Minerals Engineering 96–97 (2016) 83–93

Contents lists available at ScienceDirect

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

Effect of Mg2+ and Ca2+ as divalent seawater cations on the floatability of

molybdenite and chalcopyrite
Tsuyoshi Hirajima a,⇑, Gde Pandhe Wisnu Suyantara b, Osamu Ichikawa a, Ahmed Mohamed Elmahdy a,c,
Hajime Miki a, Keiko Sasaki a
a
Department of Earth Resources Engineering, Faculty of Engineering, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan
b
Graduate School of Engineering, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan
c
Central Metallurgical Research & Development Institute, PO Box 87, Helwan, Cairo, Egypt

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

Article history: Seawater flotation has been applied to mineral processing in areas located far from fresh water resources.
Received 8 December 2015 However, as seawater has a detrimental effect on molybdenite floatability under alkaline conditions
Revised 16 May 2016 (pH > 9.5), its application in the conventional copper and molybdenum (Cu-Mo) flotation circuit is hin-
Accepted 21 June 2016
dered. A fundamental study of the effect of two divalent cations in seawater, Mg2+ and Ca2+, on the floata-
Available online 2 July 2016
bility of chalcopyrite and molybdenite is presented in this paper. Floatability tests showed that both
MgCl2 and CaCl2 solutions depress the floatability of chalcopyrite and molybdenite at pH values higher
Keywords:
than 9. Furthermore, Mg2+ exerts a stronger effect than Ca2+ owing to the adsorption of Mg(OH)2 precip-
Flotation
MgCl2
itates on the mineral surfaces, as indicated by dynamic force microscopy images. The floatability of chal-
CaCl2 copyrite was significantly depressed compared with that of molybdenite in a 10
2 M MgCl2 aqueous
Seawater solution at pH 11. This phenomenon is likely due to the adsorption of hydrophilic complexes on the min-
Depression eral surface, which reduces the surface hydrophobicity. A reversal of the zeta potential of chalcopyrite in
Dynamic force microscopy MgCl2 and CaCl2 solutions at pH 11 and 8, respectively, indicated the adsorption of precipitates onto the
surface. In contrast, the zeta potential of molybdenite decreased continuously under the same conditions.
The floatability test of chalcopyrite and molybdenite in mixed systems showed that selective separation
of both minerals should be possible with the addition of emulsified kerosene to a 10
2 M MgCl2 solution
at pH 11. A mechanism is proposed to explain this phenomenon.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction Flotation is a physicochemical mineral separation process that

Flotation is a water-intensive process; however, due to the scar-
city of fresh water resources, increasing social demand for fresh
water access, and stringent regulation of industrial water use,
flotation units need to use groundwater, recycled water, or seawa-
ter, which have high electrolyte concentrations (Wang and Peng,
2014). Moreover, the use of seawater for flotation is more suitable
for mining operations located near the seashore or in areas that
lack access to ground and fresh water. However, from an economic
perspective, seawater flotation is only feasible if the production
output (i.e., concentrate grades and mineral recoveries) is compa-
rable to those of conventional fresh water flotation units.
Abbreviations: DFM, dynamic force microscopy; PAX, potassium amyl xanthate;
DTP, sodium diisobutyl dithiophosphate; EK, emulsified kerosene.
⇑ Corresponding author.
E-mail address: hirajima@mine.kyushu-u.ac.jp (T. Hirajima).
utilizes the differences in the surface properties of valuable miner-
als and unwanted gangue minerals (Wills and Napier-Munn, 2006).
In froth flotation, bubbles are used to separate hydrophobic min-
eral particles, as the hydrophobic particles stick to the bubbles,
float to the surface, and are recovered as a froth. The main chal-
lenges facing the use of seawater in copper-molybdenum sulfide
flotation are achieving a high recovery of not only copper but also
its valuable by-products (Mo and Au), managing pyrite depression
in seawater at lower pH, reducing the excess of lime consumption
in seawater, and obtaining a high copper concentrate grade for
high-pyrite copper ores (Castro, 2012). Moreover, seawater and
saline water are usually considered detrimental for the flotation
of copper-molybdenum ores.
However, the presence of some inorganic salts (e.g., KCl, NaCl,
Na2SO4, MgCl2, MgSO4, and CaCl2) in solutions and seawater is
reported to inhibit bubble coalescence and reduce bubble size,
thus improving bubble stability and providing adequate frothing

