Hydrometallurgy 173 (2017) 115124

Contents lists available at ScienceDirect

Hydrometallurgy
journal homepage: www.elsevier.com/locate/hydromet

The eect of grinding chemistry on cyanide leaching of gold in the presence MARK
of pyrrhotite
A. Rabieh, J.J. Eksteen, B. Albijanic
Department of Mining Engineering and Metallurgical Engineering, Western Australian School of Mines, Curtin University, GPO Box U1987, Perth, WA 6845, Australia

A R T I C L E I N F O A B S T R A C T

Keywords: In this paper, the role of grinding media type on cyanide-based gold leaching in the presence of pyrrhotite was
Grinding media investigated. Pyrrhotite commonly occurs in many gold ores together with pyrite and/or arsenopyrite. A syn-
Gold cyanidation thetic ore was prepared by mixing pyrrhotite with gravity gold concentrate and quartz. Pyrrhotite is of interest to
Pyrrhotite and galvanic interaction grinding and leaching chemistry because this sulde mineral is the highest cyanide- and oxygen-consuming iron
sulde and, contrary to pyrite, tends to behave as a reactive anode. The results showed that when the grinding of
the ore was performed using the ceramic media, the gold extraction process was up to three times more ecient
than that when the forged steel media was used. The reason is that the galvanic interactions between the forged
steel media and pyrrhotite resulted in the formation of iron hydroxide; the formed iron hydroxide converted free
cyanide to ferrocyanide that cannot dissolve gold (i.e. 75% of free cyanide was converted to ferrocyanide).
Additionally, the galvanic interactions also signicantly reduce the dissolved oxygen content and thus cyanide
gold leaching is more dicult. In contrast, when grinding was performed with the ceramic or high chromium
(30%Cr) media, an insignicant amount of free cyanide was decomposed (i.e. 1% of free cyanide was converted
to ferrocyanide while 2.5% of that was converted to thiocyanate). For that reason, cyanide gold leaching is very
ecient when ceramic or high chromium media was used during the grinding of the ore. It can be expected that
the leach behaviour would be dierent if leaching is preceded by pre-oxidation, as is often done for pyrrhotite
bearing ores. In addition, a theoretical model was developed to describe the correlation between the galvanic
current and the grinding media oxidation, as measured by using EDTA extraction technique and dissolved sulfur
analysis by HPLC during grinding. The electrochemical data were correlated with the amount of the grinding
media oxidation during grinding with dierent media type. A linear relationship between the galvanic current
and grinding media oxidation was obtained. It was found that the higher the galvanic current, the higher the
amount of iron hydroxide was produced during grinding. Bench scale cyanide leaching can therefore serve as a
useful diagnostic to determine the extent of galvanic interactions between media and minerals in grinding mills
(similar to bench scale otation which has also been used as diagnostic tool).

1. Introduction
It is well known that grinding media type might aect downstream
operations such as otation and leaching. Specically, during grinding
with steel grinding media, mineral particles become coated with dif-
ferent species of iron, changing the surface properties of the minerals. For
example, it was found that the higher the percentage of iron in grinding
media, the lower the otation response of sulde minerals (Martin et al.,
1991; Cullinan et al., 1999; Rabieh et al., 2016). The decrease in otation
response was probabbly due to the presence of a hydrophilic layer of
oxidised iron species on the surface of sulde minerals.
Furthermore, during grinding of sulde ores due to the galvanic
current between forged steel grinding media and sulde minerals,
electrons are transferred from the anode (i.e. grinding media) to the
cathode (i.e. cathodic mineral), causing the change of the potential
of both materials from their rest potential to the mixed potential
(Adam et al., 1984; Martin et al., 1991). As a result, the corrosion of
the less cathodic steel grinding media (low rest potential) increases
and the corrosion of the more cathodic material (high rest potential)
decreases (Fontana, 1987). Therefore, grinding media oxidation
increases.
While gold-bearing sulde ore cyanidation can be challenging there
is a great variation in the extent to which sulde minerals aect the
leaching process. Deschenes et al. (2003) suggested that all sulde
minerals could be divided into three groups depending on their eect
on gold leaching:

Corresponding author.
E-mail address: jacques.eksteen@curtin.edu.au (J.J. Eksteen).

