated at 500 rpm did not show any signicant
dierence. It has been evidenced that the surface processes
are controlling the dissolution reactions mentioned
above. It is generally observed that the rotation of the
electrode increases the current density and enhances the
diusion of cyanide ions. There was not a signicant
dierence observed between the stationary or rotating
electrode for the pure gold disc. Both curves show that
pure gold was passivated in cyanide solution and the
passive layer was stable even at high overpotentials. This
passive layer changed the gold surface color to faintly
brown. These curves are similar to the one presented by
Cathro and Koch (1964) at low cyanide concentration.
As it was stated earlier, either hydroxide ion
adsorption or AuCN monolayer formation is responsi-
ble for passivation of the gold surface. It has also been
suggested that gold passivation at potentials greater
than 700 mV with respect to SHE, could be explained by
oxygen adsorption on the electrode surface (Woods,
1976; Dorin and Woods, 1991). This potential may be
aected by the presence of impurities in the system or on
the electrode surface resulting in oxygen adsorption at
lower potentials. In the present work, such a reaction at
0.2
(b)
0
Potential (V) vs Eref
-0.2
(a)
-0.4
-0.6
-0.8
0 0.0005
Fig. 1. Anodic polarization plots of sample one and two for various rotating speeds: (a) sample one, 0 rpm, (b) sample one, 500 rpm, (c) sample two,
0 rpm, (cs ) simulated data of test c, (d) sample two, common conditions: 500 rpm, pH 10.5, 24 C, cyanide conc. 0.01 M, scan rate 0.2 mV/s.
M.M. Aghamirian, W.T. Yen / Minerals Engineering 18 (2005) 89102
those tested potentials is most probably not responsible
for gold passivation. The Tafel slopes at the active area
of curves (a) and (b) are 0.067 and 68 mV/decade
respectively. These values are close to 59 mV/decade,
which has been postulated to be the evidence that
Reaction 7 is the rate-determining step in the gold dis-
solution process (Kirk et al., 1980a,b).
The low activity of pure gold toward the anodic
reaction has also been reported by Mussatii et al. (1997).
In their investigation, they did not detect any anodic
current when the gold potential was less than )110 mV
vs. SCE. They believe that the anodic current detected
by other researchers at lower potential could be due to
trace heavy-metal impurities.
Curve (c) shows typical currentvoltage behavior for
the unpassivated gold electrode. This curve clearly
indicates that changing the size of the gold crystals and
the presence of minor amounts of silver in the gold
electrode had a remarkable eect on stopping the pas-
sivation phenomena. To prove that this eect was not
simply due to the selective dissolution of silver from the
gold electrode, silver and gold assays were measured.
The gold and silver were assayed as 8.02 and 0.06 mg/l
respectively. The results show that silver has not been
dissolved selectively and the electron transfer was due to
gold dissolution behavior. The shape of this curve is
very similar to those obtained by Kudryk and Kellogg
(1954) who demonstrated that the gold dissolution rate
at a high potential range is controlled by cyanide diu-
sion to the electrode surface. The Tafel slope at the ac-
tive area of this curve is equal to 122 mV/decade which
indicates that Reaction 6 is the rate-determining step
(Kirk et al., 1980a,b). The shape of the curve may sug-
gest that cyanide adsorption on the gold surface follows
a Langmuir adsorption isotherm. Further investigation
is required to conrm this conclusion.
Curve (d) shows that rotation speed has a pro-
nounced eect on the gold anodic behavior of this
(cs) (d)
(c)
0.001 0.0015 0.002 0.0025
Current Density (A/cm2)
 M.M. Aghamirian, W.T. Yen / Minerals Engineering 18 (2005) 89102 93