http://dx.doi.org/10.1016/j.mineng.2016.06.023
0892-6875/Ó 2016 Elsevier Ltd. All rights reserved.
84 T. Hirajima et al. / Minerals Engineering 96–97 (2016) 83–93
characteristics (Craig et al., 1993a, 1993b; Laskowski et al., 2013;
Quinn et al., 2014). For example, the flotation circuit in the Raglan
concentrator (Xstrata Nickel) in northern Quebec, Canada, can be
operated without the addition of frothers because the high salt
content of the process water reduces the bubble size and stabilizes
the froth layer (Quinn et al., 2007). Similar results were obtained
by Laskowski et al. (2013) and Castro et al. (2013), who showed
that fine bubbles can be obtained in seawater, even without the
addition of frothers.
Several sulfide minerals are processed in flotation plants using
saline water and seawater (Drelich and Miller, 2012; Moreno et al.,
2011; Wang and Peng, 2014). Bore water with high ionic strength
is used in nickel flotation plants (Mt Keith mine, Leinster mine,
and Kambalda nickel concentrator operated by BHP Billiton) in
Western Australia. In Chile, the Las Luces concentrator (Las Cenizas
Mining Group) employs seawater in a copper-molybdenum plant in
Taltal. Furthermore, the Batu Hijau concentrator (Newmont) in
Sumbawa Island, Indonesia, utilizes seawater to process a gold-
rich porphyry copper ore (chalcopyrite-bornite) (Castro, 2012).
In addition, seawater has been used to separate copper minerals
(e.g., chalcopyrite) from molybdenum minerals (e.g., molybdenite)
in copper-molybdenum (Cu-Mo) flotation units. Sodium hydrosul-
fide (NaHS) is employed to depress chalcopyrite in this process,
allowing molybdenite to be collected in the froth layer (Moreno
et al., 2011). However, NaHS produces toxic hydrogen sulfide gas
(H2S) at low pH values. Therefore, safer methods are needed to sep-
arate chalcopyrite and molybdenite.
In the absence of chalcopyrite depressants, molybdenite exhi-
bits poor floatability in seawater at high pH values, whereas chal-
copyrite recovery is slightly lower under the same conditions
(Castro, 2012; Laskowski et al., 2013; Ramos et al., 2013). This
behavior limits the application of seawater in conventional Cu-
Mo flotation circuits at high pH values. It was suggested that this
phenomenon might be caused by the precipitation of Mg(OH)2
on the bubble surface, which strongly affects bubble–particle inter-
actions. Moreover, the precipitation of Mg(OH)2 on the molybden-
ite surface may account for the depression of molybdenite
floatability. However, these studies employed ore samples contain-
ing various minerals (e.g., chalcopyrite, bornite, chalcocite, pyrite,
and molybdenite) and flotation reagents (e.g., collectors and
frothers), which, to some extent, may affect mineral floatability
and the mechanism involved during the flotation process. In addi-
tion, calcium and magnesium ions in seawater can form colloidal
hydroxides, carbonates, and sulfates, which can depress some min-
eral species, such as molybdenite (Castro, 2012). Therefore, a fun-
damental study of the effect of divalent cations in seawater (e.g.,
Mg2+ and Ca2+) on the floatability of pure chalcopyrite and molyb-
denite in the absence of flotation reagents is needed to understand
the basic mechanism of this process.
On the other hand, the floatability of molybdenite significantly
increases with increasing NaCl concentration due to the effect of
inorganic electrolytes on the compression of electrical double lay-
ers. Thus, the zeta potentials of the bubbles and particles are
reduced, and repulsive interactions are decreased (Lucay et al.,
2015). Furthermore, the increase in the floatability of molybdenite
was found to be independent of pH above 0.5 M ionic strength.
Similar results were obtained by Zhao and Peng (2014), who stud-
ied the effect of electrolytes (e.g., NaCl, LiCl, NaF, KCl, and NaI) on
the flotation of secondary copper minerals, such as chalcocite, in
the presence of clay minerals (e.g., kaolinite). An increase in chal-
cocite floatability with an increase in the electrolyte concentration
was attributed to the electrolytes reducing the electrostatic attrac-
tion between kaolinite and chalcocite, thus preventing the coating
of kaolinite on the chalcocite surface. Similarly, Raghavan and Hsu
(1984) demonstrated that CaCl2 improves the floatability of molyb-
denite in the presence of fuel oil as a molybdenite collector by
decreasing the surface potential of the molybdenite. Moreover,
Kurniawan et al. (2011) reported that inorganic electrolytes
enhance coal flotation performance, depending on the type and
concentration of the salts.
The enhancement of mineral floatability may be associated with
the effect of salt on destabilizing the hydration layer on the min-
eral surface and improving the surface hydrophobicity (Wang
and Peng, 2014). Troncoso et al. (2014) showed that the contact
angle of silica increases with an increase in the concentration of
various electrolytes (e.g., NaCl, CaCl2, and AlCl3), which indicates
that the mineral surface may become more hydrophobic in saline
solution. However, the stability of the hydration layer on chalcopy-
rite and molybdenite surfaces in MgCl2 solution improves at higher
pH values (Suyantara et al., 2016). A bubble–particle interaction
study demonstrated that the increase in the hydration layer stabil-
ity lengthens the bubble–particle induction time in a MgCl2 aque-
ous solution at pH 11 owing to the adsorption of Mg(OH)2
precipitates on the mineral surface, which reduces the surface
hydrophobicity.
Seawater contains cations (e.g., Na+, Mg2+, Ca2+, and K+) and
anions (e.g., Cl
, SO2



4 , HCO3 , and Br ), which affect the electric
charge at the bubble/aqueous solution and particle/aqueous solu-
tion interfaces. The mineral surface charge is affected by the elec-
trolyte type and solution ionic strength or concentration. At
constant surface charge density, the surface potential decreases con-
tinuously as the electrolyte concentration increases (Israelachvili,
2011). Li and Somasundaran (1991) and Han et al. (2004) showed
that the bubble/aqueous solution interface is negatively charged in
10
2 M NaCl and 10
2 M CaCl2 solutions at all pH values and posi-
tively charged in a 10
2 M MgCl2 solution in the pH range of 9.5–
10.5. In addition, Israelachvili (2011) demonstrated that relatively
small amounts of divalent ions substantially lower the magnitude
of the surface potential. On particle/aqueous interfaces, Ca2+ and
Mg2+ ions are reported to increase the magnitude of the original pos-
itive surface charge of aged dolomite particles and alter the original
negative surface charge of aged calcite particles (Kasha et al., 2015).
Moreover, the surface of sulfide minerals (e.g., chalcopyrite, tennan-
tite, and molybdenite) are negatively charged in 10
3 M KCl and
5  10
3 M KNO3 solutions in the pH range of 3–10 (Hirajima
et al., 2014; Petrus et al., 2012; Raghavan and Hsu, 1984).
In the flotation process, collectors physically adsorb on mineral
surfaces and alter the surface hydrophobicity. However, collectors
and inorganic ions may compete for adsorption sites on mineral
surfaces, affecting the flotation recovery in seawater or electrolyte
solutions. This effect can be seen in corundum flotation using
sodium dodecyl sulfate as a collector in sodium chloride and
sodium sulfate solutions. Fuerstenau and Pradip (2005) demon-
strated that corundum depression is greater at higher salt concen-
trations. On the other hand, an activation effect was observed when
sulfate ions were added to a solution containing sodium dodecyl
chloride. These results indicate the complexity of the interaction
between collectors and electrolyte solutions on a mineral surface.
In addition, the presence of inorganic electrolytes might affect the
adsorption of collectors on the mineral surface. Wang et al.
(2015) showed that the contact angle of sphalerite treated with
potassium amyl xanthate (PAX) decreases with increasing NaCl
concentration, which indicates a decrease in the adsorption of
PAX on the sphalerite surface owing to the competitive adsorption
of salt ions. The adsorption mechanism of collectors and inorganic
ions on mineral surfaces is driven by the mineral surface charge.
Therefore, it is important to understand the mineral surface charge
in electrolyte solutions or in the presence of collectors.
In the present work, the effect of divalent cations in seawater
(Ca2+ and Mg2+) on the flotation of pure molybdenite and chalcopy-
rite was intensively investigated. The effect of these ions on the
zeta potentials of molybdenite and chalcopyrite was also studied.
 T. Hirajima et al. / Minerals Engineering 96–97 (2016) 83–93 85