http://dx.doi.org/10.1016/j.hydromet.2017.08.013
Received 6 April 2017; Received in revised form 4 August 2017; Accepted 12 August 2017
Available online 14 August 2017
0304-386X/ 2017 Elsevier B.V. All rights reserved.
A. Rabieh et al.
A group that reduces gold leaching (stibnite > orpiment >
pyrrhotite > chalcocite > arsenopyrite > chalcopyrite);
A group has little impact on gold leaching (pyrite and sphalerite);
And a group that improves gold leaching (galena).
The eect of sulde minerals on gold cyanide leaching is governed
by amount of sulde in the ore, as well as solubility and reactivity of
sulde minerals (Deschenes et al., 2003). Although the eects of sulde
minerals and their dissolved ions on the gold leaching process have
been investigated (Ntemi et al., 2013), the inuence of galvanic inter-
actions occurring between grinding media and sulde minerals on gold
cyanide leaching has not yet been systematically evaluated. For that
reason, the objective of this work is to investigate the inuence of
pyrrhotite (highly reactive sulde mineral) and the grinding media on
cyanide gold leaching. In this paper, pyrrhotite was selected because
this sulde mineral is the highest cyanide- and oxygen-consuming iron
sulde mineral. With regards to grinding media type, the following
grinding media were used: the forged steel, 21% chromium, 30%
chromium and ceramic media. It should be noted that in the case of free
milling gold ores, pre-treatment options are not required before cyanide
gold leaching; thus, this work is relevant for free milling gold ores and
not for refractory gold ores. The research in this paper is also performed
to establish a baseline for comparison to other iron sulde minerals
such as arsenopyrite and pyrite at the same mass % in the ore (Rabieh
et al., 2017a, 2017b). It is realised that industrial gold-bearing pyr-
rhotite ores may often be preoxidised which may change the outcomes
of gold leaching. However, the preoxidation falls outside the scope of
the research reported here.
2. Materials and methods
2.1. Synthetic ore
The synthetic ore was obtained by mixing 12 g of pyrrhotite (99%
purity), 0.5 g of the gravity gold concentrate (which contains 6% gold
concentrate analysed by re assay), and 587.5 g of quartz (99.5%
purity). Therefore, the synthetic gold ore had 2% pyrrhotite and
50 ppm gold. Table 1 shows the XRD analysis of the gold concentrate.
The 2% level of pyrrhotite was chosen to correspond to the mass per-
centage used for arsenopyrite and pyrite (Rabieh et al., 2017a, 2017b).
2.2. Grinding system
The modied grinding system is designed and built to replicate
plant mill performance at a laboratory scale. The feed ore was prepared
to a P80 of 75 m by wet grinding at 34% w/w solids in the modied
grinding system as seen in Fig. 1. The four dierent grinding media
were used: the forged steel, 21% chromium, 30% chromium and
ceramic media. The surface area of the grinding chamber is 0.2 m2.
Table 1
XRD analysis of the gold concentrate.
Mineral Mass (%)
Clinochlore 6
Talc <1
Biotite 27
Muscovite 25
Amphibole 1 paran oil. The carbon paste-pyrrhotite was packed into the hole of the
Clinopyroxene 3
Rutile 1
Albite 3
Alpha quartz 22
Calcite <1
Dolomiteankerite 4
Malachite 4
Magnetite 3 Potentiostat's electrodes were placed inside the mill. The three-elec-
116
Hydrometallurgy 173 (2017) 115124
Table 2 shows the dimension, surface area, chemical composition of
dierent grinding media types used in this project.
The grinding system (see Fig. 2) was built not only to grind the ore
but also to monitor the chemical conditions (pH, Eh, and dissolved
oxygen) during grinding of the ore samples. The changes of chemical
conditions during grinding of the ore were recorded every minute by
the controller system. In addition, every 20 min around 35 mL of the
slurry was collected for the EDTA iron extraction and sulfur measure-
ments. The amount of dissolved sulfur was measured by high perfor-
mance liquid chromatography (HPLC) method; SO42 ions were the
only detected sulfur species.
2.3. Cyanide leaching test procedure
After grinding, the mill discharge was transferred into the 3 L glass
reactor (Fig. 3), mechanically agitated along with a sucient quantity
of Perth tap water to generate slurry samples at 22% (w/w) solid ratio.
Around 12 mL of caustic soda solution (1 M NaOH) was added in order
to maintain a pH of approximately 10.5, and then mixed for another
5 min; then the solid sodium cyanide was added to the slurry in order to
keep the cyanide solution strength at 500 ppm. During the leaching
time, approximately 20 mL of the slurry samples were collected after 2,
4, and 24 h. These samples were ltered to analyse the solution for gold
content by atomic absorption (AAS) technique, the cyanide speciation
by high performance liquid chromatography (HPLC), and the cyanide
solution strength by silver nitrate titration. After 24 h, the leach residue
solids were ltered, thoroughly washed, dried, weighed, and submitted
for gold analysis by re assay.
2.4. Determination of iron hydroxide content by EDTA
An ethylene diamine tetra acetic acid (EDTA) extraction method
was employed to determine the amount of extractable iron in the
samples taken from the mill slurry after every 20 min of grinding of the
ore. In this method, a 250 mL of 3%wt. of EDTA solution at pH of 7.5
was prepared and transferred into a laboratory leaching vessel with a
magnetic stirrer and purged with nitrogen for 5 min to remove the
dissolved oxygen. The mill slurry samples (25 mL) were then added.
The experiments were carried out for 5 min by purging with nitrogen.
The slurry was then ltered through a 0.2 m millipore membrane and
the solution was analysed by inductively-coupled plasma optical
emission spectrometry (ICP-OES) for iron content, and then the amount
of iron in the aqueous phase was calculated.
2.5. Electrodes preparation and electrochemistry apparatus
2.5.1. Grinding media electrodes preparation
The forged steel and 30% chromium media were used to make the
working electrodes. A grinding media sample was cut to the size of
15 15 3 mm to ll a Teon tube. The electrode was connected to a
screw by a conductive epoxy resin, and non-conductive epoxy was used
to ll in between the Teon tube and electrodes for isolation. Before
each experiment, the electrode surface was gently polished by a set of
polishing discs and alumina slurries.
2.5.2. Mineral electrode preparation
The carbon paste-pyrrhotite electrode was prepared by thoroughly
mixing 4 g of graphite powder with 3 g of pure pyrrhotite and 3 g of
electrode body and smoothed on clean paper until it had a shiny surface.
2.5.3. In situ mill electrochemistry measurements
All in situ mill electrochemistry measurements were performed
using a portable Potentiostat/Galvanostat system in which the
trode setup consisted of the grinding media electrode or carbon paste-
A. Rabieh et al. Hydrometallurgy 173 (2017) 115124