sample. This remarkable increase of current densities at Reactions 6 and 7 appear to become faster. The cover-
the same potentials as compared with curve (c) is an ages denoted by h1 and h2 become smaller and the total
important sign that the reaction rate is mass transfer surface area available to react with cyanide ions be-
controlled. The similarity between curves (a) and (b) and comes larger. In addition, cyanide ions in competition
the lower activity of the gold anodic reaction under with hydroxide ions may be adsorbed faster on gold
these conditions indicates that the reaction is controlled surfaces at higher overpotentials in comparison to lower
chemically. This observation suggests that passivation is overpotentials. Both mechanisms may account for the
probably due to AuCN monolayer formation on the sharp increase of the current after passing the minimum.
gold surface, since the pH in all those tests was constant. The ButlerVolmer equation can be used to study the
However, the adsorption of hydroxide ions on the gold charge transfer kinetics of the gold anodic reaction in
silver alloy could be dierent from that of the pure gold, cyanide solution as shown below:
which could cause the observed dierences. If the for-   
Cs anF
mation of the AuCN monolayer is responsible for pas- ia i0   exp g
Cb RT
sivation, the presence of silver atoms in the goldsilver  
Cos 1
a  n  F
alloy results in a reduced gold dissolution activation
 exp
g 10
energy with respect to the pure gold system. This leads Cob RT
directly to facilitation of the formation of gold-cyanide where ia is the current density, i0 the exchange current
complexes from the intermediate product, AuCN, and density, a the symmetric coecient, n the number of
inhibits passivation. On the other hand, the formation electrons involved in the gold dissolution which is equal
of silver cyanide complexes in the solution can be con- to 1, g the overpotential, F the Faraday constant, R the
sidered as an explanation for enhancement in gold an- gas constant, T the absolute temperature. Cs , and Cb are
odic behavior and elimination of passive layer. This was the cyanide concentrations on the gold surface and in
conrmed by Wadsworth (2000) who suggested that in solution respectively, Cos and Cob the oxygen concen-
the presence of sliver cyanide complexes in the solution, trations on the gold surface and in bulk solution
a tight passive surface lm composed of AuCN may not respectively.
be formed on gold surface. In the presence of Ag(I), the Since the extent of backward reaction at high positive
passive layer is more permeable due to the inclusion of overpotentials is negligible, the second term of Eq. (10)
AgCN in it, which increases the lm diusivity. can be omitted. If a is 0.5 at ambient temperature, an
Curve (d) also shows that the anodic current density overpotential of 118 mV is large enough to make the
rises very steeply with increasing potential followed by error of neglecting the second term less than 1%.
the occurrence of a maximum and a minimum. After the Ficks rst law is:
minimum point, the current density increases continu-
n  F  D  Cb
Cs
ously with increasing potential. The extent of bare gold ia 11
zd
surface has an important impact on gold dissolution
according to Reaction 5. The surface coverage of the where z is the stoichiometric number of cyanide ions
adsorbed intermediate such as AuCN
which is equal to )2 and d is the diusion layer thick-
ads and AuCNads
can be indicated by h1 and h2 , respectively. A third ness. From this equation, it can be concluded that the
factor in the reactions is the coverage of hydroxide ions, maximum possible current density is equal to:
h3 . The total bare surface, ht , available for CN
n  F  D  Cb
ilim 12
adsorption can be written as: d
ht 1
h1
h2
h3 8 Combining Eqs. (10)(12), and ignoring the backward
reaction, the following equation is obtained:
The potential of zero charge of gold in alkaline solution  
RT
is known to be 0 V vs. SHE (Kirk et al., 1978). As the i0  ilim  exp g
aF
potential falls below the potentials corresponding to the ia   13
RT
maximum current of curve (d), the anions (e.g. 1 i0  exp g
hydroxide ions), have greater diculty adsorbing onto aF
the gold surface. Thus Eq. (8) can be simplied as fol- This equation has also been used by Li and Wadsworth
lows: (1993) for modeling the anodic electrochemical data of a
ht 1
h1
h 2 9 silver electrode. This model can be used to t the
experimental data, curve (c) of Fig. 1, if a, i0 , and ilim are
If instead the potential is increased, hydroxide ion predetermined. The a value is around 0.5. From Eq.
adsorption on the gold surface is facilitated, which (13), the overpotential can be calculated by
competes with the cyanide ion reaction. For this reason  
the current decreases with increasing potential. Above a RT ia  ilim
g  Ln 14
specic potential, in this case )0.14 vs. SCE, the rates of aF ilim  i0
ia  i0
94
This equation can be used to calculate i0 . An alternative
way is to use the Tafel equation with the assumption
that the concentrations of the reactants on the gold
surface and in solution are almost equal. Such an
assumption is only true when the anodic current is not
more than 10% of the limiting current.
RT RT
g  Lni0  Lnia
anF anF
By using Eq. (15), the apparent exchange current density
at 24 C is estimated to be 6.463 10
1 lA/cm2 .
Limiting current density can be estimated from the
plateau part of the curve (c) in Fig. 1 which was found to
be 485 lA/cm2 . Substituting this value into Eq. (12) and
using the proposed value for the diusion constant of
cyanide ions (Thompson, 1947), D 1:75  10
5 cm2 /s,
the diusion layer thickness, can be estimated. The
thicknesses of this layer was estimated to be 0.348 mm
which is in the range of diusion layer thickness for an
unstirred solution (Brockris and Drazic, 1972). Esti-
mating the value of i0 and ilim , we are now able to use
Eq. (13) to t curve (cs ), to the experimental curve (c).
Comparing curves (c) and (cs ) leads us to the conclusion
that the calculated current density and the experimental
results agree well with each other. This conrms that the
dissolution reaction was controlled primarily by electron
transfer kinetics, and to a lesser extent by cyanide dif-
fusion. The number of electrons transferred, as ex-
pected, was equal to one and the symmetric coecient
was very close to 0.5.
The eect of cyanide concentration on the gold an-
odic polarization prole at stationary condition is indi-
cated in Fig. 2. From the results, it can be seen that the
maximum in the dissolution of gold at high overpoten-
tial is a function of cyanide diusion. Once again, under
these circumstances there is no indication of passivation
0.100
(d)
-0.100
Potential (V) vs Eref
-0.300
-0.500
-0.700
-0.900
0.0000
Fig. 2. Eect of cyanide concentration on gold anodic polarization plots: (a) 0.05 M, (b) 0.025 M, (c) 0.01 M, (d) 0.005 M, common conditions:
0 rpm, pH 10.5, 24 C, scan rate 0.2 mV/s.
M.M. Aghamirian, W.T. Yen / Minerals Engineering 18 (2005) 89102
in the gold anodic behavior. When the cyanide con-
centrations were adjusted at 0.05, 0.025, 0.01 and 0.005
M, the limiting current was about 280, 150, 50, and 23
lA/cm2 respectively. From Eq. (12), it can be concluded
that there is a linear relationship between limiting cur-
rents and cyanide concentrations if cyanide diusion
controls the reaction. Fig. 3 shows a linear relationship
15 between limiting current density and cyanide concen-
tration. The square of the regression correlation coe-
cient was 0.996, strong evidence for the linear relation
between cyanide concentrations and limiting currents.
The results also demonstrate that the applied value of
cyanide diusion coecient, D 1:75  10
2 cm2 /s, is
relatively accurate.
3.2. Eect of sulde and antimony ions
It is generally believed that due to the high electro-
negativity of gold, anions such as S2
are almost com-
pletely discharged and can form adsorption layers of
high coverage up to a monolayer (Wierse et al., 1978). In
fact, the deposition of sulfur ions on gold surface has
been reported before (Wierse et al., 1978; Buckley et al.,
1987; Hamilton and Woods, 1983; Briceno and Chan-
der, 1990). In cyanide solution, the gold open circuit
potential is relatively low. Consequently, it is interesting
to determine if sulde ions are still able to deposit on a
gold surface and make it passive. The eect of sulde
ions on gold anodic behavior is shown in Fig. 4. Curve
(b) indicates that sulde ions have a very negative eect
on gold anodic reaction and the gold surface was pas-
sivated. Even at high overpotentials, gold is still pas-
sivated which suggests that a highly protective lm is
formed on gold surface. This result is not in agreement
with what has been reported by previously Kudryk and
Kellogg (1954). On the other hand, Buckley et al. (1987),
(c) (b) (a)
0.0005 0.0010 0.0015 0.0020 0.0025 0.0030
Current Density(A/cm2)
 M.M. Aghamirian, W.T. Yen / Minerals Engineering 18 (2005) 89102 95