In addition, the possibility of selective separation of molybdenite solution (150 mL) at various concentrations (0–10
2 M) and pH
and chalcopyrite in these ionic solutions is assessed. values (4–11) under ultrasonication for 1 min. The pH of the result-
ing suspension was maintained for 10 min by adding KOH or HCl
solution, and then the solution was transferred into the Hallimond
2. Materials and methods
tube. The flotation test was started by injecting nitrogen gas
through a glass frit at a flow rate of 20 mL/min for 1 min. The float
2.1. Materials
and tailing fractions were collected separately, dried in an oven at
105 °C for 12 h, and then weighed. Kerosene was emulsified by
Chalcopyrite from Miyatamata Mine, Akita Prefecture, Japan,
mixing with a 0.01 M MgCl2 solution (150 mL) in a food blender
and molybdenite from Wolfram Camp, Queensland, Australia, were
(Palcookin Tescom TM5) at 12,000 rpm for 1 min. The emulsified
used in this study. CaCl2 and MgCl2 solutions of various concentra-
kerosene was then used for mineral treatment.
tions (10
4–10
2 M) were used as models for divalent cations in
seawater. Seawater typically contains 0.411 g/L (10
2 M) Ca2+
2.4. Contact angle measurement
and 1.278 g/L (5  10
2 M) Mg2+ (Dittmar, 1884). KOH and HCl
were used for solution pH adjustment. MilliporeÒ ultra-pure water
The contact angles of various mineral samples were measured
with a resistivity of 18.2 MXcm was used in all experiments. All
using a goniometer (Dropmaster 300, Kyowa Interface Science
reagents employed in this study were analytical-grade, with the
Co., Ltd.). The captive bubble method was used in these measure-
exception of kerosene.
ments. After immersing the mineral sample in the treatment solu-
For surface characterization using dynamic force microscopy
tion, a 0.1-lL air bubble was generated and attached to the mineral
(DFM) and contact angle measurements, the chalcopyrite was cut
surface using a microsyringe. The contact angle was measured
to provide a flat surface and polished using 400- to 4000-grit
automatically for 50 s using FAMAS version 3.7.2 software (Kyowa
emery papers, followed by 3- and 1-lm diamond sprays on DP-
Interface Science Co., Ltd.). This measurement process was
Mol (Struers, Germany) cloth and a DP-Nap (Struers, Germany) fine
repeated 3 times, and the average values of the contact angle read-
polishing cloth mounted on a polishing machine plate (Doctor Lap
ings are reported here.
ML 182, Maruto, Japan). To prepare the molybdenite samples, the
molybdenite was cut, and the cleaved surface was exposed and
2.5. Zeta potential measurement
polished consecutively with 3-, 1-, and ¼-lm diamond spray on
cotton. Following polishing, each mineral surface was rinsed with
Samples of each mineral powder (0.1 g) were suspended in
ethanol and dried with nitrogen gas; these steps were repeated 5
MgCl2 or CaCl2 solution (50 mL) at various concentrations (10
4–
times. The surface was then rinsed with ultra-pure water and dried
10
2 M) and pH values (2–11) under ultrasonication for 1 min.
with nitrogen gas; these steps were repeated 5 times.
The pH of the resulting suspension was maintained for 10 min by
The chalcopyrite sample was ground using an agate mortar and
adding KOH or HCl solution. The zeta potential measurements
pestle and then dry-screened. In the case of chalcopyrite, particle
were conducted using a zeta potential analyzer (Zetasizer Nano-
size fractions of +75
106 lm were employed for flotation experi-
zs, Malvern Co., Ltd.). At least three measurements were conducted
ments, and particle size fractions of
38 lm were used for zeta
for each pH condition. Precipitation was observed in MgCl2 and
potential measurements. The molybdenite fine powder (<30 lm)
CaCl2 solutions at high pH values (9–12), and the zeta potential
sample was supplied by Sumitomo Metal Mining, Co., Ltd. The min-
measurements for the precipitates were conducted in a similar
eral sample powders were prepared by adding the mineral powder
way as those for the mineral samples.
(5 g) to 1 M HNO3 (50 mL) under ultrasonication (Yamato 3510,
Branson) for 1 min. The resulting suspension was filtered and
2.6. Species diagrams for calcium and magnesium
rinsed with ultra-pure water until the filtrate pH was 5. The min-
eral powder was then immersed in acetone under vacuum condi-
Species diagrams were constructed for calcium and magnesium
tions and freeze-dried (FDU 1200, Eyela) for 12 h to remove
to describe the chemical species that may be present over the
water and acetone from the powder.
entire range of experimental conditions. The following reactions
were used to construct the calcium species diagram.
2.2. Dynamic force microscopy (DFM)
CO2ðgÞ þ H2 O $ H2 CO3ðaqÞ K1 ¼ 10
1:47 ð1Þ
Scanning DFM was conducted to obtain the surface topography H2 CO3ðaqÞ þ OH
ðaqÞ $ HCO
3ðaqÞ þ H2 O K2 ¼ 10 7:65
ð2Þ
and phase images continuously. The measurements were con-
ducted under ambient conditions of 27.7 ± 1.0 °C and 68.4 ± 6.8% HCO
3ðaqÞ þ OH
ðaqÞ $ CO2

3ðaqÞ þ H2 O K3 ¼ 10 3:67
ð3Þ
relative humidity. The scanning was conducted using an CaCO3ðsÞ $ Ca2þ 2

ðaqÞ þ CO3ðaqÞ K4 ¼ 10
8:34 ð4Þ
aluminum-coated silicon cantilever (SI-DF20, Epolead, Japan) with
a spring constant of 14 N/m and a resonance frequency of 129 kHz. Ca2þ
ðaqÞ þ HCO
3ðaqÞ $ CaHCOþ3ðaqÞ K5 ¼ 10 0:82
ð5Þ
The polished mineral sample was immersed in the treatment solu- Ca2þ
ðaqÞ þ OH
ðaqÞ $ CaOHþðaqÞ K6 ¼ 10 1:40
ð6Þ
tion (50 mL) for 10 min. Subsequently, the mineral surface was
CaOHþðaqÞ þ OH
ðaqÞ $ CaðOHÞ2ðaqÞ K7 ¼ 10 1:37
ð7Þ
dried with nitrogen gas. Surface images (256  256 pixel resolu-
tion) were collected for the untreated and treated samples using CaðOHÞ2ðsÞ $ Ca2þ
ðaqÞ þ 2OH
ðaqÞ K8 ¼ 10
5:22
ð8Þ
a NanoNavi S-image atomic force microscope (Seiko Instruments
Inc., Japan) at a scan rate of 0.96 Hz. The images were processed The equilibrium constants for calcium and carbonate compiled
by image flattening and 3rd-order image tilting prior to analysis by Somasundaran and Agar (1967) were used in this study. The
using SPIP software (Image Metrology, Denmark). CO2 partial pressure used in this calculation was 4  10
4 atm.
To construct the magnesium species diagram, the following reac-
tions were used.
2.3. Flotation study