Fig. 1. Schematic diagram of the modied grinding system.

pyrrhotite electrode as the working electrode, an Ag/AgCl in 3 mol/L (Eqs. (2) and (3)) and hydrolysis of Fe 2 + and Fe 3 + (Eqs. (4) and (5))).
KCl as a reference electrode and a platinum wire as a counter electrode. Cathodic reaction on sulde minerals

1 2 O2 + H2 O + 2e 2(OH) K = 3.687 1013 (1)

3. Results and discussion
Anodic reaction on grinding media (steel media oxidation):
3.1. Grinding Fe Fe2 + + 2e K = 6.598 1013 (2)

3.1.1. Pulp chemical environment during the grinding of the ore Fe Fe3 + + 3e K = 1.025 1013 (3)
The pulp chemical measurements during grinding of the ore with
Hydrolysis of iron ions to form precipitation on the grinding media:
four dierent grinding media are given in Fig. 4. As seen in Fig. 4,
changes from the forged steel grinding media to the high chromium and Fe2 + + 2H2 O Fe(OH)2 + 2H+ K = 2.534 1012 (4)
ceramic grinding media shifted the grinding conditions from a moder-
ately reducing to an oxidising environment. The reason for the drop of Fe3 + + 3H2 O Fe(OH)3 + 3H+ K = 3.654 105 (5)
pulp potential during grinding with the forged steel media is due to the
Pyrrhotite oxidation:
galvanic interactions between the forged steel media and pyrrhotite,
causing the increase in the hydroxide ion production as shown in Eq. 4Fe7 S8 + 69O2 + 74H2 O 28Fe(OH)3 + 32SO4 2 + 64H+ K = 1 10308
(1). Thus, the pulp potential decreased while the pulp pH increased. (6)
Furthermore, according to Marsden and House (1992), in neutral and
alkaline solutions, and in the presence of dissolved oxygen, pyrrhotite
can be oxidised by the reaction presented in Eq. (6). The pyrrhotite 3.1.2. EDTA and dissolved sulfur measurements during grinding
consumed the oxygen in the slurry, resulting in the drop of the dis- The EDTA iron extraction technique was performed on the samples
solved oxygen content. collected every 20 min during grinding with dierent grinding media.
In addition, under the above conditions, the following electro- EDTA-extracted iron can originate from the oxidation of grinding media
chemical reactions may be happened on the surface of sulde minerals and pyrrhotite.
(cathodic reaction, Eq. (1)) and grinding media (i.e. anodic reaction So, the amount of EDTA iron extraction can be shown as:

Table 2
Properties of grinding media.

