Minerals Engineering 55 (2014) 87–95

Contents lists available at ScienceDirect

Minerals Engineering
journal homepage: www.elsevier.com/locate/mineng

The influence of pyrite content on the flotation

of chalcopyrite/pyrite mixtures
Clement Owusu ⇑, Susana Brito e Abreu, William Skinner, Jonas Addai-Mensah, Massimiliano Zanin
Ian Wark Research Institute, The ARC Special Research Centre for Particle and Material Interfaces, University of South Australia, Mawson Lakes, Adelaide, SA 5095, Australia

a r t i c l e i n f o a b s t r a c t

Article history: In the flotation of copper ores, several processing plants report that copper recovery is affected by the
Received 2 July 2013 proportion and reactivity of pyrite in the ore, with the effect becoming more intense when the feed par-
Accepted 27 September 2013 ticles are finer as a result of regrinding. In this work, a mixed model mineral system consisting of chal-
Available online 20 October 2013
copyrite (CuFeS2) and pyrite (FeS2) with varying pyrite content (20–80 wt.%) was used to investigate the
effect of pyrite on the pulp chemistry and chalcopyrite recovery. Flotation tests showed that chalcopyrite
Keywords: flotation rate, recovery and grade, as well as the pulp oxidation potential, decreased with increasing pyr-
Chalcopyrite
ite content whilst pyrite recovery increased. Surface analysis (XPS, ToF-SIMS and EDTA) indicated that
Pyrite
Pulp redox potential
copper activation of pyrite increased with increasing pyrite content, facilitating pyrite recovery. The
Dissolved oxygen decrease in chalcopyrite recovery can be attributed to increased surface oxidation.
Flotation Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction regrinding to very small particle size is commonly performed

The most abundant mineral resource of copper is the sulphide
minerals. Examples of such minerals include chalcopyrite (CuFeS2),
bornite (Cu5FeS4), chalcocite (Cu2S) and covellite (CuS). In recent
times, it is very rare to find copper deposits with head grade aver-
aging more than 2% Cu (Davenport et al., 2002). In both porphyry
and skarn ores, the copper grade is lowered by the presence of
other sulphide and non-sulphide gangue minerals such as pyrite
(FeS2), sphalerite (ZnS), and quartz (SiO2) (Aplan and Chander,
1988; Kydros et al., 1994). Chalcopyrite, a major resource for cop-
per production, is not exceptional; it is often found in association
with iron sulphides, particularly pyrite, and minor amounts of
sphalerite and molybdenite (Habashi, 1978; Wang and Forssberg,
1991). Recent studies indicate that selective separation of chalco-
pyrite from pyrite is very difficult due to several electrochemical
interactions that occur at the minerals surfaces during the grinding
and conditioning processes (Finkelstein, 1997). For instance, gal-
vanic interactions between pyrite and chalcopyrite surfaces prior
to flotation may lead to the oxidation and dissolution of the chal-
copyrite and copper activation of the pyrite. The formation of
hydrophilic surface coatings of iron hydroxide which depress chal-
copyrite flotation may also occur. Copper activation of pyrite fa-
vours collector adsorption on to the surface, enhancing its
floatability and subsequently lowering the overall chalcopyrite
grade (Xu et al., 1995; Peng et al., 2003). To improve mineral liber-
ation and increase chalcopyrite flotation recovery and grade,
⇑ Corresponding author. Tel.: +61 8 8302 3714; fax: +61 8 8302 3683.
E-mail address: owucy001@mymail.unisa.edu.au (M. Zanin).
(Gaudin, 1939; Sutherland, 1989). However, in most cases, an
increase in valuable mineral recovery is not observed upon
regrinding, implying that size and liberation changes are not the
only key influential factors determining the flotation performance
(Frew et al., 1994). Ye et al., 2010 showed that the surface chemis-
try of minerals (which impacts on the optimum mineral separa-
tion) may change significantly after regrinding, depending upon
the type of mill and grinding media used. The effect of grinding
media and environment on chalcopyrite flotation has been re-
ported in previous studies (Van Deventer et al., 1991; Grano,
2009; Bruckard et al., 2011; Peng and Grano, 2010). These studies
revealed that grinding with mild steel media yields lower chalco-
pyrite flotation recovery compared with stainless steel, which is
attributed to an increase in the formation of iron hydroxide species
on the chalcopyrite surface. Arguably, it is difficult to deconvolute
the effect of pyrite and steel grinding media on chalcopyrite as
both serve as sources of iron ions released into the pulp solution.
Recently, Pease et al. (2006) showed that when fully inert ceramic
medium is used in the regrinding of the ore, the effect of iron con-
tamination was noticeably reduced, improving the selectivity
against sulphide gangue minerals and increasing chalcopyrite flo-
tation recovery. Recent reports from various processing plants
show that chalcopyrite flotation recovery after regrinding is af-
fected by the proportion and type of pyrite in the ore, the effect
becoming more pronounced when the feed ore particles are fine
(<20 lm). Therefore, the aim of this work is to extract the exact
influence of pyrite content on the chalcopyrite recovery, grade,
and selectivity after regrinding using an IsaMill with fully inert
ceramic medium. Model mineral systems comprising of pyrite