MgðOHÞ2ðsÞ $ Mg2þ
ðaqÞ þ 2OHðaqÞ K9 ¼ 10
10:71 ð9Þ
Mineral floatability tests were conducted using a Hallimond
MgOHþðaqÞ $ Mg2þ
ðaqÞ þ OH
ðaqÞ K10 ¼ 10
2:445
ð10Þ
tube. Mineral powder (0.5 g) was suspended in MgCl2 or CaCl2
86 T. Hirajima et al. / Minerals Engineering 96–97 (2016) 83–93


MgðOHÞ2ðaqÞ $ Mg2þ
ðaqÞ þ 2OHðaqÞ K11 ¼ 10
0:019
The equilibrium constants for reactions (9)–(11) were calcu-
lated from the free energies of formation of each possible species
(Wagman et al., 1982). The given equilibrium constants were used
to calculate the species diagrams.
3. Results and discussion
The effect of CaCl2 and MgCl2 on the floatability of chalcopyrite
and molybdenite was examined. In the absence of CaCl2 and MgCl2,
the recovery of chalcopyrite remained constant with increasing pH
(Fig. 1). Moreover, at pH values lower than 9, particularly at pH 7–
8, there was no significant difference in the floatability of chalcopy-
rite in the presence or absence of salts. However, MgCl2 signifi-
cantly depressed the recovery of chalcopyrite at concentrations
higher than 10
3 M and pH values higher than 9, whereas CaCl2
only had a marginal impact on the floatability of chalcopyrite at
the same concentrations and pH values. Similar detrimental effects
of MgCl2 on the floatability of chalcopyrite were reported by
Nagaraj and Farinato (2014), whereas CaCl2 was reported to have
no significant effect on the floatability of chalcopyrite, which is
likely due to the addition of sodium diisobutyl dithiophosphate
(DTP) as a chalcopyrite collector.
This decrease in the floatability of chalcopyrite at pH values
higher than 9 contradicts several studies reporting that chalcopy-
rite floatability is not sensitive to pH in seawater solutions
(Castro, 2012; Laskowski et al., 2013; Ramos et al., 2013). This con-
tradiction may be a result of the higher pulp density (35%)
employed in the previous studies, which increases the number of
mineral particles available in the flotation system and reduces
the concentration of adsorbed hydrophilic precipitates on the par-
ticle surfaces. In the present study, the pulp density was very low
(0.3%). Therefore, the concentration of adsorbed precipitates on
the mineral surfaces was higher, and the hydrophilic precipitates
had a more profound effect on the floatability.
The different behaviors observed may also be the result of the
presence of other electrolytes in seawater (e.g., NaCl) that mitigate
the detrimental effect of MgCl2 on the floatability of chalcopyrite at
high pH values by reducing the repulsive interactions between the
bubbles and particles. Furthermore, the addition of copper-
selective collectors (e.g., Sascol-95 (isopropyl ethyl thionocarba-
mate), Matcol TC-123 (n-butanol and pentanol isomers), Matcol
D-101 (modified dithiocarbamate), IsobX, MX-7017 (modified
dithiocarbamate), and MX-945 (modified dithiocarbamate)) in
the previous study could improve the floatability of chalcopyrite.
Although seawater precipitates might affect the adsorption of
100
(A)
90
80
70
Recovery, %
60
50
40 CaCl2 0 M
-4
30 CaCl2 10 M
-3
CaCl2 10 M
20 -2
CaCl2 10 M
10
0 0
3 4 5 6 7 8
pH
Fig. 1. Recovery of chalcopyrite in (A) CaCl2 and (B) MgCl2 solutions at various concentrations and pH values.
ð11Þ these collectors on chalcopyrite surfaces, this requires further
investigation and is beyond the scope of this work. Moreover, the
presence of multi-ion components in seawater and the mixed min-
erals (copper, iron, and molybdenum minerals) employed in the
previous study may lead to the involvement of different mecha-
nisms during the flotation process. In contrast, the floatability test
in the present work was conducted without any collectors or
frothers and focused on the effect of divalent cations on the natural
floatability of single minerals (i.e., chalcopyrite and molybdenite).
In addition, Castro (2012) and Smith and Heyes (2012) demon-
strated that chalcopyrite floatability is not sensitive to the pres-
ence of salts. As Smith and Heyes conducted their work at pH 8,
their results are in agreement with the chalcopyrite floatability
presented in Fig. 1. Meanwhile, Castro et al. employed collectors
in their experiments that, to some extent, improve the floatability
of chalcopyrite in seawater. Therefore, the presence of collectors
makes it difficult to determine whether the floatability of chal-
copyrite was sensitive to the presence of salts. On the other hand,
in the present work, the floatability of chalcopyrite was deter-
mined in the absence of flotation reagents (i.e., collectors and
frothers); thus, the effect of CaCl2 and MgCl2 on chalcopyrite and
molybdenite floatability could be observed directly.
Similar trends were observed for the effects of CaCl2 and MgCl2
on molybdenite floatability (Fig. 2), with the molybdenite recovery
decreasing with increasing concentrations of CaCl2 and MgCl2, par-
ticularly at concentrations higher than 10
3 M and pH values
higher than 9. Nagaraj and Farinato (2014) demonstrated a similar
detrimental effect of MgCl2 on the floatability of molybdenite.
However, in contrast to the present work, they showed that Ca2+
ions had no impact on molybdenum and copper recoveries at
either pH 9.5 or 10.8. This difference in behavior might be caused
by the effect of pulp density on the floatability of molybdenite, as
discussed above, and the effect of the addition of the copper-
selective collector DTP on the floatability of chalcopyrite. Further-
more, the particle size employed in Nagaraj and Farinato’s work
was P80 of 106 lm after milling, whereas the molybdenite particle
size employed in this work was <30 lm. Raghavan and Hsu (1984)
reported that the floatability of molybdenite improves with
increasing molybdenite particle size. This result may be caused
by an increase in the natural floatability of molybdenite with
increasing face/edge aspect ratio of the particles.
As shown in Fig. 1, the chalcopyrite recovery at pH 11 was sig-
nificantly lower in the 10
2 M MgCl2 solution compared with that
in the 10
2 M CaCl2 solution. The chalcopyrite recovery was 60%
and 18% in 10
2 M CaCl2 and MgCl2 solutions, respectively, at pH
11. Meanwhile, under the same conditions, molybdenite recovery
only decreased slightly in a 10
2 M MgCl2 solution (Fig. 2). The
100
(B)
90
80
70
Recovery, %
60
50
40 MgCl2 0 M
-4
30 MgCl2 10 M
-3
MgCl2 10 M
20 -2
MgCl2 10 M
10
9 10 11 12 3 4 5 6 7 8 9 10 11 12
pH
 T. Hirajima et al. / Minerals Engineering 96–97 (2016) 83–93 87