Media Diameter (mm) Weight (kg) Surface area (m2) Chemical composition of media

Forged steel 25 6 0.3 98% Fe, 0.8% C, 0.3%Cr, 0.2% Si and 0.7% Mn
21% chromium 25 6 0.3 75% Fe, 2.8% C, 21% Cr, 0.45% Si, and 0.75% Mn
30% chromium 25 6 0.3 67% Fe, 2.3% C, 30% Cr, 0.45% Si and 0.45% Mn
Ceramic 20 6 0.4 97% ZrO2, 3% MgO, < 0.3% SiO2, < 0.2% Fe2O3, < 0.3% CaO, < 0.2% TiO2, < 0.2% Al2O3

117
A. Rabieh et al. Hydrometallurgy 173 (2017) 115124

Fig. 2. Modied grinding system.

mill which was made of steel and thus there were the galvanic inter-
actions between the inner wall of the mill and pyrrhotite particles.

3.2. Cyanide gold leaching

3.2.1. Pulp chemical environment during the cyanide gold leaching

Although free cyanide was added to dissolve gold only (see Eq.
(10)), free cyanide also reacts with pyrrhotite (Eqs. (11), (12) and (13))
and iron hydroxide as presented in Eq. (14) (Marsden and House,
1992). It should be noted that the presence of iron hydroxide can be
explained by Eqs. (4), (5), (6) and (13).
4Au + 8CN + O2 + 2H2 O 4Au(CN)2 + 4OH K = 5.618 106
(10)
Fe7 S8 + CN 7FeS + SCN K = 1.028 107 (11)

2FeS + 12CN + 5O2 + 2H2 O + 4e

2Fe(CN)6 4 + 2SO4 2 + 4OH K = 1 10308 (12)

Fig. 3. Schematic view of the leaching apparatus. FeS + CN + 0.75O2 + 1.5H2 O

QEDTA for Iron = QFe from media oxidation + Q Fe from pyrrhotite (7)
It is hypothesised that pyrrhotite was oxidised by the reactions
shown in Eq. (6). The equation shows that the amount of dissolved
sulfur is 8/7 of the EDTA-extractable iron formed from pyrrhotite oxi-
dation (Eq. (6)). Therefore, the oxidation amount of forged steel is
given by the following equation:
QFe from pyrrhotite oxidation = 7 8 QSO4
(8)
QFe from media oxidation = Q EDTA for Iron 7 8 Q SO4 (9)
Figs. 5 and 6 show the amount of iron hydroxide and dissolved
sulfur in the slurry as a function of grinding time and grinding
media.
The calculated amount of iron due to the oxidation of grinding
media is presented in Fig. 7. The forged steel produced the highest
amount of iron hydroxide in comparison with the 30% chromium and
ceramic media. The results show that most of the EDTA-extractable iron
in the system originated from the grinding media oxidation. In the case
of the ceramic media, the iron oxidation appeared to have occurred
despite that the ceramic media is non-conductive and not iron-bearing.
The reason for that is ascribed to the contribution of inner wall of the
Fe(OH)3 + SCN K = 1.307 1057 (13)
Fe(OH)3 + 6CN + e Fe(CN)6 4 + 3OH K = 6.660 1018
(14)
Fig. 8 shows EhpH diagram for the FeSCNH2O system at 25 C.
As seen in Fig. 8, the most dominant ion at conditions occurring during
the cyanide gold leaching (i.e. pH = 10.5, Eh = 0.3 to 1 V vs SHE) is
ferrocyanide. The formation of ferrocyanide can be explained by Eqs.
(12) and (14). It should be noted that other factors can also aect the
consumption of free cyanide signicantly, including the presence of
impurity, foreign ion inclusions, lattice dislocations in the mineral
crystal structure, and the interaction of the mineral with other reactive
or conductive minerals and metals that can result in galvanic interac-
tions (Marsden and House, 1992).
Fig. 9 shows the changes of pulp chemical variables (pH, DO and Eh)
during the cyanide gold leaching process. The amount of iron hydroxide
was higher when the forged steel or 21% chromium grinding media
were used, compared to when 30% chromium media or ceramic media
were used. For that reason, using of the forged steel or 21% chromium
media resulted in the production of a strong reducing environment
during the rst 2 h of leaching probably due to the reaction between
iron hydroxide and free cyanide (see Eq. (14)). Dissolved oxygen was
consumed by the reaction of free cyanide with pyrrhotite (Eqs.