3.0

2.5

Limiting Current Density (mA/cm2)

2.0

1.5

1.0

0.5

0.0
0.00 0.01 0.02 0.03 0.04 0.05 0.06
Cyanide Conc. (M)

Fig. 3. Variation of limiting current density with cyanide concentration, conditions same as Fig. 2.

0.2

(b) (c) (a)

0
Potential (V) vs Eref

-0.2

-0.4

-0.6

-0.8
0 0.0005 0.001 0.0015 0.002 0.0025

Current Density(A/cm2)

Fig. 4. Gold anodic voltammograms in the absence and presence of sulde and antimony ions: (a) no antimony or sulde ions, (b) sulde ion 10
ppm, (c) antimony ions 50 ppm, common conditions: pH 10.5, rpm 500, cyanide conc. 0.01 M, scan rate 0.2 mV/s.

using X-ray photoelectron spectroscopy, conrmed that ions (Hamilton and Woods, 1983). Sulde ions form S0
sulde ions could be reduced on the gold electrode in on the gold surface according to Reactions 16 and 17.
alkaline solution at a potential of )200 mV vs. SHE. As The involvement of two electrons in sulde ion deposi-
they pointed out, this is the potential where the initial tion has been conrmed by Wierse et al. (1978).
underpotential deposition layer is formed and below Au HS
! AuHS
ads 16
that at which elemental sulfur can be formed. Due to
sulfur deposition on the gold surface, a strong layer AuHS
ads ! AuS H 2e 17
bound to the gold is formed substrate which can be best Increasing the potential resulted in a gradual struc-
described as a gold sulde compound. The formation of tural change of such a chemisorbed layer (Briceno and
such a layer has a strong impact on lowering the gold Chander, 1990). A two dimensional sulfur layer with the
activity toward anodic reaction. These results are in general formula of Sx forms at this region according to
close agreement with those reported by Jerey and Reaction 18 and 19, which acts as a strong barrier for
Breuer (2000). cyanide ion diusion.
Curve (b) indicates that gold was passivated at lower AuS xHS
! AuSxHS
ads 18
potentials than expected, )444 mV vs. SCE. The im-