molybdenite recovery was 60% and 50% in 10
2 M CaCl2 and by Li and Somasundaran (1991). The calcium species diagrams
MgCl2 solutions at pH 11, respectively. These mineral recovery show that CaCO3 precipitate occurs at pH values higher than 8.3
results indicate that, in comparison with CaCl2, MgCl2 significantly and 9.3 in 10
2 M and 10
4 M CaCl2 solutions, respectively
depresses chalcopyrite floatability but only slightly reduces molyb- (Fig. 3A and B). The formation of CaCO3 precipitate indicates that
denite recovery at high pH, suggesting that the selective separation the slight decrease in the floatability of both minerals at pH values
of these two minerals should be possible in 10
2 M MgCl2 solution higher than 9 might be caused by the presence of CaCO3 in the
at pH 11. solution.
The effect of solution concentration on mineral floatability can In the magnesium species diagrams (Fig. 3C and D), Mg(OH)2
be explained using magnesium and calcium species diagrams precipitate is observed at pH values higher than 9.4 and 10.4 in
(Fig. 3). A similar magnesium species diagram has been reported 10
2 M and 10
4 M MgCl2 solutions, respectively. MgCl2 exhibits

100 100
(A) (B)
90 90
80 80

70 70
Recovery, %

Recovery, %
60 60

50 50

40 CaCl2 0 M 40 MgCl2 0 M
-4 -4
CaCl2 10 M MgCl2 10 M
30 -3 30
-3
CaCl2 10 M MgCl2 10 M
20 -2 20 -2
CaCl2 10 M MgCl2 10 M
10 10

0 0
3 4 5 6 7 8 9 10 11 12 3 4 5 6 7 8 9 10 11 12
pH pH

Fig. 2. Recovery of molybdenite in (A) CaCl2 and (B) MgCl2 solutions at various concentrations and pH values.

0 0
-2 CO32- -4 CO32-
+ (A) 10 M CaCl2 - + (B) 10 M CaCl2 -
H OH H OH
-2 2+
-2
Ca
CaCO3 (s) Ca
2+ CaCO3 (s)
-log(Concentration)

-log(Concentration)

-4 -4
CaCO3 (aq) CaCO3 (aq)
-6 H2CO3 -6 H2CO3
CaHCO3+

CaHCO3+
-8 -8
+ +
CaOH CaOH
-10 - -10 -
HCO3 HCO3

-12 -12
Ca(OH)2 (aq) Ca(OH)2 (aq)

-14 -14
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
pH pH

0 -
0 -
+ -2 + -4
H (C) 10 M MgCl2 OH H (D) 10 M MgCl2 OH
-2 -2
2+ 2+
Mg Mg(OH)2 (s) Mg Mg(OH)2 (s)
-log(Concentration)

-log(Concentration)

-4 -4

-6 -6

-8 -8
+ +
MgOH MgOH
-10 -10

-12 Mg(OH)2 (aq) -12 Mg(OH)2 (aq)

-14 -14
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
pH pH

Fig. 3. (A and B) Calcium and (C and D) magnesium species diagrams at 10
2 M and 10
4 M.
88 T. Hirajima et al. / Minerals Engineering 96–97 (2016) 83–93

detrimental effects on chalcopyrite and molybdenite recovery at cipitate in the seawater precipitate obtained by titration with
pH values higher than 9, indicating that the depression of mineral NaOH.
floatability is likely due to the presence of Mg(OH)2 precipitate in To understand the wettability of chalcopyrite and molybdenite,
the solution. At lower concentrations (i.e., 10
4 M MgCl2), the con- contact angle measurements were conducted. The results are
centration of Mg(OH)2 is low and thus cannot significantly affect shown in Figs. 4 and 5 for chalcopyrite and molybdenite, respec-
the floatability of both minerals. Similar findings were reported tively, in CaCl2 and MgCl2 solutions at various concentrations and
by Nagaraj and Farinato (2014), who only observed Mg(OH)2 pre- pH values. The contact angle profiles in both figures show that

90 90
(A) (B)
80 80

70 70
Contact angle, deg

Contact angle, deg

60 60

50 50

40 40

30 CaCl2 0 M 30 MgCl2 0 M
-4 -4
CaCl2 10 M MgCl2 10 M
20 -3 20 -3
CaCl2 10 M MgCl2 10 M
10 -2 10 -2
CaCl2 10 M MgCl2 10 M
0 0
3 4 5 6 7 8 9 10 11 12 3 4 5 6 7 8 9 10 11 12
pH pH

Fig. 4. Contact angles of chalcopyrite in (A) CaCl2 and (B) MgCl2 solutions at various concentrations and pH values.

90 90
(A) (B)
80 80

70 70
Contact angle, deg

Contact angle, deg

60 60

50 50

40 40

30 CaCl2 0 M 30 MgCl2 0 M
-4 -4
CaCl2 10 M MgCl2 10 M
20 -3 20 -3
CaCl2 10 M MgCl2 10 M
-2 -2
10 CaCl2 10 M 10 MgCl2 10 M

0 0
3 4 5 6 7 8 9 10 11 12 3 4 5 6 7 8 9 10 11 12
pH pH

Fig. 5. Contact angles of molybdenite in (A) CaCl2 and (B) MgCl2 solutions at various concentrations and pH values.