118
A. Rabieh et al. Hydrometallurgy 173 (2017) 115124

Fig. 5. The amount of EDTA extractable iron as a function of grinding time: (solid line)
forged steel media; (dash line) 21% chromium media; (dash dot) 30% chromium; (dot
line) ceramic media.

Fig. 6. The amount of dissolved sulfur (SO42 ) in the slurries as a function of grinding
time: (solid line) forged steel media; (dash line) 21% chromium media; (dash dot) 30%
chromium; (dot line) ceramic media.

Fig. 4. Pulp chemical conditions as a function of grinding time: (solid line) forged steel
media; (dash line) 21% chromium media; (dash dot) 30% chromium; (dot line) ceramic
media.

(11)(13)). For that reason, the gold leaching in the presence of pyr-
rhotite is problematic for the cyanidation process, because dissolved
oxygen is used not only for cyanide gold leaching (Eq. (10)) but also for
the pyrrhotite oxidation (Eqs. (6), (12) and (13)). In other words, there
is a competition for dissolved oxygen between pyrrhotite and gold. It is
important to highlight that free cyanide was also consumed by iron
hydroxide which is especially important when the forged steel or 21%
chromium media were used (Eq. (14)). Additionally, when the forged
steel or 21% chromium grinding media were employed in the rst 2 h
of leaching, hydroxide ion is produced; therefore, the pulp pH increased
to 11.6 until the free cyanide or iron hydroxide were consumed.

3.2.2. Cyanide gold leaching performance

Fig. 10 shows the inuence of grinding media on cyanide-based gold
leaching. As seen in Fig. 10, when the forged steel grinding media was
used, the gold dissolution decreased dramatically maybe due to the Fig. 7. The amount of iron oxidised from media as a function of grinding time: (solid line)
signicant galvanic interactions between pyrrhotite, gold and the forged steel media; (dash line) 21% chromium media; (dash dot) 30% chromium; (dot
line) ceramic media.
forged steel media. The consequences of the galvanic interactions were