paired anodic dissolution is mainly due to sulde ion AuSxHS ads ! AuSx1 xH 2xe 19
adsorption on the gold surface. In basic solution the Consequently, the gold dissolution current decreased
sulde ion is believed to be in the form of hydrosulde considerably when potential was increased beyond 0 V
96
vs. SCE. The required extent of surface coverage to
make the gold passive depends highly on overpotential.
At a higher overpotential, gold can be passivated at a
lower surface coverage (Buckley et al., 1987). This may
provide an explanation for more intensive passivation of
gold at the higher overpotential.
In addition, such a deposited layer has a more
impermeable structure which almost totally retards gold
anodic reaction. Measuring the surface area under the
anodic curve, such as curve (b) in Fig. 4, and having the
scan rate, we can estimate the maximum expected gold
assay in solution using Faradays law, Eq. (20):
itw
m 20
nF
where w is the gold atomic weight and t the time. By
using this equation the maximum gold assay is estimated
to be 1.18 ppm and the measured assay was 1.06
ppm.The dierence is partially due to sulde ion depo-
sition. More details about sulfur ion deposition on gold
electrode, the structure of such a deposited layer in basic
media, and the eect of lead ion addition were reviewed
elsewhere (Briceno and Chander, 1990; Woods et al.,
1989; Gao et al., 1992; Jerey and Breuer, 2000; Tshi-
lombo and Sandenbergh, 2001).
The eects of antimony ions, dissolved from an
antimony uoride compound, on gold anodic behaviors
can be seen in curve (c) of Fig. 4. At low potential re-
gions, the gold anodic polarization curve did not change
considerably in the presence of antimony ions. The
major eect of antimony ions was on increasing open
circuit potential, which apparently results from the
adsorption of those ions on the gold surface or double
layer. The adsorption of these ions on gold surface was
not stable and gold anodic behavior was restored by
increasing the potential beyond a certain limit, about
)600 mV vs. SCE. Higher than this potential, gold an-
odic current density increased with potential until it
0.200
0.000
Potential (V) vs Eref
-0.200
-0.400
-0.600
-0.800
0.0000
Fig. 5. Eect of metal ions on gold anodic polarization: (a) no metal ion, (b) lead (from Pb(NO3 )2 ) 0.77 ppm, (c) Fe (from FeK4 (CN)6 3H2 O) 200
ppm, (d) copper (from CuCN) 50 ppm, common conditions: pH 10.5, rpm 500, cyanide conc. 0.01 M, scan rate 0.2 mV/s.
M.M. Aghamirian, W.T. Yen / Minerals Engineering 18 (2005) 89102
reached a maximum current density. Increasing the
potential beyond this level resulted in the reduction of
current density. Although passivation in this case was
not completed, it is likely that the adsorption of anti-
mony ions (probably in hydroxide form (Larger and
Forssberg, 1990)) on the gold surface caused the surface
to be partially passivated. After this minimum point, the
current density arose again with increasing potential.
These results show that the adsorbed antimony ions
were replaced by cyanide ions and the surface passiv-
ation was not completed. It was observed that antimony
ion concentration in the solution was reduced from 50 to
13.7 ppm. This decrease in antimony concentration can
be an explanation for diminishing the negative eect of
such ions at higher potentials.
Sulde ions even at the lowest concentration tested in
direct gold dissolution tests, 10 mg/l, had serious
retarding eect on gold dissolution kinetic. Therefore
the sulde ion concentration was adjusted at this level in
this study. Whereas the concentration of antimony ions
was adjusted at 50 mg/l since at this concentration, gold
dissolution rate decreased signicantly in direct gold
dissolution experiments.
3.3. Eect of lead, iron and copper ions
Direct leaching experiments indicated that lead ions
at low concentration and ferrocyanide ions at high
concentration have a signicant positive eect on gold
dissolution. Copper cyanide at moderate concentration
tended to have a slightly negative eect on the gold
dissolution. The eect of these ions at the same con-
centrations on gold anodic dissolution has been studied
to see if there is any correlation between the results of
leaching tests and gold anodic behavior. Linear sweep
voltammograms for gold sample 2 in the presence of
iron, copper and lead ions are presented in Fig. 5. These
voltammograms are fairly similar to each other.
(a)
(c) (b)
(d)
0.0005 0.0010 0.0015 0.0020 0.0025 0.0030
Current Density (A/cm2)
 M.M. Aghamirian, W.T. Yen / Minerals Engineering 18 (2005) 89102
In the presence of lead ions, the peak current density,
which occurred at )400 mV vs. SCE, increased to about
1.9 mA/cm2 . The strong similarity between gold anodic
behavior in the presence and absence of lead ions on the
logarithmic scale suggests that the exchange current
density did not change by any signicant amount in the
presence of lead ions. Consequently, electrocatalytic
eects may not be considered an important reason for
the lead enhancing eect on the gold dissolution rate, at
least under these circumstances. Curve (b) has a small
plateau at a potential of )625 mV vs. SCE, which is
attributed to the lead stripping from the gold electrode.
The potentiodynamic scan was started at )800 mV and
the scan rate was 0.2 mV/s. Consequently, the gold
electrode potential was lower than the lead ion bulk
deposition potential for more than 14 min.
The Nernst equation for lead bulk deposition is
(Pourbaix, 1974):
E0
0:126 0:0295  logPb2
2
where (Pb ) is the lead ion concentration and E0 is the
standard emf of the reaction. At the concentration of
0.77 ppm Pb2 , E0 is equal to )0.54 mV vs. SCE. The
discrepancy between this potential and the stripping
potential may be rationalized by the complexation of
lead cations with anions such as OH
. The stripping
potential from the experimental result, and Eq. (21),
both indicate that in gold leaching practice where the
gold potential is about )400 mV vs. SCE and lead
concentration is in the range of some milligrams per li-
ter, lead ion deposition on gold can only occur through
underpotential deposition. Such a deposition has been
postulated to have a bi-functional eect. It accelerates
both the cathodic reduction (Eisenmann, 1978) and the
anodic dissolution reactions. It is likely that the for-
mation of such a surface layer is energetically more
favorable for various stages of the oxidation Reactions
57, as opposed to the bare surfaces. Optimum results
0.200
0.000
Potential (V) vs Eref
-0.200
(a)
-0.400
-0.600
-0.800
0.0000
Fig. 6. Eect of lead ions on pure gold anodic polarization, Pb (from Pb(NO3 )2 0.75 ppm, common conditions: pH 10.5, rpm 500, cyanide conc. 0.01
M, scan rate 0.2 mV/s.
97
were achieved for lead ion concentration of 0.72 mg/l in
direct gold dissolution experiments. Based on these
ndings the concentration of lead ions was adjusted to
0.77 mg/l.
A similar result was also observed in the presence of
ferrocyanide and cuprous cyanide ions. However, in the
presence of these ions the polarization curves illustrated
almost the same behavior as that of the gold at higher
potentials. In conclusion, in the presence of lead ions at
low overpotential (which is more important as a prac-
tical point of view) has virtually no eect on the gold
anodic behavior. Under the same conditions, it is spec-
ulated that copper and iron ions have slightly negative
eect on gold anodic reaction.
Optimum gold dissolution rate was achieved for a
concentration of 200 mg/l ferrocyanide ions in direct
gold dissolution experiments under the similar experi-
mental conditions mentioned above. Thus the concen-
tration of ferrocyanide ions was xed at the
21 aforementioned concentration. The concentration of
copper ions in the cyanide solution was xed at 50 mg/l
as the dissolution of chalcopyrite (16 g/l) in cyanide
solution even at long retention time, 8 h, never intro-
duces more than 40 mg/l of copper ions into cyanide
solution.
These results are contradictory with those of other
investigators (Pan and Wan, 1979; Cathro, 1965) but
may be attributed to the presence of the slight amount of
silver in the gold electrode. Fig. 6 shows the eect of lead
ion on anodic behavior of pure gold. The result is sim-
ilar to the one reported by Cathro (1964). From this
gure, it can be concluded that the eect of lead ions on
the polarization of pure gold, which is apparently to
remove the passive lm, was disturbed in the case of
goldsilver alloy. A comparison of curves (a) and (b) in
Fig. 6, indicates that in the presence of lead ions the
maximum current increased substantially, however,
passivation still occurred at higher overpotentials.
(b)
0.0002 0.0004 0.0006 0.0008 0.0010 0.0012 0.0014 0.0016
Current Density(A/cm2)
98 M.M. Aghamirian, W.T. Yen / Minerals Engineering 18 (2005) 89102