0892-6875/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.mineng.2013.09.018
88 C. Owusu et al. / Minerals Engineering 55 (2014) 87–95
and chalcopyrite in different proportions (20–80 wt.%) are used for
the study. Surface species formed on the mineral surfaces during
grinding and flotation were determined using ToF-SIMS, XPS and
EDTA extraction to provide fundamental understanding on the spe-
cies responsible for the flotation performance of the two minerals.
2. Experimental
2.1. Materials and reagents
Chalcopyrite (Cp) (Mannum minerals) and pyrite (Py) (Peruvian
mine) minerals of particle size ranges 400–2000 lm and
600–2000 lm, respectively were used for this study. The chemical
composition of the two samples is reported in Table 1. It is noted
that, from stoichiometric calculations, the Cp sample contains
about 75% Cp, 4% Si and about 9% of non-sulphide iron minerals.
The Py sample contains about 97% Py and virtually no copper sul-
phide. Blends of the two samples, in different ratios, have been
used in this study.
Ceramic beads (Ø 3.5 mm) (Xstrata Technology, Australia) were
employed as the regrinding media for the IsaMill. Sodium isopro-
pyl xanthate (SIPX) and Dowfroth 250 (polypropylene glycol) were
used as a collector and a frother, respectively. Analytical reagent
grade lime and demineralised water were used throughout the
whole experiment for pH/pulp conditioning and grinding purposes,
unless otherwise stated.
2.2. Experimental procedures
2.2.1. Mineral grinding and flotation
Samples of 200 g of Cp single mineral or Cp/Py mixtures com-
prising different proportions of Py and Cp (20, 50 and 80 wt.% Py
addition) were mixed with 0.3 dm3 of demineralised water and
lime and ground with grinding medium (stainless steel rods,
10  22 mm and 6  16 mm) in a Galigher laboratory mill to ob-
tain a d80 of 105 lm at pH 10. The mill product was transferred
into a 1.5 dm3 flotation cell (Agitair Model LA-500R), pH adjusted
to 10.5 with lime and conditioned with xanthate collector
(200 g/t SIPX) for 2 min. The conditioned pulp was reground to
a d80 of 20 lm in a 1 dm3 laboratory attrition mill (IsaMill). After
regrinding, the ground pulp was quickly removed from the mill
and transferred back into the 1.5 dm3 Agitair flotation cell (in or-
der to retain the pulp potential and the dissolved oxygen level at
or close to the value in the mill) and pH adjusted to 10.5. The
pulp was then conditioned with 50 g/t of SIPX and frother
(25 g/t Dowfroth 250) for 2 and 1 min, respectively. A schematic
diagram of the flotation procedure is presented in Fig. 1. Four
concentrates were collected at time intervals of 1, 2, 3 and
4 min. The dry masses of the four concentrates and tailings were
determined and chemical assays were undertaken for determina-
tion of their elemental compositions.
The impeller speed (1000 rpm), pH (10.5) and air flow rate
(2.5 dm3 min1) were maintained during conditioning and flota-
tion in all experiments. Pulp oxidation potential (Eh) and dissolved
oxygen level (DO) after IsaMilling prior to collector addition were
measured at room temperature using a TPS 90-FLMV meter
(Fig. 1). The pulp pH was maintained at 10.5 to mimic plant
Table 1
Chemical composition of chalcopyrite and pyrite samples used for the study.
Mineral Elements (mass %)
Zn Fe S Ca Cu Si Pb Mn
Chalcopyrite 0.01 28.9 29.0 1.76 26.4 4.2 0.34 0.02
Pyrite 0.18 44.9 53.5 0.44 0.13 0.7 0.02 0.02
conditions. The pulp Eh was expressed relative to the standard
hydrogen electrode (SHE). In some tests the amount of collector
added was kept constant regardless of the amount of Cp in the
mineral mixture (200 g/t after primary grinding and 50 g/t after
regrinding), while in other tests, it was scaled to the amount of
Cp in the mixture, (200 g/t of Cp after primary grinding and 50 g/
t of Cp after regrinding). This is discussed further below.
2.2.2. Flotation data analysis
After chemical analysis, the elemental assays for Cu and Fe for
each sample were converted into mineral compositions using the
following mathematical relations based on the stoichiometric com-
positions of Cp and Py:
ð%CpÞ ¼ %Cu  ð183:3=63:5Þ ðChalcopyriteÞ ð1Þ
%Fe ¼ %Fe  %Cu  ð55:8=63:5Þ ðIron present as pyriteÞ ð2Þ
%Py ¼ %Fe  ð1119:8=55:8Þ ðPyrite in sampleÞ ð3Þ
The metallurgical balance of each test was also performed with
respect to Cp and Py, and recovery versus time data were fitted to a
first order rate equation:
R ¼ Rmax ð1  ekt Þ ð4Þ
where R is the recovery at time t, Rmax is the maximum recovery at
infinite time and k is flotation rate constant. From Eq. (4), Rmax and k
were inferred for the two minerals.
2.2.3. Subtraction of entrainment
In this work, mineral particles recovered by true flotation and
entrainment were determined using the method developed by
Ross (1991). A simplified solution of the method developed by Ross
can be obtained by dividing the total flotation time in time inter-
vals (corresponding to the collection of the four concentrates). In
a short time interval Dt, the recovery by entrainment, as a percent-
age of the feed, can be written as:
MiðDtÞ
EiDt ¼ XðDtÞ  RW ðDtÞ  ð5Þ
Mifeed
where RW(Dt) is the recovery of water in the time interval, Mifeed is
the mass of mineral in the flotation feed and Mi(Dt) is the average
mass of mineral in the cell during the time interval. X(Dt) is a trans-
fer or entrainment factor, which is particle size dependent (Ross
and Van Deventer, 1988). The entrainment factor was estimated
from Eq. (5) with the assumption that the recovery of the hydro-
philic mineral (which in this study is pyrite) in the last concentrate
(i.e. the last Dt) is purely by entrainment, EPy(Dt) = RPy(Dt). This
method is a discretisation of the method developed by Ross
(1991), which would require numerical solution. It has been tested
and proved to show negligible differences from the Ross method.
Recovery by entrainment was subtracted from the chalcopyrite
and pyrite recovery data to give the recovery by true flotation.
2.3. Techniques
2.3.1. EDTA extraction
Ethylenediamine tetra acetic acid disodium salt (EDTA)
extraction followed by inductive coupled plasma (ICP) analysis
was performed to measure the amount of extractable metal
oxide/hydroxide Cu and Fe at the mineral particle surface. A vol-
ume of 190 cm3 of a 3 wt.% solution of AR grade EDTA was pre-
pared and the pH was adjusted to 7.5 using AR grade sodium
Mg
hydroxide solution. The solution was then placed in a continu-