50 50
(A) (B)
40 40
30 -2 30 -2
Molybdenite in 10 M MgCl2 Chalcopyrite in 10 M MgCl2
Zeta potential, mV

Zeta potential, mV

-3 -3
20 Molybdenite in 10 M MgCl2 20 Chalcopyrite in 10 M MgCl2
-4 -4
10 Molybdenite in 10 M MgCl2 10 Chalcopyrite in 10 M MgCl2

0 0

-10 -10

-20 -20

-30 -30

-40 -40

-50 -50
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
pH pH

Fig. 6. Zeta potential of (A) molybdenite and (B) chalcopyrite in MgCl2 solution at various concentrations.
 T. Hirajima et al. / Minerals Engineering 96–97 (2016) 83–93 89

50 50
(A) (B)
40 -2 40 -2
Molybdenite in 10 M CaCl2 Chalcopyrite in 10 M CaCl2
-3 -3
30 Molybdenite in 10 M CaCl2 30 Chalcopyrite in 10 M CaCl2

Zeta potential, mV

Zeta potential, mV
-4 -4
20 Molybdenite in 10 M CaCl2 20 Chalcopyrite in 10 M CaCl2

10 10

0 0

-10 -10

-20 -20

-30 -30

-40 -40

-50 -50
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
pH pH

Fig. 7. Zeta potential of (A) molybdenite and (B) chalcopyrite in CaCl2 solution at various concentrations.

the mineral contact angles decrease with increasing concentration compared with CaCl2, MgCl2 has a stronger effect on surface
of CaCl2 and MgCl2, although concentration has no significant hydrophobicity at high pH values. Indeed, the increase of surface
effect on the contact angle of molybdenite and chalcopyrite in wettability stabilizes the liquid layer on the surface and lengthens
the low pH range. Moreover, the decrease in the contact angle the induction time (i.e., the time required for bubbles to remove
becomes more profound with increasing pH. the intervening liquid layer on particle surfaces and form bub-
In contrast, Troncoso et al. (2014) showed that the contact angle ble–particle attachments), resulting in poor mineral floatability.
of quartz increases with increasing salt concentration (e.g., NaCl, Suyantara et al. (2016) demonstrated that the induction time is
CaCl2, and AlCl3). This difference in the effect of salt concentration higher on chalcopyrite and molybdenite surfaces in a 0.01 M MgCl2
on mineral contact angle may be caused by a difference in the solution at pH 11.
mechanism involved at the solid surface. In the present work, the The zeta potentials of chalcopyrite and molybdenite as a func-
decrease in contact angle is caused by the adsorption of Mg(OH)2 tion of MgCl2 and CaCl2 concentration are shown in Figs. 6 and 7,
and CaCO3 precipitates on the surface, which alters the surface respectively. The zeta potentials in the molybdenite profile are
hydrophobicity, as will be discussed later. Meanwhile, in the previ- negative over the entire pH range in both CaCl2 and MgCl2 solu-
ous work, the increase in contact angle is likely due to a weakening tions. A similar zeta potential trend was observed for chalcopyrite
of the electrostatic field by the counter ions adhered to the solid at lower pH values. However, the zeta potential was reversed at pH
surface (i.e., compression of the electrical double layer that 11 and 8 in 10
2 M MgCl2 and CaCl2 solutions, respectively. The
decreases repulsive interactions with increasing concentration). decrease in the magnitude of the zeta potential of both minerals
Thus, fewer hydrogen bonds are formed between water molecules with increasing salt concentration (Figs. 6 and 7) indicates a
and the hydroxyl groups (silanol) of the quartz surface. decrease in the electrostatic repulsion between solid surfaces,
The observed contact angles (Figs. 4 and 5) confirm the mineral which should enhance the floatability of the minerals. Therefore,
floatability results depicted in Figs. 1 and 2. Indeed, the contact the decrease in the floatability of the minerals cannot be attributed
angle profiles show that the chalcopyrite and molybdenite surfaces to the decrease in zeta potential and is instead likely caused by the
became more hydrophilic with increasing CaCl2 and MgCl2 concen- increase of surface wettability in MgCl2 and CaCl2 solutions, as
trations and pH values. Therefore, the low floatability of both min- indicated by the contact angle measurements (see Figs. 4 and 5).
erals was due to an increase of the surface wettability. Moreover,
100
50 90
40 80
30 70
Recovery, %

20
Zeta potential, mV

60 Chalcopyrite
10 50 Molybdenite
Cu (mixed minerals)
0 40 Mo (mixed minerals)
-10 30
-20 20
-2
Precipitate from 10 M CaCl2
-30 10
-2
Precipitate from 10 M MgCl2
-40 0
Fresh
-10 0 10 20 30 40 50
-50 water
7 8 9 10 11 12 13 14 Kerosene concentration, mg kerosene/g mineral
pH
Fig. 9. Effect of kerosene addition on the floatability of chalcopyrite and molyb-
Fig. 8. Zeta potential of precipitates obtained from 10
2 M MgCl2 and CaCl2 denite (single and mixed minerals with a chalcopyrite-to-molybdenite ratio of 1:1)
solutions. in 10
2 M MgCl2 solution at pH 11.
90 T. Hirajima et al. / Minerals Engineering 96–97 (2016) 83–93

100 100 -3
-3
MoS2 in 10 M KCl CuFeS2 in 10 M KCl
80 -2
(A) 80 -2
(B)
MoS2 in 10 M MgCl2 CuFeS2 in 10 M MgCl2
-3 -3
60 MoS2 and EK in 10 M KCl 60 CuFeS2 and EK in 10 M KCl
-2 -2
Zeta potential, mV MoS2 and EK in 10 M MgCl2 CuFeS2 and EK in 10 M MgCl2

Zeta potential, mV
40 -3
40 -3
EK in 10 M KCl EK in 10 M KCl
-2
20 -2
EK in 10 M MgCl2 20 EK in 10 M MgCl2

0 0

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
pH pH

Fig. 10. Zeta potential of (A) molybdenite (MoS2) and (B) chalcopyrite (CuFeS2) with emulsified kerosene (EK) in 10
2 M MgCl2 and 10
3 M KCl solutions.