3.4. Eect of quartz
In order to study the eect of dierent sulde min-
erals on gold anodic behavior, the eect of inert solid
particles was studied. Fig. 7 shows the gold anodic
behavior in the presence of quartz particles. As can be
seen from this gure, quartz has almost no eect on gold
anodic behavior. This is in agreement with what was
expected from inert solid particles and conrms the idea
that the physical eect of inert solid particles on gold
anodic behavior is negligible. Thus, the main reason for
the negative or positive eects of dierent minerals on
gold anodic behavior is due to chemical or galvanic
interactions.
3.5. Eect of stibnite
Gold anodic behavior in the presence of stibnite is
illustrated in Fig. 8. The gold open circuit potential
shifted signicantly toward the anodic potential in the
presence of stibnite. Film formation on the gold surface,
which started at low potentials, about )800 mV vs. SCE,
acted as a strong barrier for the gold oxidation reaction.
Such a layer on the gold surface and in the presence of
stibnite under leaching conditions was conrmed by
XPS experiment. In the presence of such a surface lm,
cyanide diusion to react with gold was prohibited and
the open circuit potential shifted toward the anodic
direction. Above this potential, )450 mV vs. SCE,
cyanide ions started to diuse through the surface layer
and react with gold. The increase of the potential caused
an increase in the cyanide diusion which reached a
maximum level under this set of conditions. This po-
tential was )330 mV vs. SCE (Fig. 8, curve (b)).
The current density reached a maximum level and
was maintained at that level until the overpotential ex-
ceeded )200 mV vs. SCE. Above this potential, the
nature of the surface lm probably started to change. At
the beginning of these structural changes, the current
density increased but it fell down to less than 20 lA at

0.200

0.000 (b)
Potential (V) vs Eref

(a)

-0.200

-0.400

-0.600

-0.800
0.0000 0.0005 0.0010 0.0015 0.0020 0.0025

Current Density(A/cm2)

Fig. 7. Gold anodic polarization in the absence and presence of quartz: (a) no quartz, (b) quartz 8 g/l, common conditions: pH 10.5, rpm 500,
cyanide conc. 0.01 M, scan rate 0.2 mV/s.

0.200

0.000 (a)
(b) (c)
Potential (V) vs Eref

-0.200

-0.400

-0.600

-0.800
0.0000 0.0005 0.0010 0.0015 0.0020 0.0025

Current Density(A/cm2)

Fig. 8. Gold anodic polarization in the absence and presence of stibnite: (a) no stibnite, (b) stibnite 4 g/l, pH 10.5, (c) stibnite 4 g/l, pH 9.0, common
conditions: cyanide conc. 0.01 M, rpm 500, scan rate 0.2 mV/s.
 M.M. Aghamirian, W.T. Yen / Minerals Engineering 18 (2005) 89102
higher overpotential. This is a sign of the formation of a
very impenetrable passive layer on the gold disc surface.
At higher overpotential, beyond 100 mV vs. SCE, the
current density increased slightly, probably making this
layer stable but more permeable. Consequently, the gold
dissolution reaction began at a low rate. To conrm the
nature of the anodic reaction, especially in the last
current region, and to determine whether it was due to
gold anodic dissolution or any other reactions, the area
under curve (b) was measured and the maximum ex-
pected gold assay was calculated by using Eq. (20). The
assay was measured to be 1.42 ppm, whereas the maxi-
mum expected gold assay was 1.45 ppm. This conrms
that gold anodic dissolution was the main reaction
throughout this curve.
Finkelstein (1972) pointed out that stibnite in the
alkaline solution initially oxidizes to thio-antimonite,
SbS2

3 and sulde ions then nally antimonate and sul-
fate will be formed. Similar to sulde ions, thio-antim-
onite may adsorb on the gold surface according to the
following reaction:
Au SbS2
2

3 ! AuSbS3 ads
These adsorbed ions may undergo oxidation according
to the following reaction:
AuSbS2

3 ads ! AuSbS3 2e
In this equation, SbS3 may be considered as a chemi-
sorbed species on the gold surface. This compound is the
most likely to passivate gold in the presence of stibnite.
Although the exact nature of this layer was not deter-
mined, XPS experiments conrmed the presence of
antimony in this layer. No stibnite particles were ob-
served on the gold surface under microscopic study since
the gold surface was washed several times with distilled
water. Cyanide ions do not form any stable compound
with antimony ions and it seems that the only eect
of such ions is to accelerate the oxidation of the thio-
0.200
0.000
Potential (V) vs Eref
-0.200
-0.400
-0.600
-0.800
0.0000
Fig. 9. Eect of galena on gold anodic polarization: (a) no galena, (b) galena 8 g/l, common conditions: pH 10.5, rpm 500, cyanide conc. 0.01 M, scan
rate 0.2 mV/s.
99
antimonite ions to antimonate ions through the fol-
lowing reaction (Hedley and Tabachnick, 1950):