0.68
ously-stirred reaction vessel and purged with nitrogen for
0.08
10 min. 10 cm3 of slurry was collected immediately after
 C. Owusu et al. / Minerals Engineering 55 (2014) 87–95
Fig. 1. Flowsheet of the experimental flotation procedure.
regrinding with the IsaMill and placed in the EDTA solution for
5 min (while purging with nitrogen). The sample was then filtered
through a 0.45 lm millipore membrane and the extract analysed
using inductively coupled plasma (ICP) atomic emission spectro-
photometry. The percentage of EDTA extractable iron or copper,
an indicator of mineral surface oxidation, was expressed as the
ratio of EDTA soluble copper or iron to the total amount of copper
or iron in their insoluble forms in the samples (wt./wt.% Cu or
wt./wt.% Fe).
2.3.2. X-ray photoelectron spectroscopy (XPS)
The surface atomic concentration of species on the mineral sur-
faces was examined by XPS. Samples of regrinding mill (IsaMill)
discharge with d80 of 20 lm for the 20, 50 and 80 wt.% Cp/Py mix-
ture were subjected to analysis. A Kratos Axis Ultra X-ray photo-
electron spectroscope (XPS) was employed with an Al Ka
monochromated X-ray source (1486.8 eV) operating at 130 W. At
the initial stage, the samples were examined in a survey mode to
identify all the elements on the surface to determine their atomic
concentration after which various elemental regions were scanned
for information on the chemical bond and oxidation states. The re-
sults are presented as surface weight concentration and XPS
spectra.
2.3.3. Time-of flight secondary ion mass spectrometry (ToF-SIMS)
Samples of flotation concentrates (con. 1–4) from the test of 20
and 80 wt.% Py additions were washed with NaOH solution (pH
10.5) to remove any suspended material (i.e., very fine and/or col-
loidal particles). The mineral particles (chalcopyrite/pyrite) were
wet mounted on to indium foil on a silicon plate and introduced
in the fore-vacuum chamber of the ToF-SIMS instrument. A ToF-
SIMS instrument of the model PHI TRIFT V nanoToF from Physical
Electronics Inc. was used in this study. A pulsed liquid metal 79Au+
primary ion beam operating at 30 kV was used for the analysis. The
analysis was conducted under the high imaging resolution mode
and charge neutralisation was provided by an electron gun and a
gun for Ar+ ions. ‘Raw’ data files for positive (e.g. including Cu+,
FeOH+, etc.) and negative (e.g. including O, C2 H5 OCS  
2 , SO3 , SH ,
etc.) secondary ions were acquired for images of 256  256 pixels
and raster size 75–100 lm. Data were acquired for 2 min with the
ion dose below the static limit (<1012 ions/cm2). The positive and
negative mass spectra of regions of interest (i.e., particles) were ex-
tracted offline using the WinCadenceN software. The intensity of
the secondary ion signals was normalised to the total ion yield of
selected peaks.
3. Results and discussion
3.1. Flotation results
3.1.1. Effect of pyrite content on chalcopyrite flotation
The flotation experiments for the various Cp and Py blends were
repeated 3 times and the standard deviation on the recovery values
89
was found to be approximately 2%. All recoveries are indicated
after subtraction of entrainment (refer to Section 2.2.3). The effect
of Py content on Cp flotation recovery after IsaMilling is shown in
Fig. 2 (tests at constant SIPX addition rate) and 3 (tests with SIPX
addition rate scaled to the amount of Cu in the mixture). This
was done to deduce the effects of the following two mechanisms:
(i) surface modifications on chalcopyrite (oxidation) and reduced
floatability and (ii) copper activation of pyrite and subtraction of
collector available for chalcopyrite.
Figs. 2 and 3a show that the amount of pyrite in the feed has
significant impact on the chalcopyrite recovery, grade and selectiv-
ity. For the tests with constant collector addition rate (Fig. 2a), an
increase in Py content from 0 to 80 wt.% resulted in a decrease in
Cp recovery (after 10 min of flotation) from 96% to 82% and a
decrease in the flotation rate constant from 0.94 to 0.45 min1
(Table 2). In the tests at scaled collector addition rate (Fig. 3a),
Cp recovery and flotation rate constant decreased from 96% to
56% and from 0.94 to 0.39 min1, respectively. Similarly, the recov-
ery of Py gradually decreased as the amount of Py present in the
mixture increased (Figs. 2 and 3b). For example, in Fig. 2b, Py
recovery decreased from 33% to 9% when the Py content in the feed
was increased from 20 to 80 wt.%. The Cp grades for the various
Cp/Py mixtures are presented in Table 2. The maximum achievable
concentrate grade and recovery decreased with increasing Py
content in the flotation feed. This was associated with a decrease
in Cp flotation kinetics.
The major difference between the results in Figs. 2 and 3 is the
lower Cp recoveries obtained in the tests where the collector addi-
tion rate was scaled to the amount of Cu in the mineral mix, espe-
cially at high percentages of pyrite added. The second difference is
that Py recoveries are generally lower and almost independent of
the percentage of Py in the feed when the collector addition rate
is scaled to the amount of Cu in the ore mix. This may suggest that
Py competes with Cp for adsorption of collector, and that collector
dosage is therefore critical. These results are in good agreement
with those obtained in most copper processing plants where the
recovery of copper after regrinding is affected by the proportion
of pyrite in the ore. The results also correlate well with those ob-
tained by Ekmekci and Demirel (1997) who also showed that the
collectorless flotation behavior of Cp and Py mixtures was con-
trolled by the proportion of both minerals in the mixture.
3.2. The role of pulp redox potential (Eh) and dissolved oxygen (DO)
Pulp electrochemistry measurements at the start of condition-
ing (Table 2) show that an increase in Py content leads to a
decrease of the pulp redox potential (Eh) after fine grinding. For
example, at 20 wt.% Py addition (SIPX scaled to Cu in the feed)
the Eh was 203 mV (SHE) whilst at 80 wt.% Py addition the Eh
decreased to 176 mV (SHE). Although the decrease in Eh is not
large, the dissolved oxygen in the pulp was remarkably reduced
with increased pyrite in the mix. The decrease in the pulp DO/Eh
suggests the creation of a more reductive environment as more
90 C. Owusu et al. / Minerals Engineering 55 (2014) 87–95