Fig. 11. Dynamic force microscopy images of chalcopyrite surfaces: (A) Untreated surface, (B) surface treated in a 10
2 M MgCl2 solution at pH 11, and (C) surface treated in
10
2 M MgCl2 solution containing emulsified kerosene at pH 11. Area scanned = 2 lm  2 lm. Left: 3D topography images, center: 2D topography images, right: phase
images.
 T. Hirajima et al. / Minerals Engineering 96–97 (2016) 83–93 91

The poor floatability of chalcopyrite at high pH values is likely
caused by an improvement of the surface wettability owing to
the adsorption of hydrophilic complexes on the chalcopyrite
surface, as indicated by the reversal of the zeta potential of
chalcopyrite in 10
2 M MgCl2 and CaCl2 solutions. Moreover, the
precipitates obtained from 10
2 M MgCl2 and CaCl2 solutions
exhibited positive zeta potentials at high pH values (Fig. 8). Hence,
the reversal of the zeta potential of chalcopyrite is caused by the
adsorption of precipitates on the surface. Under the same condi-
tions, although the adsorption of precipitates altered the
hydrophobicity of the molybdenite surface, the floatability of
molybdenite was higher than that of chalcopyrite because of a
decrease in electrostatic repulsion owing to the adsorption of
positively charged precipitates and an increase in solution
concentration.
The flotation results (Figs. 1 and 2) indicate the potential to selec-
tively separate molybdenite and chalcopyrite in a 10
2 M MgCl2
solution at pH 11. To achieve higher selectivity, kerosene was added
to this solution as a flotation collector by an emulsification process.
Kerosene was chosen because it is usually used together with diesel
oil as a molybdenite collector (Bulatovic, 2007; Hirajima et al., 2014)
and has a long-chain carbon structure that is not significantly
affected by electrolyte ions. The mineral recoveries presented in
Fig. 9 show the effect of kerosene addition on chalcopyrite and
molybdenite flotation in a 10
2 M MgCl2 solution at pH 11. Single-
mineral flotation results suggest that the floatability of molybdenite
increased with the addition of emulsified kerosene (16 mg/g min-
eral), whereas the floatability of chalcopyrite remained low. The
flotation results for mixtures of molybdenite and chalcopyrite
(1:1) showed similar trends to those for the single mineral systems.
The zeta potentials of molybdenite, chalcopyrite, and kerosene
in a 10
2 M MgCl2 solution are presented in Fig. 10. Kerosene
had no significant effect on the zeta potential of chalcopyrite in
10
3 M KCl in the high-pH range, indicating low kerosene adsorp-
tion on the chalcopyrite surface. However, the reversal of the
zeta potential of chalcopyrite with emulsified kerosene in a