SbS2
3

3 CN 3O2 ! CNS SbO3 24
Since in practice it is preferred to decrease the pH to 9
or 9.5 to diminish the negative eect of stibnite, it would
be interesting to study the eect of stibnite at the lower
pH on gold anodic polarization. Fig. 8c is a gold anodic
voltammogram in the presence of stibnite at pH 9. Once
more, it was observed that the gold open potential
shifted about 140 mV toward the anodic direction, lower
than the one observed at pH 10.5. In the foot of this
curve, the current density increased slowly with poten-
tial. The current density increased sharply with potential
when the potential was increased beyond a certain level,
)400 mV vs. SCE. From this it can be concluded that
the gold surface was not passivated. This eect is mainly
due to the lower solubility of stibnite in alkaline solution
at lower pH. The antimony ion concentration in cyanide
solution, measured by atomic adsorption spectroscopy,
was 16.75 ppm at pH 10.5, reduced to 4.67 ppm at pH 9.
Stibinite concentration in this experiment was adjusted
22
at 4 g/l since at such a concentration gold dissolution
rate in direct leaching experiments was almost retarded.
3.6. Eect of galena
23
Gold anodic behavior in the presence of galena is
illustrated in Fig. 9. At low overpotentials, the results
are consistent with those obtained in the presence of
lead ions. At a higher potential range, the current passed
through two maxima and two minima. The rst maxi-
mum was attributed to the maximum cyanide diusion
under these conditions. In the presence of galena, the
maximum current density, which was limited by the
cyanide diusion limit, was enhanced due to dissolved
lead ions. Such an eect is also demonstrated in Fig. 5
for the case of lead ions. Under this condition, the
(b) (a)
0.0005 0.0010 0.0015 0.0020 0.0025
Current Density(A/cm2)
100
surface concentration of cyanide ions on the gold sur-
face approached zero and the concentration of
(AuCN
)ads also approached zero. Therefore, the cur-
rent density is reduced. At higher potential, the surface
concentration of cyanide ion was no longer zero and the
rate of formation of (AuCN
)ads controlled the gold
anodic dissolution. Furthermore, an increase in surface
concentration of (AuCN)ads promoted the nal reaction
with cyanide ions. At higher potential, about 100 mV vs.
SCE, cyanide ion concentration at the gold surface
dropped to a low level which caused the current density
to decline. Such a complex behavior has also been ob-
served by Eisenmann (1978), however cyanide concen-
tration and temperature were higher in his work (0.5 M
and 55 C respectively).
Once again we can observe that, in the presence of
lead ions or galena, gold anodic polarization did not
change signicantly at low and moderate overpotentials.
As a result, it is reasonable to believe that under these
conditions a signicant positive eect of lead ions or
galena on gold dissolution rate, which was observed in
direct gold disc dissolution experiments under the simi-
lar lead or galena concentration, is not mainly due to the
enhancement on gold anodic reaction.
3.7. Eect of chalcopyrite, pyrite and pyrrhotite
Linear sweep voltammogram curves for gold anodic
dissolution in the presence of pyrite, pyrrhotite or
chalcopyrite are presented in Fig. 10. There are several
interesting points to be considered in this gure. In pure
cyanide solution or in the presence of pyrite, the gold
open circuit potential was about )780 mV vs. SCE. On
the other hand, chalcopyrite and pyrrhotite slightly
shifted the gold open circuit potential, about 40 mV, in
the anodic direction. Consequently, it is reasonable to
conclude that the eect of those sulde minerals on the
gold open circuit potential was not signicant. However,
0.200
0.000
Potential (V) vs Eref
-0.200
-0.400
-0.600
-0.800
0.0000
Fig. 10. Gold anodic polarization in the absence and presence of sulde minerals: (a) no sulde mineral, (b) pyrite, (c) chalcopyrite (d) pyrrhotite,
common conditions: pH 10.5, cyanide conc. 0.01 M, rpm 500, sulde mineral conc. 8 g/l, scan rate 0.2 mV/s.
M.M. Aghamirian, W.T. Yen / Minerals Engineering 18 (2005) 89102
there is a potential region at the foot of the voltammo-
grams in which current density did not increase to any
signicant value in comparison with curve (a). After
increasing the potential above a certain level, current
density started to rise sharply with potential. All of these
curves did not reach any maximum point and current
density increased continuously with potential.
Pyrite and pyrrhotite are slightly soluble in cyanide
solution and introduce various sulfur specious and fer-
rous cyanide ions into the solution. In addition to these
ions, chalcopyrite also dissolved to a much greater ex-
tent, and formed various cuprous cyanide compounds
(Shantz and Reich, 1978; Rees and Van Deventer, 1999).
It has been shown in Fig. 5 that iron and copper ions,
even at relatively high concentration, did not have any
signicant negative eect on gold anodic behavior. Since
the concentration of copper and iron released from these
minerals were on the order of milligrams per liter, we
can conclude that such ions are not expected to have any
signicant negative eect on gold anodic behavior.
Other than these ions, sulde ions are also expected to
be in solution. It has been shown, in Fig. 4, that sulde
ions had very serious stable retarding eect on the gold
anodic behavior, which caused the gold to be passivated.
Consequently since such a passivation was not observed
under these circumstances, it is reasonable to believe
that dissolved sulde ions from those sulde minerals
did not reach levels sucient enough to cause the gold
to be passivated. In addition, some sulde minerals such
as galena and pyrite are active in reducing the sulde