(a) 100 (b) 100

80 80 20 wt.% Py add.
50 wt.% Py add
80 wt.% Py add

% Py recovery
%Cp recovery

60 60

40 0 wt.% Py add. 40
20 wt.% Py add.
50 wt.% Py add.
80 wt.% Py add.
20 20

0 0
0 2 4 6 8 10 0 2 4 6 8 10

Time ( t, min) Time (t, min)

Fig. 2. Cp (a) and Py (b) recovery versus time as a function of the pyrite content in the mix ([SIPX] = 250 g/t in all tests).

(a) 100
(b) 100

80 80
% Cp recovery

% Py recovery

60 60
20 wt.% Py add.
50 wt.% Py add.
80 wt.% Py add.
40 40
0 wt.% Py add.
20 wt.% Py add.
50 wt.% Py add.
20 20
80 wt.% Py add

0 0
0 2 4 6 8 10 0 2 4 6 8 10
Time (t, min) Time (t, min)

Fig. 3. Cp (a) and Py (b) recovery verse time as a function of the pyrite content in the mix ([SIPX] = 250 g/t of Cp in the feed).

Table 2
Summary of pulp electrochemistry and flotation results.

Experiments
Mixtures (Cp-Py) Eh DO Rate constant (min1) Recovery (%) Grade (%)
pH = 10.5 (mV, SHE) ppm Cp Py Cp Py Cp Py
Equal amount of collector addition after IsaMilling (250 g/t SIPX)
0% Py 248 7.42 0.94 – 96 – 88 –
20% Py 240 7.03 0.90 0.48 94 33 68 23
50% Py 195 5.60 0.82 0.32 90 27 63 33
80% Py 190 4.00 0.45 0.14 80 9 41 52
Collector addition after IsaMilling scaled to copper (250 g/t SIPX of Cp)
20% Py 203 8.4 0.85 0.26 91 11 82 14
50% Py 190 6.1 0.64 0.21 72 8 65 25
80% Py 176 4.5 0.39 0.13 56 5 47 44

Py is introduced into the system. Pyrite is oxidative in nature;
therefore, increasing pyrite content subsequently increases the
oxidative reactions occurring at the mineral surface, thereby
decreasing dissolved oxygen levels in the flotation pulp, as shown
in Eq. (6). This significantly lowers the oxidation and adsorption
rate of the collector on to the valuable mineral surface, thereby
impacting on its flotation recovery (He et al., 2006; Holmes and
Crundwell, 2000). The decrease in Cp recovery with increasing Py
content (Figs. 2 and 3a) is associated with the decrease in the pulp
DO/Eh; it has been shown in previous studies that maximum Cp
flotation is reached at Eh values around 250–275 mV (SHE) (e.g.
Richardson and Walker, 1985; Goktepe and Williams, 1995; Kuo-
panportt et al., 2000; He et al., 2006).
FeS2 þ 15=4O2 þ 7=2H2 O () FeðOHÞ3 þ 2H2 SO4 ð6Þ
To prove that the change in Cp recovery results from changes in
the pulp DO/Eh associated with increasing Py content in the min-
eral mixture, an experiment was performed in which the pulp
 C. Owusu et al. / Minerals Engineering 55 (2014) 87–95
was deliberately purged with air to increase the pulp DO/Eh in one
of the tests (80 wt.% Py addition) prior to flotation. Aeration of the
pulp for 48 min increased the oxygen concentration from 4.5 to
5.9 ppm and Eh from 176 to 335 mV (SHE). As a result, the Cp flo-
tation rate increased from 0.39 to 0.66 min1 and the recovery in-
creased from 54% to 72% (Fig. 4) while Py recovery decreased by
1%. The increase in Cp flotation with Eh may be due to xanthate
adsorption/dixanthogen formation at the Cp surface (Richardson
and Walker, 1985; He et al., 2006).
3.3. Solution and surface analysis
3.3.1. EDTA extraction and XPS study
Feed samples obtained after regrinding (IsaMill product) for all
the Cp/Py mixtures were analysed using EDTA extraction and XPS
to identify and quantify species formed on the minerals surfaces.
Table 3 shows that the relative amount of EDTA extractable Cu ions
(CuEDTA/CuT) increases from 5.1% to 8.7% with increasing Py con-
tent in the mixture from 20 to 80 wt.%, while the relative amount
of EDTA extractable Fe remains approximately constant at 2.5%.
The increase in EDTA extractable Cu suggests oxidation of Cp
(e.g. formation Cu(OH)2 on the surface). The decrease in DO/Eh as
the amount of Py increases is related to the reduction of more oxy-
gen on the pyrite surface during oxidation (Eq. (6)). The mineral
surface and bulk weight percentage of elements for different
Cp/Py mixtures as measured by XPS are shown in Table 4. To have
an idea of the alterations in the composition of elements on the
minerals surface, surface to bulk ratio calculations for Cu, Fe and
S were determined (Table 4). The surface to bulk ratio of Cu in-
creased from 1.2% to 2.6% whilst that for Fe and S remained similar
as the Py content in the blend increased from 20 to 80 wt.%. The
over-abundance of Cu on the surface with respect to the bulk is
most likely the result of the transfer of Cu ions from Cp on to Py
surface (copper activation of Py). The data suggest that copper acti-
vation of pyrite increases with increasing pyrite content in the
mineral mix. Copper activation of pyrite promotes the adsorption
of xanthate collector on to the pyrite surface, enhancing its flota-
tion under alkaline conditions (Finkelstein, 1997; Xu et al., 1995;
Shen et al., 2001)
The Cu2p spectra of the XPS analysis (Fig. 5a) show little evi-
dence of any intensity in the region around 945 eV where charac-
teristic Cu (II) ‘‘shake-up’’ satellites are found (due to energy loss
incurred by the outgoing Cu2p photoelectrons interacting with,
and promoting, valence electrons into vacant Cu d states) (Skinner
et al., 1996; Prestidge et al., 1997). The surface copper is mainly in
100
80 No aeration
With aeration
Cp recovery (%)
60
40
20
0
0 2 4 6 8 10
Time (min)
Fig. 4. Effect of pre-aeration on Cp flotation recovery (80 wt.% Py, [SIPX] = 250 g/t of
Cp in the feed).
91
a cuprous state with location of most intense peak around 932 eV
representing Cu (I) (Fairthorne et al., 1997). The lack of Cu (II) on
the surface, combined with the EDTA extraction Cu result, suggests
that all the Cu (II) is in solution. The Fe2p spectra for the various
Cp/Py mixtures (Fig. 5b) are very similar in shape, with a sharp
peak at 706.5 eV, characteristic of low-spin Fe (II) in pyrite, and a
broader contribution extending to about 714 eV. The latter in-
cludes the iron oxy/hydroxide species on the mineral surface as
well as a contribution from Fe in unaltered chalcopyrite (Fair-
thorne et al., 1997). Although the Fe2p spectra of the XPS suggest
oxy/hydroxide species formation, it is difficult to identify the exact
species formed on the individual minerals as an average of the ex-
posed mixed Cp/Py minerals surfaces is measured.
3.3.2. ToF-SIMS study
To understand the surface chemistry differences occurring on
the chalcopyrite and pyrite surfaces when they are present in dif-
ferent proportions during grinding and flotation, and their impact
on the flotation behaviour of the two minerals, ToF-SIMS surface
analysis was performed on flotation concentrates from the tests
at 20 and 80 wt.% Py additions. The information for the different
mineral phases in the samples was extracted individually for Cp
and Py particles. The presence of the two minerals can be visual-
ised by imaging for copper (Cu) and iron (Fe) signals. A typical
ToF-SIMS image of a concentrate sample is shown in Fig. 6, display-
ing the total ion (a), Cu (b) and Fe (c) imaging signals. Fig. 6(d) is an
overlay image showing the presence of Cu (green) and FeOH (blue)
on pyrite surfaces. The presence of Cu ions on Py particles confirms
that the Py was copper activated in the presence of Cp.
A statistical set of Py and Cp particles in each concentrate
stream (concentrates 1–4) was investigated. A large number of
grains (28 particles on average) was analysed to derive statistically
significant results. The intensities of the ToF-SIMS secondary ions
of total oxygen (O), copper (Cu), sulphur (in the form of SH) and
xanthate collector fragment (C2H5OCS2) were found to differ be-
tween the Cp and Py mineral surfaces. The trends of these signals
in each of the phases are presented in Fig. 7, for the mixed blends
Cp/Py: 80/20 and Cp/Py: 20/80. One should note that in the test
with 20 wt.% Py addition, the data for Py is presented only for con-
centrate 4; it was not possible to identify a statistically-significant
number of Py particles in the earlier concentrates. The distribution
of Ca ions was analysed with both XPS and ToF-SIMS but interest-
ingly, only a little amount of Ca ions was found on the minerals
surface.
Fig. 7 shows that the intensities of hydrophobic sulphur (in the
form of SH) and copper ion signals are higher on the Cp surfaces
than on Py surfaces. This behaviour is observed for the different
Py additions, although, when Py content is increased from 20% to
80%, the intensity of these signals on the Cp surfaces decreases
considerably. The intensity of the copper signal on Py particles is
approximately 10 times lower than that on Cp particles (for the
test with 80 wt.% Py addition), providing an idea of the extent of
copper activation of Py from contact with Cp in the pulp. The inten-
sity of the copper signal on Cp surfaces shows an interesting trend
throughout the flotation concentrates: decreases initially from C1
to C2, increasing from C2 to C3 and decreasing again from C3 to
C4. This behaviour may, possibly, be due to initial dissolution of
Cu ions into the solution following precipitation of copper species
(e.g. copper hydroxides) and re-dissolution. This trend is observed
for both 20 and 80 wt.% Py additions but the effect is more pro-
nounced for the test with 20 wt.% of Py. Although a speculation,
further investigation into this behaviour may help to better under-
stand the underlying mechanisms occurring during the flotation of
mixed minerals, especially when galvanic interactions occur. There
is a complex interplay between the species in the solution and on
the minerals surfaces and this varies during flotation.
92 C. Owusu et al. / Minerals Engineering 55 (2014) 87–95