119
A. Rabieh et al.
Fig. 8. EhpH diagram for the FeSCNH2O system at 25 C (Marsden and House, 1992).
mainly the forged media oxidation (see Fig. 7), resulting in the for-
mation of iron hydroxide. The ore had a small amount of pyrrhotite
(2%) and thus even though pyrrhotite might have oxidised sig-
nicantly, Figs. 5 and 6 show that pyrrhotite oxidation did not change
the amount of iron hydroxide in the slurries. Iron hydroxide might have
precipitated on gold surfaces, reducing the gold leaching performance.
Additionally, a signicant amount of the free cyanide (i.e. 75%) was
converted to ferrocyanide (see Fig. 11 and Eq. (14)) which could not
dissolve gold.
When the ceramic or high chromium media was employed, the gold
dissolution in the presence of cyanide was the highest probably due to
the lowest galvanic interactions. Nevertheless, the gold extraction was
about 50% only, even though 1% of free cyanide was converted to
ferrocyanide and 2.5% of free cyanide was converted to thiocyanate i.e.
the free cyanide did not decompose signicantly (see Fig. 11). The
possible explanation might be due to the passivation of gold grains by
pyrrhotite which signicantly reduced the leaching performance of
gold. It means that the understanding of cyanide gold leaching me-
chanism is challenging due to the complex interactions of various fac-
tors such as galvanic interactions, cyanide consumptions, and gold
passivation.
It is well known that pyrrhotite-containing gold ores are often
preoxidised prior to cyanidation, which will change and probably
dampen the dierences in leach response noted in this research due to
the formation of iron hydroxide and gypsum which can coat residual
suldes and gold. However, it serves as a useful baseline to which one
can compare less reactive iron suldes such as arsenopyrite and pyrite.
Like otation (Rabieh et al., 2017a), cyanidation serves as a diagnostic
of media-mineral interactions.
3.2.3. Cyanide speciation during gold leaching
The cyanide speciation was performed on the solution samples
collected during leaching (at 2, 4, and 24 h). As can be seen in Fig. 11,
when the forged steel or 21% chromium media were used, around 75%
and 44% of the free cyanide was consumed by iron hydroxide and
pyrrhotite to form ferrocyanide, respectively (see Eq. (14)). For that
reason, the gold cyanide leaching was less ecient when the forged
steel or 21% chromium media were used. It should be noted that a small
amount (< 1%) of free cyanide was converted to thiocyanate as seen in
Fig. 11.
By contrast, when the ceramic media or 30% chromium grinding
media were used, only 1% of free cyanide was converted to ferrocya-
nide because the amount of produced iron hydroxide during grinding
120
Hydrometallurgy 173 (2017) 115124
with ceramic and 30% chromium grinding media were insignicant.
The amount of converted free cyanide to thiocyanate was also negli-
gible. As a result, when the ceramic or 30% chromium grinding media
were used, the free cyanide almost did not decompose and thus cyanide
gold leaching was more ecient than that when grinding was per-
formed with the forged media or 21% chromium media.
3.3. Electrochemical properties of grinding media and pyrrhotite during
grinding
3.3.1. Open circuit potentials (OCP)
The open circuit potential is one of the important electrochemical
properties which determine the role of grinding media and sulde
minerals in the galvanic couple when exposed to the same environment.
The open circuit potential of the grinding media (i.e. forged steel or
30% chromium grinding media) and pyrrhotite electrodes were mea-
sured in situ in the mill. Fig. 12 shows the open circuit potentials of the
forged steel, 30% chromium grinding media and pyrrhotite electrodes
during grinding of the gold-pyrrhotite synthetic ore. The forged steel
showed much lower open circuit potential than pyrrhotite, demon-
strating that the forged steel behaved as anode with regards to pyr-
rhotite. A similar observation was seen in the case of the 30% chro-
mium grinding media. However, the 30% chromium grinding media
had slightly higher open circuit potential than the forged steel grinding
media.
As seen in Fig. 12, the open circuit potential of grinding media and
pyrrhotite electrodes decreased in the following order:
Forged steel> 30%chromium > pyrrhotite.
3.3.2. Polarization measurements
Due to the dierence in open circuit potentials of grinding media
and pyrrhotite, a galvanic interaction occur during grinding. In this
study, the galvanic current between grinding media and pyrrhotite was
determined from polarization curves of the electrodes. Fig. 13 shows
how to determine the galvanic current from the polarization curves of
the forged steel and 30% chromium during grinding. The dash line
show the galvanic current through pyrrhotite couple with dierent
grinding media during grinding. It can be seen that pyrrhotite had a
signicantly higher galvanic current with the forged steel than that
with the 30% chromium; the presence of chromium in the grinding
media reduced the galvanic current between the grinding media and
pyrrhotite.
3.4. Correlation between oxidised iron and galvanic current measurements
3.4.1. Theoretical model
It is well known that the galvanic interactions between sulde mi-
nerals and grinding media during grinding increase the corrosion rate
of grinding media (Huang and Grano, 2005; Rabieh et al., 2017a,
2017b). The higher the corrosion rate of grinding media, the more iron
hydroxide in the mill discharge are produced.
Huang and Grano (2006) have developed a model to describe the
correlation between the galvanic current between grinding media and
sulde minerals and the iron hydroxide (i.e. the EDTA extractable iron).
According to this model, in a grinding system, the amount of EDTA
extractable iron (QFe(t)) in the slurry is produced from the grinding
media and the sulde mineral. Additionally, Q0 is also included in the
model; Q0 represents iron hydroxide identied on the surface of
grinding media and the mineral samples before grinding of the ore. The
total amount of oxidised iron (QFe(t)), for a given time t, represents the
sum of the oxidation amount of grinding media and the mineral, and is
expressed by:
A. Rabieh et al. Hydrometallurgy 173 (2017) 115124

Fig. 9. Pulp chemical conditions during cyanide leaching as a

function of grinding time.