Fig. 12. Dynamic force microscopy images of molybdenite surfaces: (A) Untreated surface, (B) surface treated in a 10
2 M MgCl2 solution at pH 11, and (C) surface treated in
10
2 M MgCl2 solution containing emulsified kerosene at pH 11. Area scanned = 2 lm  2 lm. Left: 3D topography images, center: 2D topography images, right: phase
images.
92 T. Hirajima et al. / Minerals Engineering 96–97 (2016) 83–93
10
2 M MgCl2 solution at high pH values (Fig. 10) indicates that
kerosene could not prevent the adsorption of precipitates on the
surface of chalcopyrite.
On the other hand, the magnitude of the zeta potential of
molybdenite increased with the addition of emulsified kerosene
in a 10
3 M KCl solution. In other words, more kerosene was
adsorbed on the molybdenite surface than on the chalcopyrite
surface. Moreover, kerosene also increased the magnitude of the
zeta potential of molybdenite in the presence of precipitates in a
10
2 M MgCl2 solution at high pH values, although the zeta poten-
tial of kerosene in the absence of molybdenite reversed under the
same conditions, indicating the formation of kerosene and precip-
itate aggregates. This phenomenon indicates that the adsorption of
kerosene on the molybdenite surface reduces the adsorption of
precipitate on the surface, resulting in an increase of the magni-
tude of the zeta potential of molybdenite.
DFM measurements confirm this phenomenon by showing that
more precipitates are deposited on the chalcopyrite surfaces
(Fig. 11) than on the molybdenite surface (Fig. 12). The DFM
images (Fig. 11C) show a significant amount of precipitate on the
chalcopyrite surface, even after the addition of emulsified kerosene
to a 10
2 M MgCl2 solution at pH 11. In contrast, the coverage of
precipitate on the molybdenite surface is lower (Fig. 12C). The
phase images (Figs. 11 and 12) indicate a phase lag due to surface
elasticity, adhesion, and friction. Brighter colors in the phase
images correspond to greater phase lags (more hydrophilic sur-
faces). The phase image of chalcopyrite in the presence of kerosene
(Fig. 11C) shows higher phase lag compared with the untreated or
treated surface in 10
2 M MgCl2 at pH 11 (Fig. 11A and B, respec-
tively), indicating higher surface adhesion and a more hydrophilic
surface. However, the phase image of molybdenite under the same
conditions (Fig. 12C) has a darker color than the treated surface in
10
2 M MgCl2 at pH 11 (Fig. 12B), which indicates lower surface
adhesion and a less hydrophilic surface, resulting in higher
floatability.
From the zeta potential results and the surface characterization
using DFM, the following mechanism can be proposed. First, pre-
cipitates of Mg(OH)2 are formed in a 10
2 M MgCl2 solution at
pH 11. The precipitates adhere to the chalcopyrite and molybden-
ite surfaces, altering the inherent hydrophobicity of the minerals.
However, the molybdenite surface is covered by fewer precipitates
than the chalcopyrite surface, as indicated by the pH-dependent
behavior of the zeta potential, which remains negative for molyb-
denite, but shows a reversal for chalcopyrite. This phenomenon is
confirmed by the DFM images. In the presence of kerosene, the pre-
cipitates adhere to emulsified kerosene and form aggregates.
Under this condition, the precipitates remain on the chalcopyrite
surface owing to the low kerosene adsorption on the surface. How-
ever, in the case of molybdenite, kerosene is adsorbed on the
molybdenite surface, which reduces the adsorption of precipitates
on the surface, as indicated by the increase in the magnitude of the
zeta potential of molybdenite in the presence of kerosene in a
10
2 M MgCl2 solution at pH 11. As a result, the enhanced molyb-
denite floatability can be attributed to the effect of kerosene on
reducing the adsorption of precipitates on the molybdenite surface,
thus improving the molybdenite surface hydrophobicity. On the
other hand, the floatability of chalcopyrite remains low due to
the limited kerosene adsorption on its surface.
4. Conclusions
The effect of Ca2+ and Mg2+ as representative divalent cations in
seawater on chalcopyrite and molybdenite floatability was investi-
gated in this study. Both CaCl2 and MgCl2 significantly reduce the
floatability of chalcopyrite and molybdenite at pH values higher
than 9 owing to the adsorption of Mg(OH)2 and CaCO3 precipitates
on the mineral surfaces, which reduces the surface hydrophobicity.
Compared with CaCl2, MgCl2 reduces the floatability of both miner-
als to a greater extent. The depression caused by both CaCl2 and
MgCl2 is proportional to their concentrations. Therefore, control
of the concentration of both salts is important in seawater floata-
tion. In addition, MgCl2 and CaCl2 solutions reduce both the
hydrophobicity and surface charge of chalcopyrite and molybden-
ite. However, the zeta potential of chalcopyrite reverses in
10
2 M MgCl2 and CaCl2 solutions at pH 11 and 8, respectively,
owing to the adsorption of precipitates on the chalcopyrite
surface. The selective separation of molybdenite and chalcopyrite
should be possible with the addition of emulsified kerosene to a
10
2 M MgCl2 solution at pH 11. The mechanism proposed based
on DFM surface characterization and zeta potentials suggest that
the molybdenite surface is covered by fewer precipitates in the
presence of kerosene, which is likely the reason for the increased
molybdenite floatability.
Acknowledgements
This work was supported by a Grant-in-Aid for Science Research
(JSPS KAKENHI Grant No. 15H02333) from the Japan Society for the
Promotion of Science (JSPS), Japan, the Sumitomo Metal Mining
Co., Ltd., Japan, and the Ministry of Education, Culture, Sports,
Science, and Technology (MEXT), Japan.
References
Bulatovic, S.M., 2007. Handbook of Flotation Reagents. Elsevier Science &
Technology Books, Amsterdam.
Castro, S., 2012. Challenges in flotation of Cu-Mo sulfide ores in sea water. In:
Drelich, J. (Ed.), Water in Mineral Processing. Society for Mining, Metallurgy,
and Exploration, Colorado, USA, pp. 29–40.
Castro, S., Miranda, C., Toledo, P., Laskowski, J.S., 2013. Int. J. Miner. Process. 124, 8–
14.
Craig, V.S.J., Ninham, B.W., Pashley, R.M., 1993a. Nature 364, 317–319.
Craig, V.S.J., Ninham, B.W., Pashley, R.M., 1993b. J. Phys. Chem. 97, 10192–10197.
Dittmar, W., 1884. Report on researches into the composition of ocean water,
collected by the HMS Challenger. Challenger Reports, Physics and Chemistry.
Drelich, J., Miller, J.D., 2012. Induction time measurements for air bubbles on
chalcopyrite, bornite, and gold in seawater. In: Presented at the Water in
Mineral Processing, Society for Mining, Metallurgy, and Exploration,
Englewood, CO, USA, pp. 73–85.
Fuerstenau, D.W., Pradip, 2005. Adv. Colloid Interface Sci. 114–115, 9–26
(Dedicated to the Memory of Dr. Hans Joachim Schulze).
Han, M.Y., Ahn, H.J., Shin, M.S., Kim, S.R., 2004. Water Sci. Technol. 50, 49–56.
Hirajima, T., Mori, M., Ichikawa, O., Sasaki, K., Miki, H., Farahat, M., Sawada, M.,
2014. Miner. Eng. Froth Flotat. 66–68, 102–111.
Israelachvili, J.N., 2011. Intermolecular and Surface Forces, third ed. Academic
Press, Burlington, USA.
Kasha, A., Al-Hashim, H., Abdallah, W., Taherian, R., Sauerer, B., 2015. Colloids Surf.
A 482, 290–299.
Kurniawan, A.U., Ozdemir, O., Nguyen, A.V., Ofori, P., Firth, B., 2011. Int. J. Miner.
Process. 98, 137–144.
Laskowski, J.S., Castro, S., Ramos, O., 2013. Physicochem. Probl. Miner. Process. 50,
17–29.
Li, C., Somasundaran, P., 1991. J. Colloid Interface Sci. 146, 215–218.
Lucay, F., Cisternas, L.A., Gálvez, E., Lopez-Valdivieso, A., 2015. Miner. Metall.
Process. 32, 203–208.
Moreno, P.A., Aral, H., Cuevas, J., Monardes, A., Adaro, M., Norgate, T., Bruckard, W.,
2011. Miner. Eng. 24, 852–858.
Nagaraj, D.R., Farinato, R., 2014. Chemical factor effects in saline and hypersaline
waters in the flotation of Cu and Cu-Mo ores. In: Presented at the XXVII
International Mineral Processing Congress, Santiago, Chile.
Petrus, H.T.B.M., Hirajima, T., Sasaki, K., Okamoto, H., 2012. Int. J. Miner. Process.
102–103, 116–123.
Quinn, J.J., Kracht, W., Gomez, C.O., Gagnon, C., Finch, J.A., 2007. Miner. Eng. 20,
1296–1302.
Quinn, J.J., Sovechles, J.M., Finch, J.A., Waters, K.E., 2014. Miner. Eng. 58, 1–6.
Raghavan, S., Hsu, L.L., 1984. Int. J. Miner. Process. 12, 145–162.
Ramos, O., Castro, S., Laskowski, J.S., 2013. Miner. Eng. 53, 108–112.
Smith, L.K., Heyes, G.W., 2012. The effect of water quality on the collectorless
flotation of chalcopyrite and bornite. In: Water in Mining. Presented at the 3rd
International Congress on Water Management in the Mining Industry, Santiago,
Chile.
Somasundaran, P., Agar, G.E., 1967. J. Colloid Interface Sci. 24, 433–440.
 T. Hirajima et al. / Minerals Engineering 96–97 (2016) 83–93 93

Suyantara, G.P.W., Hirajima, T., Elmahdy, A.M., Miki, H., Sasaki, K., 2016. Colloids
Surf. A 501, 98–113.
Troncoso, P., Saavedra, J.H., Acuña, S.M., Jeldres, R., Concha, F., Toledo, P.G., 2014. J.
Colloid Interface Sci. 424, 56–61.
Wagman, D.D., Evans, W.H., Parker, V.B., Schumm, R.H., Halow, I., Bailey, S.M.,
Churney, K.L., Nuttall, R.L., 1982. J. Phys. Chem. Ref. Data 11, 2–260.
Wang, B., Peng, Y., 2014. Miner. Eng. Froth Flotat. 66–68, 13–24.
Wang, J., Xie, L., Liu, Q., Zeng, H., 2015. Miner. Eng. 77, 34–41.
Wills, B.A., Napier-Munn, T.J., 2006. Mineral Processing Technology, seventh ed.
Elsevier Science & Technology Books.
Zhao, S., Peng, Y., 2014. Miner. Eng. Froth Flotat. 66–68, 152–156.