Table 3
Results of EDTA extractable metal ions for the various Cp/Py mixtures after IsaMilling at pH 10.5.

Mixture (Cp–Py) Eh., mV (SHE) Cu Fe

Cu in sol., mg/l CuEDTA, mg/l CuEDTA/CuT, % Fe in sol., mg/l FeEDTA, mg/l FeEDTA/FeT,%
20% Py 241 0.39 53 5.1 0.28 42 2.6
50% Py 223 0.12 27 6.4 0.11 32 2.5
80% Py 210 0.04 23 8.7 <0.01 50 2.4

Table 4
Bulk and surface average weight percentage and surface to bulk ratio of elements on mineral surface for different Cp/Py mixtures.

Mixtures Bulk weight (%) Surface weight (%) Surface to bulk ratio (%)
Cu Fe S Cu Fe S Cu Fe S
20 wt.% Py 21.9 34.8 32.3 25.6 21.2 29.2 1.2 0.6 0.9
50 wt.% Py 13.1 38.6 40.3 19.2 21.0 32.3 1.5 0.5 0.8
80 wt.% Py 5.3 42.4 48.2 13.7 20.6 38.8 2.6 0.5 0.8

(a) x102 (b) x102

18
40 Cu (I) S Fe (II) S
16
Fe-Ox
35
14
30 12
CPS

CPS

25 10

8
20
6
15
4
965 960 955 950 945 940 935 930 925 735 730 725 720 715 710 705 700
Binding Energy (eV) Binding Energy (eV)
Fig. 5. Cu2p and Fe2p XPS spectra for the various Cp/Py mixtures (pH = 10.5; SIPX concentration scaled to % Cp in the feed (20 wt.% Py, 50 wt.% Py, 80 wt.% Py).