AMd t 1 t grinding media and the cathodic current (the last two terms in the Eq.
QFe (t) = Q0 +
n1F
0 Ig,Md (t)dt +
n2 F
0 A 0,Mn I0,Mn (t)dt
(15)) are very low and can be neglected. Therefore, Eq. (15) becomes:
AMd t 1 t
+
n1F
0 i c,Md (t)dt +
n2 F
0 A g,Mn ia,Mn (t)dt
(15) QFe (t) = Q0 +
AMd
0
t
Ig,Md (t)dt +
1
0
t
A 0,Mn I0,Mn (t)dt
n1F n2 F (16)
where Ig,Md (t) is galvanic current of grinding media, Ag,Mn is the con-
Moreover, the average amount of EDTA extractable iron (QFe(t))
tacted surface area of a sulde mineral, AMd and A0,Mn are the surface
area of grinding media and the particles, respectively. I0,Mn(t) is the
exchange current density on the mineral surface. ic,Md (t) is the cathodic

during grinding can be measured from the average galvanic current

( IgAMd ), the average exchange current density of the mineral
( IA 0,Mn ) and the grinding time (tgrind).
current on the grinding media surface while ia,Mn(t) is the anodic cur-
rent on the mineral surface; n1 and n2 are the number of moles of AMd A 0,Mn
electrons transferred per mole of iron and pyrrhotite oxidation, re- QFe (t) = Q0 t grind + IgAMd + IA 0,Mn
n1F n2 F (17)
spectively. F is the Faraday constant i.e. 96,500C/mol.
Huang and Grano (2006) suggested that during grinding, the oxi- As can be seen in Fig. 7, the oxidised iron from the grinding media
dation rate of the sulde mineral due to galvanic interaction with govern the EDTA extractable iron in the mill discharge. Therefore, the