ions on their surface (Paul, 1984). However, there is not
enough evidence to prove that this mechanism is oper-
ative in an oxygen depleted solution.
It has been reported previously (Lorenzen and van
Denventer, 1992), that the galvanic interaction between
gold and sulde minerals plays an important negative
eect on gold dissolution behavior. In our case, how-
ever, gold and sulde minerals were liberated from each
(a) (c) (d)
(b)
0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 0.0035
Current Density(A/cm2)
 M.M. Aghamirian, W.T. Yen / Minerals Engineering 18 (2005) 89102
other completely and oxygen concentration in the
solution was reduced to almost zero, the galvanic eect
might not play such an important role.
Comparing curve (a) with all other curves in Fig. 10, it
is possible to conclude that the gold anodic curves in the
presence of sulde minerals were dislocated toward the
more anodic potentials. On the other hand, a steady-state
polarization curve for oxygen reduction (cathodic curve)
on the gold surface showed a continuous increase of
current with potential, when the potential was less than
)300 mV vs. SCE. This is in agreement with the ndings
of other researchers (Li, 1991). In conclusion, the inter-
section of the oxygen reduction prole (cathodic curve)
with the anodic curves indicates that mixed potentials
shift toward the more anodic potential in the presence of
those sulde minerals, and the corrosion current would
consequently decrease. A shift of the free corrosion po-
tential toward the more noble values of pure gold in the
presence of sulde minerals was reported by Lorenzen
and van Denventer (1992). This can explain the negative
eect of those minerals on the gold dissolution rate.
Since in direct leaching tests, the concentrations of the
minerals were changed form 2 to 16 g/l, an average min-
eral concentration (8 mg/l) was selected for preliminary
electrochemical experiments. Further experiments under
various concentrations are planned in the next stage.
4. Conclusions
The anodic behavior of pure gold and gold contain-
ing a minor amount of silver, 0.31%, in de-aerated
cyanide solution and in the presence of dierent ions
and sulde minerals have been studied. From the results
presented in this paper, the following conclusions can be
drawn:
1. The anodic polarization of pure gold showed the sur-
face of the gold became passivated in cyanide solu-
tion. In the case of goldsilver alloy, passivation
area was not observed. In addition, this disc was
more active towards the anodic reaction. It is postu-
lated that the presence of a minor amount of silver in
the gold disc may decrease the activation energy for
gold cyanide reaction. Further investigation is re-
quired to justify this postulation.
2. In the presence of trace amounts of sulde ions, the
gold anodic reaction was retarded. This was due to
the formation of a passive layer on the gold surface.
Cuprous or ferrocyanide ions did not change the gold
anodic polarization to any signicant degree. Lead
ions had a pronounced eect on the peak current of
the pure gold at potential )240 mV vs. SCE, however
at higher potentials, the gold surface was still passiv-
ated. On the other hand, in the case of goldsilver
alloy, it left the anodic characteristic almost unchanged.
101
3. In the presence of stibnite, the gold anodic reaction
decreased to a very low level. The formation of
thio-antimonates on the gold surface as a strong pas-
sive layer is believed to be the reason for such behav-
ior. At a lower pH of 9, the gold anodic polarization
did not exhibit any passive region and no passive
layer was observed on the gold disc surface after ano-
dic polarization.
4. In the presence of pyrite, pyrrhotite and chalcopyrite,
gold anodic polarization curves were moved some-
what toward the anodic direction, although the gold
open circuit potential did not change signicantly.
This was mainly due to the minor increase of current
with potential at the foot of anodic waves. As a result
of such displacement, mixed potential will move to-
ward the anodic direction. Consequently gold corro-
sion current, or on the other word gold dissolution
rate, will decrease too.
Acknowledgements
Financial assistance from the Ministry of Culture and
Higher Education of Iran is gratefully acknowledged.
The authors would like to thank Dr. M. Shirkhanzadeh
and Dr. M. Tulmin and for helpful discussions which
inuenced the direction of this work. Appreciation is
also expressed to Mr. A. Pindred and Mr. C. Cooney for
their technical help.
References
Adzic, R.R. et al., 1980. Oxygen reduction on electrode surface
modied by foreign metal ad-atoms: lead ad-atoms on gold.
Electroanal. Chem. 114, 3751.
Briceno, A., Chander, S., 1990. Oxidation of hydrosulphide ions on
gold. J. Appl. Electrochem. 20, 506517.
Brockris, J.OM., Drazic, D.M., 1972. Electrochemical Science. Barnes
and Noble Books, New York. p. 137.
Buckley, A.N., Hamilton, I.C., Woods, R., 1987. An investigation of
the sulphur (II)/sulphur (0) system on gold electrodes. J. Electro-
anal. Chem. 216, 213227.
Cathro, K.J., 1964. The eect of oxygen in cyanide process for gold
recovery. Aust. Inst. Min. Metall. Proc. 207, 181205.
Cathro, K.J., 1965. The eect of thallium on the rate of extraction of
gold from pyrites calcine. Aust. Inst. Min. Metall. Proc. 208, 127
137.
Cathro, K.J., Koch, D.F.A., 1964. The anodic dissolution of gold in
cyanide solutions. J. Electrochem. Soc. 111, 14161420.
Cordova, R., Martins, M.E., Arvia, A.J., 1979. Multiplicity of anodic
and cathodic current peaks observed during potentiodynamic
perturbations of gold/sodium hydroxide solution interfaces. J.
Electrochem. Soc. 126 (7), 11721176.
Deschenes, G. et al., 1998. Eect of the compostion of some sulphide
minerals on cyanidation and use of lead nitrate and oxygen to