Another interesting finding is the amount of collector adsorbed
on the particles surfaces: a higher intensity of collector signal was
observed on Py particles compared to Cp for both Cp/Py mixtures.
This result shows that Cu-activated pyrite competes with chalco-
pyrite for collector adsorption, and therefore, explains the in-
creased Py recovery when the collector dosage was kept constant
during the flotation test. The collector adsorbed on Cp particles re-
mained approximately constant throughout the flotation concen-
trates in both mineral mixtures, though it appears to be slightly
higher for the test with 80 wt.% of Py. This observation is consistent
with the anecdotal observation that minimum levels of oxidation
are required for the collector to adsorb on to the mineral surface.
Since the collector adsorbed on chalcopyrite particles remains con-
stant, it does not explain, alone, the observed decrease in Cp recov-
ery when Py content is increased. The overall hydrophilic loading
on the particles surfaces appears to have a dominant effect on
the flotation behaviour of chalcopyrite.
The oxidation on the mineral surfaces was also examined. The
Py particles were found to exhibit higher exposure of oxygen per
unit surface area compared to Cp particles, regardless of their pro-
portion in the mineral mixture. By increasing the Py content in the
mixture, the oxidation of Cp is enhanced, due to the galvanic inter-
actions occurring between the two minerals. When Cp and Py min-
erals come into contact, a galvanic cell is established and current
flows from Cp to Py. Pyrite is the mineral with the highest rest po-
tential of all sulphides (less electrochemically active), and acts as
the cathode while chalcopyrite acts as the anode (Ekmekci and
Demirel, 1997). Hence, galvanic interactions will favour the reduc-
tion of oxygen (cathodic reaction, Eq. (7)) at the pyrite surface with
formation of hydroxyl species (Eq. (6)), and the oxidation of chal-
copyrite (anodic reactions), with formation of covellite and con-
comitant production of iron hydroxide and elemental sulphur
(Eq. (8)). High pH and Eh conditions in the pulp, will promote fur-
ther oxidation of copper and sulphur on the chalcopyrite surface
(Eq. (9)).
H2 O þ 1=2O2 þ 2e $ 2OH ð7Þ
CuFeS2 þ 3H2 O $ CuS þ FeðOHÞ3 þ S0 þ 3Hþ þ 3e ð8Þ
CuS þ 2H2 O $ CuðOHÞ2 þ S0 þ 2Hþ þ 2e ð9Þ
The oxidation of sulphide surfaces is a complex electrochemical
process, involving a number of intermediate electrochemical reac-
tions with the formation of sulphoxy species: SO, SO2, SO3 and SO4
(Rimstidt and Vaughan, 2003). In alkaline solutions, a combination
of these sulphur species may be expected on the surface of sul-
phide minerals. The sulphur oxidation on the surfaces of the Cp
and Py particles was examined in this study by following the SO 3
secondary ion signal. The exposure of SO3 ion on the surfaces of
these minerals and its variation throughout flotation is presented
in Fig. 8a. It is evident from the figure that SO3 species form in
higher amounts on Py particles compared to Cp, regardless of their
 C. Owusu et al. / Minerals Engineering 55 (2014) 87–95
Fig. 6. ToF-SIMS image of total ion (a), Cu (b) and Fe (c) secondary ion signals and overlay (d) (scale bar = 10 lm).
proportion in the mineral system. When the content of Py in the
feed increases (Py > Cp), the oxidation of Cp is enhanced, as evi-
denced by the greater exposure of SO 3 species on the Cp particles.
An interesting observation is that the formation of SO 3 on the sur-
faces of Cp and Py follows a similar trend throughout the flotation
concentrates, increasing in the course of flotation. This finding sug-
gests that the oxidation of sulphur follows the same mechanism/
rate on both minerals and is in agreement with the proposal that
oxidation of sulphide surfaces occurs via the same electrochemical
mechanism (Rimstidt and Vaughan, 2003). Increasing the content
of Py in the system will promote sulphur oxidation on both Cp
and Py particles surfaces.
Oxidation of Cp follows an oxidative dissolution mechanism;
the Cu and Fe ions dissolve into the solution leaving a hydrophobic
sulphur-rich or metal-deficient layer and/or elemental sulphur on
the surface. This phenomenon is generally accepted to impart sur-
face hydrophobicity and induce collectorless flotation of chalcopy-
rite. The dissolved Cu ions in solution will, in turn, adsorb on to the
pyrite surface, i.e., copper-activation of pyrite. It has been shown
that Cu (I) species interact with either a S2 or S on the pyrite sur-
face by chemical adsorption and that no iron dissolution from pyr-
ite occurs during this process. The overall reaction (adsorption and
oxidation) for the Cu-activation of pyrite can be represented by Eq.
(10) (von Oertzen et al., 2007). The Fe2+ ions released into the solu-
tion will rapidly, under strong alkaline and electrochemical envi-
ronments, oxidise to Fe3+ and form Fe(OH)3 precipitates on the
93
Cu
Py
mineral surface (Eq. (6)). The presence of Fe3+ ions in the solution
may also act as oxidant species, enhancing oxidation of pyrite. The
presence of two possible sources of electron acceptor species in the
oxidation of pyrite (O2 and Fe3+) may explain the higher amounts
of sulphoxy species, SO3 , observed on the pyrite particles surfaces
compared with chalcopyrite particles.
FeS2 þ 3xCu2þ þ 4xH2 O $ Cu3x FeSð2xÞ þ xSO2
4 þ 8xH
þ
ð10Þ
Iron hydroxy/hydroxide species (FeOOH, Fe(OH)3) form on the
surface of these minerals under alkaline aerated solutions, Fe(OH)3
being the most likely abundant species. The presence of these spe-
cies on the individual mineral phases was examined in the flota-
tion concentrates (Fig. 8b). The exposure of FeOH+ species on Cp
particles was found to be approximately constant in the course
of flotation (C1–C4), in both tests; the effect was enhanced by
increasing the content of Py in the mineral mixture. In contrast,
Py particles were found to exhibit increasing exposure of FeOH+
throughout flotation. Previous studies of pyrite have shown that
a layer of Fe(OH)3 forms on the surface and that oxidation of pyrite
proceeds upon formation of this layer. Due to the porous structure
of the iron hydroxide layer, the oxygen can diffuse through the
layer, reach the surface and be reduced (Eq. (6)), further oxidising
pyrite, which explains the observed increase of FeOH+ on the pyrite
surfaces during flotation (Chander and Briceno, 1987; Ekmekci and
Demirel, 1997; Ahlberg and Broo, 1996).
94 C. Owusu et al. / Minerals Engineering 55 (2014) 87–95