121
A. Rabieh et al.
Fig. 10. Eect of grinding media on gold-leaching in the presence of pyrrhotite: (solid
line) forged steel media; (dash line) 21% chromium media; (dash dot) 30% chromium;
(dot line) ceramic media.
amount of oxidised iron due to the current density on the mineral
surface can be neglected. As a result, the amount of oxidised iron is
linear with the galvanic current as seen in Eq. (18):
AMd
QFe (t) = Q0 t grind + IgAMd
n1F
3.4.2. Empirical correlation between iron oxidation and galvanic current
The theoretical model (see Eq. (18)) was used to explain the cor-
relation between iron oxidation and the galvanic current. Fig. 13 shows
how the galvanic current inuences the oxidation of grinding media
(i.e. forged steel and 30% chromium grinding media).
Fig. 14 shows the correlation between the galvanic current and iron
oxidation in the mill discharge during grinding of pyrrhotite. The oxi-
dation rate of grinding media is proportional to the galvanic current. It
means that the higher the galvanic current between the forged steel and
pyrrhotite, the larger amount of oxidised iron was produced. During
grinding, it was also observed that the linear relationship between iron
oxidation and galvanic current with the forged steel media was dif-
ferent to that with the 30% chromium media. The coecient of the
galvanic current with the forged steel media is 2 times higher than that
with the 30% chromium media. The higher coecient of galvanic
Fig. 11. The amount of free cyanide, ferrocyanide and thiocyanate in slurry as a function of leaching time: (solid line) forged steel media; (dash line) 21% chromium media; (dash dot)
30% chromium; (dot line) ceramic media.
Hydrometallurgy 173 (2017) 115124
Fig. 12. Open circuit potentials of forged steel, 30% chromium and pyrrhotite electrodes
as a function of time.
current with the forged steel grinding media may be related to dierent
physico-chemical properties of the grinding media.
4. Conclusions
In this paper, the inuence of the grinding media on cyanide gold
leaching in the presence of pyrrhotite was investigated. The oxidation
of grinding media was quantied using EDTA extraction method and
dissolved sulfur measurements while cyanide species was determined
using high-performance liquid chromatography. A 2% pyrrhotite con-
(18) tent in ore was used to compare to similar weight percentage levels of
other iron sulde minerals in other research (Rabieh et al., 2017a,
2017b). The following results have been observed:
i. Pulp chemistry data revealed that grinding with forged steel media
resulted in a greater reducing environment in comparison with the
high chromium and ceramic media. The reason is probably the
galvanic interactions between the forged steel media and pyrrhotite.
The degree of oxidation of forged steel media was the highest while
the opposite is true in the case of the ceramic grinding media as
conrmed with the EDTA extraction and dissolved sulfur measure-
ments.
ii. When the ceramic media was used, the gold extraction was about
50% even though free cyanide levels and dissolved oxygen levels of
around 8 ppm were maintained. The possible explanation for re-
duced leaching performance is probably the passivation of gold
grains by pyrrhotite or the precipitation of iron hydroxide on gold
122
A. Rabieh et al.
Fig. 13. Polarization curve of forged steel, 30% chromium and pyrrhotite at dierent times. Measured times at (a) 20 min; (b) 40 min; (c) 60 min; (d) 80 min.
Fig. 14. Correlation between the oxidation of grinding media and galvanic current be-
tween pyrrhotite and grinding media (QFe is the average amount of grinding media
oxidation, mol/min; Ig is the average of galvanic current A/cm2).
particles. It should be noted that iron hydroxide was produced
during grinding of the ore.
iii. When the forged steel or 21% chromium media were used, the cy-
anide gold leaching was less ecient than that when the ceramic
media or 30% chromium was used. It is thought that this is due to
the formation of iron hydroxide during grinding of the investigated
synthetic ore with the forged media; iron hydroxide reacted with
free cyanide and thus most of the free cyanide was converted to
ferrocyanide which cannot dissolve gold. Industrially, it is expected
that ores containing signicant amounts of pyrrhotite will be pre-
oxidised which is expected to alter the impact of the grinding media
123
Hydrometallurgy 173 (2017) 115124
on the gold leaching. However it was not the objective in this re-
search to look into the altering eects of preoxidation upon
leaching, but rather to provide a benchmark for comparison to other
iron suldegrinding media interactions.
iv. In many ways, the 30% chromium media behaved more like the
ceramic media than like the 21% chromium media with regards to
the downstream impacts on leaching.
v. The selection of appropriate grinding media is far from obvious as
Rabieh et al. (2017a) demonstrated in which case it was found that
forged media gave marginally better otation recoveries of gold and
pyrite in contact with gold-bearing pyrite ores. The leaching or
otation response of gold-bearing iron sulde minerals is therefore
strongly dependent on the grinding media metallurgy and the mi-
neralisation of sulde minerals in the ore.
Acknowledgement
Financial support from the AMIRA International P420F Project and
the member companies (AngloGold Ashanti, Australian Gold Reagents,
Barrick Gold Corporation, Gekko Systems, Lhoist, Newcrest Mining,
Newmont Asia Pacic, Northern Star Resources, Orica Australia, CRC
ORE, Pionera and Vega Industries) is gratefully acknowledged, as well
as scholarship support from Curtin University.
References
Adam, K., Natarajan, K.A., Iwasaki, I., 1984. Grinding media wear and its eect on the
otation of sulde minerals. Int. J. Miner. Process. 12 (13), 3954.
Cullinan, V.J., Grano, S.R., Greet, C.J., Johnson, N.W., Ralston, J., 1999. Investigating
ne galena recovery problems in the lead circuit of Mount Isa mines lead/zinc con-
centrator, part 1: grinding media eects. Miner. Eng. 147163.
Deschenes, G., Hall, D., Fulton, M., 2003. Cyanidation of a pyrrhotite-bearing gold ore.
Eur. J. Miner. Process. Environ. Prot. 3, 353361.
Fontana, M.G., 1987. Corrosion Engineering, Third edition. McGraw-Hill International
A. Rabieh et al.
Edition, New York.
Huang, G., Grano, S., 2005. Galvanic interaction of grinding media with pyrite and its
eect on otation. Miner. Eng. 18 (2005), 11521163.
Huang, G., Grano, S., 2006. Galvanic interaction between grinding media and arseno-
pyrite and its eect on otation. Part I. Quantifying galvanic interaction during
grinding. Int. J. Miner. Process. 78 (2006), 182197.
Marsden, J.O., House, C.I., 1992. The Chemistry of Gold Extraction. Ellis Horwood,
Chichester, England.
Martin, C.J., McIvor, R.E., Finch, J.A., Rao, S.R., 1991. Review of the eect of grinding
media on otation of sulphide minerals. Miner. Eng. 4 (2), 121132.
124
Hydrometallurgy 173 (2017) 115124
Ntemi, A., Terytze, M., Schneider, M., 2013. An Evaluation of the Current Situation of
Cyanide Waste Disposal and Treatment Methods. PhD Thesis Freie Universitt, Berlin.
Rabieh, A., Albijanic, B., Eksteen, J.J., 2016. A review of the eects of grinding media and
chemical conditions on the otation of pyrite in refractory gold operations. Miner.
Eng. 94, 2128.
Rabieh, A., Eksteen, J.J., Albijanic, B., 2017a. Inuence of grinding media and water
quality on otation performance of gold bearing pyrite. Miner. Eng. 112, 6876.
Rabieh, A., Eksteen, J.J., Albijanic, B., 2017b. Galvanic Interaction of Grinding Media
With Arsenopyrite and Pyrite and Its Eect on Gold Cyanide Leaching. Minerals
Engineering (Under Review).