alleviate their impact. Hydrometallurgy 50, 205221.
Deschenes, G. et al., 2000. Eect of lead nitrate on cyanidation of gold
ores: progress on the study of the mechanisms. Miner. Eng. 13 (12),
12631279.
102 M.M. Aghamirian, W.T. Yen / Minerals Engineering 18 (2005) 89102
Dorin, R., Woods, R., 1991. Determination of leaching rate of
precious metals by electrochemical techniques. J. Appl. Electro-
chem. 21, 419424.
Eisenmann, E.T., 1978. Kinetics of the electrochemical reduction of
dicyanoaurate. J. Electrochem. Soc. 125, 717723.
Finkelstein, N.P., 1972. Gold Metallurgy in South Africa, p. 284.
Fleming, C.A., 1992. Hydrometallurgy of precious metals recovery.
Hydrometallurgy 30, 127162.
Gao, X. et al., 1992. Observing surface chemical transformations by
atomic-resolution scanning tunnelling microscopy: sulde elect-
rooxidation on Au(III). J. Phys. Chem. 96, 41564159.
Guan, Y., Han, K.N., 1994. An electrochemical study on the
dissolution of gold and copper form gold/copper alloys. Met.
Trans. 25B, 817827.
Hamilton, I.C., Woods, R., 1983. An investigation of the deposition
and reaction of sulphur on gold electrodes. J. Appl. Electrochem.
13, 783794.
Hedley, N., Tabachnick, H., 1950. Chemistry of cyanidation. Mineral,
Dressing Notes, No. 17, American Cyanamid Co.
Jerey, M.I., Breuer, P.L., 2000. The cyanide leaching of gold in
solutions containing sulphide. Miner. Eng. 13 (1011), 10971106.
Kameda, M., 1949. Fundamental studies on dissolution of gold in
cyanide solution. III. Eect of alkalis, lead acetate and some
impurities contained in foul cyanide solution. Sci. Rep. RITU, Ser.
A 1, 133138.
Kirk, D.W., Foulkes, F.R., 1980. Anodic dissolution of gold in
aqueous alkaline cyanide solution at low overpotentials. J. Elect-
rochem. Soc. 127 (9), 19931997.
Kirk, D.W., Foulkes, F.R., Graydon, W.F., 1978. A study of anodic
dissolution of gold in aqueous alkaline cyanide. J. Electrochem.
Soc. 125 (9), 14361443.
Kirk, D.W., Foulkes, F.R., Graydon, W.F., 1980a. The electrochem-
ical formation of Au(I) hydroxide on gold in aqueous potassium
hydroxide. J. Electrochem. Soc. 127 (9), 10691076.
Kirk, D.W., Foulkes, F.R., Graydon, W.F., 1980b. Gold passivation
in aqueous alkaline cyanidation. J. Electrochem. Soc. 127 (9),
19621969.
Kudryk, V., Kellogg, H.H., 1954. Mechanism and rate controlling
factors in the dissolution of gold in cyanide solution. J. Metals 6,
541548.
Larger, G., Forssberg, K.S.E., 1990. Equilibrium model with reference
to stibnite otation. IMM Trans., Jan.Apr. 99, C62C70.
Li, J., 1991. Rate Process of Silver Cyanidation. Ph.D. Thesis, The
University of Utah.
Li, J., Wadsworth, M.E., 1993. Electrochemistry study of silver disso-
lution in cyanide solutions. J. Electrochem. Soc. 140, 19211927.
Liu, G.Q., 1993. Cyanidation of Gold and Silver in Oxygen Enriched
Solution. M.Sc. Thesis, Queens University.
Lorenzen, L., van Denventer, J.S.J., 1992. Electrochemistry interac-
tions between gold and its associated minerals using cyanidation.
Hydrometallurgy 30, 177.
Mac Arthur, D.C., 1972. A study of gold reduction and oxidation in
aqueous solution. J. Electrochem. Soc. 119, 672677.
McIntyre, J.D.E., Peck, W.F., 1976. Electrodepositing of gold.
J. Electrochem. Soc. 123 (12), 18001813.
Mills, T., 1951. An Introduction Study of the Dissolution of Gold and
Silver in Cyanide Solution. Ph.D. Thesis, University of Melbourne.
Mussatii, D. et al., 1997. Electrochemical aspects of the dissolution of
gold in cyanide electrolytes containing lead. In: Dreisinger, D.B.
(Ed.), Proceeding of Aqueous Electrotechnologies: Progress in
Theory and Practice. The Minerals, Materials Society, New York,
pp. 247265.
Nicol, M.J., 1980. The anodic behaviour of gold. Gold Bull. 13, 105
111.
Omar, I.H., Pauling, H.J., J uttner, K., 1993. Underpotential deposi-
tion of copper on Au(III) single-crystal surface. J. Electrochem.
Soc. 140, 2187.
Pan, T.P., Wan, C.C., 1979. Anodic behavior of gold in cyanide
solution. J. Appl. Electrochem. 9, 653655.
Paul, R.L, 1984. The role of electrochemistry in the extraction of gold.
J. Electroanal. Chem. 168, 147162.
Pourbaix, M., 1974. Atlas of Electrochemical Equilibrium in Aqueous
Solutions, p. 485.
Rees, K.L., Van Deventer, J.S.J., 1999. The role of metal-cyanide
species in leaching gold from a copper concentrate. Miner. Eng. 12
(8), 877892.
Sawaguchi, T., Yamada, T., Okinaka, Y., 1995. Electrochemical
scanning tunneling microscopy and ultrahigh-vacuum investigation
of gold cyanide adlayers on Au(III) formed in aqueous solution.
J. Phys. Chem. 99, 1414914155.
Shantz, R., Reich, J., 1978. A review of copper-cyanide metallurgy.
Hyrometallurgy 3, 99109.
Sun, X., Guan, Y.C., Han, K.N., 1995. The electrochemical behavior
of the dissolution of goldsilver alloys in cyanide solutions. Prepr.
Soc. Miner. Engrs. AIME 95-8.
Thompson, P.F., 1947. The dissolution of gold in cyanide solutions.
Electrochem. Soc. Trans. 91, 4171.
Tshilombo, A.F., Sandenbergh, R.F., 2001. An electrochemical study
of the eect of lead and sulphide ions on the dissolution rate of gold
in alkaline cyanide solutions. Hydrometallurgy 60, 5567.
Wadsworth, M.E., 2000. Surface processes in silver and gold cyani-
dation. Int. J. Miner. Process. 58, 351368.
Wadsworth, M.E., Zhu, X., 2003. Kinetics of enhanced gold dissolu-
tion: activation by dissolved silver. Int. J. Miner. Process. 72, 301
310.
Wierse, D.G. et al., 1978. Electrochemical properties of sulfur
adsorbed on gold electrodes. J. Eletroanal. Chem. 92 (2), 121131.
Woods, R., 1976. Gold. In: Bard, A.J. (Ed.), Electroanalytical
Chemistry, vol. 9. Marcel Dekker, New York, pp. 119125.
Woods, R., Constable, D.C., Hamilton, I.C., 1989. A rotating ring disc
electrode study of the oxidation of sulfur (II) specious on gold and
sulde minerals. Int. J. Miner. Process. 27, 309326.