(a) 0.6 (b) 0.25

Cu+ SH -
ToF-SIMS Intensity
0.5 0.2

ToF-SIMS Intensity
0.4
0.15
0.3
0.1
0.2

0.05
0.1

0 0
Con 1 Con 2 Con 3 Con 4 Con 1 Con 2 Con 3 Con 4
Flotation stream Flotation stream

Cp in Cp/Py: 20/80 Py in Cp/Py: 20/80 Cp in Cp/Py: 20/80 Py in Cp/Py: 20/80

Cp in Cp/Py: 80/20 Py in Cp/Py: 80/20 Cp in Cp/Py: 80/20 Py in Cp/Py: 80/20

(c) 0.003
C2H5OCS2-
(d) 0.4

0.35
O-
0.0025
ToF-SIMS Intensity

ToF-SIMS Intensity
0.3
0.002
0.25

0.0015 0.2

0.15
0.001
0.1
0.0005
0.05

0 0
Con 1 Con 2 Con 3 Con 4 Con 1 Con 2 Con 3 Con 4
Flotation stream Flotation stream
Cp in Cp/Py: 20/80 Py in Cp/Py: 20/80 Cp in Cp/Py: 20/80 Py in Cp/Py: 20/80
Cp in Cp/Py: 80/20 Py in Cp/Py: 80/20 Cp in Cp/Py: 80/20 Py in Cp/Py: 80/20

Fig. 7. ToF-SIMS secondary ion signals of Cu (a), SH (b) collector fragment C2H5OCS2 (c) and O (d) on Cp and Py surfaces in the concentrates (C1–C4) of the flotation tests with
20 and 80 wt.% Py additions.

0.06 0.025
- FeOH
+
SO3
0.05 0.02
ToF-SIMS Intensity
ToF-SIMS Intensity

0.04
0.015

0.03
0.01
0.02

0.01 0.005

0 0
Con 1 Con 2 Con 3 Con 4 Con 1 Con 2 Con 3 Con 4
Flotation stream Flotation stream
Cp in Cp/Py: 20/80 Py in Cp/Py: 20/80 Cp in Cp/Py: 20/80 Py in Cp/Py: 20/80
Cp in Cp/Py: 80/20 Py in Cp/Py: 80/20 Cp in Cp/Py: 80/20 Py in Cp/Py: 80/20

Fig. 8. ToF-SIMS signals of hydrophilic oxidation products SO3 and FeOH on chalcopyrite and pyrite surfaces in the concentrates (C1–C4) of the flotation tests with 20 and
80 wt.% Py additions.

4. Conclusions performance of both minerals, using a fully inert ceramic grinding

medium. After IsaMilling, chalcopyrite flotation kinetics and final
The overall aim of this study was to understand the effect of recovery decreased with increasing pyrite content in the mixture.
pyrite content in chalcopyrite/pyrite mixtures on the flotation The chalcopyrite grade of the concentrate also decreased as a result
 C. Owusu et al. / Minerals Engineering 55 (2014) 87–95
of higher pyrite recovery. The competition between Cp and Py for
collector adsorption was confirmed by the presence of collector
on both minerals surfaces from ToF-SIMS analysis. The pyrite
was copper-activated by Cu ions dissolved from chalcopyrite, as
evidenced by ToF-SIMS and XPS surface analyses. The galvanic
interactions occurring between the two minerals were enhanced
by increasing the pyrite content in the mixture (i.e. increasing
the cathode surface area), accelerating the oxidation of chalcopy-
rite but also of pyrite. ToF-SIMS surface analysis on flotation con-
centrates showed a decrease of the intensities of copper and
sulphur secondary ion signals together with an increase of oxygen,
iron hydroxide and sulphate species on Cp particles when the Py
content in the feed was increased. The cathodic reactions on the
pyrite surfaces (i.e., reduction of O2 to OH) were also favoured
by increasing the pyrite content in the mineral system, in turn, pro-
moting the formation of iron hydroxides on the surface, which was
found to increase in the course of flotation. The formation of sul-
phur species, SO 3 , on the surfaces of chalcopyrite and pyrite was
found to follow a similar rate during flotation, increasing with Py
content in the mixture. The exposure of hydrophilic oxidation
products, FeOH+ and SO
3 , was found to be greater on pyrite parti-
cles compared with chalcopyrite, regardless of their proportion in
the mineral mixture. The increase in the oxidation of the minerals
with increasing pyrite content in the feed is also supported by the
decrease of dissolved oxygen in the flotation pulp.
Based on the experimental evidence, it can be concluded that
the decrease in chalcopyrite recovery is a result of the formation
of hydrophilic oxidation species on the surface, which is enhanced
by the galvanic interactions with pyrite. The overall hydrophilic
loading on the surface plays a dominant role in the floatability of
chalcopyrite. Collector addition rate was not a key player in the flo-
tation of chalcopyrite, since there was sufficient on the particles
surfaces, but contributed to the flotation of pyrite.
The amount of dissolved oxygen and Eh of the pulp is critical,
since oxygen/Eh controls xanthate adsorption on chalcopyrite.
During processing of chalcopyrite-pyrite mixtures, it is likely that
different oxygen flow rates are required for optimum mineral sep-
aration, depending on the relative abundance of the two minerals
in the feed, and on pyrite reactivity. These preliminary findings
suggest the necessity of developing characterisation tools and pro-
cedures to assist plant operators in setting the optimum parame-
ters for mineral separation according to the feed mineralogy.
Acknowledgements
The financial support from the University of South Australia,
Australian Research Council (ARC), AMIRA international and Ghana
government, and discussion with Michael Young (Xstrata Technol-
ogy), Daniel Fornasiero and John Denman (UniSA) are gratefully
acknowledged.
References
Ahlberg, E., Broo, A.E., 1996. Oxygen reduction at sulphide minerals. The effect of
surface pre-treatment on the oxygen reduction at pyrite. Int. J. Miner. Process.
47, 49–60.
Aplan, F.F., Chander, S., 1988. Collectors for sulphide minerals flotation. In:
Somasundaran, P., Moudgil, B.M. (Eds.), Reagents in Mineral Technology.
Marcel Dekker Inc., NY, pp. 335–369.
95
Bruckard, W.J., Sparrow, G.J., Woodcock, J.T., 2011. A review of the effects of the
grinding environment on the flotation of copper sulphides. Int. J. Miner. Process.
100, 1–13.
Chander, S., Briceno, A., 1987. Kinetics of pyrite oxidation. Miner. Metall. Process.
33, 171–176.
Davenport, W.G., King, M., Schlesinger, M., 2002. Extractive Metallurgy of Copper,
Tech & Eng., Oxford OX5IGB, UK.
Ekmekci, Z., Demirel, H., 1997. Effects of galvanic interaction on collectorless
flotation behavior of chalcopyrite and pyrite. Int. J. Miner. Process. 52 (1), 31–
48.
Fairthorne, G., Fornasiero, D., Ralston, J., 1997. Effect of oxidation on the
collectorless flotation of chalcopyrite. Int. J. Miner. Process. 49 (1–2), 31–48.
Finkelstein, N.P., 1997. The activation of sulphide minerals for flotation: a review.
Int. J. Miner. Process. 52 (2–3), 81–120.
Frew, J. A., Davey, K. J., Glen, R. M., 1994. Effect of fine grinding on flotation
performance: distinguishing size from other effects. In: Fifth Mill Operators’
Conference, pp. 263–270.
Gaudin, A.M., 1939. Principles of Mineral Dressing. McGraw-Hill Book, New York.
Goktepe, F., Williams, K.P., 1995. Electrochemical effects in flotation of a Turkish
complex sulfide ore. Miner. Eng. 8 (9), 1035–1048.
Grano, S., 2009. The critical importance of the grinding environment on fine particle
recovery in flotation. Miner. Eng. 22, 386–394.
Habashi, F., 1978. Chalcopyrite: its chemistry and metallurgy. McGrand-Hill, USA.
He, S., Skinner, W., Fornasiero, D., 2006. Effect of oxidation potential and zinc
sulphate on the separation of chalcopyrite from pyrite. Int. J. Miner. Process. 80,
169–176.
Holmes, P.R., Crundwell, F.K., 2000. The kinetics of the oxidation of pyrite by ferric
ions and dissolved oxygen: an electrochemical study. Geochim. Cosmochim.
Acta 64 (2), 263–274.
Kuopanportt, H., Suorsa, T., Dahl, O., Niinimaki, J., 2000. A model of conditioning in
the flotation of a mixture of pyrite and chalcopyrite ores. Int. J. Miner. Process.
59, 327–338.
Kydros, K.A., Gallios, G.P., Matis, K.A., 1994. Modification of pyrite and sphalerite
flotation by dextrin. Sep. Sci. Technol. 29 (17), 2275–2663.
Pease, J.D., Curry, D.C., Barns, K.E., Young, M.F., Rule, C., 2006. Transforming flow
sheet design with inert grinding media-the IsaMill, paper 16. In: 38th Annual
Meeting of the Canadian Mineral processors, pp. 231–249.
Peng, Y., Grano, S., 2010. Effect of grinding media on the activation of pyrite
flotation. Miner. Eng. 23, 600–605.
Peng, Y., Grano, S., Fornasiero, D., Ralston, J., 2003. Control of grinding conditions in
the flotation of chalcopyrite and its separation from pyrite. Int. J. Miner. Process.
69 (1–4), 87–100.
Prestidge, C.A., Skinner, W.M., Smart, R., St, C., 1997. Copper (II) activation and
cyanide deactivation of zinc sulphide under mildly alkaline conditions. Appl.
Surf. Sci. 108 (3), 333–344.
Richardson, P.E., Walker, G.W., 1985. The flotation of chalcocite, bornite,
chalcopyrite, and pyrite in an electrochemical-flotation cell. In: XV Int. Min.
Process. Congress, pp. 198–210.
Rimstidt, D.J., Vaughan, D.J., 2003. Pyrite oxidation: a state-of-the-art assessment of
the reaction mechanism. Geochim. Cosmochim. Acta 67 (5), 873–880.
Ross, V.E., 1991. Comparison of methods for evaluation of true flotation and
entrainment. Trans. Inst. Miner. Metall. (Sect. C: Miner. Process. Extr. Metall.):
Inst. Min. Metall. 100, C121–C126.
Ross, V.E., Van Deventer, J.S.J., 1988. Mass transport in flotation column froths. In:
Sastry, K. (Ed.), Proceedings of an International Symposium. Littleton, CO: SME.
Shen, W.Z., Fornasiero, D., Ralston, J., 2001. Flotation of sphalerite and pyrite in the
presence of sodium sulphite. Int. J. Miner. Process. 63 (1), 17–28.
Skinner, W.M., Prestidge, C.A., Smart, R., St, C., 1996. Irradiation effects during X-ray
photoelectron spectroscopy studies of copper (II) activated zinc sulfide. Surf.
Interface Anal. 24 (9), 620–626.
Sutherland, D.N., 1989. Batch flotation behavior of composite particles. Miner. Eng.
2 (3), 351–367.
Van Deventer, J.S.J., Ross, V.E., Dunne, R.C., 1991. The effect of milling environment
on the selective flotation of chalcopyrite from a complex sulphide ore. Int.
Miner. Proc. Cong. Dresden 17, 129–140.
von Oertzen, G.U., Skinner, W.M., Nesbitt, H.W., Pratt, A.R., Buckley, A.N., 2007. Cu
adsorption on pyrite (100): Ab initio and spectroscopic studies. Surf. Sci. 601,
5794–5799.
Wang, X.H., Forssberg, K.S.E., 1991. Mechanisms of pyrite flotation with xanthates.
Int. J. Miner. Process. 33 (1–4), 275–290.
Xu, M., Finch, J.A., Rao, S.R., Lin, D., 1995. Reverse flotation of pyrite from a zinc-
concentrate using nitrogen. Miner. Eng. 8 (10), 1159–1173.
Ye, X., Gredelj, S., Skinner, W., Grano, S.R., 2010. Regrinding sulphide
minerals-Breakage Mechanisms in milling and their influence on surface
properties and flotation behaviour. Powder Technology 203 (2), 133–147.